MXPA98002392A - Method and modeling device through deposition for the formation of objects and standardstridimension - Google Patents

Abstract

Data manipulation methods in a thermal sterolitography apparatus, characterized in that the data represent a plurality of start / stop transitions, to facilitate the calculation of Boolean operations

Description

METHOD AND DEVICE OF MODELING BY SELECTIVE DEPOSITION FOR THE FORMATION OF OBJECTS AND SUPPORTS THREE-DIMENSIONALS DESCRIPTION OF THE INVENTION This invention relates to the techniques for forming three-dimensional objects (3D) and supporting those objects during the formation; more particularly, it relates to the techniques for use in the Prototype and Rapid Manufacturing Systems (RP &M); and more particularly with the construction and support methods and apparatus for use in the Es tereol i togra fía Thermal system (TSL), Modeling system by means of Deposition Fused (FDM), or other Modeling by Selective Deposition (SDM) system. Several procedures for the production of three-dimensional automatic or semi-automatic objects or Prototype and Rapid Manufacturing have been available in recent years, characterized because each one proceeds by integrating 3D objects from 3D computer data descriptive of the objects and in a way additive from a plurality of sheets formed and adhered. These sheets are sometimes called cross sections of the object, layers of structure, layers of the object, or simply layers (if the context clarifies that reference is made to the solidified structure of the appropriate form). Each sheet represents a cross section of the three-dimensional object. Commonly, the sheet is formed and adheres to a stack of sheets formed and adhered previously. In some RP & M technologies, techniques have been proposed which deviate from a strict layer-by-layer accumulation process where only a portion of an initial sheet is formed and prior to the formation of the portion (s) ) remaining (s) of the initial sheet, at least one subsequent sheet is formed at least partially. According to such a procedure, a three-dimensional object is accumulated or integrated by the application of successive layers of material that can flow, not solidified to a work surface and then the selective exposure of the layers to the synergistic stimulation in desired shapes or geometric figures , to cause the layers to harden selectively in the sheets of the object which adhere to the previously formed sheets of the object. In this procedure, the material is applied to the work surface to areas which are not part of a sheet of the object and to the areas which become part of the sheet of an object. Common to this procedure is Stereolithography (SL), as described in U.S. Patent No. 4,575,330, to Hull. According to a modality of the Es t ereol i t ogra f í a (SL), synergistic stimulation is radiation from a ÜV laser and the material is a photopolymer. Another example of this method is Selective Sintering by Laser (SLS), as described in US Patent No. 4,863,538 issued to Deckard, in which the synergistic stimulation is by IR radiation of a CO2 laser and the material is a powder Sinterable This first procedure can be called stereolithography based on photons. A third example is a Three-dimensional Printing (3DP) and Direct Current Production Casting (DSPC), as described in US Patents Nos. 5,340,656 and 5,024,055, to Sachs et al., In which the synergistic stimulation is a chemical binder. (for example, an adhesive) and the material is a powder consisting of particles which are joined together after the selective application of the chemical binder. According to such a second method, an object is formed by selectively cutting the cross sections of the object having shapes and sizes of sheets of material to form sheets of the object. Normally, in practice, sheets of paper are stacked and adhered to previously cut sheets before being cut, but cutting before stacking and adhesion is also possible. Typical of this procedure is the Manufacture of Laminated Objects (LOM), as described in U.S. Patent No. 4,752,352, issued to Feygin in which the material consists of paper, and the means for cutting the sheets to the desired shapes and sizes is a CO2-based laser - U.S. Patent No. 5,015,312 issued to Kinzie also discusses the integration of the object with LOM techniques. According to such a third method, the sheets of the object are formed by the selective deposition of a material that can flow, not solidified, on a work surface into desired geometric figures in areas which become part of the sheets of an object. After or during selective deposition, the selectively deposited material solidifies to form a subsequent sheet of the object which adheres to the previously formed and stacked sheets of the object. Then, these steps are repeated to successively integrate the sheet object into sheet. This technique of object formation can be generically called Modeling by Selective Deposition (SDM). The main difference between this procedure and the first procedure is that the material is deposited only in those areas which will become parts of the sheets of an object. Typical of this procedure is Modeling by Fused Deposition (FDM), as described in US Patents Nos. 5,121,1329 and 5,340,433, to Crump, in which the material is supplied or distributed in a state that can flow to an environment which is at a temperature lower than the fluid temperature of the material, and which then hardens after it is allowed to cool. A second example is the technology described in U.S. Patent No. 5,260,009, to Penn. A third example is the Manufacture of Ballistic Particles (BPM), as described in U.S. Patent Nos. 4,655,492 / 5,134,569 and 4,216,616, issued to Masters, in which the particles are directed to specific sites to form cross-sections of the object. A fourth example is Thermal Stereolithography (TSL) as described in U.S. Patent No. 5,141,680, to Almquist et al. When the SDM technique is used (also like other construction techniques RP &ampM), the convenience of various methods and apparatus for the production of useful objects depends on a variety of factors. Since these factors can not normally be optimized simultaneously, a selection of an appropriate construction technique and associated method and apparatus involve transactions or exchanges depending on the specific needs and circumstances. Some factors to be considered may include 1) cost of equipment, 2) cost of operation, 3) speed of production, 4) accuracy of the object, 5) finished surface of the object, 6) material properties of the objects formed, 7) use anticipated of the objects, 8) availability of secondary processes to obtain different properties of the material, 9) ease of use and restrictions of the operator, 10) required or desired operating environment, 11) security and 12) time and effort of postprocessing. In this regard there has been a need for a long time to simultaneously optimize as many of those parameters as possible to more effectively construct three-dimensional objects. As a first example, there has been a need to improve the production speed of the object when constructing objects using the third procedure, SDM, as described above (for example, Stereolithography to Thermal) while simultaneously maintaining or reducing them. the cost of the equipment. As a second example, there has been a long-felt need for a low-cost RP &M system that can be used in an office environment. In the SDM process, also like the other RP & M procedures, the commonly accurate formation and placement of the work surface is required in such a way that the outward facing cross-section regions can be accurately formed and positioned. The first two procedures naturally provide work surfaces on which layers of material and sheets can be formed. However, since the third method, SDM, does not necessarily provide a work surface, it suffers from a particularly acute problem of forming and accurately positioning the subsequent sheets, which contain regions not fully supported by the previously stocked material, such as regions that include faces facing away from the object in the direction of the previously stocked material. In typical construction processes, where the subsequent sheets are placed above the previously formed sheets this is particularly a problem for the faces facing down (face down portions of the sheets) of the object. This can be understood by considering that the third procedure theoretically only deposits material in those areas of the work surface which will become part of the corresponding sheets of the object. Thus, nothing will be available to provide a work surface or to support any surface facing down that appears on a subsequent cross section. The regions face down, as well as the cross-sectional regions of face up and continua, as described in Stereol and Photon-based Photography, but as applicable to other RP & M technologies in which include SDM, are described in detail in U.S. Patent Nos. 5,345,391 and 5,321,622 issued to Hull et al., and Snead e al., respectively. The previous sheet is non-existent in the face down regions and thus is not available to carry out the desired support function. Similarly, the non-solidified material is not available to carry out the function of the support since, by definition, in the third process, such material is not commonly deposited in the areas which do not become part of a cross section of the object. The problem that results from this situation can be termed as the problem of "lack of work surface". The problem of "lack of work surface" is illustrated in Figure 1, which shows two sheets, identified with the numbers 1 and 2, integrated when using a three-dimensional model method and apparatus. As shown, the sheet 1, which is located above the sheet 2, has two faces facing downwards, which are shown with shading and are identified with numbers 3 and 4. By employing the SDM method described above, the non-solidified material is never deposited on the volumes directly below the surfaces facing down, which are identified with the numbers 5 and 6. Thus, with the SDM procedure, there is nothing that provides a work surface for, or to support the two surfaces facing downwards. Several mechanisms have been proposed to address this problem, but so far none has proven to be completely satisfactory. One such mechanism, suggested or described in U.S. Patent No. 4,247,508 issued to Housholder, U.S. Patent Nos. 4,961,154; 5,031,120; 5,263,130; and 5,386,500, issued to Pomerantz, et al., US Patent No. 5,136,515, issued to Helinski; U.S. Patent No. 5,141,680, issued to Almquist, et al .; U.S. Patent No. 5,260,009, issued to Penn; the North American Patent no. 5,287,435, issued to Cohen, et al .; U.S. Patent No. 5,362,427, issued to Mitchell; U.S. Patent No. 5,398,193, issued to Ounghills; US Pat. Nos. 5,286,573 and 5,301,415 issued to Prinz, et al., involves filling the volumes below the surfaces face down with a support material different from that used to construct the object and which is allegedly they easily separate from it (by means of a higher melting point for example). In relation to Figure 1, for example, the volumes identified with the numbers 5 and 6, would be filled with the support material before the time when the material used to form the surfaces facing down 3 and 4 is deposited. A problem with the two-material process (that is, different construction material and support material) is that it is expensive and annoying due to inefficiencies, heat dissipation requirements, and costs associated with the handling or handling and supply of the product. or the second material For example, a mechanism for handling and assorting the separate material for the support material has to be provided Alternatively, means have to be provided to coordinate the handling and supply of the materials by means of a single system. Another procedure, described in U.S. Patent No. 4,99,143, issued to Hull, et al., U.S. Patent No. 5,216,616, issued to Masters, and U.S. Patent No. 5,386,500, issued to Pomerantz, et al, is to integrate in general the support structures spaced from the same material as the one used to build the object, there have been a multitude of problems with this pro cedimiento. A first problem has involved the inability to elaborate arbitrary high support structures insofar as they are simultaneously assured that they are easily separated from the object. Secondly, a problem has been found with respect to the inability to obtain an easy separation capacity between the object and the supporting structure, while simultaneously maintaining an effective work surface for the construction of, and the support of the surfaces facing outwards. A third problem involves the inability to accumulate the support structure in the direction perpendicular to the planes of the cross sections (for example, the vertical direction) at approximately the same speed, as that to which the object accumulates. A fourth problem has involved the inability to ensure an easy separation capacity and minimal damage to the surfaces facing up, when the supports have to be placed on them in order to support surfaces facing downwards above them. which are part of subsequent layers. A fifth question has involved the desire to increase the performance of the system. To illustrate the objective of obtaining easy separability, determine that the surface area on which each support comes into contact with the object is kept as small as possible. On the other hand, the objective of accumulating a support in the Z direction at a speed approaching that of the object accumulation determines that the cross sectional area of each support is as large as possible to provide a large area ratio perimeter, to minimize by this the loss of material to accumulate in the Z direction due to slippage, dispersion, misplacement and the like, by allowing a large target area to compensate for any inaccuracy in the deposition process and limit the capacity of the material to disperse horizontally instead of vertically accumulating. In addition, the object of obtaining minimal damage to the surface facing down determines that the spacing between the supports is kept as large as possible, in order to minimize the contact area between the supports and the object. On the other hand, the objective of providing an effective work surface for the construction of the surface facing down determines that the spacing or separation is kept as small as possible. As is evident, there is a conflict to simultaneously obtain these obj ective. This problem is illustrated in Figure 2 in which, compared to Figure 1, like elements are named with like numbers. As shown, the downward facing surface 3 is supported by means of column supports 7a, 7b, and 7c, while the downward facing surface 4 is supported by means of column supports 8a, 8b, 8c and 8d. The column supports 7a, 7b and 7c are widely spaced from each other in order to minimize damage to the downward facing surface 3. In addition, they are configured to contact the surface face down in a relatively small surface area to improve separability. On the other hand, due to its surface area of small cross-section, it can not have the capacity to accumulate in the vertical direction, sufficiently quickly with the growth velocity of the object, and, due to its wide spacing or separation, can not having the ability to provide an effective work surface for the construction of, and the support of the downward facing surface 3. In contrast, the column supports 8a, 8b, 8c and 8d are more closely spaced apart from each other with the In order to provide a more effective work surface for the construction or integration and the support of the surface 4 facing downwards, each is also configured with a larger surface area in order to allow them to grow at a speed approximately that Unfortunately, due to its narrower spacing and the larger cross-sectional area, these supports They will cause more damage to the surface face down when they separate. All patents referred to above in this section of the specification are incorporated by reference herein in their entirety. 3. Attachment Appendices and Patents and Related Solutions Appendix A is appended hereto and provides details of the preferred thermal tereoli ty materials for use in some preferred embodiments of the invention. Appendix B is attached hereto and provides details of the control techniques of the preferred data manipulation system in a preferred SDM system. This is a copy of the US Patent Application No. 08 /, filed concurrently with the present, and a file of the 3D System No. USA.143. The following applications are incorporated herein by reference in their entirety: According to the techniques of It tereoli tografí to Thermal and some techniques of Modeling by Deposition in Molten State, a three-dimensional object is integrated from layer to layer from a material which is heated until it has the ability to flow and which is then supplied with a spout . The material may be supplied as a flow of semicontinuous material from the dispenser or may alternatively be supplied as individual drops. In the case where the material is supplied as a semi-continuous flow, it is conceivable that less stringent work surface criteria are acceptable. A prior embodiment of the thermal engineering technique is described in US Pat. No. 5,141,680. Thermal Stereolithography is particularly appropriate for use in an office environment due to its ability to use non-reactive, non-toxic materials. Furthermore, the process of forming objects using these materials does not need to involve the use of radiation (for example, UV (ultraviolet) radiation, IR (infrared) radiation, visible light and / or other forms of laser radiation), heating of the materials at combustible temperatures (eg, burning of the material along the boundaries of the cross-section as in some LOM techniques), reactive chemical components (eg, monomers, photopolymers) or toxic chemical components (eg. solvents), complicated cutting machinery and the like, which can be noisy or present significant risks if handled improperly. Instead of this, the formation of the object is obtained by heating the material to a temperature at which it can flow and then selectively assorting the material and allowing it to cool. U.S. Patent Application No. 08 / 534,447, mentioned above, is directed to data transformation techniques for use in the conversion of 3D object data into support data and the object for use in a preferred Selective Stroke Modeling system (SDM) based on SDM / TSL principles. This referenced application is also directed to various manipulation techniques, data control and system control to control the SDM / TSL system described hereinafter. Some alternative data manipulation techniques and control techniques are also described for use in SDM systems as well as for use in another RP & M system. U.S. Patent Application No. 08 / 535,772, as mentioned above, is directed to the preferred material used by the preferred SDM / TSL system described herein. Some alternative materials and methods are also described. The North American Patent Application No. 08 / 534,447, mentioned above, is directed to some particularities of the preferred SDM / TSL system. Some alternative configurations are also discussed.

The assignee of the present application, 3D Systems, Ine., Also owns a variety of other US Patent Applications and US Patents in the field of RP & M and in particular in the portion of Tereolithography based on photons of that field. These patents include descriptions which can be combined with the teachings of the present application to provide enhanced SDM object formation techniques. The US Patent Applications and US Patents pertaining to each other are incorporated herein by reference as summarized below herein: The present invention implements a variety of techniques (methods and apparatus) that can be used alone or in combination to treat a variety of problems associated with the construction and support of objects, formed by using Selective Deposition Modeling techniques. Although directed primarily to SDM techniques, the techniques described hereinafter can be applied in a variety of ways (as will be apparent to those of skill in the art reading the present disclosure) to other RP &amp technologies.M as described hereinabove to improve the accuracy of the object, the surface finish, the construction time and / or post-processing stress and the post-processing time. In addition, the techniques described herein can be applied to the selective deposition modeling systems that use one or more construction and / or support materials, wherein one or more are selectively supplied and in which others can be supplied non-selectively and in where elevated temperatures may or may not be used for all or part of the materials to assist in their deposition. The techniques can be applied to the SDM systems, where the construction material (eg paint or ink) is flowed for assortment purposes by the addition of a solvent (eg, alcohol, acetone, paint thinner or other solvents suitable for the specific construction, where the material is sol idible after or during assortment by causing the separation of the solvent (for example, by heating the assorted material, by supplying the material to a partially evacuated construction chamber ( this is subjected to vacuum), or simply by allowing enough time for the solvent to evaporate.) Alternatively or additionally, the construction material (eg, paint) can be thixotropic by nature, in which an increase in strength Cutting on the material could be used to assist in its assortment or distribution or the thixotropic property can simply be used to help The material to retain its shape after it is supplied or dispensed. Alternatively, and / or additionally, the material may be reactive in nature (e.g., a photopolymer, thermal polymer, a one or two part epoxy material, a combination material, such as one of the aforementioned materials in combination with a wax or plastic or thermal material, or at least insurmountable sol when combined with another material (eg, paris gypsum and water) where, after assortment, the material is reacted by appropriate application of the prescribed stimulation. (eg, heat, EM, electromagnetic radiation [visible, IR-, UV, X-rays, etc.], a reactive chemical component, the second part of a two-part epoxy, the second or multiple part of a combination) So that the material of construction and / or combination of materials solidifies, for example, Thermal materials and assortment techniques can be used alone or in combination with the previous alternatives. In addition, various assortment techniques may be used, such as the assortment by single or multiple ink jet devices including, but not limited to, thermal fusion ink jets, bubble jets, etc., and nozzles or single-hole or multi-orifice extrusion heads, continuous or semi-continuous flow.

Thus, it is a first object of the invention to provide a method and apparatus for producing objects of higher accuracy. A second object of the invention is to provide a method and apparatus for the production of objects with less distortion by controlling the thermal environment during the formation of the object. A third object of the invention is to provide a method and apparatus for the production of objects with less distortion by controlling the manner in which the material is dispensed. A fourth object of the invention is to provide a method and apparatus for improving the speed of production of objects. A fifth object of the invention is to provide a support structure method and apparatus that allows for supports of the object of arbitrary height to be formed. A sixth object of the invention is to provide a supporting structure method and apparatus that provides a good surface area of t or j or. A seventh object of the invention is to provide a method and apparatus that forms a support structure that easily separates from the downward facing surfaces of the object. An eighth object of the invention is to provide a supporting structure method and apparatus that results in minimal damage to the downward facing surfaces of the object after separation therefrom. A ninth object of the invention is to provide a method and apparatus for separating supports from the object. A tenth object of the invention is to provide a supporting structure method and apparatus that vertically supports supports at a speed of approximately the vertical construction speed of the object. A thirteenth object of the invention is to provide a method and apparatus that forms a support structure that is easily detached from the face-up surfaces of the object. A twelfth object of the invention is to provide a method and apparatus of support structure that results in minimal damage to the face-up surface of the object after separation therefrom.

A thirteenth object of the invention is to provide a method and apparatus for producing supports that are separated from the vertical surfaces of the object. A fourteenth object is to provide support structures that can be combined with other RP & M technologies for improved object training. It is proposed that the above objects can be obtained separately by different aspects of the invention and that additional objects of the invention will involve various combinations of the above independent objects, so that the combined benefits can be obtained from the combined techniques. Other objects of the invention will become apparent from the description that is provided hereunder. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 illustrates surfaces facing downwards of an object; Figure 2 illustrates two categories of support structures for supporting the downface surfaces of Figure 1; Figure 3 is a diagram of the main functional components of the Deposition Modeling System Select i va / Es t ereol i togra ph y Thermal, preferred; Figures 4a and 4b illustrate the orifice plate of the print head of Figure 3, at different orientations to the scanning or scanning direction; Figure 5 is a more detailed drawing of the flattening device of Figure 3; Figure 6 illustrates the relative spacing between the adjacent nozzles on the orifice plate and the adjacent raster lines; Figure 7 illustrates the pixel grid which defines the resolution of system data; Figure 8 illustrates two perpendicular examples of the raster line orientation; Figure 9 illustrates two examples of deposition propagation in the scan direction or secondary scan; Figures 10a and 10b illustrate two examples of the propagation of the deposition in the main scan or scan direction; FIGS. 11 and 11b illustrate an example of the cross-linking of scan or scan lines; Figures 12a and 12b illustrate an example of the array of location of the droplets along several scan or scan lines; Figures 13a and 13b illustrate a further example of the location lattice of the drop along several scan or scan lines; Figure 14 illustrates a geometric figure of the deposition on the chessboard in a single pixel; Figure 15 illustrates a 3 x 3 column support pixel geometric figure that forms a preferred support structure; Figures 16a-16d illustrate various overprint schemes; Figures 17a and 17b illustrate a malfunction or match problem that can be made when an overprint technique is used; Figure 18 illustrates the resulting deposition regions when the pixels of Figure 15 are exposed when using an overprint scheme; Figure 19 illustrates an alternative pixel geometric figure for column supports; Figure 20 illustrates a Hybrid support structure; Figures 21a and 21b illustrate arch supports; Figures 22a-d show a cross-linking mode for the deposition of the material during the construction of an object; Figures 23a-h illustrate a construction mode which uses horizontal and vertical pixel shifts; Figures 24a-d illustrate a deposition mode that reduces the risk of bonding between the regions separated by a space; Figures 25a-e illustrate a construction technique in which the object is separated into pieces, integrated or separately constructed and then adhered together; Figure 26 illustrates a preferred two-step screen scanning and graduation configuration; Figures 27a-27e show various combinations of work surface and pointing positions; Figure 28a shows a side view of a mode of attachment supports; Figure 28b shows a side view of another embodiment of joint supports; Figures 29a-29e show a top view of the tie layers for a mode of attachment supports; Figures 30a-30m show a top view of the tie layers for another embodiment of joint supports; Figures 31a-31c show a top view of tie layers for another embodiment of the tie supports; Figures 32a-31d show a top view of the tie layers of another embodiment of the joint supports. As discussed previously, the present application is directed to the support techniques and construction techniques appropriate for use in a Selective Deposition Modeling (SDM) system. In particular, the preferred SDM system is a Tereoli Thermal Relief (TSL) system. The Description of the Preferred Modalities will begin with a description of the preferred TSL system. A more detailed description of the preferred system, data manipulation techniques, system control techniques, formulation and material properties, and various alternatives are described in the US Patent Applications referenced above and incorporated herein. 08 / 534,813; 08 / 534,447; 08 / 535,772; and 08 / 534,477; and the United States Archives of 3D Nos. 143, presented concurrently with this. In addition, alternative systems are discussed in a variety of previously incorporated applications and patents, especially those referred to as directly related or applicable to SDM, TSL or Modeled by Fused Deposition (FDM). As such, the support structures and building styles described later herein, should be interpreted as being applicable to a variety of SDM, TSL and FDM systems and not limited by the system examples described herein. In addition, as previously indicated, these support structures and construction styles have utility in the other RP & M technologies.

A preferred embodiment of an apparatus for carrying out the SDM / TSL is illustrated in Figure 3. The apparatus comprises an assortment platform 18, a distributor head 9 (e.g., multi-orifice ink jet head), wherein the distributor head 9 is located on the assortment or distribution platform 18, a flattening device 11 and a platform 15 of partial construction. The distributing platform 18 is a horizontal element which has the capacity to support the flattening device 11 and the distributing or assortment head 9. The assortment or distributor platform 18 is slidably coupled to a stage X by means of a coupling element 13. Stage X 12 is preferably controlled by a -. computer or control microprocessor (not shown) and controllably moving the distributor platform 18 back and forth in the X direction or in the main scan or scan direction, either on one side or the other of the platform 18, fans or fans (not shown) are mounted to blow air vertically downwards to help cool the distributed or assorted material 14 and -The partially constructed platform 15, so as to maintain the desired construction temperature. Suitable mounting arrangements for fans or fans, and / or other cooling systems include, but are not limited to, mist devices for directing vaporizable liquids (eg, water alcohol or solvents) on the object surface, cooling devices by forced air with fans or fans mounted between the flattening device 11 and the distributor head 9 and cooling devices forced air ventilation with fans or stationary or mobile fans mounted in a displaced manner from the distribution platform. Cooling systems can include active or passive techniques for separating the heat which can be controlled by computer, in combination with devices that detect the temperature to keep the material previously distributed within the range of the desired construction temperature. Other methods for cooling include, but are not limited to, sanitizing the material with a substance, which functions as a blackbody radiator, especially at IR frequencies, such that heat is more readily radiated from the object during the construction process. . Additional alternative methods include, but are not limited to, the addition of a conductive substance to the material every few layers, the addition of a solvent to the material, construction parts with passages or cooling passages or with an embedded substrate (such as wires). cross-linked) for cooling or construction on a glass plate or sheet or thousand sheet. Other embodiments for cooling the material or at least keeping the material supplied or distributed at an appropriate temperature, could involve the use of the direction of a gas that moderates the temperature (e.g., a cooling gas such as air) on the surface. of the partially formed object, as discussed above, but may additionally include controlled techniques for removing the cooling air from the surface. Such techniques could involve the use of blowing and suction devices and the alternative positioning of blow ducts (gas insertion ducts) and suction ducts (gas separation ducts). These ducts can allow the cooling gas to be separated before excessive heating of the gas causes a loss in the effective cooling rate. The gas directed at the surface can be introduced in a cooled state, introduced at room temperature or introduced at some other appropriate temperature. If properly configured, these alternative insertion and separation ducts can allow a speed, scanning or sweeping faster than currently permissible due to turbulence or wind distortion of brittle structures such as supports. These conduits could be configured to provide air flow in the direction opposite to the movement of the print head, thereby reducing the speed of the net wind coming into contact with the partially formed object. The blowing or suction associated with the individual ducts can be reversed, activated or deactivated depending on the direction of movement of the print head.

The print head 9 is a commercial print head configured for thermal jet fusing inks, such as, for example, thermal or wax-like plastic materials and modified for use in a three-dimensional modeling system, where the print head undergoes movements and accelerations from back to front. Modifications to the print head include the configuration of any container on the board, in such a way that the accelerations result in a poor minimum placement of the material in the container. A preferred embodiment includes a commercial 96-jet print head, model No. HDS 96i., Sold by Spectra Corporation, Nashua, Hew Hampshire including container modifications. The printhead is supplied with material in a state that can flow from a Packing and Material Handling Subsystem (not shown), which is described in the North American Patent Application referenced above No. 08 / 534,447 . In the preferred embodiment, all of the 96 jets or nozzles on the head are computer controlled to selectively trigger droplets through the orifice plate 10 when each orifice (ie, the jet), "is appropriately located for fill or distribute the drops on the desired sites. In practice, approximately 12,000 to 16,000 orders per second have been sent to each jet selectively triggering each shot (assortment or distribution of a drop) not firing (not stocking or distributing a drop) depending on the position of the jet and the sites desired for the deposition of the material. Also, in practice, firing orders have been sent simultaneously to all the jets. Since, the above-mentioned preferred print head contains nearly 100 jets, the firing rates indicated above, result in the need to send approximately 1.2 to 1.6 million firing commands to each head in every second. Thus, the head is controlled by a computer to selectively fire the jets and cause them to simultaneously emit drops of the molten material through one or more holes in the orifice plate 10. Of course, it will be appreciated that in alternative preferred embodiments, heads with different numbers of jets may be used, different firing frequencies are possible, and in appropriate circumstances it is possible to trigger the jets in a non-simultaneous manner. To construct a three-dimensional object more effectively, it is desirable that all jets fire correctly. To ensure that all jets are firing correctly or at least maximize the number of which are firing correctly, several techniques can be used. One such modality involves the inspection of the jets after the formation of each sheet. This technique includes the steps of: 1) forming a sheet; 2) inspect the jets by printing a geometric figure of test lines on a piece of paper, with all the jets firing; 3) optically detect (by means of barcode scan or the like) if a jet is firing defectively; 4) jet desobturation; 5) Separate the entire newly distributed or assorted layer (for example, by machining using a preferred flattener device to be described later herein, and 6) Rebuild the layer with all the jets including the unobstructed jet .

A second embodiment involves the following preferred stages: 1) forming a layer; 2) optically detect a poorly fired jet; 3) Re-scan the lines on the layer that must have been formed by the poorly fired jet; 4) stop the use of the poorly fired jet in the rest of the construction process; and 5) explore the subsequent layers while compensating for the poorly fired jet (this is, do extra steps with a jet of work to cover the lines corresponding to the jet fired poorly). Optionally, the poorly fired jet can be inspected periodically to see if it has started to work again. If so, this jet is put back into operation. Another option involves putting a faulty jet through a reactivation routine to see if it can be done - operational. This could occur during the construction process or during system service. As a further alternative, it may be possible to determine whether or not a jet is firing correctly by tracking the electrical characteristics of the piezoelectric element as the shot occurs. A third embodiment could involve the use of a flexible element to clean excess material from the bottom of the print head. This modality involves the firing of all the jets, followed by a cleaning of the orifice plate with a heated rubber blade (for example, VITON). Preferably, the sheet is positioned in such a manner as to contact the orifice plate as they move relative to one another to thereby cause a cleaning action to remove the excess material from the orifice plate and revitalize in a hopeful way any jet which was not behaving properly. It is further preferred that the orifice plate and the sheet or sheet are positioned at an angle to each other such that at any time during their contact, only a portion of the orifice plate is in contact with the cleaning sheet, for minimize by this the force that the blade exerts on the orifice plate. The orifice plate 10 is mounted on the distribution platform 18, in such a way that drops of material are allowed to be emitted from the underside of the distributor platform 18. The orifice plate 10 is illustrated in Figures 4a and 4b. In a preferred embodiment and as shown in Figure 4a, the orifice plate 10 (ie, the line of holes) is mounted approximately perpendicular to the sweep direction or main scan (X direction) and is configured with N = 96 individually controllable holes (called 10 (1), 10 (2), 10 (3) ...(10 (96)) Each hole is equipped with a piezoelectric element that causes a pressure wave to propagate through the material when an electrical trigger pulse is applied to the element. The pressure wave causes a drop of material from the hole to be emitted. The holes 96 are controlled by the control computer which controls the speed and synchronization of the trigger pulses applied to the individual orifices. With reference to Figure 4a, the distance "d" between the adjacent holes in the preferred embodiment is approximately 26.27 mils or 0.677 mm (8/300 inch). Thus, with 96 holes, the effective length "D" of the orifice plate is approximately (N x8 / 300 inch) = (96 x 8/300 inches) = 2.56 inches (65.02 mro-). To accurately construct an object, the printhead must shoot in such a way that the droplets reach particular "desired drop sites," that is, sites where the drop is proposed to settle. The desired drop sites are determined from a data map or pixel sites, which describes the object as a series of site points relatively spaced. In order for the droplets to settle at the desired drop sites, the print head must trigger the droplets from a "desired firing site" or a "desired firing time", which is based on the relative position of the head of the droplet. printing to the desired site of the drop, the speed of the print head and the ballistic characteristics of the particles after they are ejected. In a preferred embodiment, the scan or frame scan is used to position the print head 9 and the holes in the desired trigger sites. The printing process for each sheet is carried out by a series of relative movements between the head 9 and the desired drop or firing sites. Printing commonly occurs as the head 9 moves relatively in a sweep direction or main scan. This is followed by a typically minor increase in movement in a secondary scan direction as long as the distribution does not occur, which in turn is followed by a scan or inverse sweep in the sweep direction or main scan in which the presents the assortment or distribution. The process of alternating sweeps or main scans and secondary sweeps occurs repeatedly until the sheet is completely deposited. Alternative preferred embodiments may perform small secondary scan movements while the main scan is presented. Due to the commonly large difference in net scanning speed along the main and secondary directions, such that an additional alternative results in deposition along scan lines which are almost parallel to the direction of main scan and perpendicular to the secondary scan direction. Additional alternative preferred modes may use vector scanning techniques or a combination of vector scanning and frame scanning techniques. It has been found that the drops, immediately after being distributed from the jet orifice, have an elongated shape, in comparison to their width. The ratio of the length to width of the drop can be defined as the aspect ratio of the drop. It has further been found that the aspect ratio of these drops becomes smaller as the droplets travel from the jet orifice (that is, they become more spherical in shape). It should be appreciated that, in some embodiments, the spacing or separation between the orifice plate 10 and the work surface is preferably large enough, such that the emitted drops thereof have become semicircular in shape when they make an impact. with the work surface. On the other hand, it should also be appreciated that this separation, which determines the distance that the drops should leave during the printing process before impact must be minimized in order to avoid accuracy problems which may occur as the travel time or travel increases. In practice, it has been found that these two conditions are satisfied satisfactorily when at least 90% of the emitted drops from the orifice plate have obtained an aspect ratio (ie, the ratio formed by the width of the drop divided by its length) which is preferably less than about 1.3, more preferably less than about 1.2 and more preferably which is between about 1.05 and 1.1. In alternative preferred embodiments, the print head 9 can be mounted at an angle not perpendicular to the main scan or scan direction. This situation is shown in Figure 4b, where the print head 9 is mounted at an angle "a" to the main scanning or scanning direction (eg, the "X" direction). In this alternative situation, the spacing between the holes is reduced from d to d '= (d x without a) and the effective length of the print head 9 is reduced to D' = (D x sin a). When the spacing d 'is equal to the desired print resolution in the secondary scan direction (direction approximately perpendicular to the main scan direction), the angle a is considered the "knowing angle". If the dod 'spacing is not found at the desired secondary print resolution (that is, the print head is not at the know angle) then for optimum efficiency in the printing of a layer, the desired resolution should be selected from Such a way to make "dod" an integer multiple of the desired resolution.Similarly, when printing with a? 90 °, there is a separation between the jets in the main scanning direction as well as the secondary scanning direction.This spacing or separation is define by d "= dx cosine of a. This in turn determines that the optimization of the printing efficiency will be presented when the print resolution in the desired main direction is selected so that it is an integral divider of d "(this assumes that the firing sites are located in a rectangular grid.) Another way of expressing this is that the angle a is selected in such a way that d 'and / od "when divided by the appropriate integers M and P produce the desired secondary and principal resolutions of exploration. An advantage of using the preferred print head orientation (a = 90 °) is that it allows any desired print resolution in the main scanning direction, while still maintaining optimum efficiency. In other preferred embodiments, multiple heads may be used which extend from end to end (extend in the secondary scan direction) and / or which are stacked from back to back (stacked in the main scan direction). When stacked from back to back, the print heads may have holes aligned in the main scanning direction, so that they print on the same lines or alternatively they may be offset from each other to supply or distribute material throughout. of different main exploration lines. In particular, it may be desirable to have the print heads from back to back moved to each other in the secondary scan direction by spacing the desired screen line to minimize the number of major scanning or scanning steps that must be presented. . In other preferred embodiments, the deposition sites that define the data may not be located by the pixels configured in a rectangular grid, but may instead be located by pixels configured in any other geometric figurine (eg, a configuration displaced or staggered). More particularly, the deposition sites can be varied fully or partially from layer to layer in order to carry out a displacement of the partial pixel drop site for an entire layer or for a portion of a layer based on the particularities of a region to be sprinkled by jets. Preferred printing techniques in the present involve deposition resolutions of 300, 600 and 1200 drops per inch in the main scanning direction and 300 drops per inch in the secondary scanning direction. With reference to Figures 3 and 5, the flattening device 11 includes a heated rotating cylinder (eg, rotating at 2000 rpm), 18a with a textured surface (eg, knurled). Its function is to melt, transfer and separate the portions of the previously distributed layer or sheet of material in order to smooth it, to fix a desired thickness for the last layer formed and to adjust the net upper surface of the last layer formed to a level wanted. The number 19 identifies a layer of material which was recently deposited by the printhead. The rotating cylinder 18a is mounted on the dispenser platform, in such a way that it allows it to project from the underside of the platform by a sufficient amount in the Z direction, such that it is brought into contact with the material 19 at a level wanted. Importantly, the rotating cylinder 18a is mounted to project at a desired distance below the sweep plane by the underside of the printhead or orifice plate. In the case where the orifice plate projects itself underneath the distribution platform 18, the rotary cylinder 18a will also be projected below the distribution platform 18. In a preferred embodiment, the projection below the orifice plate in the z direction is in the range of 0.5 mm to 1.0 mm.

The extent to which the roller extends below the distributing platform 18 is determinant of the spacing or separation between the orifice plate 10 and the work surface. Thus, in some preferred embodiments, it is preferred that the extent to which the flattening device 11 extends below the orifice plate 10 does not conflict with the condition described above in relation to the relationship to the aspect ratio of the drop, in which 90% of the droplets have obtained an aspect ratio by impact preferably less than about 1.3, more preferably less than about 1.2, and more preferably between about 1.05-1.1. The rotation of the cylinder distributes the material from the newly deposited layer, identified in the figure with the number 21, which comes out on its smooth surface 20. The material 21 adheres to the knurled surface of the cylinder and is displaced until it is brought into contact with the cleaner 22. As shown, the cleaning element 22 is arranged to effectively "scrape" the material 21 from the surface of the cylinder. The cleaning element is preferably manufactured from VITON, although other materials, such as TEFLON®, are capable of scraping the material from the surface of the cylinder, they are also suitable. Preferably, the scraper material is non-wetting with respect to the liquefied construction material and is durable enough to come into contact with the rotating cylinder 18a, without rapidly wearing down. The separated material is expelled under suction via an umbilical tank to the heated waste (not shown), where it is either disposed of or recycled. The waste tank of the flattening device is constantly kept under vacuum in order to continuously separate the material from the flattening cylinder. When the tank fills, the system automatically reverses the vacuum for a few seconds to purge the waste material from a unidirectional valve to a larger waste tray. Once empty, the vacuum is restored and the waste continues to be extracted from the flattening device. In practice, it has been observed that approximately 10-15% of the assortment material is separated by the flattening device. Although the most preferred embodiments use a combination of rotation, fusion and scraping to carry out planarization, it is believed that other modalities could use any of these three elements or any combination of two of them. In the present implementations, the cylinder 18a rotates (e.g., at approximately 2000 rpm) in a single direction as the head moves back and forth, either in one direction or the other. In alternative embodiments, the cylinder 18a can be rotated in opposite directions based on the forward or backward direction in which the platform 18 sweeps as it moves in the main scanning direction. Some embodiments could include that the axis of rotation of the cylinder 18a is an axis offset relative to the axis of orientation of the print head. In other embodiments, more than one cylinder 18a could be used. For example, if two cylinders are used, each could be caused to rotate in different directions and can also be vertically postoably to allow a selected one to participate in flattening during any given sweep.

When a single printing head 10 and cylinder 18a are used, the flattening is effectively presented only at each second step of the print head, although the deposition occurs at each step (ie the flattening always occurs at the same direction) . Under these conditions, flattening occurs when the sweeping direction points along the same direction as an arrow pointing from the cylinder to the printhead. In other words, flattening occurs when the scanning direction is such that the cylinder follows the print head as the elements travel the layer in the main scanning direction. Other preferred embodiments could use a single cylinder, but use one or more localized print heads on either side of the cylinder, such that flattening occurs effectively when sweeping in both directions. Other alternative modes could uncouple the movement of the print head (s) and the flattening cylinder. This decoupling could allow polarization and independent distributor activity. Such decoupling could involve that the directions of the print head sweep (e.g., X direction) and cylinder sweep (e.g., Y direction) are different. Such decoupling could also allow multiple layers or lines to be formed from a single layer to be deposited between the stages of flattening. With reference to Figure 3a, the part forming platform 15 is also provided. The three-dimensional object or part, identified in the figure with the reference number 14, is integrated or built on the platform 15. The platform 15 is slidably coupled to the stage Y 16a and 16b which makes the platform move controllably from back to front in the Y direction (that is, graduation direction or second scan direction) under computer control. The platform 15 is also coupled to the stage Z, 17, which makes the platform controllably move from top to bottom (typically progressively downward during the integration process) in the Z direction under computer control.

To construct a cross section, the sheet or layer of a part, the stage Z is directed to move the platform 15 of construction of parts in relation to the print head 9, in such a way that the last integrated cross section of the part 14 is located at an appropriate amount below the hole plate 10 of the print head 9. Then, the printing head 9 in combination with the Y-stage 16a, 16b is caused to bar one or more times over the construction region XY (the head sweeps back and forth in the X direction, while the Y-step 16a) , 16b moves the partially formed object in the Y direction). The combination of the last formed cross section, sheet or layer of the object and any support associated with it defines the working surface for the deposition of the next sheet and any support associated therewith. During the translation in the XY directions, the jet orifices of the printing head 9 are fired coincidently or correspondingly with the layers previously distributed to deposit the material in a geometric figure and sequences desired for the construction of the next sheet of the object .

During the distribution process, a portion of the distributed material is separated by the flattening device 11 in the manner discussed above. The movements X, Y and Z are distributed, flattened and repeated to build or integrate the object from a plurality of layers distributed and selectively adhered. In addition, the platform 15 can be graduated either in the Y or Z direction while the address of the distributing platform 18 is in the process of being inverted after the completion of a scan. In a preferred embodiment, the material deposited during the formation of a sheet has a thickness at or somewhat greater than the desired thickness of the layer. As described above, the excess material deposited is removed by the action of the flattening device. Under these conditions, the actual accumulation thickness between the layers is not determined by the amount of the material for each layer, but instead is determined by the vertical descending increase made by the platform after the deposition of each layer. If it is desired to optimize the construction speed and / or minimize the amount of wasted material, it is desirable to trim as little material as possible during the deposition process. The less material is cut out, the thicker each sheet is and the faster the object is built. On the other hand, if it is done at the thickness of the layer, this is an increase in the z direction, too large, then the amount of accumulation associated with at least some drop sites will begin to fall behind the desired level. This delay results in the actual physical work surface being. it finds a different position from the desired work surface and probably results in the formation of a non-planar work surface. This difference in position can result in poor XY placement of the drops due to a longer flight time than expected and may also result in the vertical defective placement of the characteristics of the object starting or ending at the layers in which the actual work surface is poorly positioned. Accordingly, in some embodiments it is desirable to optimize the increase of the layers in the vertical direction.

To determine an optimal increment in the Z example, an accumulation diagnostic part can be used. This technique preferably involves constructing layers of one or more test parts in successively larger Z increments, measuring the height of the formed characteristics and determining which Z increments give rise to formation heights (i.e., vertical accumulation) of the correct amount and which Z increments give rise to formation heights which were delayed behind the desired amount. It is expected that the increments in layers (that is, Z increments) up to a certain amount (that is, the maximum acceptable quantity) would produce accumulation levels for the object equal to those predicted by the product of numbers of layers and the thickness of each cap. After the increase of the layer exceeds the maximum acceptable amount, the level of accumulation of the object would fall shortly predicted by the product of the number of layers and the thickness of each layer. Alternatively, the plain of the upper surface of the diagnosed part (s) may be lost (indicating that some drop sites may be receiving sufficient material while others may not). When inspecting the diagnostic part (s), the maximum acceptable amount of increase Z can be determined empirically. Then, the optimum Z increment amount can be selected as this maximum acceptable amount or can be selected in some thickness somewhat less than this maximum amount. Since it is known that different styles of construction and support accumulate in the vertical direction at different speeds, the above test can be carried out for each style of construction and support style, from where the optimum increase Z for a combination of different Styles can then be selected in such a way that they are not thicker than any of the maximum quantities determined for each style individually. In addition, the distributing head, by following a given scan line, can only maintain a substantially constant velocity over part of the scan line. During the rest of the exploration, the head 9 will be in acceleration or deceleration. Depending on how the firing of the jets is controlled, this may or may not cause a problem with excessive accumulation during the acceleration and deceleration phases of the movement. In the case where the speed changes can cause a problem in the speed of accumulation, the construction of the support part can be confined to the portion of the scan line on which the print head has a substantially constant speed. Alternatively, as described in the North American Patent Application, concurrently filed corresponding to the 3D File, No. USA-143, a scheme of the shot control can be used, which allows the exact deposition during the acceleration or deceleration portions of a scanning line. As indicated above, in some preferred embodiments, the print head 9 is directed to track a raster geometry. An example of this is shown in Figure 6. As shown, the geometric shape figure configuration consists of a series of raster lines (or scan lines), R (l), R (2), ... (R (N)), which run in the X direction or main scan direction and are arranged (ie, spaced) along the Y direction (i.e., graduation direction or scan direction or secondary scan). Weft lines are spaced apart by a distance d, which, in a preferred embodiment is about 83.8 microns or 1/300 of an inch (approximately 3.3 mils.) Since the holes of the print head 9 are spaced apart by a distance d, which, as discussed above, is preferably 0.6774 μm (26.67 mils) and since the desired number of raster lines can be extended in the graduation direction by a distance greater than the length of the orifice plate 10 or, approximately 65.02 mm (2.56 inches), the print head 9 must be traversed on the work surface by means of multiple steps in order to track all the desired raster lines. This is preferably done by following a two-step process. In the first stage, the print head 9 is passed 8 times on the work surface in the main scan or scan direction, with the Y step 16a, 16b, which is graded by the amount d in the secondary scan direction after each step in the main scanning direction. In the second stage, the Y stage 16a, 16b is graded by a distance equal to the length of the orifice plate 10 (2.5600 inches + dr (0.0267 inches) = 2.5867 inches (65.70 mm) .Then, this two stage process it is repeated until all the desired weft lines have been tracked In a first step, for example, the print head 9 could be directed to track R (l) frame lines (via hole 10 (1 in Figure 4) ), R (9) (via orifice 10 (2)), R (17) (via orifice 10 (3)), etc. The stage Y 16a, 16b would then be directed to move the construction platform 18 the distance d , (a weft line) in the graduation direction. In the next step, the print head 9 could be directed to track R (2) (via 10 (1)), R (10) (via 10 (2 )), R (17) (via 10 (3)), etc. Then, six more steps would be carried out with the Y stage 16a, 16b graduated by the distance d, after each step, until they have taken a total of 8 steps. In order to carry out the first stage (consisting of 8 steps) the second stage is carried out if more plot lines are to be tracked. The second stage consists in directing the stage Y to moving the construction platform by an amount equal to the full length of the plate 10 of the hole + dr, 2.5867 inches (65.70 mm). As necessary, another set of 8 steps, comprising the first stage, are carried out followed by another second stage. Then, the two-step process described above would be repeated until all the branch lines have been tracked. An example of this two-stage process is shown in Figure 26 for a printhead consisting of two jets and wherein the jets are separated from each other by 8 raster spacings. The scanning of the cross sections begins with the first jet positioned in the position 201 and the second jet placed in the position 301. The first stage of the scanning process begins with the scanning of the frame lines 211 and 311 in the direction indicated by the first and second jets respectively. As part of the first step, the initial scan of the frame lines 211 and 311 is followed by a graduation increment of a frame line width as indicated by elements 221 and 321. Continuing as part of the first stage, the initial screen scan and the increment of graduation are followed by s-ite more screen scans (shown by the pairs of lines 212 and 312, 213 and 313, 214 and 314, 215 and 315, 216 and 316, 217 and 317 and 218 and 318) separated by six more increments of 1-frame line width graduation (shown with element pairs 222 and 322, 223 and 323, 224 and 324, 225 and 325, 226 and 326 and 227 and 327) . Immediately after the scanning of the frame line pairs 218 and 318, the second stage of the process is taken, wherein the head is graduated in the Y direction according to the direction and lengths of the frame lines 228 and 229. The length of this graduation is equal to the width of the head (that is, in this example widths of 8 weft lines) plus the width of 1 more weft line. After this large increase, the first stages and the second stages are repeated as many times as necessary to complete the exploration of the particular cross section that is formed. It will be evident to those of ordinary experience in the art, that this two-stage exploration technique can be implemented in other ways in alternative modalities. For example, the second stage may consist instead of the positive increment increment in Y as defined by elements 228 and 328, which consists of the large negative increase in Y as indicated by element 330 (ie, three head widths less a line width). It can be summarized that this preferred embodiment includes the following characteristics: 1) the spacing along a graduation direction between the adjacent jets is an integral multiple (N) of the desired spacing (dr) between the adjacent deposition lines which extend in a printing direction, which is approximately perpendicular to the graduation direction; 2) the first step includes carrying out a variety of steps (N) in the printing direction, wherein each step is shifted in the direction of graduation by the desired spacing (dr) between the adjacent deposition lines; and 3) the second step includes the displacement of the printing head 9 in the direction of graduation by a large amount, such that the jets can deposit material in other N steps, where the successive steps are separated by graduation increments of a plot line and after that, another large graduation increment will be made as necessary. In the most preferred embodiments, the amount of graduation of the second stage will be equal to the sum of the spacing between the first jet and the last jet plus the desired spacing between the adjacent deposition lines (this is N x J + dr, where J is the number of jets on the printhead 9). As indicated in the previous example, other graduation quantities of the second stage are possible. For example, negative second stage increments (opposite direction to the graduation increments used in the first stage) equal to the sum of the width of the head plus the product of twice the width between the successive streams minus the width of a separation of the plot line. In other modalities, it is possible to use the graduation quantities of the second stage which vary or which alternate from back to front between positive and negative values. In these embodiments, the amount of increase of the second stage has the common characteristic that it is greater than the individual graduation amounts used in the first stage. In other preferred embodiments, other single-stage or multi-stage graduation configurations may be used, increments in the graduation direction could generally be made in such a way that they include increments involving both negative and positive movements along the Y axis. This could be done to explore raster lines that were initially omitted. This will be described additionally in association with a technique called "lattice." In some preferred embodiments, the firing of the ink jets is controlled by a rectangular bit map, i.e., pixel sites, maintained in the control computer or other memory device. The bit map consists of a grid of memory cells, in which each memory cell corresponds to a - pixel of the work surface and in which the rows of the grid extend in the main scanning direction (X direction) ) and the columns of the grid extend in the secondary scan direction (Y direction). The width of (or distance between) the rows (spacing along the Y direction) may be different from the width (or length of, or distance between) of the columns (spacing along the Y direction) that determines that different data resolutions may exist along the X and Y directions. In other preferred embodiments, the non-uniform pixel size is possible within a layer or between layers where one or both of the width or length of the pixel is varies by the position of the pixel. In still other preferred embodiments, other pixel alignment configurations are possible. For example, the pixels in the adjacent rows may be displaced in the main scanning direction by a fractional amount of the spacing between the pixels, such that their center points do not align with the center points of the pixels in the neighboring rows. This fractional amount can be 1/2 such that its center points are aligned with the pixel boundaries of the adjacent rows. It can be 1/3, 1/4 or other amount, in such a way that they take two or more intermediate layers before the pixel configurations are realized over the subsequent layers. In alternative modes, the alignment of the pixel could be dependent on the geometry of the structure of the object or support that is stocked. For example, it may be desirable to shift the alignment of the pixel when forming a portion of a support configuration that is supposed to join a space between the support columns or when a face-down portion of an object is formed. These and other alternative pixel alignment schemes can be implemented by modifying the pixel configuration or alternatively, by defining an array of higher resolution pixels (in X and / or Y) and by using pixel trigger settings that do not trigger in each pixel site, but instead, they fire on selected spaced pixel sites which may vary according to a predisposition configuration of the predetermined or random object. The resolution of data in the main scan direction can be defined in terms of the pixels of the Main Address (MDP). MDP can be described in terms of the pixel length or in terms of the number of pixels per unit length. In some preferred embodiments, MDP = 300 pixels per inch (26.67 mi / pixel or 667.4 μ / pixel). In other preferred embodiments, MDPO = 1200 pixels per inch. Of course, any other MDP value can be used as desired. Similarly, the resolution of data in the secondary scan direction can be defined in terms of the Secondary Address Pixels (SDP) and the SDP can be described in terms of the pixel width or in terms of the number of pixels per unit length. In some modalities SDP = MDP = 300 pixels per inch (26.67 thousandths of an inch / pixel or 677.4 μm / pixel). The SDS may or may not be equivalent to the spacing between the raster lines and the MDP may or may not be equivalent to the spacing between the successive drop sites along each raster line. The spacing between the successive grid lines can be defined as secondary drop sites (SDL), while the spacing between the successive drop sites along each grid line can be defined as the Main Drop Sites. (MDL). Similar to SDP and MDP, SDL and MDL can be defined in terms of droplets per unit length or drop spacing.

If SDP = SDL is in a one-to-one correspondence between data and droplet sites along the secondary scan direction and the pixel spacing is equal to that of the raster line spacing. If MDP = MDL there is a one-to-one current between the data and the drop sites along the main scan direction. If SDL and / or MDL is greater than SDP and / or MDP, respectively, more drops will be needed to be fired than those for which there is data, so each pixel will need to be used to control the drip of more than one drop. The supply of these extra drops can be done either by supplying the droplets at intermediate points between the centers of the successive pixels (that is, intermediate drip "ID") or alternatively, directly above the pixel centers (this is , direct drip, "DD"). In any case this technique is called "overprint" and results in a faster accumulation of material and facilitates mechanical design constraints that involve maximum scanning speeds and acceleration speeds, since the same Z accumulation can occur in both which moves to the print head and / or the object more slowly. The difference in overprint ID versus no overprint and over impr is DD ion, is shown in Figures 16a to 16d. Figure 16a shows a single drop 60 which is deposited and an associated solidified region 62, which surrounds it when the print head moves in the direction 64. On the other hand, Figure 16b shows the same region that is cured or solidified at the same time. use the ID overprint technique where two drops 60 and 66 are deposited in association with the single data point when the head moves in the 64th direction. The deposition zone filled by the two drops is shown by region 68. Figure 16c shows a similar situation for a four-drop ID overprint scheme, where the drops are indicated by the numbers 60, 70, 66 and 72 and the deposition zone is illustrated at 76 and where the scanning direction is illustrated still by the number 64. Figure 16d shows a similar situation for a line of pixels 78, 80, 82, 84, 86 and 88, where the number 90 shows the length of the deposition zone without overprint and the number mere 92 shows the length of the deposition zone when an overprint technique is used - four drops. The above can be generalized by saying that the overprint of ID add to about 1/2 to less than 1 length of an additional pixel to any region where it is used. Of course, the more overprint drops are used, the more vertical growth a pixel region will have. If SDL and / or MDL is less than SDP and / or MDP, respectively, drops will be fired at fewer sites than those for which data exists, at least for a given step of the printhead. This situation of the data can be used to implement the displacement pixel and / or non-uniform pixel techniques discussed above. A grid of N rows per M columns is shown in column 7. As shown, the rows in the grid are designated as R (l), R (2) ... R (N), while the columns in the grid are named with C (l), C (2) ... C (M). The pixels that make up the grid are also shown. These are referred to as P (l, l), P (l, 2) ... P (M, N).

To build or integrate a cross section, the bitmap is first loaded with the data representative of the desired cross section (also as any support which you wish to integrate or build). Assuming, as with some preferred embodiments, that only one construction and support material is used. If it is desired to deposit material in a given pixel site, then the memory cell corresponding to that location is indicated appropriately (that is, it is loaded with a binary "1") and if no material is to be deposited an opposite indication or indicator is used (for example, a binary "0"). If multiple materials are used, the cells corresponding to the deposition sites are indicated in an appropriate manner to indicate not only the location sites of the drop, but also the type of material to be deposited. For ease of data handling, the compressed data defining a region of the object or support (e.g., RLE data which defines activation-deactivation location points along each raster line as described in the Patent Application North American presented concurrently No. corresponding to the 3D Systems File No. USA.143) can be calculated with Bolean geometry with a description of the fill configuration (for example, Style file information as described in Document USA.143) to be used for the particular region to derive a representation from binal bit map used to shoot the assortment jets. The actual control of the jets may be determined by a modified bitmap subsequently, which contains data which have been falsified or otherwise modified to allow more efficient data passage to the trigger control system. These considerations are further discussed in the North American Patent Application based on 3D Systems of File Number USA.143. The weft lines that make up the grid are then assigned to the individual holes in the manner described above. Then, a particular orifice is directed to shoot or not in shooting sites corresponding to the desired drop sites or pixel sites depending on how the corresponding cells in the bitmap are indicated.

As discussed above, the printhead 9 has the ability to deposit drops at many different resolutions. In some preferred embodiments of the present invention, SDP = SDL = 300 pixels and drops per inch. Also in some preferred embodiments, MDL is allowed to take three different values, while MDP remains fixed 1) MDL = 300 drops per inch and MDP = 300 pixels per inch; 2) MDL = 600 drops per inch; and MDP = 300 pixels per inch or 3 MDL = 1200 drops per inch and MDP = 300 pixels per inch. When the MDL to MDP ratio is greater than one, drops are caused per extra pixel at intermediate sites (ID overprint) between the centers of the pixels. With the printhead and the material presently preferred, the volume per drop is about 80 to 100 picoliters, which produces approximately droplets having a diameter of 50.8 μm (2 mils). With the currently preferred print head, the maximum trigger frequency is approximately 20 Khz. By way of comparison, a firing rate of 1200 dpi at 13 ips involves a firing frequency of approximately 16 Khz, which is within the allowable limit. In some preferred embodiments, construction styles are defined separately from the object data for ease of data manipulation, transfer and memory loading. In this regard, as indicated above, the descriptive data of the object is subjected to boolean transformation (that is, it intersects) together with the descriptive information of a construction style on a pixel-in-pixel basis, to produce a pixel representation in pixel of the deposition configuration in any given site. For example, if a completely solid configuration is to be completed in two steps (for example, a two-stage configuration), the object's data would first be subjected to Boolean transformation (for example, inter-seperated) with a first configuration of the style of construction that represents the portion of the pixels in which the drops are to be deposited (or because of the ease of terminology can be said "exposed" in analogy to the selective solidification that is used in the photon-based theory) . The resulting modified pixel data could then be used to control the firing of the jets. Right away, the data of the object would be subjected to the Boolean calculation (for example, seperate integers) with the configuration of the complementary construction style to produce the modified pixel data to control a second shot of the jet. In other preferred embodiments, the object data and the support data can be correlated immediately to construct style data of this derivation. In additional preferred embodiments, the construction style information could also include pixel shift information, pixel size information, overprint information, scanning direction preferences for deposition on each pixel site, flattening direction and preferences rotational and the like. The construction styles described herein improve system performance by: 1) improving the construction speed; 2) improve the accuracy of the formed object; 3) improve the surface finish; 4) reduce the effort on the object and / or distortion of the object; or 5) a combination of one or more of these simultaneously.

A significant problem with the Selective Deposition Modeling Systems involves ensuring the reliability of the deposition of the material and more particularly obtaining the uniform thickness of the deposited cross sections. Another problem involves obtaining a uniform thickness for all construction styles. In ink jet systems, this reliability problem can take the form inter alia, of the wrong shot or the non-firing of the jets. In a multiple jet system, there are additional problems with respect to the non-uniformity of the direction of the jet shot, the non-uniformity of the volume assorted between the jets and to a smaller extent, the non-uniformity of the volume assorted from a single jet with the pass of the time. The problem of non-uniformity of the thickness of the cross-section can also result in other phenomena. As an example, once a drop comes out of a jet there is a flight time before the drop meets the work surface. When it exits the jet, the drop is triggered with an initial downward velocity component at a distance from the jet, since the jet moves in the main scanning direction, the drop also has a horizontal velocity component. Once the drop exits the jet, it undergoes several external and internal forces, which include the force of gravity, the viscous drag forces and the surface tension. These initial conditions and forces in turn lead to the conclusion that the drop can not and probably will not settle directly on the work surface below the position from which it was fired. Instead of this, the drop will settle a little further away from this theoretical drip point, usually in the direction of travel of the printhead. In other words, the location of the shot and the location of the impact (or drop) will not have the same XY coordinates, but instead will be displaced from each other. The displacement in the horizontal distance that is presented depends on the factors indicated above, but also on the distance between the hole plate 10 and the vertical position (eg, position "Z"), of the working surface in each horizontal site (for example, position X and / or Y). As indicated above, variations in the vertical position may occur for a variety of reasons. For example, the variations may result from differences in geometry between the different portions of a cross section (more or less dispersion of the material results in a less or less thickness of deposition). As another example, the variations may result from the ordering or temporary deposition of a given spatial configuration (a previously deposited material) on an adjacent pixel site, may limit the ability of the material to be dispersed in that direction). As indicated previously, the preferred system for implementing this invention uses flattening to bring each cross section deposited to a uniform height, wherein the net thickness of the layer results from the difference in level Z between the levels of flattening of two. consecutive layers. In turn, if it is desired that the leveling step forms a completely smooth and uniformly leveled layer, the Z increment between the flatnesses should be at or below the minimum deposition / accumulation thickness for each point throughout the layer. If a jet shoots weakly (or does not trigger), the buildup of the minimum thickness can result in a much smaller net cap thickness (ie, close to zero or zero) than the desired and therefore times of construction many more. Several techniques for dealing with these deposition / accumulation problems are described herein. Other preferred embodiments could involve the use of flattening on periodic layers instead of on each layer. For example, the flattening can be used in every second, third or other layer spaced of higher order. Alternatively, the determination of which layers or portions of layers for flattening may be based on the geometry of the object. Correction of Flight Time As indicated above, a difficulty in ensuring that the drops collide with the desired sites on the work surface involves the time the drops are in flight (ie, the time of flight of the drops) . If the flight times were always the same and if the direction and the amount of displacement were always the same there would be no question of time of flight, since the only effective would be a displacement between the firing coordinates and the coordinates of the arrangement. However, when forming three-dimensional objects, it is usually desirable to shoot the material when the head is traveling in the positive and negative main scanning directions (and may still involve, for example, toggling the definitions of primary and secondary scan directions) . This results in a change in the direction of travel (eg, reversal of the direction of travel) between scans due to relative movement that occurs in different directions (eg, opposite direction). This problem can easily be addressed by causing trigger signals to occur before the head actually reaches the point directly above the desired deposition site. This correction to the firing time is known as the "correction of flight time". The flight time can be corrected by using a correction factor applied to each scan in each separate direction or alternatively a single correction factor can be used to bring the deposition of a scan direction into register or match the scans not corrected made in the other direction. Flight time correction can be implemented in a variety of ways. One way, for example, is to properly define the initial trigger site (X position) at the beginning of each raster line, which initial trigger site will be used to set the firing sites for all other pixels along the line. line of branch. Figures 27a-27e illustrate the relationships between the firing site, the drip site and the flight time, where like elements are named with like numbers. Figure 27a illustrates the situation where the firing sites 404a and 404b are both coincident with the desired drip site 402 (ie, no flight time correction factor is used). The element 404a represents the firing site when the head passes in the positive X direction, represented by the element 406a and the element 404b represents the firing site when the head passes in the negative X direction, represented by the element 406b. The elements 408a and 408b represent the nominal path followed by the drops after they leave the firing sites 404a and 404b, respectively. The nominal trajectories 408a and 408b direct the droplets to the actual drip sites 410a and 410b, where the droplets impact the surface and form the impacted droplets 412a and 412b. The crossing point (that is, focal point) for the shots fired, in fool that is explored in both directions, is shown with the number 414. The plane defined by crossing points for the entire layer can be called as the plane focal. Elements 416a and 416b represent the flight time factor used in terms of an X offset between the trigger sites and the desired drip site. Whether the actual drip sites coincide or not with the desired drip site determines the convenience of correction factor. In Figure 27a it can be seen that the droplets move in divergent directions and that the impacted droplets do not overlap on the work surface, to result in a minimum accumulation in Z and an inaccurate XY placement of the material. Figure 27b represents the situation where small flight time correction factors 416a and 416b are used, which result in a focal point located above the desired work surface and at a narrower spacing of the drops 412a and 412b in comparison with that shown in Figure 27a. If the correction of the flight time were longer, the accumulation in Z would increase due to the overlap or overlap of the impacted drops 412a and 412b. Figure 27c depicts a situation where the correction factors of the flight time used result in the most accurate placement of the impacted droplets 412a and 412b (the thickness of the impacted droplet 412a is assumed to be small compared to the distance 418 with the drop and that the angle of incidence is not too great). If the optimum flight time correction factor is based on the maximum Z-accumulation, then Figure 27c represents the optimal situation. Figure 27d represents the situation where the flight time correction factors 416a and 416b are slightly larger than those used in Figure 27c, but still result in an accumulation of Z based on the superposition of both drops. Placements in the X direction of the drops are still reasonably accurate and the focal point 414 of the assortment is somewhat below the desired work surface (and actual work surface). Figure 27e represents the situation where even larger time-of-flight correction factors are used, such that the Z-accumulation is reduced to a minimum amount and where the focal point is still further below the surface of desired work. If the dragging effects and the gravitational effects on the flight time are ignored, the value (time) of correction of the flight time would be equal to the distance (length) separating the hole from the work surface divided by the downward speed (length / time) to which the drop is discharged. However, it is believed that trawling is an important factor. For example, in some preferred embodiments, the print head scan speed is approximately 13 inches per second, the distance from the orifice plate to the work surface is approximately 0.020 inches, and it is believed that the speeds of vertical shot initials are on the order of 200 to 360 inches per second. If the drag forces or other frictional forces are ignored, under these initial conditions, a displacement between the firing sites and drip sites of approximately 0.8 to 1.3 mils would be expected. However, under these conditions, in practice, displacements in the principal scanning direction between the firing site and the drip site of approximately 2 mils have been observed. The appropriate correction value can be easily determined empirically by attempting to deposit drops at a single X site when scanning in both directions and withdrawing the expense with different correction values until the two drops settle at the same point. As indicated above, in some preferred embodiments, the most appropriate time of flight correction value is in which the drops strike in the same position. In terms of the previous example, if the drag forces are ignored, flight time correction factors of approximately 60 to 100 μS would be expected. In practice it has been found that correction factors of approximately 150 to 200 μS are also more appropriate. In other preferred embodiments, the optimum flight time correction factor is not set at a value which produces the most accurate pointing (ie, the focal point is not on the work surface) but instead is fixed at a value which would produce the most accurate pointing or pointing at a distance below the actual work surface (this is, the focal point is located below the work surface). These modalities can be called "modalities of displaced surface pointing". In this context, it is considered that the most accurate pointing or pointing occurs when the vertical accumulation speed is the highest and probably when the X position is more precisely impacted. Figure 27d shows an example of pointing or pointing for these displaced pointing modes of the surface. It is believed that these shifting patterns of the surface are particularly useful when the construction is presented without the use of additional components to maintain the desired and actual working surface at the same level (eg, without a flattening device or without additional elements such as a device for detecting the level of the surface and mechanisms or adjustment schemes). A characteristic of these modalities of displaced pointing of the surface is that the accumulation in Z is autocor regulated or au compensated. While the increments in Z between the deposition of successive layers are within an appropriate range and the deposition configuration allows the horizontal dispersion of the assortment material instead of only the vertical accumulation, the accumulation in excess Z over a layer causes a reduction in the accumulation of Z over one or more subsequent layers to cause the net accumulation to maintain the focal point somewhat below the actual work surface. On the other hand, again as the increments in Z between the deposition of successive layers are within an appropriate range and the deposition configuration allows the horizontal dispersion of the assortment material instead of only the vertical accumulation, an accumulation in Z too low on one layer causes an increase in Z accumulation on one or more subsequent layers, thereby causing the net accumulation to maintain the focal point somewhat below the actual work surface. The preferred Z increase range is discussed further below in the present. This aspect of autocorrection can be understood by studying and comparing Figures 27c, 27d and 27e. When the deposition begins (for example, on the platform) the flight time correction factor (s) is chosen so that the focal point is somewhat below the actual surface as it is. shown in Figure 27d (ie, the focal point must be adjusted to an appropriate position such that the situations shown in Figures 27c and 27e are not presented). If, when the first layer is formed, too little material is deposited, for the given increment Z, used, the actual surface will be lower compared to the repositioned focal plane (but it will still be higher while the increase in Z does not it's too big). This results in a more optimally focused deposition when the next layer is formed, this in turn results in an increase in the thickness of the deposition as shown in Figure 27c. If the accumulation in net Z --- that results from the deposition of the second layer is still too low (compared to the two increases in Z made), then the next layer when deposited will have a real surface which is closer to the plane of the optimal focus than the original surface. This approach closest to the optimal positioning results in an increased Z-accumulation, which will again lead to the accumulated net thickness towards that required by the Z increments. On the other hand, if the net accumulation of the deposition of the second layer is greater than that determined by the two increments in Z, then the actual work surface will be more separated from the focal plane and less accumulation in Z, after the formation of the next layer, will be presented, to lead by this to the net accumulation towards the amount required in the increments in Z. This is the situation represented in Figure 27e. When the focal plane is appropriately below the actual work surface, when the amount of increase in Z is appropriately selected to roughly match the deposition rates and when the objects / supports are formed in a non-uniform manner. solid (not all pixel sites are directly deposited, the system is stabilized and both supports and objects can be formed with exact vertical dimensions without the explicit need for a flattening device.) Of course, a flattening device can still be used if desired For the optimal operation of these modalities, it is preferred that the increase in Z should be selected to be between the average amount accumulated per layer during the optimal aiming (for example, Figure 27c) and the accumulated average amount when it is not presented the overlap (for example, Figure 27e). It is also preferred that the thickness of the layer be significant. better than the distance separating the optimal focus plane (eg, Figure 27c) from the plane where the overlap is no longer present (eg, figure 27d). As indicated above, in some of these embodiments, objects can be formed in such a way as to allow regions of the material to be horizontally dispersed instead of just vertically accumulating, based on the level of optimization of the aiming and thereby enabling the autofocus. Or, the accumulation in Z. A modality as such would involve the formation of the object as a combination of alternating solid layers and layers of chessboard. Other modalities as such could involve the formation of solid outward faces and checkerboard, displaced chessboard or other open structures in internal regions of the object. Other appropriate construction configurations can be determined empirically by construction test parts and analysis. In some of these modes of pointing the displaced surface, the position of the most preferred initial focal plane / target surface is selected to be approximately in half of the situations shown in Figures 27c and 27e. One way to accomplish this is to ignore hypothetical focal points and focus instead of flight time values. The flight time correction values may be selected such that they are greater than the correction values of the optimum flight time (as discussed above) and less than the correction values of the flight time which produce impact zones. immediately adjacent but not overlapping (this is not overlapping). More preferably, the selected flight time values would be taken as approximately the average of these two extremes. Some embodiments of pointing the displaced surface could be used to simultaneously form different portions of objects and / or supports, such that their upper surfaces are intentionally at different heights after the formation of any given layer. These different height modalities could benefit from the use of data manipulation techniques, such as the SMLC techniques discussed in the North American Patent Application referred to previously No. 08 / 428,951 as well as some of the patents and applications. of US Patent to which reference is made above. In addition to . The flight time issues indicated above, other issues arise that can be corrected by using the modified flight time correction factors. For example, when using ID overprint techniques to cause more buildup, the characteristics on the scan lines' which are scanned in opposite directions will lose their alignment since the characteristic will be extended in one direction on one line and in the other direction on the other line. This situation is shown in Figures 17a and 17b. Figure 17a shows two points 60 and 100 pertaining respectively to scan lines scanned in the directions 64 and 104. Regions 62 and 102 show the extensions of the deposited material associated with points 60 and 100 respectively. Figure 17b shows the same points 60 and 100 where the formation of jets occurs using four times the overprint (that is, deposition of four drops per pixel). The deposition extensions are shown with the numbers 76 and 106 respectively. As you can see, due to the different overprint directions, the coincidence or registration between the physical characteristics on the two lines is lost. The mismatch or incorrect previous correspondence can be corrected by means of an additional flight time correction factor, which can be determined empirically or also theoretically to cause the realignment of the characteristics on different scan lines. Of course, this form of correction does not take into account any extra length added to the characteristics of the object along the scan lines. A different form of correction that can avoid both problems is proposed, which involves the recognition that a given pixel is not limited on its far side, in the scanning direction, by an adjacent pixel that also determines the deposition of the material . Based on this recognition, no overprinting is used on such an unrestricted pixel. As another alternative, the length of the extra line could be compensated for by using a form of drop width compensation similar to the line width compensation used in the photon-based photometry and as described in the Requests. US Patent Nos. 08 / 475,730 and 08 / 480,670, but applied only to the points along each scan line representing a transition from deposition to non-deposition. As an approximate correction, these "endpoints" could simply be canceled from the deposition settings since they will be in the range of 1/2 to fully covered in the use of the ID overprint of the immediately adjacent pixels. Another variant involves the use of displaced flight time correction data to implement the deposition of subpixels. The flight time correction factors can also be used in varying ways for purposes somewhat opposed to those described above. In these embodiments, the flight time correction factors can be used to deposit material at intermediate pixel sites (ie, subpixel) for the implementation of improved construction techniques. These improved construction techniques could involve the formation of face-down surfaces, formation and placement of supports, improved vertical material accumulation, improved resolution and the like. In the preferred embodiments, improved object formation can be achieved in single-step or multi-step implementations. Drop Width Compensation: In some situations, it may be desirable to modify the object data when performing drop width compensation (that is, offset compensation). Compensation (by shifting inward toward solid or fuller pixel widths) can be used to obtain improved accuracy if the drop width is at least somewhat larger than the width and / or length of the pixel. This technique can be used in combination with any of the modalities described above or any modality described herein below. As the width of the drop approaches or is greater than twice the width (and / or length) of the pixel, better accuracy can be obtained by a single or multiple pixel shift. The drop width compensation may be based on techniques such as those described in US Patent Applications Nos. 08/475, 730 and 08 / 480,670. Alternatively, they can involve pixel-based erosion routines. In some modalities, pixel-based erosions could involve multiple steps through a bitmap, where the "solid" pixels that meet certain criteria would be converted to "hollow" pixels. Some modality could involve the following stages, where each edge of the bitmap is: 1) in a first step through the bitmap all the "solid" pixels which are limited on their right side by a "hollow" pixel "they become" hollow "pixels; 2) in a second step all the "solid" pixels which are limited on their left side by a "hollow" pixel are converted to "hollow" pixels; 3) in a third step, all the "solid" pixels which are limited in their upper side by a "hollow" pixel are converted to "hollow" pixels; and 4) in a fourth step all the "solid" pixels which are limited in their lower side by a "hollow" pixel are converted to "hollow" pixels. Other modalities could change the order of stages (1) to (4). If an erosion of more than one pixel is required, steps (1) to (4) can be repeated so many times until the correct amount of reduction is obtained. These modalities can carry out a reasonable drop width compensation; however, they suffer from the disadvantage that pixels in solid corner regions (either the corner of an object or the edge of an object that does not run parallel to either the X or Y axis) separate at a faster rate that the pixels in which they represent border regions that are parallel to either the X or Y axes. Other modalities which try to treat these differentials in erosion speed could involve stages such as the following: 1) in a first step through the bitmap all the "solid" pixels which are limited on their right side by a "hollow" pixel "and all other sides by" solid "pixels are converted to" hollow "pixels; 2) In a second step, all the "solid" pixels which are limited on their left side by a "hollow" pixel and on all other sides by "solid" pixels are converted to "hollow" pixels; 3) In a third step all the "solid" pixels which are limited, on at least their upper side, by a "hollow" pixel are converted to "hollow" pixels and 4) In a fourth step all the pixels "solid" "- which are limited, on at least its underside, by a" hollow "pixel are converted to" hollow "pixels. Other modalities could change the order of stages (1) to (4) or the conditions on which the conversion will be based. If more than one pixel erosion is required, steps (1) to (4) can be repeated so many times until the correct amount of reduction is obtained. These modalities do a better job of minimizing the reduction of excess in the corner regions. Other modalities could involve adjusting erosion conditions based on whether two, three or all four sides of a pixel are limited or not by "hollow" pixels. Other modes can vary the erosion conditions, depending on how many times the bitmap has passed. Other modalities may use a combination of erosions and Boolean comparisons with the original cross section or other partially compensated bitmaps to derive the final bitmap representations of the pixels to be exposed. Numerous other modalities and algorithms for subjecting pixels to erosion, while emphasizing the reduction or maintenance of certain characteristics of the object will be apparent to those of skill in the art, in view of the teachings herein. In situations where the dimensions of pixels X and Y are significantly different, the compensation of the drop width may only be necessary along one axis instead of both axes. In these situations, modalities similar to those described above can be used, wherein only the portion of the stages will be carried out by erosion. It is anticipated that the deposition width compensation schemes can also be used by using subpixel displacement quantities in either or both of the X and Y dimensions. Alßa ori z tion One technique (method and apparatus) known as Scrambling or scrambling can be employed in the construction process. This technique can be used in combination with any of the modalities described above or any modality described hereinafter. According to this technique, the way of distributing the material in each site - for two consecutive cross sections it is varied. This can lead to a more uniform accumulation of material through a layer (ie, sheet) to give or result in the ability to potentially use thicker layers, to thereby improve construction or integration time. This technique also minimizes the effects of any single jet or plurality of jets that may not be properly firing. The variation of the deposition can be presented in several ways. For example, the variation can be presented by: 1) varying the jet which deposits the material on a given portion of a layer, in relation to the jet of material deposited on the corresponding portion of the immediately preceding layer, 2) varying the temporal order or spatial order of assortment over any given portion of the layer in relation to any other portion of the layer; and 3) a combination of these, such as by varying the main scanning operation or direction and / or varying the orientation or secondary scanning direction. The variation of deposition from layer to layer can be presented in a completely randomized manner or it can be presented in a periodic or planned manner. A similar technique has been used in photon-based photography, but for a completely different purpose (see Alternate Sequence in the North American Patent Application referenced above No. 08 / 473,834). Now specific modalities will be made to vary the deposition. The preferred present scrambling technique maintains the operation of the primary and secondary scan directions but uses a different jet (e.g., jet) to deposit the material along corresponding scan lines between two layers. In other words, a first spout is used to scan a particular main scan line on a first layer and a second spout can be used to scan that particular main scan line on a subsequent layer (the one immediately above the scan line). particular on the first layer). In some preferred embodiments, a particular scanning line is exposed (ie, deposited on) from layer to layer, using a different stream until 96 layers have been deposited and each of the 96 jets has been deposited on the line of particular exploration, after which the process is repeated. These modalities are examples of "full head" randomization. In other preferred embodiments, the reading of "half head" is preferred. The randomization of half a head can reduce the number of steps that must be carried out on the cross section depending on the geometry of the object. Based on the currently preferred 96-head head construction, half-head randomization involves scanning over any given site with the randomized or disordered assortment that occurs either from jets 1 to 48 or jets 49 to 96. To explain the full head randomization modes in more detail, reference is made to Figures 4a and 6. For a particular layer, the hole 10 (1) could be used to follow the scanning line R (l) -R (8); the hole 10 (2), lines R (9) -R (16); hole 10 (3), lines R (17) -R. { 25); hole 10 (4), lines R (26) -R (33), etc. However, on the next layer, these assignments are changed, such that a given hole ~ does not track the same scan line on the next layer. For example, the following new assignments can be made: hole 10 (1); lines R (257) -R (264); orifice 10 (2), lines R (265) -R (272); hole 10 (3), lines R (273) -R (280), etc. Another embodiment could involve relatively rotating the partially formed object and / or the print head by some amount (for example 30 °, 60 ° or 90 °) between the deposition for two layers, such that the main scanning orientations and secondary are changed from their previous orientations. This results in the deposition of the material in a present layer, (ie sheet) of any jet, which occurs mainly above the material which was deposited by other jets on the previous layer. This is shown in Figure 8, where - the scan lines associated with a first layer are represented by the lines Rj (l), Rl (2), R? (3), R? (4),. . . R 1 (li-3), R 3 N 2), R (Nl), R (N) while the scanning lines associated with a subsequent layer are represented by the line R2 (D # R2 <), R2O), R2 (4), ...

R2 (N-3), R2 (N-2), R2IN-1), R2 < N), which are rotated through 90 ° with respect to the scanning lines of the first layer. The amount of rotation may vary between subsequent layers or may be a constant amount. The angles can be chosen in such a way that if the rotation is continued by a sufficient number of layers, identical jets will deposit material on identical scanning lines where the deposition occurred on the previous layers. Alternatively, the angles can be chosen in such a way that no replenishment of the jet to the identical scan line is present. Additional modalities could involve changing the order of progress from one scan line to another (in the secondary scan direction). This is shown in Figure 9, where, for a first layer, the order of deposition of the material on the main scan lines begins on the upper scan line. R3 (l) and proceeds to the scan lines R3 (2), R3 (3) ... R3 (N-2), R3 (N-1), and ends with the lower scan line R3 (N). The order of progress of the scan lines is shown by the arrow R3p. The deposition of the material on the scan lines for a subsequent layer starts on the lower scan line, R (l) and proceeds to the scan lines R (2), R4 (3) • •. R4 (N-2), R4 (N-1) and ends with the upper scan line R4 (N). The order of advancement of the scanning lines on this subsequent layer is in the opposite direction to that of the lines on the first layer and is shown by the arrow R4p. Additional modalities are shown in the Figure 10a and 10b, wherein the scanning direction along corresponding scan lines is inverted between two subsequent layers. Figure 10a shows the scanning directions for the scan lines on a first layer, where the scan lines R5 < 1) and R5 (3) are scanned from left to right and the scan line R5 (2) is scanned from right to left. Figure 10b shows that the scanning directions are inverted on a subsequent layer, where the scan lines Rg (l), R6 (2) and 6 (3) overlap Rsd), Rs (2) and R5O), respectively, and where the scan lines Rg (l) and Rg (3) are scanned from right to left and the line of lll Rg scan (2) is scanned from left to right. Many other randomization configurations are possible, in which combinations of the techniques described above are included. Depending on the chosen scrambling technique, the scrambling process may cause an overall increase in the deposition time of the layer, since it may result in the need to carry out additional main scanning steps. However, it is believed that this possible disadvantage is overcome by the improvement in the uniform accumulation or construction of the layer. Additionally, since heat removal is a significant problem when using high assortment temperatures (as used to make the fluid material), these extra steps can be used effectively to allow additional cooling to occur before the deposition of a subsequent layer. Displacement of the Drip Site: As indicated above, some construction techniques can be improved by the use of displaced scan lines and / or displacement of drip sites along with scan lines. These displacement techniques could be used in combination with the randomization techniques indicated above, although it should be understood that the corresponding lines and drip sites on the successive layers may be displaced from each other. These techniques can also be used in combinations with other embodiments described hereinbefore or hereinafter. In some preferred embodiments, this displacement can be up to the line spacing or drip spacing of 1/2. One use of displaced pixel formation could involve depositing the material on a downward face portion of a cross section to help unite the space between the adjacent support elements. In effect, the face down region can be cured in multiple steps wherein the progressive or alternating movement between the successive steps is used to join a wide space between the support elements. In these embodiments, any portion that is not facing down the cross section could be exposed by using one or more depositions and the displaced and undisplaced pixels and deposition in any portion face down could be presented by multiple depositions ( or exposures) where the pixel zones partially overlap. The overall deposition height, in the preferred embodiments, could be made uniform by trimming to an appropriate level by flattening. In some embodiments, the displacement of the pixels and consequently, the dripping sites could occur during the formation of the support structure to improve the formation of arc-like structures, bridges or branching structures (e.g. a tree) . In some embodiments, the displacement of the pixels could occur during the formation of the object to improve the construction of the sections of the object which protrude for a limited amount beyond the borders of the sheet of the immediately preceding object. The protruding supports and portions of the object can be formed without the use of the displaced pixel, but it is believed that the displaced pixels can be useful to assist in the formation of such structures, where less material can sink to the regions below the levels. of the erapas in which it was stocked. The modalities can involve the displacement of the pixels on each layer or alternatively they can involve the displacement of the pixels only in periodic layers. In this last alternative, the material is deposited according to the same pixel positions on a variety of layers. According to this alternative, the stabilization of the pending regions may be allowed to occur by the accumulation of multiple layers, above an initial pending region, before attempting the formation of a subsequent slope portion. The displacement of the pixels, for example, to construct branching supports or to taper the structures of the object outwards, - results in the formation of structures which branch out over the empty space. The extension of this branch is limited to something less than a drop width per layer. If each layer extends beyond the boundary of its immediate preceding layer or if multiple layers accumulate one on top of the other followed by periodic extensions beyond the boundary of an immediately preceding layer, an extension angle can be defined on the basis of to the average extent over a diversity of layers. The maximum extension angle depends in part on the speed at which the nearby material and in the extended portion solidifies, which in turn depends on the amount of the material stocked nearby and on the extended portion. The layers can be accumulated or constructed at any angle where the material solidifies fast enough and is able to hold the next layer of material. In some preferred modalities, extension angles near 30 degrees have been obtained. It is believed that extension angles approaching or even greater than 45 degrees are possible. Due to the cooling rates of the material, it is preferred that the formation of the pending solid object portions be obtained in multiple steps. In a preferred embodiment, the extension region is deposited in one or more initial steps and the fully supported regions are exposed in one or more subsequent steps. This embodiment allows the material in the extension regions to cool and solidify without the added delays that could be associated with the absorption of heat from the material stocked in the internal regions. In another preferred embodiment, the interior of the layer is initially exposed, and the extension regions are exposed in one or more subsequent steps. This mode allows the time for the material on the inner portions to cool somewhat before filling the extension regions, thereby reducing the risk of the extension material remaining fluid for too long a time. For a given set of construction parameters, usable extension angles can be determined empirically by construction and examination test parts. The displacement of the pixels can be used in combination with multiple steps on a given portion of a layer, to allow the accumulation of material around a given geometric characteristic in a prescribed order and displacement configuration. For example, displacement can occur on one side of a feature, such that a fraction of a pixel displaced in its position is presented far away from that side of the feature, while a different displacement could be used for such that the same fractional displacement could be obtained in the opposite direction on the opposite side of the characteristic. An alternative to displaced pixels is simply to construct or integrate objects using higher resolution data and associated configurations or construction styles that produce the desired drip density, which may be less than inherently provided by the data, but which produces still the formation of solid structures or other desired characteristics. Intertwining of the Exploration Line: Interlacing is another technique that can be used to improve the formation of objects. As with all other embodiments, described herein, the modalities of this section may be combined with those other modalities described herein. As discussed previously, if the head is not oriented to the angle of knowledge, spatially between the jets is not equal to the desired resolution and thus is not equal to the desired spacing of the main or raster scan lines. As such, by its nature, an interlacing form of the main scan line should be used if you really want to deposit material along all the main scan lines. However, additional interlacing can be performed for a variety of reasons (for example, to improve the cooling of the layer and / or the accumulation of the material). A variety of interleaving configurations of the scan line can be used, whether or not the print head is oriented to the angle of knowing, whether or not the preferred plot scanning technique is used, whether or not use a vector scanning technique, whether or not any other scanning technique or combination technique is used. In a preferred embodiment, as previously described, the head is oriented perpendicular to the main scanning direction and a resolution of 300 scan lines per inch is used. In this configuration, the successive jets are spaces at 8/300 of an inch apart. The head is made to perform 8 main explorations, the first 7 of which are followed by a secondary scan of width equal to the spacing between the weft lines (frame width) and eight of which are followed by a secondary scan of width equal to the effective head width. plus the frame width. Repeats of the previous scan settings are made, until the width incremented in the secondary scan direction becomes equal to or greater than the width of the construction or integration region. Alternative modalities could limit the X range of the main scan to that which is sufficient to effectively cover the work region required by the object, through the cross section of the particular object that is scanned, by each segment of the length of the scan. object required to perform the 8 main explorations closely spaced or by other schemes which lead to a reduction in the exploration time. Similarly, the positioning along the secondary scan axis could be limited to itself to the width and position of the object, the cross section that is scanned, the particular portion of a cross section that is scanned or the like. In preferred embodiments, the use of scrambling may increase the amount of grading that needs to be carried out in such a manner that appropriate jets can track the appropriate main scanning lines. Other modalities may limit the main exploration to trajectories which actually include active drip sites. As a first preferred alternative interlacing technique, the non-adjacent scan lines would be left unexposed after at least a first step, after which in one or more subsequent steps the intermediate lines would be exposed. In other preferred embodiments, it is desired that the intermediate weft lines be traced before the deposition of the material, either on an adjacent weft line or another or after deposition of the material on both adjacent lines. Examples of this type of modality are illustrated in Figures a, IIb and 22a-22d. Figures 1a and 1b illustrate the situation where one line is omitted _and another not in a first step. Figure B1 shows four scan lines where two lines are going to be exposed (that is, the drip sites to be used) in a first step. Figure 11b shows the same four scan lines where the other lines are going to be exposed (that is, the drip sites to be used) in a second step. Additional examples of interlocking configurations are shown in Figures 22a-22d. In these figures, two arrow heads 30 represent the main scanning direction, the spacing dr, represents the spacing between successive grid lines and for clarity the start points and end points of the lines are displaced, although in practice, the lines would have the same starting and ending points. Figure 22a illustrates a series of raster lines to be scanned in the main scanning direction. Figure 22b illustrates the first weft lines 32 to be exposed in a first step and second weft lines 34 to be formed in a second step according to the Example of Figures Ia and llb. Figure 22c illustrates the weft lines 32, 34, 36 and 38 to be exposed in first, second, third and fourth steps respectively. Figure 22d illustrates raster lines 32, 34, 36, 38, 40 and 42 to be displayed in first, second, third, fourth, fifth and sixth steps respectively. In the example of Figure 22d, other raster line scan commands could also be used as long as it is ensured that when intermediate lines are laid down they are not limited either on one side or the other or are limited on both sides. sides by adjacent weft lines previously deposited. For example, other useful exploration orders could be 32, 34, 38, 36, 40 and 42; 32, 34, 40, 38 and 42; or similar. In a preferred system, to fully implement these modalities in a generalized manner by using a minimum number of steps, an odd number of weft lines would need to exist between the line scanned by one of the jets (e.g., a first jet). ) and the line scanned by an adjacent jet (eg, a second jet). In other words, the spacing number d between the successive jets would have to be uniform; which requires by this that two adjacent jets must be positioned to explore lines of frame M and M + N where M and N are integers and N is even. In the case where the spacing between the jets is not appropriate (for example, not uniform),. it is always possible to scan only appropriate plot lines (for example, those associated with a jet and another no) in a first step and then expose the remaining scanning lines in one or more subsequent steps. Since the deposition width can be significantly wider than the spacing of the frame line, other preferred modalities could not be based on the omission of one scan line and another not in a first step, but instead of this be based on the selection of the scanning lines for the deposition (ie, exposure) in the first step, so that the lines of the deposited material do not come into direct contact with each other, and then the filling in any line of the frame omitted in one or more subsequent exposures. This first alternative entanglement technique can be implemented fully or approximately even when the adjacent jets are inappropriately positioned for the "resolution of the desired scan line (ie, the jet positions and the scan line resolution are such). so there is a uniform number of weft lines between the line scanned by one of the jets and the line scanned by an adjacent jet.) This can be done in at least three ways: 1) each jet is used to to explore a plot line and another not between its initial position and the position of the line formed initially by the adjacent jet, except at least two adjacent weft lines to be scanned by each jet will be left unexposed, until at least a second step when the remaining raster lines will be exposed, 2) each stream is used to scan one raster line and another, not until also explores the weft line adjacent to the first line scanned by the adjacent jet, after which the remaining unexposed lines will be selectively exposed in a second step; and / or 3) only one jet if and one other is not used in the scanning process, to thereby ensure that there is an odd number of weft lines between any two adjacent jets.

In these modalities, also like the previous modalities, it is preferable to expose the alternating lines for the total layer before beginning a second step to expose the intermediate lines; however, it is possible to consummate the exposure of all the scanning lines between the starting points of some or all of the adjacent jets before being even a first step on other portions of the layer. Numerous other interlacing modalities will be apparent to those of ordinary skill in the art when studying this description. For example, the interlacing with higher step numbers can be used or the interlacing where there is some contact between the exposed lines in a first step. Of course, any combination of interleaving with the randomization techniques described previously can also be used. The additional exposure of a subsequent layer can change the scanning order of the various sets of lines and / or the scanning directions of the lines themselves (for example, invert the scan order of the first, second, and order sets more high) .

Additional modalities could involve the completion of the interlacing exposures for a first layer while exposing the regions during the formation of one or more subsequent layers. Interlocked Drip Site: As with the interlacing of the scanning line, the formation of the object can use the interleaving of the drip site along the individual scanning lines. In this case, each scan line would be exposed by at least two steps, where a first step would expose a number of drip sites and after that, in one or more subsequent steps, the remaining drip sites would be exposed. As an example of two steps (ie, step), in a first step a drip site if and one would not be exposed, while in a second step the intermediate drip sites would be exposed. This situation is illustrated in Figures 12a and 12b. Figure 12a illustrates four scan lines each with 9 drip sites where one drip site and another will not be exposed in a first step, while Figure 12b shows the same lines and sites, but instead from this only illustrates that the complementary drip sites are going to be exposed in a second step. As a second example of two stages, each two sites will be exposed in a first step, while in a second step both intermediate sites, between them, would be exposed. As a third example, a first step could expose four sites yes and four not starting with the first site, then in a second step every four sites would be exposed starting with a third site and finally, in a third step one site would be exposed and another not starting with the second site. As with all other embodiments described herein, the embodiments of this section are combinable with the other embodiments described herein. In these interleaving techniques, successive scan lines can be exposed using different or shifted interleaving configurations, so that two-dimensional interleaving configurations can be developed (displaced pixels can also be used). For example, a two-stage interleaving configuration may be used in each scan line, where the starting points on successive lines are offset by-- one pixel, such that a configuration of a first board step is formed. Of chess. Figures 13a and 13b illustrate this example. Figure 13a illustrates the chessboard configuration of the first step, while Figure 13b illustrates the complementary chessboard configuration that is set forth in a second step. Similarly with the interlacing of the scanning line, the interleaving of the drip site can complete all the steps on individual lines before exposing the subsequent lines, although it is preferred that all the lines are exposed with each step before initiating the subsequent steps on the partially exposed lines. In addition, the completion of all the steps on portions of individual lines, can be obtained before starting the exhibition on the remaining portions of those lines. A third interlacing technique involves interlacing sensitive to characteristics. In this technique, the order in which a given drip site is exposed depends on the geometry of the immediate cross section alone or on multiple cross section geometries. The feature-sensitive interlacing may involve the interlacing of the scanning line and the interleaving of the drip site. For example, in a single layer embodiment one can determine the boundary regions of the cross sections and ensure that the boundary regions are exposed in a first step, some interior portions of the cross section could also be exposed in the first step or alternatively the exposure of all interior portions may be delayed until one or more subsequent steps are made. For example, the interior portions can be exposed using a chess board interlacing configuration in a first step, in combination with all the exposed regions that are also exposed in the first step. Then, in a second step, the remaining interior portions would be exposed. It is also possible that a wide border width could be defined for exposure in a first step, such that more than one border of the width of the drip site can be placed around the cross section before carrying out the subsequent steps. This wide border region could be implemented using erosion routines such as those described above in association with Drop Width Compensation. As a further alternative, it can be ensured that only one of the boundary sites of the scan line or the border sites of the drip site (boundaries along the lines in the secondary scan direction) are exposed in the first step. As a further alternative, the internal regions can be exposed in whole or in part before supplying the material in the border regions. It is believed that the assortment of border regions could lead first to improved accumulation in the vertical direction and that exposure of the border regions could eventually lead to improved horizontal accuracy of the object. An additional alternative could involve the assortment of a nearby border region initially, followed by the assortment of deeper inner regions of the cross section and finally followed by the assortment of the internal cross-section boundary itself.

Examples of a feature-sensitive interleaving technique, of multiple cross sections, could involve initially exposing those sites which are part of the present cross section, but which were internal boundary or solid object regions over the previous cross section. Solid boundary and internal regions over the previous cross section could include solid boundary regions and internal regions of support structures as well as object structures. When using this modality, the deposition in the regions of the object of face downwards less critical (this is important) does not appear in the first step, unless these regions of face downwards are really supported by some structure of some nature (for example, a support column directly below). In one or more subsequent exposures, the material is dispensed to form unsupported face down features. Since the deposition width is usually wider than the width of the pixel, it is more likely that a drop which is dispensed will settle on a pixel site adjacent to the previously stocked material on that cross section, the drop will collide and adhere to -easily to the neighbor deposited material as opposed to continuously falling down to a cross section below which it was proposed. Further, since in the preferred embodiments, the support structures commonly consist of no more than one pixel in separation, when the exposure of the unsupported face-down regions is presented, the assortment material will most likely be wedged between the material already Assortment on the present layer as opposed to being wedged between the material assortment on a previous layer. However, since the diameter of the drop is usually less than the deposition diameter (ie diameter of the impacted drop) and since it may be less than the width of the pixel, the material deposited at an adjacent pixel site it may not extend sufficiently to the trajectory of a falling drop to ensure a collision and stop the particle. In another preferred embodiment, the drip sites would be displaced by a fraction of pixel width (preferably about 1/2 pixel width) along the main and / or secondary scan directions (preferably both) when the unsupported face-down regions and preferably the adjacent regions, in such a way that a drop is more likely to be supported at least partially by the previously supplied material than if the droplets were deposited in perfect alignment. It is preferred that the droplets on the partially unsupported regions are filled in a subsequent step of those dispensed over the fully supported regions. However, it is possible to depend only on the superposition with the previous cross section (and not any additional benefit associated with adhesion to the material previously distributed over the given cross section) to ensure reasonable vertical placement of the drops in the partially unsupported regions . In this modality at least the regions of supports (for example, column) on the current layer would not be displaced. This ensures that correspondence or layer-by-layer matching occurs. It is further preferred that the wide spaces be closed by progressive inward (ie, multistage) work deposition sites of the unsupported sides of the space by using multiple steps on the cross section, where each step is partially offset from the passageway. immediately preceding, to ensure proper overlap of the drops to limit any placement of material beyond the required vertical level. In addition, in a preferred embodiment, simultaneous multilayer curing techniques, as described in US Patent Application No. 08 / 428,951, are used in order to move the critical face down data to one or more layers, such that the deposition material forming the face layer down will be located at the correct level. An example of this modality of horizontally and vertically multistage displacement utilizing a horizontal displacement of 1/2 pixel and vertical displacement of the thickness of 1 layer is shown in Figures 23a-23h. Figure 23a shows a side view of an object 120 to be formed. Figure 23b shows the object 120 as it would normally be divided into layers 122, 123, 126, 128 and 130. Figure 23c illustrates the object 120 as it is to be divided into layers 122, 124, 126, 128 'and 130'. The layer 128 'is different from the layer 128 in that the face down portion of the layer has been separated, since it is anticipated that it will be created during the deposition of the material on the next layer using a series of successively displaced exposures. Layer 130 'is similar to layer 130, except that a different deposition configuration could be used in its formation. Figure 23d again illustrates the layers 122, 124, 126 and 128 ', but also illustrates the deposition sites or pixel positions, 132-137, in which the material can be deposited during the formation of the layer 130 *. Figure 23e is similar to Figure 23d, except that instead of showing the drip sites 132-137, the drip sites 140-146 are shown. As can be seen from the relative positions of the drip sites, sites 132-137 and 140-146 are offset from each other by a width of 1-2 pixel. Figure 23f illustrates the deposition pattern formed from a first step of the print head in the formation of the layer 130 '. Drops 150, 151, 152 and 153 are deposited at drip sites 141, 145, 142 and 144, respectively. It can be seen that drops 152 and 153 were only partially supported by layer 128 'and that as a result it is assumed that they partially extend (as illustrated) to the region originally belonging to layer 128. Figure 23g illustrates the configuration of deposition of the first step in the formation of the layer 130 ', also as of the additional material deposited in a second step. The regions 160 and 162 were deposited in the first step and are represented in Figure 23f as in the regions 150, 152, 151 and 153. The deposition in the second step is presented according to the array of pixels illustrated in Figure 23d. Drops 155 and 156 are deposited at drip sites 132 and 137. In practice, the assortment of drops 155 and 156 would initially result in excess material being applied over a portion of regions 160 and 162, but this excess would be cut during the flattening process. Drops 157 and 158 are deposited at drip sites 134 and 135, but since these sites are not fully bounded from below by the previously deposited material, it is assumed that a portion of the assortment material will extend downward to the original part region. of layer 128. The displaced assortment of drops 152, 153, 157 and 158 results in the formation of the downward face portion of layer 128 which was separated from layer 128 '. In a third step and final step, the drop 164 is deposited on the drip site 143, to complete the formation of the layer 130 '. In other preferred embodiments, several aspects can be changed to the previous example. For example, the extension of the material to the lower regions of the layer (assumed to occur when droplets or drip sites are highly supported in a partial manner) could take on values different from the thickness extension of a described layer. The extension may be less than the thickness of 1 layer or at least different from an integral number of thicknesses of the layer. Perhaps, the extension would be an integral number of thicknesses of the layer (for example, thicknesses of 2 to 5 layers or more). In such a case, for a more accurate formation, it would be desirable to transform the initial representation of the object to a modified representation, as described in US Patent Application 08 / 428,951, (either before or after the generation of the data of the cross section) such that when the material is distributed according to the modified representation, the bottom of the face down feature is appropriately located. Other variations could use the deposition based on the geometry, in multiple steps, along with different displacement values such as 1/4 of a pixel (so that 3/4 of the drip zone would not be supported) or 3/4 of pixel (in such a way that only 1/4 of the drip site would not be supported). These different displacement amounts could lead to greater control over the extension amounts to the previous layer regions. Other variations could use different deposition orders, different amounts of overprint or uniform amounts of deposition per drop. Still other variations may not use the displaced pixels, but instead use higher resolution pixels, possibly in combination with deposition settings that produce the correct droplet density. An additional interlacing technique comb: 1) sensitivity to the characteristics, and 2) selective direction of the scan when object characteristics are exposed. In this embodiment, the geometry of the cross section (e.g., cross-sectional boundary information) of a present layer and possibly the geometry of the cross section (i.e., the cross-sectional boundary information) of the immediately preceding layer they would be used to determwhat the scanning direction would be when different regions of the cross section are exposed. For example, when exposing the leftmost portion of a solid region of a cross section, it may be advantageous to scan the head (ie, the jet used to expose the lto be formed) from left to right if it is desired that the drop not one or not one partially some small space. On the other hand, if it is desired that some union be present it may be advantageous to ensure that the scan is in the opposite direction. Similarly, when exposing the right back portion of a solid region of the cross section, it may be advantageous to scan from right to left (not to join) or from left to right (to join). By controlling the direction of exploration when border regions are deposited, it can be assured that the moment or amount of horizontal movement of the drops does not contribute to join spaces or improve the union of the spaces. An example of the non-union technique is illustrated in Figures 24a-24d. Figures 24a-d illustrate side views of two columns as they are formed and how they are cut in a plane XZ. The Z direction is perpendicular to the planes of the cross sections and the X direction is the main scanning direction. The reference number 108 indicates that the cross section is formed and the reference numerals 100, 102, 104 and 106 refer to the previously formed transverse sections. Figure 24a illustrates the cross section 108 with a dashed lsince the deposition of the material has not been carried out. Figure 24b indicates that the scanning direction 110 is to the right and that the drops 112 are deposited on the left most side of each column in a first step. Figure 24c indicates that the scanning direction 124 is on the left and that the drops 114 are deposited on the rightmost side of each column in a second step. Figure 24d indicates that the scan may be presented either in one direction or another 126 and that the drops 116, 118, 120 and 122 are deposited to consummate the formation of the cross section in a third step. In contrast to the illustrated three-step modality, a two step modality could be used where droplets 116, 118, 120 and 122 can be deposited to their respective sites during one or both of the first or second steps when droplets 112 and 114 were deposited. It is anticipated that the object could be relatively reoriented (e.g., one or more rotations about the vertical axis) with respect to the relative scan direction of the print head (ie jets) such that the edges of any characteristic of The desired cross-section can be exposed by moving the print head relatively in a desired direction to improve or decrease the likelihood of joining the small spaces. As indicated above, the distance from the orifice plate to the work surface is too small, the droplets will have an elongated shape (this is a large aspect ratio) as they collide with the work surface. In the case of construction with elongated droplets, it is anticipated that the scanning directions indicated above for deposition on the edges of the solid features could produce opposite results from those indicated above. Other interlacing techniques could involve bidirectional printing of adjacent weft lines or non-adjacent weft lines. The construction techniques described above can be applied to the formation of solid objects or in combination with other techniques for the formation of partially or semi-solid hollow objects. In an original design of an object, the portions of the object are assumed to be solid (i.e., formed of solidified material) and the portions are assumed to be hollow (i.e., empty or hollow regions). Actually, these proposed hollow (or empty) regions that are not supposed to be part of the object since, by definition whenever there is an object it is supposed to consist of material. In the context of the present invention a hollow or semi-solid non-solid object is an integrated object or to be integrated according to the teachings of some preferred embodiments wherein a portion of what must be a solid object has been removed. A classic example of this could be the cupping, partial cupping or honeycomb formation of what was originally a solid structure of the object. These originally solid structures are referred to as object walls, regardless of their spatial orientation. Some preferred construction styles form completely solid objects, while other construction styles form solid surface regions of the objects but hollow or partially hollow interior regions. For example, interior portions of an object could be formed in a checkerboard, striped, hexagonal, shingled or honeycomb manner (these and other construction styles useful herein, as implemented in the tereoli tography). base photons are described in the patents and applications referred to above). You can consider the previous non-solid deposition configurations, structures of internal object supports. As such, the other support structures described herein may also be used as internal object support structures. Such non-solid objects would be lighter in weight than their solid counterparts, would use less material, could still be formed more quickly depending on the details of the specific construction parameters and could be formed with less risk of encountering heat dissipation problems, since that much less heated material is deposited during its formation. These objects could be useful as investment casting patterns due to the decrease in the possibility of cracking of the molds. Temperature Control: The modalities of formation of additional objects, involve the formation of the object where the partially formed object is maintained within a desired temperature range as it is formed or maintained at least in such a way that the differential in temperature across - the part (or the temperature difference gradient) is small. If during the formation of the object, the different portions of the object are allowed to be at different temperatures, the object will experience a differential amount of shrinkage as it cools to room temperature or as it is brought to its temperature of use (the temperature at which In use) . This differential in contraction could lead to the development of efforts within the object and associated distortions or even fractures of the object. It is preferred that the temperature differential remain within a range which is effective in maintaining the distortion of the object within a reasonable limit. The temperature difference across the object is preferably maintained within a range of 20 ° C, more preferably within a range of 10 ° C, and even more preferably within a range of 5 ° C and more preferably within a range of 3 ° C. In any case, the desired temperature can be estimated by taking into consideration the coefficient of thermal expansion of the material and the differential in contraction (or expansion) that could occur during cooling (or heating) of the formed object at a uniform temperature. If the contraction differential results in an error outside of a desired tolerance range, the temperature ranges mentioned above can be adjusted. In the formation of objects, the initial data of the object can be scaled to take into account the dimensional changes in the object that will appear as the object cools from its jet temperature (approximately 130 ° C in the preferred mode) to its solidification temperature (approximately 50 ° C-80 ° C with a maximum DSC energy transfer temperature of approximately 56 ° C at its construction temperature (approximately 40 ° C-45 ° C) and finally at its use temperature ( for example, room temperature - approximately 25 ° C.) This scaling factor could be used to expand the initial design of the object by an appropriate thermal shrinkage compensation factor, such that it would be sized appropriately at its temperature of use. fur anticipates that one or more contraction factors dependent on the geometry or at least dependent on the axes could be used to compensate for at least partially the critical regions of the object by the expected variations in the temperature of the object during construction. The temperature of the previously formed sheets and the rate of increase of the sheet that is formed have been found to be important parameters for the formation of objects with reduced distortion and in particular with reduced corrugation distortion. Presently preferred materials undergo about 15% shrinkage when cooled from their solidification temperature to room temperature. This contraction provides a tremendous motivating force to cause wavy distortion, the accumulation of internal stresses and associated post-processing distortions (these distortions are described with respect to photon-based tereoligraphy in applications and patents to referred to above, wherein many of the construction techniques described therein can be used effectively in the practice of SDM and TSL in view of the teaching found in the present application). It has been found that if the construction temperature of the object and in particular if the temperature of the last formed layer is maintained at a temperature higher than the ambient temperature during the construction process, the distortion of the corrugation will be reduced. It is preferred that the temperature of the partially formed complete object be maintained above room temperature and more particularly that its temperature remain within a narrow tolerance band due to the differential concentration considerations discussed above. For an effective formation of the object, it is evident that the construction temperature of the partially formed object must be maintained below the melting point of the material. Additionally, the construction temperature must be maintained at a temperature lower than a temperature that allows the solidified material to have sufficient resistance to shear and compression and even tensile strength (especially if lateral or face-facing object forming modalities are used). above) to allow the object to form exactly as long as it experiences the typical forces associated with the construction process (for example, the forces of inertia associated with the accelerations that the object will experience, drag forces or vacuum forces associated with the flattening device and in the print head which is brought into contact or passing close to the object, the air pressure forces associated with any air flow used to cool the object and the gravitational forces on the object due to its own weight Some of these forces are dependent on the mass of the object and increase so with the depth in the part. Thus, a slightly negative temperature gradient from the higher to lower layers (that is, decrease in temperature of the layers formed more recently to the layers formed above), can provide increased resistance in the regions that need it. which simultaneously allows the layer or layers to finally be at a sufficiently high temperature to result in minimal undulation and other distortions. One could use a calculation of the simple gravitational force added with a calculation of the force of inertia for one or more positions of the part (based on the mass of the part and the acceleration in the Y direction that - experience) as an approximation of the minimum necessary cut resistance of the solidified material. This, in combination with an empirical determination of the variation of the shear strength of the material with the temperature can be used to estimate the approximate upper construction temperature limit for any position on the object. Of course, it is preferred that additional considerations be taken into account, especially near the last formed sheet of the object, since dynamic thermal effects occur at the interface of the partially formed object and the resulting material that would involve reflow phenomena and phenomena. of thermal capacity which are dependent on the geometry parameters of the object, temperature differentials and cooling technique. Thus, the actual global maximum construction temperature will probably be lower than the previous estimated amount. On the other hand, as indicated above, corrugation and other distortions can be significantly reduced when building at elevated temperatures, where the higher the temperature, the lower the distortion. It is postulated that this reduction in distortion results from a combination of the improved ability of the material to flow at elevated temperatures and its lower capacity to withstand shear loads which allow some distribution of the material to be present to thereby reduce the stress, which causes the distortion. It is further postulated that near work, or preferably at higher temperatures of solid state phase change (e.g., crystallization temperature or vitreous transition temperature) will result in faster and potentially more significant reductions in stress and distortion. Since these base changes normally occur over a wide range, it is postulated that several levels of benefit are presented depending on where the working temperature lies in these ranges and the time of the process is allowed. Melting temperatures and / or solidification temperatures and solid state transition temperatures can be determined by using Differential Scanning Calorimetry (DSC) techniques, which in turn can be used in the determination of the appropriate construction temperature ranges. Additionally, the appropriate construction temperature ranges can be determined empirically. It has been determined that some benefit can be gained by working at any temperature higher than the ambient temperature and it is anticipated that the closer it moves to the melting temperature and / or the solidification temperature the greater the benefit. Thus, the working temperature range could be set as a percentage of the distance along the temperature differential between the ambient temperature and the melting or solidification temperature or the ambient temperature and the minimum cut resistance temperature Dear. Alternatively, the working temperature can be selected to be a temperature at which the material has a certain percentage of its cut resistance at room temperature. For example, it may be desirable to set the working temperature (construction) in such a way that the cut resistance is 75%, 50%, 25% or even 10% of its maximum ambient temperature value. Improvement of the Surface Finish Additional construction modalities - useful to improve the surface finish of the object involve taking the advantage of the aesthetically pleasing face up surface which results from the practice of the preferred SDM techniques. In these embodiments, the number of effective face-up surfaces (e.g., the overall area) is increased while the effective face-down surface number is reduced from that defined by the design of the original object. This involves dividing the object into two or more pieces and changing the orientation of the separated pieces, so that as many critical surfaces as possible are made as face-up surfaces, vertical surfaces or face-up / vertical combined surfaces. , where no truly external surface or only less critical surfaces remain as face-down surfaces. These separate object components are then integrated independently of each other, each with the appropriate orientation. Then, the supports are removed and the resulting components are combined by gluing or the like, in such a way that a complete object is formed mainly from surface regions facing upwards and verticals. If rough or rough surfaces are desired instead of smooth surfaces, the prior art can be used to ensure that the critical surfaces are formed as face-down surfaces.

As an alternative to the up-facing surfaces which are to be wrinkled or roughened they can be formed simply with supports extending therefrom. An example of this construction technique is illustrated in Figures 25a-e. Figure 25a illustrates the configuration of an object 60 to be formed using SDM (i.e., the design of the desired object). If the object is formed directly from this design, the object will be formed with features or surfaces facing upwards (5O, 52 and 54) and features or surfaces face down (56 and 58). As discussed previously, the formation of face-down features requires the prior formation of a support structure which acts as a work surface on which the material forming the face-down features is distributed. After the formation of the object and the separation of the supports, it has been found that the surfaces facing downwards are left with a rough and irregular surface finish. If the surface facing downward is desired to be smooth, the object must experiment with an additional post-processing, which may require sanding or detailed filling. Figure 25b illustrates the first step in the practice of the prior art. This first stage involves the division of the design of the original or desired object into two or more components. The division is carried out in such a way that all the critical features of the object can be formed either as vertical surfaces or facing surfaces upwards (preferably as faces facing upwards and more preferably as faces facing upwards which they do not have face-down surfaces above them, such that no supports will be formed which start from and damage the surfaces facing up). Additional details about the formation of supports and issues associated with them will be described later in the present. In the present example, all surfaces 50, 52, 54, 56 and 58 are considered critical and must be formed as faces facing upwards. Figure 25b shows that the object 60 is divided into two portions 62 and 64. The portion 62 includes original outward facing features 50, 52 and 54 and new temporary outward face features 72 and 74. Portion 64 includes features of original or desired outward facing 56 and 58 and new or temporary outward facing features 72 'and 74'. Figure 25c shows the preferred orientation (the right side upwards) of the portion 62 during forming, such that the surfaces 50, 52 and 54 are formed as face-up features. Figure 25d illustrates the preferred (downward facing) orientation of the portion 64 during forming, such that the surfaces 56 and 58 are formed as face-up features. After the formation of each portion 62 and 64 the supports are removed and the pairs of temporary surfaces 72 and 72 'and 74 and 74' are prepared for coupling. Figure 25e shows the joining of the portions 62 and 64 to form the object 60, wherein all the critical outward facing portions (i.e., original surfaces 50, 52, 54, 56 and 58) have good surface finish. Additional Construction Styles: Other construction styles may include one or more of the following: 1) the use of the highest resolution distribution or assortment in the scanning directions; 2) the use of a higher drip density per unit area in the formation of upper surfaces facing downwards than in the formation of interior regions of the object; 3) the use of surface regions facing downwards which extend to at least N layers (for example 5 to 10) above the surfaces facing down; 4) the use of a higher drip density per unit area when upper surface surfaces are formed facing upwards than in the formation of interior regions of the object; 5) the use of surface regions facing upwards which extend to at least N layers (e.g., 5 to 10) below an upward facing surface; 6) the use of a higher drip density per unit area when border regions of an object are formed than when forming inner regions, boundary regions which extend to at least L drop widths (eg, 2 to 4) ) to the interior of an object; and 7) formation of interior regions of the object by means of raster scanning and vector exploration across boundary regions.

Support Styles: The following portion of the application is directed primarily to the formation of media. However, it should be appreciated that since the supports are formed from the deposited material, all the construction techniques mentioned above are applicable to the construction process of supports. In addition, as will be appreciated all aspects of the process of construction of the support are applicable to the construction of the object as well. The supporting structures must serve several needs which may be opposite: 1) preferably they form a good work surface on which the sheets of the object are integrated and even successive support sheets; 2) are preferably easily separable from the downward facing surface they support; 3) if parts of an upwardly facing surface of the object are preferably easily separable therefrom; 4) when they are separated, the supports preferably cause only minimal damage to the surfaces facing downwards and facing upwards, and preferably have a finish at least tolerable to good on those surfaces; ) are preferably integrated at a reasonable speed per cross section in the vertical direction (eg, Z direction); 6) are preferably formed using a minimum number of steps per layer; and 7) their training is preferably reliable. A variety of different support styles have been developed or proposed, which obtain different balances between these needs. To optimize the construction speed, vertical accumulation is important and as such, it is desirable to have integrated supports at approximately the same speed as the object. In particular, it is preferred that the vertical accumulation of the supports (eg, from a single step per layer) is at least as large as the desired layer thickness set by the use of the flattener device. The closer the support accumulation is to the accumulation of the object, the thicker the usable layer is and the less material will be separated during flattening, which increases the efficiency of the construction process. For a material and data apparatus, the vertical accumulation of the material from different styles of support and construction can be determined empirically, as previously described, by constructing or integrating test parts for each style or deposition configuration by using different layer thicknesses (leveling levels) and thereafter measuring the parts to determine when the buildup of material was delayed behind the anticipated thickness as determined by the number of layers deposited and the expected layer thickness. From this information, either the thickness of the layer (level of flattening) can be adjusted to an appropriate amount for a desired combination of construction and support styles or the required required support and construction styles can be set to obtain the desired layer thickness. Some preferred style of support styles emphasize the speed of formation, maintain easy separation, but have a rough surface finish in the regions where the supports have been separated. This style of support involves the formation of solid columns, which are separated by small spaces. In particular, in a preferred system, the data is supplied at 300 pixels per inch in both X and Y directions and the object and supports are formed using four times the ID overprint in the X direction (main scan direction). Each support layer includes areas of three by three pixels where the support material is to be supplied with the columns separated by two pixel areas of a deposition of no defined pixel along the main scanning direction (X direction) and the zone of a pixel of no pixel deposition defined in the secondary scan direction (Y direction). The location of the data defining these pixel zones is illustrated in Figure 15a. The "X" in the figure illustrates pixels which contain drop data while the "O" in the figure illustrate pixels which contain "no drop" data. The 50 squares have been inscribed around the zones to highlight the shapes of the deposition zones --- however, due to the ID overprint in the X direction, the spaces of two pixels are actually narrowed considerably (by a width of almost one pixel) when the actual deposition is presented. Thus, the actual resulting deposition configuration approximates more closely to columns of 4 by 3 pixels wide (12-14 thousandths of an inch by -9-10 thousandths of an inch) although with rounded corners, which are separated by 1 space of one pixel wide in X and Y (3.3 thousandths). This situation is illustrated approximately in Figure 18. In the practice of object construction, it has been found that the supports of the previous configuration accumulate at approximately the same speed as the object and thus a single step of the head on each site The drop can be used to form the supports and the object on each layer. It has also been found that the above support structure is easily detachable from the object but that a poor face-down surface finish results. Therefore, in terms of speed of. -construction, the previous style is preferred, but in terms of the surface finish, a significant room temperature persists to improve. One variant involves using multiple steps of the distributor head to form a support portion of a cross section. Another alternative periodically distributes a cross section of extra support in order to equalize the accumulation of vertical material between the supports and the object. Another variant involves allowing the formation of the support to be delayed behind the formation of the object by one or more layers to eliminate or minimize the flattening problems that may arise in the case where fragile supports are constructed. The problem is that the flattening device can cause these supports to distort if the support portions of a cross section are supplied during the same or the same steps as the corresponding object portion of the cross section. By allowing a delay of one or more layers to occur, excessive contact between the supports and the flattening device can be avoided and it is anticipated that the resulting distortion of the supports will be minimized. Other support structures similar to a column are possible, in which columns of different dimensions or shapes are included. For example, data format and overprint techniques could be combined to produce physical columns of approximately 3 by 3 pixels (9-10 thousandths by 9-10 thousandths of an inch) in size of 2 by 3 or 3 for 2 pixels (these can result in less vertical accumulation), size 2 by 2 pixels (6-7 thousandths of an inch by 6-7 thousandths of an inch) probable loss in vertical accumulation speed), size of 4 by 4 pixels (12-14 thousandths of an inch by 12-14 thousandths of an inch) (may be more difficult to separate and may cause additional damage to the object's surfaces or even larger sizes.) Other columns of cross-sectional shape may also be used. These could include structures formed in a more circular fashion (eg, octagonal or hexagonal), cross-like structures, structures with different length-to-width aspect ratios, or combinations of structures that can be intermixed.Other alternatives could include moving columns support alternatives in one or both of the main scan and sub-scan directions, for example, one support column and another could not be moved in the secondary scan direction by 1/2 the separation between the columns. Figure 19, A wider spacing of support columns is possible, particularly if some technical The arc or junction support is used to narrow the space between the support columns before encountering a face down face of the object. Two examples of arc-like supports are shown in Figures 21a and 21b, where different amounts of pixel shift (or at least drop drop control) are used. Branching Supports: As described in several sites hereinbefore, some preferred embodiments use supports that can be described as branching supports. The arc-like supports discussed above are the example of a type of branching support. The branching supports or branch supports are a support structure which are integrated in such a way that the portions of some sheets extend outwards in a cantilevered manner from solidified regions on the immediately preceding sheets. These outward extensions may be based on identical (this is fixed) pixel positions of layer-in-layer. Alternatively, these outward extensions may be based on fractional displacements of the pixel width at pixel positions between some or all of the layers. Additional alternatives may be based on changing pixel configurations among some or all of the layers. Some embodiments of branching supports produce more individual support structures to a surface to be supported than the number of support structures from which the branching supports originate in a lower layer. In addition to the various embodiments described previously (which in essence can be considered as branching supports), Figures 28a, 28b, 29a-3, 30a-m, 31a-c, 32a-d show additional examples of support structures of branching preferred. Figure 28a shows a side view of the column supports 504, 506 and 508 starting at the surface 500 and working to the surface 502. These column supports are joined together by branch members 510, 512, 514 and 516. Figure 28b illustrates a side view of a type of branching-type supports working from the surface 500 to the surface 502. It is shown that the supports are branched every two layers. In this two-dimensional view, such a branch appears to be in a two-trajectory fork-like configuration, while other branches simply branch along a single path. The same support structure illustrated in Figure 28b is viewed from a different view in Figures 31a-c and 32a-d. Other preferred branching configurations are illustrated in the example of Figures 29a-e. Figures 29a-e illustrate top views of successive branch cross sections for a single support tree using X only and Y branches only and results in a total of four support branches from a single support line. Figure 29a illustrates a single support structure that will be branched into a plurality of structures. This structure of a single support can be called the "main line" of the tree or support structure. As will be explained later in this, for ease of data manipulation, the main line can be considered to consist of four separate but identical components, which maintain their separate identity, but can be subjected to Boolean calculation together to produce the scanning configuration for any given layer. Of course, in practice, a real region to be supported could require a plurality of these main line elements spaced appropriately with each other. Figure 29b illustrates a first branch in the X direction. As with the other Figures that follow, the shaded solid regions, as illustrated, represent the regions of deposition for the present cross section, where the region (s) illustrated (s) with dashed lines represents (n) the immediately pertinent branch. This way of illustrating the regions of deposition is made to make clear the correspondence between the branches. This first branch can occur after one or more layers of the main line are formed. As with the other branches to be described later herein in association with this figure and other figures that follow, branching may extend the assortment or distributed material out of the supported regions by a fraction of a pixel, an entire pixel or multiple pixels depending on the order of extraction used, the width of the pixel compared to the width of the drop, the number of identical layers to be formed above the present layer (which can compensate for the imperfections in the present layer) the capacity of the material to be partially not supported and the like. As with some of the other branches to be discussed later in the present, this branch can be viewed as a bidirectional branch (that is, one direction in the positive X direction and the other in the negative X direction) or as a unidirectional branch of two or more components initially overlapped. As will be seen from the description that follows, it can be considered that this first branch is a unidirectional branch of four initial components, where two components follow each branching direction. The actual deposition of material of these four components can be based on a Boolean union of the components, in such a way that multiple depositions on the regions of superposition are avoided. Figure 29c illustrates the next branching of the tree, where this branching may initially present one or more layers after the branch illustrated in Figure 29b. This branch of the object's components is presented in the same directions as shown in Figure 29b. Figure 29d illustrates two branches in the Y direction of each of the two branches illustrated in Figure 29c. In concept, it can be considered that this is a single branch in the Y direction of the separate components. The branch illustrated in Figure 29d is the first branch which begins the separation process of all four components. Figure 29e illustrates a final branch for this exemplary mode, where a branch is made - in the additional Y direction of each component. These end branches can be used to support the surface of an object as appropriate. If the surface of an object locates several layers above these final branches, the structures (e.g., columns) of Figure 29e can be extended until it meets the surface of the object. If the surface of the object is not at the same level for all four branches, the columns or individual portions of the columns can be extended as necessary. This extension of the support height is similar to other preferred column support embodiments discussed herein and may include the use of tie layers and the like. Of course, if different configurations are desired (for example, shapes, positions and the like) of the four-column branched support, modifications (eg, modifications to the order of branching, branch directions, extension amounts, number of layers between branches and the like) can be made to the mode shown and they will be apparent to those skilled in the art in view of the teachings herein. The main support line illustrated in Figure 29a may initially be formed on a previous cross-section of the object or an initial substrate. Alternatively, the main line may start on top of another support structure such as that illustrated in Figure 28a. In addition, if multiple trees are to be used, the branching of the trees may or may not start in the same layer and may or may not result in each branch forming after the same number of layers. The selection of where to start the branch and when to make the successive branches after this, may be based on the geometry of the object to be formed. It may be desirable to obtain the final branching configuration, for a particular tree, several layers before first encountering a surface to be supported (eg, surface of the object facing down). The branching routines carried out in association with the exemplary embodiment illustrated in Figures 29a-29e can be summarized in the following table: Component 81 Component 82 Component 83 Component 84 Construction without branching for a desired number of layers (Figure 29a) Branching in the Branch in the Branch in the Branch in the * X direction by the * X direction by the -X direction by the -X direction by the desired quantity A quantity A desired amount A desired amount A (Figure 29b) (Figure 29b) (Figure 29b) (Figure 29b) Construction without branching for a desired number of layers Branching in the Branch in the Branch in the Branch in the direction «X by the direction «X for the address -X for the address -X for the quantity A desired quantity A desired quantity A desired quantity A decade (Figure 29c) (Figure 29c) (Figure 29c) (Figure 29c) Construction without branching for a number of desired layers Branching in the Ramification in the Branch in the Branch in the direction * And by the direction -Y by the direction * And by the direction -Y by the amount A desired amount A desired Desired amount A quantity A (Figure 29d) (Figure 29d) (Figure 29d) (Figure 29d) Construction without branching for a desired number of layers Ramification in the Branch in the Branch in the Branch in the direction "Y by the direction - And by the address «And by the address -Y for the quantity A desired quantity A desired amount A desired quantity A desired Construction without branching until a new support style is increased or until it meets a surface of the object.

As desired, the various parameters summarized in the previous table can be modified. For example, the amounts of branching where they are taken as an "A" amount. As appropriate, this amount can be varied with different levels of branching or it can still be varied for different components during the same level of branching. Figures 30a-30m illustrate a branching support mode analogous to that of Figures 29a-29e, with the exception that the single main line illustrated in Figure 30a, will give rise to 16 branches as s - and indicates in the Figure 30m. For ease of compression and possibly implementation, the main line shown in Figure 30a can be considered to consist of 16 individual but identical components. Again, the displacement is carried out along only one, either of the X or Y directions during a given branch operation for a given component. All of the considerations indicated above in describing Figures 29a-3 can be applied to the exemplary embodiment illustrated in these figures as well as the exemplary embodiments that follow. Figures 31a-c illustrate an additional exemplary embodiment wherein a single main branch, as shown in Figure 31a is branched into four elements, as illustrated in Figure 31c. This embodiment differs from that in Figures 29a-29c in that the branching occurs simultaneously in the X and Y directions. As illustrated, the extension of the branch is the same in the X and Y directions but this branch extension could be vary between these directions.

Figures 32a-32d continue with the embodiment illustrated in Figures 31a-31c to produce 16 separate branched supports. These Figures further illustrate the structure illustrated in Figure 28b, where two layers are illustrated for each branch. In other preferred embodiments, other branch configurations are possible. For example, instead of producing rectangular arrays of branched supports from individual main branches, as shown in the examples described above, hexagonal arrays, triangular arrays, semicircular arrays or the like can be formed. If the obtained configurations are not adjusted together, it may be desirable to use a mixture of configurations which are alternatives in an appropriate manner to give a good fit or coupling of the final support structures, such that a surface facing downwards may be supported properly. Other preferred embodiments may use multiple logs to support individual groups of branching supports.my.

It is anticipated that these branching support modalities could produce a better face-down surface than that obtained with some of the other preferred embodiments, since it is believed that the final support structures that come into contact with the object will be spaced further apart. uniformly As indicated above, the branched support embodiments described herein could be part of a larger support structure or a hybrid support structure. Other modifications to the above embodiments will be apparent to those skilled in the art after studying the teachings herein. If the geometry and direction-sensitive interlacing techniques described above are used, it may be possible to integrate structures of smaller diameter and / or spaced more closely to provide a better work surface while still providing vertical accumulation speeds. zoned. In the preferred embodiment, the droplet diameter deposited is approximately the same as the preferred pixel diameter (approximately 2.9-3.4 thousandths of an inch). In general, however, the separation of pixels between the supports (ie, spacing between the support columns) is less critical than the separation relative to the diameter of the falling drop (e.g., 2 mils) and diameter of the drop impacted (or deposited). Preferably, the horizontal spacing between the supports (eg, support columns) is less than 6 drop diameters on the layer immediately preceding the layer containing the surface facing down to be supported. More preferably, the spacing is less than 3 diameters of falling droplet and more preferably the spacing is less than 1 to 2 diameters of the falling droplet. It has been found useful to include periodic joining elements between the support columns to limit their ability to move from their desired XY positions as they grow in height. Normally, the smaller the diameter of the support columns, the more frequently layers or joining elements are needed. These joining elements can be extended to one or more layers in height. In the preferred embodiment, it has been found that a single layer (1-2 thousandths of an inch) of bonding elements is not completely effective and that more than five layers (5-10 mils) makes the overall support structure too rigid. Thus, when the preferred 3 x 3 pixel supports are used, the tie layers are preferably between 2 layers (2-4 mils) and 5 layers (5-10 mils) in height and more preferably 3 layers (3-6 thousandths of an inch) in height. In addition, it has been found that the tie layers are repeated preferably from every 75 mils to 2 inches. More preferably every 100 to 300 thousandths of an inch and more preferably every 100 to 200 thousandths of an inch. For use with other materials, construction parameters or construction conditions, can be used in the formation and analysis of test parts to determine the effective joint thickness and the thickness of the separations. When bonding layers are used periodically they may join all the support columns together or may only join a portion of them together, where the other columns are joined in a pre-bonding use or will be joined in a subsequent bonding use. In other words, the joining elements can form a solid plane of deposited material or alternatively they can form only a partially solid plane (eg, a chessboard configuration) which joins some of the columns together. The support columns may or may not be displaced from their previous XY positions when they are restarted after the formation of the tie layers. Another preferred support structure which emphasizes easy separation and good special finishing face down on the speed of production of the object is known as a chessboard support. The cross-sectional configuration of this support structure is illustrated in Figure 14. Along each frame line, the deposition occurs when using one pixel and another one not (300 pixels / inch) and in adjacent weft lines the deposition pixels are displaced along the line by a width of one pixel. A preferred version of this support does not use ID overprint but may use DD overprint or multiple exposures to increase deposition per layer. Without overprinting of DD or multiple exposures, the thickness of the layer when this type of support is used in the preferred embodiment is limited to less than 0.4 to 0.5 mils, rather than approximately 1.3 mils that can be obtained with some preferred embodiments described previously. Instead of using DD overprint or multiple exposures with these media, it is possible not to use the preferred ID overprint of the object and simply deposit the material in thinner layers (eg, 0.3 to 0.5 mils per layer). The overprint of the object does not need to be used since the extra material would simply need to be separated during the flattening step. Because the raster scan is used and since the speed of formation of a layer is the same with or without overprinting, the construction styles according to these techniques are approximately 3 to 4 times slower than the equivalent construction styles. where overprinting is used four times. Although there is a significant increase in construction time, the improvement in surface finish can guarantee its use under certain circumstances. When constructing checkerboard supports, regular use of tie layers (eg, every 30 to 100 mils in height z) is preferred to ensure the integrity of the column. The tie layers must comprise a sufficient number of layers to ensure their effectiveness (eg, approximately the same thickness of the tie layers discussed above). A drip / non-drip chessboard configuration (in terms of drop width) is where the solidified elements are 1 drop wide (deposition width), and spacing between the center points of the successive elements is greater than 1 drop wide but less than 2 drops wide. The line supports (in terms of drop width) comprise line elements which are approximately of a droplet diameter impacted in width, wherein the spacing between the elements tangential to the orientation of the lines is less than 1 drop of width (that is, superposition), while the spacing between the elements perpendicular to the orientation of the line is greater than 1 drop in width. Preferably, the spacing between the elements perpendicular to the orientation of the line is also less than 2 drops wide. The N-by-column supports (in terms of pixels) are N-activated, preferably one or two deactivated in the main scanning direction and N-activated and preferably 1-deactivated in the gradation direction. The width of the columns and the spacing of the same can be calculated based on the knowledge of the spacing of the pixels, the diameter of the drop and any overprint used. The preferred spacing between the material deposited in adjacent columns is less than one to two drop diameters. Another possible support style involves the use of solid, or periodically discontinuous lines, which are preferably smaller - 3 pixels wide (less than 10 thousandths of an inch) and more preferably 1 to 2 pixels or smaller in width ( less than 3.3 to 6.6 thousandths of an inch) and are separated by 1 to 2 pixels or less of non-deposited material (less than 3.3 to 6.6 thousandths of an inch). These supports can run along the main scanning directions, secondary scan directions or other directions. Another type of support consists of curved line supports, which follow the border of an object. Alternatively, the support configuration may differ in different areas of the cross section. It can also be displaced from the object boundary by N pixels (or drop widths) in the scan direction or M pixels (or drop widths) in the graduation direction. Some other alternatives involve building supports from a different material than that used to form the surface or border regions of the object. Other alternatives could use a different support material only on one or more of the layers adjacent to the object. Hybrid Supports: Additional types of support structures useful for modeling by selective deposition are hybrid supports. In its simplest sense, a hybrid support is a support structure that includes at least two different types of support structure. Preferably, the structures used in a hybrid support vary depending on the height of the support and more particularly the structure at any given point may depend on the distance from that point to a surface facing upwards and / or facing downwards. of the object. In other words, the support structures are adjusted to the most appropriate structure based on the distance to the object. In an exemplary embodiment, the support configuration is changed when the point is located at a predetermined number of layers (e.g., 4 to 9) below a face-down surface. In another, the drop density per unit area or ratio of the droplet density (defined as the ratio of drops to non-drops per unit area) of the supports is decreased as you approach a face-down surface . In a variant of these embodiments, one or more layers of shelf (or intermediate) layers are used when transitioning from higher to lower drop density ratio support structures. In yet another exemplary embodiment, the ratio of the droplet density increases as it exits as an upward facing surface (e.g., 4 or more separate layers of an "upside-down" surface.) In an optional variant of this modality, one or more layers of shelves (intermediate) are used when making a transition from support structures of lowest to highest drop density ratio.It is also conceivable that support structures could vary not only on the basis of at the vertical distance of the object but also on the basis of the horizontal distance as well.For example, when the object is bounded horizontally, a different type of support may be useful than when it is some distance away from the object. illustrated in the side view of Figure 20. As shown, the structure extends from the surface 23, which may be the construction platform or may be a to face-up surface of the object being constructed, to support the face-down surface 24. As illustrated, the support structure consists of five components: (1) thin, fiber-like columns 25, which contact the surface 23 (if the surface 23 is not a face-up surface of the object, this component of the support structure can be eliminated); (2) more massive columns 26 located above the columns 25 similar to fiber; (3) intermediate layers 27 (this is a final binding layer); (4) thin, fiber-like columns 28 located above the intermediate layers and which come into direct contact with the surface 24 facing down; and (5) tie layers 29 which are used to fuse two or more of the massive columns together and which are distributed in several places between the columns 26. The thin columns 25 and 28 are 1 pixel cross section (3.3 x 3.3 thousandths of an inch) and form a "chessboard" configuration as shown in Figure 14a. The result is a series of thin fiber-like columns, which are spaced apart by 1 pixel from the adjacent columns and which are easily separated from the surfaces 23 and 24. These are equivalent to the chessboard supports discussed above, in based on the configuration of deposition of 1 active pixel, an inactive pixel of this support, the drop density ratio is about 1. If the support does not start on an upward facing surface of the columns 25 of the object they may be omitted . Columns 25 and 28 should be between 3 thousandths of an inch and 15 thousandths of an inch and preferably of about 4-6 thousandths of an inch height. The height must be kept to a minimum since it is desired that these supports are used in combination with an object which is formed with 4 times the ID overprint and since when a single step is used on support structures without overprinting they accumulate at a much slower speed than the object. On the other hand, it is desired that these supports have some height since the needle-like elements tend to fuse when the face-down surface of the object is filled thereon. Columns 26 are 3 x 3 pixels cross section (9.9 mils x 9.9 mils), and 2 pixels are spaced from adjacent columns in the scan direction and 1 pixel from adjacent columns in the graduation direction. These column supports are equivalent to the most preferred supports discussed above. As discussed above, the main reason for the extra space in the main reason is the fact that these media will receive 4 times overprint. The cross-sectional configuration formed by these columns is shown in Figure 15 and 18. The result is a series of columns more massive than the fibers-like columns 25 and 28. These columns, unlike the others, can be arbitrarily high . The reason is that the larger cross sections of these columns allow the columns to grow at approximately the same speed as the part itself (approximately 1.3 thousandth of an inch / layer). As discussed previously, it is preferred that the links 29 be used to fuse the adjacent columns of the columns 26 periodically to prevent the "migration" of these columns, which may occur after the construction by some distance. The spacing of the joints is preferably in the range discussed previously. The intermediate layers 27 represent an optional final tie layer which can function as a transition between the columns 26 and the columns 28. The reason why a transition layer is useful is that the columns 28 are approximately the same size or smaller than the spaces between the columns 26, with the result that without the transition layers, the columns 28 may fall into these spaces. In a preferred process, the intermediate layers as a whole would not be used and instead the careful placement of the columns 28 on the columns 26 would present or only the necessary portions of the intermediate layers 27 would be used., if used, these intermediate layers are of similar thickness to that of the bonding layers discussed previously. It should also be appreciated that the intermediate layers need not be between the columns 25 and the columns 26, because the columns 26 - are larger in cross section than the spacing between the columns 25. Thus, these larger columns can be integrated directly above the smaller columns without the need for intermediate layers.

Other hybrid supports are possible that make other combinations with the support elements previously described. Hybrid support structures and other support structures can also be used to form the internal portions of objects. There are additional alternatives for construction supports. For example, it is also possible to build the support from a material which is different from that used to build the part. Another possibility is to add a fluid such as water between the interstices of the support structures described above in order to provide additional support and also to assist in heat dissipation. One such method is to advantageously use a fluid having a higher density than the construction material. This will give buoyancy to drops of construction material that fall between the interstices of the columns. The material must also be chosen in such a way that its surface energy corresponds to that of the construction material in order to prevent a meniscus from forming between the fluid and the columns. An example of such material is a surfactant. Another possibility is to shoot the air jets upwards between the interstices of the columns. In this procedure, a heat dissipation and buoyancy effect are possible. Another possibility is to fill the interstices of a small number of column supports (for example, columns placed separately 0.1 to 1 inch or more) with the particles. In addition, such particles could be formed from the construction material by allowing or causing the droplets to solidify before they reach the work surface (such as by increasing the distance between the dispensing head and the work surface) or by the coating of the droplets before they settle with a sublimated material, that is, they pass directly from a solid to a gas. The supports preferably separate the object from 50 to 300 thousandths of an inch from the surface of the construction platform. Alternatively, the object can be built directly on the platform. In this alternative, the platform can be covered with a flexible sheet material that will allow easy separation of the object from the rigid platform and then from the sheet material. An electric blade can be used to separate the platform supports, in which case it is preferred that the object be placed 150 to 300 thousandths of an inch above the surface of the platform. A device similar to a thin comb with long teeth has been found to be effective in separating the supports from the platform. In this case, the thickness of the device determines the required separation between the object and the platform, usually between 50 and 200 thousandths of an inch. The supports can be separated from the object by light rubbing, brushing or by using a small probe device such as a dental tool. Another variant involves the incorporation of the modalities present in an integrated system, which includes a capacity for the automatic removal of the part and a cooling station. Other alternatives involve using a low melting point metal as a construction material, a filling of different material or materials in different weft lines or g-ota sites. Additional alternatives involve using larger drops for the construction of the support than for the construction of the part. Another alternative involves the use of powder supports, which can be formed by allowing or causing the drops to solidify before they reach the work surface, as described above. Other modes could integrate objects based on different orientations in the main scan direction (for example, Y or Z), other orientations in the secondary orientation direction (for example X or Z) or other stacking orientations (for example X or Y) . Other modes could use absolute motion schemes to obtain the desired relative movements between the object and the print head. For example, in some modalities the absolute movement of the print head could be presented in all three directions, while in other modalities, the movement of the absolute object could be presented in all three directions. In still other embodiments, the non-Cartesian movement of the printhead or the object could be used and the blasting directions may vary from layer to layer or portion of a layer to the portion of a layer. Although some modalities have been described under headings inserted in the application, these modalities should not be considered as belonging only to the topic indicated by the heading. In addition, although headers are used to improve the reading of this specification, all the description relevant to the particular topic cited by the header should not be considered as within those individual sections. All embodiments described herein are described separately or in combination with other embodiments described herein. While embodiments and applications of this invention have been shown and described, it will be apparent to those skilled in the art that many more modifications are possible without departing from the concepts of the invention herein. Accordingly, the invention is not restricted, except in the spirit of the appended claims.

Appendix A: Tables I-III Detailed preferred materials for use in some preferred embodiments Table I describes the following component formulations.

Appendix A: Tables I-III Detailed preferred materials for use in some preferred embodiments Table II-A and II-B present formulations of preferred materials for use in some preferred embodiments by component designation. TABLE II-A% by weight Appendix A: Tables I-III Detailed preferred materials for use in some preferred embodiments TABLE II-B Appendix A: Tables I-III Detailed preferred materials for use in some preferred embodiments TABLE III APPENDIX METHOD AND APPARATUS FOR THE HANDLING OF DATA AND CONTROL SYSTEM IN A MODELING SYSTEM THROUGH SELECTIVE DEPOSITION This application is a continuation in part of the North American patent application No. 08/534, 447, filed on September 27, 1996, now pending. This invention is related to the techniques to form three-dimensional objects (3D) and hold those objects during the formation; more particularly, it relates to the techniques for use in the Prototype and Rapid Manufacturing Systems (RP &M); and more particularly with the methods and apparatus of construction and support for use in the system of Thermal Tereolithography (TSL), Modeling system by Molten Deposition (FDM), or another system of Modeling by Selective Deposition (SDM). Several procedures for the production of three-dimensional automatic or semi-automatic objects or Prototype and Rapid Manufacturing have been available in recent years, characterized because each one proceeds by integrating 3D objects from 3D computer data descriptive of the objects and in a way additive- from a plurality of sheets formed and adhered. These sheets are sometimes called cross sections of the object, layers of structure, layers of the object, or simply layers (if the context clarifies that reference is made to the solidified structure in the appropriate way). Each sheet represents a cross section of the three-dimensional object. Commonly, the sheet is formed and adheres to a stack of sheets formed and adhered previously. In some RP & M technologies, techniques have been proposed which deviate from a strict layer-by-layer accumulation process where only a portion of an initial sheet is formed and prior to the formation of the portion (s) ) remaining (s) of the initial sheet, at least one subsequent sheet is formed at least partially. According to such a procedure, a three-dimensional object is accumulated or integrated by the application of successive layers of material that can flow, not solidified to a work surface and then the selective exposure of the layers to the synergistic stimulation in desired shapes or geometric figures , to cause the layers to be selectively strengthened in the sheets of the object which adhere to the previously formed sheets of the object. In this procedure, the material is applied to the work surface to areas which are not part of a sheet of the object and to the areas which become part of the sheet of an object. Common to this procedure is Es tereoli tograf i a (SL), as described in U.S. Patent No. 4,575,330, to Hull. According to a modality of Stereolithography (SL), synergistic stimulation is radiation from a UV laser and the material is a photopolymer. Another example of this procedure is Selective Laser Sintering (SLS), as described in US Patent No. 4,863,538 issued to Deckard, in which the synergistic stimulation is by IR radiation of a CO2 laser and the material is a sinterable powder . This first procedure can be called stereolithography based on photons. A third example is a three-dimensional impression (3DP) and Emptying of Direct Current Production (DSPC), as described in U.S. Patent Nos. 5,340,656 and 5,024,055, to Sachs et al., In which the synergistic stimulation is a chemical binder (e.g., an adhesive) and the material is a powder consisting of particles which join together after the selective application of the chemical binder. According to said second method, an object is formed by selectively cutting the cross sections of the object having shapes and sizes of sheets of material to form sheets of the object. Normally, in practice, sheets of paper are stacked and adhered to previously cut sheets before being cut, but cutting before stacking and adhesion is also possible. Typical of this procedure is the Manufacture of Laminated Objects (LOM), as described in US Pat. No. 4,752,352, issued to Feygin in which the material consists of paper, and the means for cutting the sheets to the desired shapes and sizes. is a CO2-based laser • US Patent No. 5,015,312 issued to Kinzie also addresses the integration of the object with LOM techniques. According to such a third method, the sheets of the object are formed by the selective deposition of a material that can flow, not solidified, on a work surface into desired geometric figures in areas which become part of the sheets of an object. After or during selective deposition, the selectively deposited material solidifies to form a subsequent sheet of the object which adheres to the previously formed and stacked sheets of the object. Then, these steps are repeated to successively integrate the sheet object into sheet. This technique of object formation can be generically called Modeling by Selective Deposition (SDM). The main difference between this procedure and the first procedure is that the material is deposited only in those areas which will become parts of the sheets of an object. Typical of this procedure is Modeling by Fused Deposition (FDM), as described in US Patents Nos. 5,121,1329 and 5,340,433, to Crump, in which the material is supplied or distributed in a state that can flow to an environment which is at a temperature lower than the fluid temperature of the material, and which then hardens after it is allowed to cool. A second example is the technology described in U.S. Patent No. 5,260,009, to Penn. A third example is the Manufacture of Ballistic Particles (BPM), as described in U.S. Patent Nos. 4,655,492; 5,134,569 and 4,216,616, issued to Masters, in which the particles are directed to specific sites to form cross sections of the object. A fourth example is the thermal tereolithography (TSL) as described in U.S. Patent No. 5,141,680, to Almquist et al. When using the SDM technique (also like other RP &M construction techniques), the convenience of various methods and apparatus for the production of useful objects depends on a variety of factors. Since these factors can not normally be optimized simultaneously, a selection of an appropriate construction technique and associated method and apparatus involve transactions or exchanges depending on the specific needs and circumstances. Some factors to be considered may include 1) cost of equipment, 2) cost of operation, 3) speed of production, 4) accuracy of the object, 5) finished surface of the object, 6) material properties of the objects formed, 7) use anticipated of the objects, 8) availability of secondary processes to obtain different properties of the material, 9) ease of use and operator restrictions, ) required or desired operating environment, 11) security and 12) postprocessing time and effort. In this regard there has been a need for a long time to simultaneously optimize as many of those parameters as possible to more effectively construct three-dimensional objects. As a first example, there has been a need to improve the production speed of the object when constructing objects that use a Selective Deposition Modeling (SDM) technique, as described above (for example, Thermal Stereolithography) as long as they are maintained or reduced in a manner simultaneous the cost of the equipment. A critical problem in this regard has been the need for an efficient technique to generate and manipulate construction or integration data. Another critical problem involves the need for an efficient technique to generate appropriate support data to support or hold an object during training. Additional problems involve the existence of control programming elements that have the ability to manipulate the massive amounts of data involved in real time, to compensate for the faulty firing of the jet or the malfunctioning of the jet, of adjustment data in such a way that they are accessible in the order required and to efficiently provide geometry-sensitive integration deposition styles and techniques. Construction or integration styles and support structures appropriate for use in SDM for which a data generation technique is needed are described in the Application U.S. Patent No. 08 / 534,813. Accordingly, there is an unfulfilled need long felt by methods and apparatus for deriving data and controlling an SDM system to overcome the disadvantages of the prior art. All patents referred to in this section of the specification are hereby incorporated by reference in their entirety. 3. Request for Related Patents The following applications are hereby incorporated by reference in their entirety: The assignee of the present application, 3D Systems, Inc., submits this application concurrently with the following related application, which is hereby incorporated by reference in its entirety: According to the techniques of Thermal Stereolithography and some techniques of Modeling by Deposition in the Molten State, a three-dimensional object is integrated from layer to layer from a material which is heated until it has the capacity to flow and which is then supplied with a spout. The material may be supplied as a flow of semicontinuous material from the dispenser or may alternatively be supplied as individual drops. In the case where the material is supplied as a semi-continuous flow, it is conceivable that less stringent work surface criteria are acceptable. A previous modality of the technique of Thermal tereoli t ografía, is described in the. U.S. Patent No. 5,141,680. Thermal Stereography is particularly appropriate for use in an office environment due to its ability to use non-reactive, non-toxic materials. In addition, the process of forming objects using these materials does not need to involve the use of radiation (for example, UV (ultraviolet) radiation, IR (infrared) radiation, visible light and / or other forms of laser radiation), heating of the materials at combustible temperatures (for example, burning of the material along the boundaries of the cross-section as in some LOM techniques), reactive chemical components (eg, monomers, photopolymers) or toxic chemical components (e.g. , solvents), complicated cutting machinery and the like, which can be noisy or present significant risks if handled improperly. Instead of this, the formation of the object is obtained by heating the material to a temperature at which it can flow and then selectively assorting the material and allowing it to cool. US Patent Application No. 08 / 534,813 is primarily directed to construction and support styles and structures which can be used in a preferred selective deposition (SDM) modeling system based on the principles of the TSL. Alternative styles and structures of construction or integration and support are also described for use in other SDM systems as well as for use in other RP & M systems. US Patent Application No. 08 / 535,772 is directed to the preferred material used by the preferred SDM / TSL system, which is described hereinafter.

Some alternative methods and materials are also described. US Patent Application No. 08 / 534,447 is an original application for the present application and is directed to data transformation techniques for use in converting data from the 3D three-dimensional object into supporting data and the object for use in a modeling system by preferred selective deposition (SDM) based on TSL principles (thermal tereolithography). The referenced application is also directed to various techniques of data manipulation, data control and system control to control the preferred SDM / TSL system described hereinafter. Data manipulation techniques and alternative control techniques are also described for use in the SDM systems as well as for use in other RP & M systems. The assignee of the present application, 3D Systems, Inc., also owns a variety of other North American Patent Applications and US Patents in the field of RP & M and in particular in the Stereolithography portion of that field. The following U.S. Patent and U.S. patent applications are incorporated herein by reference in their entirety.

The present invention implements a variety of techniques (methods and apparatuses) that can be used alone or in combination to treat a variety of problems associated with the construction and support of objects, formed by using Modeling techniques by Selective Deposition. Although directed primarily to SDM techniques, the techniques described hereinafter can be applied in a variety of ways (as will be apparent to those of skill in the art reading the present disclosure) to other RP & M technologies such as is described hereinabove to improve the accuracy of the object, the surface finish, the construction time and / or post-processing effort and the post-processing time. In addition, the techniques described herein can be applied to the selective deposition modeling systems that use one or more construction and / or support materials, wherein one or more are selectively supplied and in which others can be supplied non-selectively and in where elevated temperatures may or may not be used for all or part of the materials to assist in their deposition. The techniques can be applied to the SDM systems, where the construction material (for example, paint or ink) is flowed for assortment purposes by the addition of a solvent (eg, alcohol, acetone, paint thinner or other solvents suitable for the specific construction, where "the material is solidifiable after or during the assortment by causing the separation of the solvent (for example, by heating the assorted material, by supplying the material to a partially evacuated construction chamber (ie subjected to vacuum), or simply by allowing sufficient time for the solvent to evaporate.) Alternatively or additionally, the construction material (eg, paint) can be thixotropic by nature, in which an increase in cutting force On the material it could be used to help in its assortment or distribution or the thixotropic property can be simply used to help The material to retain its shape after it is supplied or dispensed. Alternatively, and / or additionally, the material may be reactive in nature (e.g., a photopolymer, thermal polymer, a one or two part epoxy material, a combination material, such as one of the aforementioned materials in combination with a wax or plastic or thermal material, or at least solidifiable when combined with another material (for example, paris gypsum and water) where, after the assortment, the material is reacted by the appropriate application of the prescribed stimulation (for example , heat, EM, electromagnetic radiation [visible, IR, UV, X-rays, etc.], a reactive chemical component, the second part of a two-part epoxy, the second or multiple part of a combination) such that the Construction material and / or combination of materials will solidify For example, Stereoli Thermal and Assortment techniques can be used alone or in combination with the above alternatives. In addition, various assortment techniques may be used, such as the assortment by single or multiple ink jet devices including, but not limited to, thermal fusion ink jets, bubble jets, etc., and nozzles or heads single-hole or multi-hole extrusion, continuous or semi-continuous flow.

A first object of the invention is to provide a method and apparatus for converting data from a three-dimensional object into cross-sectional data. A second object of the invention is to provide a method and apparatus for the production of objects that includes a method and apparatus for converting data of the three-dimensional object into cross-sectional data. A third object of the invention is to provide a method and apparatus for obtaining data of the support from data of the three-dimensional object. A fourth object of the invention is to provide a method and apparatus for the production of objects, which includes a method and apparatus for obtaining data from the support and using the data of the support during the formation of the object. It is proposed that the above objects can be obtained separately by different aspects of the invention and that additional objects of the invention will involve several combinations of the previous independent objects, in such a way that synergistic benefits can be obtained from the combined techniques .

Other objects of the invention will become apparent from the description herein. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a diagram of a system of It is tereolithography to Thermal preferred; Figures 2a and 2b illustrate the orifice plate of the print head of Figure 1 at two different angles; Figure 3 is a more detailed drawing of the flattening device of Figure 1; Figure 4 illustrates the relative spacing between the adjacent nozzles on the orifice plate and the adjacent raster lines; Figure 5 illustrates the pixel grid which defines the resolution of system data; Figures 6a-6d illustrate various overprint schemes; Figure 7 illustrates a first embodiment of the present invention; Figures 8a-8b illustrate the intersection of a .STL file with cutting planes; Figure 9 illustrates the effect of the Boolean extraction operation; Figure 10 illustrates a second embodiment of the present invention; Figure 1a illustrates the arrangement of the triangles in the z direction; Figures llb-llc illustrate the selection of the active triangles; Figures 12, 12b and 12c illustrate alternative ways to represent the data of the cross section; Figures 13a, 13b and 13c illustrate the placement of the transition data in lists associated with the different scan lines; Figures 14-15 illustrate in more detail the Boolean extraction operation; Figures 16-17 illustrate Boolean addition, subtraction and intersection operations; Figures 18-21 illustrate a two-step process for generating supports using intermediate layers; Figure 22 illustrates a three-step process for generating supports; Figures 23-26 illustrate a method for storing start / stop data in contiguous words; Figures 27a-27b, 28a-28b, 29a-29b, and 30a-30b illustrate a method for allocating memory to list data representing start / stop transitions; Figure 31 illustrates the property in which successive scan lines are represented by numbers of similar transitions; Figure 32 illustrates the quantization error introduced by pixelizing the start / stop data; Figure 33 illustrates the conversion of .RLE data to conversion data; Figures 34a-34c illustrate the data for use in the construction or integration of supports; Figure 35a illustrates an assumption about the part of the slope performed in the preferred embodiment of the present invention; Figures 36a-36c illustrate a ring buffer used in the present implementation of the present invention; Figure 37 illustrates a hybrid support structure; Figures 38a-38b illustrate two types of sample style; Figures 39a, 39b and 39c illustrate encounter of part / support that sometimes occurs; Figures 40a-40c represent an example of how style files are used; Figures 41a-41f represent additional style types; Figures 42-42e illustrate the falsification of data; Figure 43 illustrates the prescribed sequence of lines of the encoder; Figure 44 illustrates a resolution problem which can be presented through only one trigger counter; and Figures 45a-45b illustrate an algorithm for increasing the resolution in the scanning direction through the use of two counters. Figure 46a illustrates a side view of an object containing a space together with levels and hypothetical regions on which the formation of different support structures can be based.

Figure 46b illustrates a side view of the object of Figure 46a, wherein the space is filled with various types of support structures. Figure 47 illustrates the conceptual format of a .RLE file. As discussed above, the present application is directed to techniques and systems for data manipulation and control techniques for implementing appropriate construction techniques and techniques for use in a selective deposition modeling (SDM) system. In particular, the preferred SDM system is a Thermal Stereolithography (TSL) system. The detailed description of a preferred embodiment of the invention will begin with a description of the preferred DSL system, wherein details of the modality are described as appropriate. A more detailed description of the preferred construction and support techniques, preferably preferred formulations and material properties, that of the preferred system and various alternatives are described in previously incorporated US Patent Applications 08 / 534,447; 08 / 535,772; and 08/534477. Additional alternative systems are discussed in a variety of applications and patents previously incorporated, especially those referred to as directly related to or applicable to Modeling by Selective Deposition, Thermal Stereolithography or Modeling by Deposition in the Molten State. As such, the data manipulation techniques and system control techniques hereinafter are to be construed as being applicable to a variety of SDM, TSL and FDM systems and not inappropriately limited by the examples described in I presented. A preferred apparatus for carrying out the SDM / TL is illustrated in Figure 1. The apparatus comprises a dispensing platform or dispenser on which the dispensing head 9 (for example ink jet head) is located. multiple orifices) and flattening device 11. The dispensing platform is slidingly coupled to the stage X-12 by means of the element 13. The step X-12 makes the platform 18 move in a controlled manner distributed from back to front in the X direction , also known as the main scanning direction. The movement of stage X is under the control of a computer or control microprocessor (not shown). In addition, either on one side or the other of the platform 18 and / or between the flattening device 11 and the distributing head 9, fans (not shown) for blowing the air vertically downwards are mounted to help cool the assorted material and the substrate, in such a way that the desired construction temperature is maintained. Of course other mounting schemes for the fans and / or other cooling systems are possible, which include the use of fogging devices to direct vaporizable liquids (eg water, solvent alcohol) on the surface of the object. Cooling systems could involve active or passive heat removal techniques and can be controlled by computer in combination with devices that detect temperature to keep the material within the desired construction temperature range. printing) 9 is a commercial print head configured to emit jets of thermal fusion inks (e.g. thermoplastic or wax-like materials) and modified for use in a three-dimensional modeling system wherein the print head undergoes movements and accelerations of The forward changes of the print head involve configuring any internal container, such that accelerations result in minimal defective placement of the material in the container.In a preferred embodiment, the head is a commercial printhead of the container. 96 jets, model number H DS 96i, sold by Spectra Corporation, of Nashua, Nem Hampshire that includes modifications of the container. The print head 9 is provided with material in a flowable state from a Packing and Material Handling Subsystem (not shown). The packaging and material handling subsystem is described in the North American Patent Application referred to above No. 08 / 534,477. In a preferred implementation, all 96 overhead jets are computer controlled to selectively fire the droplets through the orifice plate 10 when each orifice (ie, jet) is properly located or positioned to distribute droplets on desired sites. In practice, approximately 12,000 to 16,000 orders per second are sent to each jet selectively, ordering each one to shoot (supply a drop) or not to shoot (do not supply a drop) depending on the position of the jet and the desired sites for the deposition of the material. Also, in practice, the firing orders are preferably sent simultaneously to all the jets. Thus, in a preferred embodiment, the head is computer controlled to selectively disperse the jets to simultaneously emit droplets of molten material through one or more holes in the orifice plate 10. Of course, it will be appreciated that in alternative modes, heads with different numbers of jets can be used, different firing frequencies are possible and under appropriate circumstances, non-simultaneous firing of the jets is possible. The orifice plate 10 is mounted on the assortment or distributor platform 18, in such a way that drops of material are allowed to be emitted from the underside of the distributor platform. The orifice plate 10 is illustrated in Figures 2a and 2b. In a preferred embodiment and as shown in Figure 2a, the orifice plate (ie, the line of holes) is mounted approximately perpendicular to the main scanning direction (eg X) and is configured with N = 96 controllable orifices. individually (called 10 (1), 10 (2), 10 (3) ... 10 (96)). Each jet (eg jets) is equipped with a piezoelectric element, which causes a pressure wave to propagate through the material when an electrical trigger pulse is applied to the element. The pressure wave causes a drop of the material to be emitted from the hole. The 96 jets are controlled by the control computer which controls the speed and synchronization of the firing impulses applied to the individual jet and consequently the speed and synchronization of the droplets that are emitted from the orifices. With reference to Figure 2a, the distance "d" between the holes in a preferred embodiment is approximately 8/300 inches (approximately 26.67 mils or 0.677 millimeters). Thus, with 96 holes the effective length "D" of the orifice plate is approximately (N x 8/300 inches) = (96 x 8/300 inches) = 2.56 inches (65.02 mm). A preferred embodiment uses the screen scan to position the print head and the holes for dispensing the material at desired drip sites. The printing process for each layer is carried out by a series of relative movements between the head and the desired drip sites. Printing commonly occurs as the head moves relatively in a main scanning direction. This is followed by a typically smaller increase in movement in a secondary scan direction, where the assortment or distribution is not presented, which in turn is followed by a reverse scan in the main scan direction in which another Once the distribution or assortment is presented. The process of alternating the main scans and the secondary scans occurs repeatedly until the lamina is completely deposited. Alternative modalities can carry out small secondary exploration movements while the main exploration is presented. Due to the normally large difference in net scanning velocity along the main and secondary directions, such alternatives still result in deposition along main almost perpendicular scanning lines (ie, the main scanning directions and of secondary exploration remain substantially perpendicular). Other alternative embodiments may use vector scanning techniques or a combination of vector scanning and raster scanning. Other alternative embodiments may use substantially non-perpendicular primary and secondary scanning directions together with algorithms that result in proper placement of the drops. In alternative embodiments, the print head may be mounted at an angle not perpendicular to the main scanning direction. This situation is shown in Figure 4b where the print head is mounted at an angle "a" to the main scanning direction. In this alternative situation, the spacing between the holes is reduced from d to d '= (d x sin of a) and the effective length of the print head is reduced to D' = (D x sin a). When the spacing d 'is equal to the desired print resolution in the secondary scan direction (direction approximately perpendicular to the main scanning direction), the angle a is considered to be the "knowing angle". If the spacing d (such as when using a preferred mode) or od '(as when using some alternative preferred mode) is not found at the desired secondary print resolution (that is, the print head is not in the know angle) ) So for optimal efficiency in the printing of a layer, the desired resolution should be selected in such a way to make dod a whole multiple of the desired resolution. Similarly, when printing with a? 90 °, there is a spacing between the adjacent jets in the main scanning direction also as the secondary scanning direction. This spacing is defined by d "d x eos a. This spacing in the main scanning direction d '', in turn, determines that optimization of printing efficiency will occur when the desired primary print resolution is selected such that it is an integer multiple of d "(assumed that the locations or shooting sites are located in a rectangular grid). In other words it can be said that the angle a is selected in such a way that d 'and / or d "(preferably both) when divided by the appropriate integers M and P produce the desired primary and secondary scan resolutions. An advantage of using the preferred printhead orientation (a = 90 °) is that it allows any desired print resolution in the main scanning direction, while still maintaining optimum efficiency. In other alternative embodiments, multiple heads can be used which are arranged end to end (extend in the secondary scan direction) and / or which are stacked from back to back (stacked in the main scan direction). When stacked from back to back, the print heads may have holes aligned in the main scanning section, so that they print on the same lines or alternatively they may be offset to distribute the material along lines of different main exploration. In particular it may be desirable to have the print heads from back to back displaced from each other in the secondary scan direction by the desired frame line spacing to minimize the number of main scanning steps that must be presented. In other alternative embodiments, the data defining the deposition sites may not be located by pixels defining a rectangular grid, but rather that they may be located by pixels disposed in some other configuration (for example displaced or stepped configuration) . More particularly, the deposition sites can be varied fully or partially from layer to layer in order to carry out the drift site displacement of the partial pixel for an entire layer or for a portion of a layer based on the particularities of a region that is going to be subjected to the jet. Currently preferred printing techniques involve deposition of 300, 600 1200 drops per inch in the main scanning direction and 300 drops per inch in the secondary scanning direction. With reference to Figures 1 and 3, a flattening device 11 consists of a heated rotary cylinder 18a, with a textured surface (eg knurled). Its function is to melt, transfer and separate the portions of the previously supplied layer of material in order to smooth it, to fix a desired thickness for the last formed layer and to fix the net upper surface of the last formed layer to a desired level (that is, the desired work surface or work level to form a next sheet of the object). The number 19 identifies a layer of material which has recently been deposited by the printhead. The rotary cylinder 18a is mounted to the dispenser platform, in such a way that it is allowed to project from the underside of the platform by a sufficient amount in the Z direction, such that it comes into contact with the material. to a desired level below the orifice plate. In a preferred embodiment, this amount is set in the range of 0.5 mm to 1.0 mm. The rotation of the cylinder 18a goes from the material of the recently deposited layer, identified in the Figure with the number 21, to exit on its smooth surface. The material 21 adheres to the knurled surface of the cylinder and is displaced until it comes into contact with the cleaning element 22. As shown, the cleaning element 22 is arranged to effectively "scrape" the material 21 from the surface of the cylinder. This material, because it still has the ability to flow, is captured by the Packing and Material Handling Subsystem, described in US Patent Application Number 08 / 534,477, (not shown), from where it is either deposited or stored. recirculates With reference to Figure 1, the integration or partial construction platform 15 is also provided. On this platform 15 the object or the three-dimensional part, identified in the Figure with the reference number 14, is integrated or constructed. This platform 15 is slidably coupled to the stage Y 16a and 16b which makes the platform controllably moveable. 15 from back to front in the Y direction (this is graduation direction or secondary scan direction) under computer control. The platform is also coupled to the stage Z-17 which makes the platform controllably move from top to bottom (usually progressively downward during the construction or integration process) in the Z-direction under the control of computer. To integrate or construct a cross section of a part, the stage Z is directed to move the platform 15 of integration or construction of the part in relation to the print head 9, in such a way that the last integrated or constructed cross section ( this is stocked or distributed and possibly flattened) of the part 14 is placed at an appropriate amount below the hole plate 10 of the print head. Then, the printhead, in combination with the lid Y, is caused to move or bar one or more times over the integration region or XY construction (the head sweeps back and forth in the X direction, while the And it moves the partially formed object in the Y direction). The combination of the last formed layer of the object and any support associated with it defines the working surface for the deposition of the next sheet and any support associated therewith. During the translation in the XY directions, the jets of the print head are fired in a manner corresponding to or coincident with the layers previously filled in order to deposit the material in a desired configuration and sequence "for the construction of the next sheet of the object. the assortment process, a portion of the assortment material is separated by the flattening device in the manner discussed above.The movements X, Y and Z, assortment and flattening are repeated to build or integrate the object from a priority of selectively stocked layers and in an alternative embodiment, the leveling step could be carried out independently of the assortment stages In other alternative modes the flattening device may not be used on all layers, but instead can be used on select or periodic covers. As previously indicated, in a preferred embodiment, the print head is directed to follow a raster configuration. An example of this is illustrated in Figure 4. As shown, the frame configuration consists of a series of frame lines, R (l), R (2), ..., R (N), which run in the X address or main scan direction and arranged along the Y direction (i.e., graduation direction or secondary scan direction). The weft lines are spaced apart by a distance d, which, in a preferred embodiment is 1/300 inches (approximately 3.3 mils or approximately 83.8 μm). Since the holes of the print head are spaced apart by the distance d, which as discussed is approximately 26.67 mils (0.6774 μm) and since the desired number of raster lines can be extended in the graduation direction For a distance greater than the length of the orifice plate, approximately 2.56 inches (65.02 mm), the print head must be swept over the work surface by means of multiple steps in order to track all the desired weft lines. This is done by following a two-stage process. In the first stage, the print head is passed 8 times on the work surface, with the step Y being graded by the amount d after each step in the main scanning direction. In the second stage, the Y stage is graduated by a distance equal to the length of the orifice plate (2.5600 inches + dp (0.0267 inches) = 2.5867 inches (65.70 mm).

Then this two-step process is repeated until all the desired frame lines have been tracked. In other words, a preferred two stage process involves a first step of alternating the steps in the main scanning direction with the movements in the secondary scanning direction of an amount equal to the desired screen line resolution until all the lines between the initial lines supplied by two adjacent jets are explored. After this a second stage is performed which includes an increase in the large graduation direction. This increase in the large graduation direction is equal to the spacing between the first and last holes of the print head plus a raster line spacing. The first and second stages are repeated until the increments in the graduation direction and the scanned lines are sufficient to deposit the material on all the weft lines required to form the cross section of the object (in which any support is included). necessary to form the subsequent cross sections).

For example, in a first step, the print head could be directed to track the raster lines R (l) (via hole 10 (1) in Figure 4), R (9) (via hole 10 (2). )), R (17) (via hole 10 (3)), etc. Then stage Y would be directed to move the construction platform to the distance of (a plot line) in the graduation direction in the next step, the print head could be directed to track R (2) (via 10 ( 1)), R (10) (via 10 (2)), R (17) (via 10 (3)), etc. Then six more steps would be carried out with the step Y graduated by the distance d after each step, until a total of 8 sweeps have been carried out. At this time, if there are more weft lines to be traced, stage Y would be directed to move the construction platform by an amount equal to the full length of the hole plate + dp, 2.5867 inches (65.70 mm). Then the two-step process described above would be repeated until all the raster lines have been tracked. In alternative modes, other Y increments could be made, including increments that involve negative and positive movements along the Y axis. This could be done in order to explore the frame lines that were initially omitted. This will be further described in the association with a technique called "interlacing". The firing of the ink jet holes is controlled by a rectangular bit map maintained in the control computer or other memory device. The bitmap consists of a grid of memory cells, in which each memory cell corresponds to a pixel of the work surface and in which the rows of the grid extend in the main scanning direction (X direction) and the columns of the grid extend in the secondary scan direction (Y direction). The width of (or distance between) the rows (spacing along the Y direction) may be different from the width (or length of or distance between) of the columns (spacing along the X direction) which determines that they can there are different data resolutions along the X and Y directions. In alternative embodiments, a non-uniform pixel size is possible within a layer or between layers where one or both of the width or length of the pixel is varied by the position of the pixel. In other alternatives, pixel alignment configurations are possible. For example, pixels that are in adjacent rows may be displaced in the main scanning direction by a fractional amount of the spacing between the pixels in the main scanning direction, such that their center points do not align with the center points of the pixels in the neighboring rows. This fractional amount can be 1/2 such that its center points are aligned with the pixel boundaries of the adjacent rows. It can be 1/3 or some other amount, such that two or more intermediate rows of pixels are located between the rows where the pixels are aligned in the main scanning direction. In additional alternatives, the pixel alignment could be dependent on the geometry of the object or support structure that is supplied. For example, it may be desirable to shift the alignment of the pixel when forming a portion of a support configuration that is supposed to join a space between the support columns. These and other alternate pixel alignment schemes can be implemented by modifying the pixel configuration or alternatively by defining an array of higher resolution pixels (in X and / or Y) and by using pixel trigger settings they do not fire at each pixel site but instead they fire at selected spaced pixel sites which can vary according to a configuration based on the object or predetermined, random. You can define the resolution of data in the main scan direction in terms of the Pixels of the Main Address (MDP). MDPs can be described in terms of the pixel length or in terms of the number of pixels per unit length. In a preferred embodiment, MDP = 300 pixels per inch (26.67 mils / pixel or 677.4 μm / pixels). In other preferred embodiments, MDP = 1200 pixels per inch. Similarly, the resolution of data in the secondary scan direction can be defined in terms of the Secondary Address Pixels (SDP) and the SDP can be described in terms of the pixel width or in terms of the number of pixels per unit length. In a preferred embodiment SDP = MDP = 300 pixels per inch (26.67 mils / pixel or 677.4 μm / pixel). The SDP may or may not be equivalent to the spacing between the screen lines and the MDP may or may not be equivalent to the spacing between the successive drip sites along each screen line. The spacing between the successive grid lines can be defined as the Secondary Drip Sites (SDL), while the spacing between the successive drip sites along each raster line can be defined as the Main Drip Sites (MDL). Similar to the SDP and MDP, the SDL and MDL can be defined in terms of droplets per unit length or droplet spacing. If SDP = SDL there is a one-to-one correspondence between the data and the drip sites along the secondary scan direction and / or pixel shift is equal to that of the spacing of the frame lines. If MDP = MDL there is a one-to-one correspondence between the data from the drip sites along the main scanning division. If SDL and / or MDL is less than SDP and MDP, respectively, more drops will be needed to be fired than those for which data exists, so each pixel will need to be used to cause more than one drop to be delivered. The assortment of these extra drops can be done in one of two ways either by supplying the drops at intermediate points between the centers of the successive pixels (this is intermediate drip, "ID") or alternatively directly above the centers of the pixels (this is direct dripping, "DD"). Either in one case or another, this technique is called "overprint" and results in faster material accumulation and facilitates mechanical design constraints that involve maximum scan speeds and maximum acceleration speeds, since the same build-up Z may be present as long as the pressure head and / or object is moved more slowly. The difference in overprint ID versus non-overprint or overprint DD is shown in Figures 6a to 6d. Figure 6a - illustrates a single drop 60 that is deposited and an associated solidified region 62, surrounding it when the print head moves in the direction 64. On the other hand, Figure 6b illustrates the same region that is cured but at using the ID overprint technique where two drops 60 and 66 are deposited in association with the single data point when the head moves in the 64th direction. The deposition zone filled by the two drops is illustrated. region 68. Figure 6c shows a similar situation for a four-drop ID overprint scheme, where the drops are indicated by the numbers 60, 70, 66 and 72 and the deposition zone is illustrated by 76 and where the scanning direction is still represented by the number 64. Figure 6d illustrates a similar situation for a line of pixels 78, 80, 82, 84, 86, and 88, where the number 90 illustrates the length of the deposition zone without over tion and the number 92 illustrates the length of the deposition zone when a four-drop ID overprint technique is used. The foregoing can be generalized by saying that the overprint of ID adds approximately 1/2 to less than one additional pixel length to any region where it is used. Of course, the more overprint droplets are used, the more vertical growth of a pixel region will occur. If SDL and / or MDL is less than SDP and / or MDP, respectively, the drops will be fired at fewer sites than those for which data exists, at least for a given step of the printhead. This data situation can be used to implement the displacement and / or pixel pixel techniques of non-uniform size discussed above. A grid of N rows per M columns are illustrated in Figure 5. As shown, the rows in the grid are designated as R (l), R (2), ..., R (N), whereas columns in the grid are named as C (l), C (2), ..., C (M). The pixels that make up the grid are also displayed. These are referred to as P (l, l), P (l, 2), ..., P (M, N). To construct or integrate a cross section, the load bit map of the first one with the data representative of the desired cross section (also as any support which you wish to integrate). Assuming, similarly to the preferred embodiment, only one construction and support material is used, if it is desired to deposit material at a given pixel site, then the memory cell corresponding to that site is suitably indicated (e.g. load with a binary "1") and if no material is going to be deposited an opposite indicator is used (for example a binary "0"). If multiple materials are used, the cells corresponding to the deposition sites are indicated appropriately to indicate not only the location sites of the drops but also the type of material to be deposited. For ease of handling or data manipulation, the compressed data defining a supporting object or region (for example active-inactive location points along each frame line) can be subjected to a Boolean calculation with a description of the configuration of fill to be used for the particular region to derive a final representation of bit map used for firing the assortment jets. Then the weft line making up the grid is assigned to the individual holes in the manner described above. Then, a particular hole is directed to shoot or not - over a pixel depending on how the corresponding cell in the bitmap is indicated. As discussed above, the print head has the ability to deposit droplets to many different generations. In the preferred embodiments of the present invention SDP = SDL = 300 pixels and drops per inch. However, MDP is allowed to take three values in the preferred mode: 1) MDL = 300 drops per inch MDP = 300 pixels per inch, 2) MDL = 600 drops per inch and MDP = 300 pixels per inch or 3) MDL = 1200 drops per inch and MDP 0 300 pixels per inch. When the ratio of MDL to MDP is greater than one, drops are caused per extra pixel at the intermediate sites (over ID printing) between the centers of the pixels. With the currently preferred printhead and material, the volume per drop is about 100 picoliters, which produces drops that are approximately 2 mils in diameter (50.8 μm). With the currently preferred print head, the maximum trigger frequency is approximately 20 Khz. By way of comparison, a firing rate of 1200 dpi at 13 ips involves a frequency of 16 Khz, which is within the allowable limit. A first preferred embodiment for producing data appropriate for the construction or integration of the part in a Modeling System by Selective Deposition (for example, a Thermal Stereolithography system), includes the generation of data representative of supports, is illustrated in Figure 7 As shown, the method starts by using a Boolean layer cutting process (represented by module 31) to convert the 30 .STL file to the 32 .SLl file. The Boolean layer cutting process, also as the .STL and .SLI formats are described in the Patents and Patent Applications referred to above (for example, US Patent Application Number 08 / 475,730 (hereinafter forward in the present '730)). Then the .SLl file is introduced to the module 33, which produces support data in the .SLl format. The data of .SLl representative of the supports, identified with the number 34, are then converted with the data - of .SLl representative of the object, identified with the number 32, in the module 35. The result is the file 36 .PFF, representative of the boundaries of the object and the support. Then the .PFF file is "grated" in the module 37, according to the determined style or the style file 38 when using the graying or shading techniques described in the '730 patent application mentioned above. The intersections between the. Shading lines and the borders of the object and the support are then used to prepare the 39 .RLE file. A problem with this modality is speed. As illustrated in Figures 8a-8b, the process involves the intersection of a 46.STL file with cut planes, such as that defined by number 47 in Figure 8a, to produce lists of segments for each cross section, such as that identified with the number 48 in the Figure 8b. Then the segments are sorted, the internal segments are removed and the appropriate endpoints are joined together to form polygons. In Figure 9, for example, the segments 48 are processed in the manner described to form the polygon 49. The process takes a long time due to the number of comparisons that must be carried out to order the segments and due to the time required to carry them. out Boolean operations on polygons. For a list of N segments such as for example the ordering stage requires N2 comparisons. In addition, the process of carrying out a Boolean operation on a polygon comprising N segments also requires N operations. For these reasons, the process of forming integration or construction data can be prohibitively long, usually several hours. However, an advantage to this procedure is that since the border elements are arranged in polyliers, the compensation of the width of the drop on these borders can be carried out analogously to the compensation routines taught in the application * 730 A second preferred embodiment designed to overcome those problems is illustrated in Figure 10. As shown, the .STL file 40 is first compressed via module 41 to the 42.CTL file. The process of compressing an .STL file into a .CTL file is described in the aforementioned US Patent Application Serial No. 08 / 428,951. Second, based on the Style information 43 provided as an entry, in module 44, the .CTL file is cut or divided in a manner similar to that described in the? 730 application, except that only one grating or surface type data are emitted in an RLE file (this is encoded run length). First, as illustrated in Figure 11, the triangles that make up the .STL file are sorted from top to bottom in the z direction. Specifically, as indicated by the identification number '50, the triangles are classified in descending order of the maximum z-value for each triangle. As shown, the order of the triangles is: A, B, C, D. The required top-down classification must be contrasted with a bottom-up classification, such as indicated by the identification number 51 in Figure a , in which the triangles are classified in ascending order of the minimum z values of the triangles. As indicated, the resulting order is: B, C, A, D. For each cut level, a list of active triangles is determined by the use of a current level indicator and an index or graduation indicator. An index or graduation indicator is advanced through the list of triangles for a given level and any triangle completely above the current level is eliminated from consideration. If a triangle is intersected by the current level indicator, it is added to the list. The process continues until the index indicator points to a triangle completely below the current level. At this point, the list of active triangles for the level is complete. Then the level indicator is changed to indicate the next lower level and the process continues. Figure 11b illustrates the process when the current level indicator is at level 52a. Indicator 53 of the index is advanced from left to right and the two triangles intersected by the current level, identified in the Figure with reference number 54a, are added to the list of active triangles. Then the process continues until the index indicator points to triangle 55a. Since that triangle is completely - below the current level, the process stops and the indicator 53 of the index points to triangle 55a. Figure 11c illustrates the process when the level indicator is advanced to level 52b. The index indicator is reset or reset to zero and then moved from left to right. Each triangle above the level is ignored and each triangle intercepted by the level is added to the list of active triangles. In the Figure, these triangles are indicated with the identification number 54b. The process is terminated when the pointer or indicator of the index points to triangle 55b, since it is the first triangle that is completely below the level. The active triangles for each cut level are intersected with that level to form a set of segments in the x-y plane. Since the triangles limit the solid and are oriented facing away from the solid region (as explained in U.S. Patent Nos. 5, 059, 359; 5, 137, 662; 5, 321, 622; and 5, 345, 391, which are incorporated herein by reference), the resulting segments also have orientation. Of these segments, without the need to order them in border circuits, the .RLE data, descriptive of the cross sections of the object, is obtained by using the same grating or shading algorithms as described in the US Patent Application '730. Figure 12a illustrates a polygonal representation of a cross section (segments arranged to form front circuits or cycles, while Figure 12b illustrates a representation of .RLE (coded length of run) of the same cross section. data, the polygonal representation is superimposed with a plurality of pixel scan or frame lines and then a list of start / stop pairs is generated at the points where the pixel or raster line intersects with the polygonal representation, with each intersection point associated with an on / off indicator For a given scan line, the on / off indicator for the points of the intersection are alternated between on or active status and off or inactive, to indicate whether the line of exploration enters or leaves a solid In Figure 12b, for example, the "on or active" portions of the successive exploration lines You are identified with the numbers 56 (1), 56 (2), 56 (3), ..., and 56 (11).

The .RLE format must be contrasted with the pixel format illustrated in Figure 12c, in which each point inside the solid is represented by a separate data point. The problem with this format of data representation is the size. At 300 DPI (dots per inch), for example a 10-inch cross section requires 9 million bits of information. The process for generating the .RLE data for the cross sections of the object is illustrated in Figures 13a-13c. As shown in Figure 13a, for each cross section, such as the cross section identified with number 57 in the Figure, a list arrangement is created, identified with the number 58, in which each list in the array corresponds to a scan line that extends to a level and given in the x direction. Then, considering each segment in the cross section, the intersections between each segment and the scan lines are indicated and the representative data of these intersections are added to the respective lists in the array. Figure 13b, for example, illustrates the additions, identified in the Figure with the number 59, to the lists by means of the -consideration of segment one. The specific data items added to the list for each site "and" contain two pieces of information: a value of the quantitative volume (QV) and the location x of the intersection. An intersection in which the segment increases in the direction and has a QV of 2. An intersection in which the segment decreases in the direction and has a QV of -2. If the segment originates or ends in a scan line, the intersection costs as a "half hit," that is, the associated QV is 1 or -1, depending on whether the segment increases or decreases in the y direction. In Figure 13d, for example, segment 1 increases in the y direction. Thus, the QV values associated with the intersection of this segment with the successive exploration line are respectively 1, 2, 2,. 2, and 2 (assuming that the scan line does not meet the tip of segment 1). In addition, the x sites of the intersections between segment 1 and the successive scan lines are respectively 126, 124, 122, 120, and 118. As shown, the data added to the array is incorporated into these values. Figure 13c illustrates additions to the array by considering segment 2. That segment is incremented in the y direction, and originates and ends in two successive scan lines. The site x of the intersection of the first exploration line is 144, while for the second it is 126. The two additions to the arrangement which incorporate these values are identified with the numbers 60 (1) and 60 (2). The purpose of half blows can be understood through the consideration of Figure 14. As shown, each scan line is associated with a total current QV which has been updated each time the scan line crosses a segment when using the QV value associated with the point of intersection with the segment. If the scan line is inside the solid, the current QV value is 2, while if it is outside the solid, the QV value is 0. Thus, when a scan line is outside the solid and crosses a border, the necessary implication is that the exploration line is now inside the solid. Then the total current QV must be updated with a value of 2 to indicate that it is now inside the solid. Printedly, if the scanning line is inside the solid and crosses a boundary, the necessary implication is that the scanning line is now outside the solid or is found with a second solid object which overlaps or overlaps with the first object. Then a value of -2 or 2 must be added when touching current to indicate the transition. If the scan line crosses a vertex, as indicated in point A in Figure 14, the scan line actually intersects with two segments when it enters the solid. Thus, each segment must contribute only a value of 1 to the total current QV. This is the reason why the value of QV associated with these vertices is already maintained at 1 or -1. It should be noted that it is possible for a scan line to cross a vertex without the current QV value changing state. As illustrated by point B of Figure 14, the segments forming the vertex have QV values of -1 and 1 respectively at the point of intersection. The result is that the total current QV associated with the scan line does not change. Additional information about the quantitative volume (QV) can be found in the American patent application '730 to which reference is made previously. After the intersections of the scan line for all segments have been added to the list, the list for each scan line is then qualified in ascending x order. Then a Boolean extraction routine is applied to extract the correct Boolean segments for each scan line. The preferred extraction routine involves maintaining a current QV count in which the QV value of each successive data point in the list is added to the running total. Any data point which has a QV of 2 this is a "start" point, when the total current QV is 0 (this is transitions from 0 to 2) any data point which has a QV of -2, this is a "stop" point, when the current total is 2 (this is transitions from 2 to 0) it is maintained. The process is illustrated in Figure 15 in which the successive stages thereof are identified with the numbers 61, 62, 63, 64, 65, 66, and 67. A pointer or indicator of the present item, identified with the number 68 is Use to point or indicate successive items. A "hold" list, identified with Figure 70, is also used to retain start and stop points which meet the prescribed conditions described above. As shown, by means of this process, only the first starting point, that is (start 20) and the last stop point, that is (stop 89) are retained. The result is the descriptive .RLE data of a line of a cross section of the object. Applying the technique to all lines for all cross sections results in a description of .RLE for the object. It should be noted that it is not necessary to classify the segments, formed by intersecting the triangles with the cutting planes in polylines (as described in the patent 622) in order to form polygonal representations of the cross sections of the object. As discussed, the classification of segments in polylysts is a time-consuming operation. In addition, it should also be appreciated that the .RLE data formed is successfully merged when the .STL file has not been joined or separated properly (this is the .STL file contains overlapping elements of the object). A benefit of the representation of .RLE with respect to the polygonal representation is that Boolean operations are much simpler and faster. The Boolean extraction algorithm has already been discussed. Several others are Boolean addition, subtraction and intersection operations. To carry out these operations more efficiently, it is advantageous to express the .RLE data in absolute terms as opposed to relative terms. For example, a line that starts at position x 100 and remains active for 30 pixels must be represented in terms of a pair of start / stop points, in which the start is at position 100, and stop at position 130. Thus, with reference to Figure 16, the .RLE data for line A identified with number 71 in the Figure and that for line B, identified with number 72 in the Figure, are represented as follows: A = [(start 20), (stop 48), (start 60), (stop 89)], B = [(start 37), (stop-a 78)]. The calculation of the Boolean addition of these two lines involves merging the two data sets, while keeping the merged list sorted in the x direction. The result is [(start 20), (start 37), (stop 48), (start 60), (stop 78), (stop 89)]. Then the merged list is submitted to the Boolean extraction algorithm discussed above where, for example, the start sites are assigned the QV values of 2 and the stop sites are assigned QV values of -2 and only those sites that have been assigned are kept. result in QV transitions from 0 to 2 (start) or from 2 to 0 (stop). The result is the data pair [(start 20), (stop 89)], which represents the Boolean addition A + B, which is identified with the number 73 in Figure 16. The calculation of the Boolean subtraction of two lines it involves the identical steps discussed above in relation to the Boolean addition operation, except that before the two lists are merged, the signs of the QV values of the list which are subtracted are inverted, so that the start transitions are convert into stop transitions and vice versa. The result of the operation A + B is identified in Figure 16 with the number 74. The calculation of the Boolean intersection of two lines involves the identical steps as the addition operation, except that the extraction routine is carried out when starting with an initial QV value of -2. The intersection between A and B is identified in Figure 16 with the number 75. Two-dimensional Boolean operations can also be easily carried out. For two-dimensional areas, each represented by a plurality of. RLE lines preferably expressed in absolute terms, Boolean operations are carried out by performing successive Boolean linear operations in each successive pair of corresponding lines in the respective areas. Figure 17 illustrates the process. The set of lines identified with the number 76 represents the area A, while the set of lines identified with the number 77 represents the area B. The Boolean addition, A + B of these two areas is identified with the number 78, in so much that the Boolean subtraction of these two areas, AB is identified with the number 79. On the other hand, a disadvantage of using the .RLE data in relation to the polygonal data, is the amount of memory required. To store each layer in .RLE form at a high resolution, it could require more than 100 MB of storage or memory for a typical part. This is too big for main memory and even having to store such large files on disk is problematic. The problem is exacerbated by the divergence between the order of construction of the part, which proceeds from the bottom up and the order of constructing the support structures, which, as described hereinafter, proceeds from top to bottom. As discussed later herein, an output file is required to construct supports in which, for each cross section, a description of .RLE is provided for that cross section, also, the Boolean sum of each cross section above the present cross section. Basically, the technique involves the calculation of the Boolean subtraction between the description of .RLE of a cross section and the representation of .RLE of the "total present" for that cross section, that is, the Boolean union of all the layers above the present layer. The pseudocode for this basic technique is shown in Figure 18, in which get_part (level) refers to a function which provides the representation of .RLE of the cross section at the prescribed level; boolean_subt ract (current_total = area A, part_for_layer = area B) refers to a function which provides the result of subtracting Boolean area A from area B; and boolean_add (area A, area B) refers to a function which provides the Boolean addition between area A and area B. An algorithm for carrying out memory management which allows the supports to be constructed without requiring the whole of the part and the total present data are stored simultaneously in memory will now be described. The preferred algorithm proceeds in two stages. In the first stage, the layers of the part are considered successively that start from the top of the part while maintaining a total run of the Boolean sum of the layers of the part. When encountering a layer, the total current or present for the layer (this is updated current total) is calculated by calculating the Boolean addition between the area of the current or present total of the previous layer and the area of the present layer. However, instead of storing or saving the total present data for all the layers, only the total data present for the intermediate layers, that is, every Nth layer where N could be 100, are stored. The rest of the total data present is discarded. This first step is illustrated in Figure 19 in relation to the part 80 and the associated supports, identified in the figure with the number 81. The generation from top to bottom of the totals present for the respective layers is identified with the number 82 and the intermediates of these are identified with the number 83. The pseudocode for this first stage is illustrated in Figure 20, in which the function get_part is that described above in relation to Figure 18 and the function boolean_addit ion is that described above in the discussion of Boolean operations.

The second step involves the selection of an intermediate layer and carrying out a top-to-bottom calculation, in the manner previously described, of those present for all layers between that intermediate layer and the next intermediate layer. The data, which consists of the data of the part and the total data present for each layer are then issued from bottom to top. When this has been carried out, the intermediate layer present and the dies between it and the next lower intermediate layer can be erased and the process repeated for the next upper intermediate layer. This second stage is illustrated in Figure 21, in which, in comparison with Figure 19, similar elements are identified with similar reference numbers. Four stages identified with the numbers 84-87 of this second stage are shown. In step 84, the running totals for all the layers between the intermediate layers 14 and 15 (for example the bottom of the part or object), identified in the Figure with the number 88, are determined and stored. Then, in step 85, the supports for these layers are determined by using the methods described hereinafter and are then emitted. The data of the part and data-totals between 14 and 15 are then deleted. Then, in step 86, the data of the part and the total data for each layer between 13 and 14, identified with the number 89 in the figure, are determined and stored. Finally, in step 87, the supports for these layers are determined and issued for integration or construction. Then the data for these layers is erased. Then the process repeats itself for each intermediate layer. It should be appreciated that this algorithm drastically reduces the memory requirements for the support generation process. If N is the number of layers between two successive intermediate layers, then the number of layers which is stored in one time is equal to the number of intermediate layers plus 2N (since the part and the total are required). If T is the total number of layers, the number of stored layers is equal to T / N + 2N. Then the optimal memory usage is obtained when N = square root of (T / 2). Thus, for a total of 5000 layers, the optimal number of intermediate layers N is 50. The total number of layers that must be stored at any time is thus 200.

The memory requirements can be further reduced by extending the aforementioned algorithm to two levels of intermediate layers. As shown in Figure 22, the algorithm proceeds in three stages, illustrated in the figure with identification numbers 90, 91 and 92. In the first stage, identified with the number 90, the first level of intermediate layers is determined. In the second stage, illustrated with the number 91, a second level of intermediaries between two of the first level of intermediaries is determined. Then, in step three, illustrated with the number 92 in the figure, the present or current totals are determined and stored for all layers between two successive second intermediaries. After calculating the supports for these layers, the data is discarded and the process is repeated for the next second intermediate level. When all the second intermediaries associated with the first intermediate level present have been processed, the next intermediate level is processed. If the number of first level intermediaries is N, and the number of intermediate seconds of level is M, then the memory requirements for this three-stage process is (T / N) + (N / M) +2. If T = 5000, N = 288, and M = 14, then, the number of layers that must be stored in one time is 66. Since this three-step process increases the calculation time, the two-stage process is preferred , unless very thin layers or very large numbers of layers are involved, in which case the three-stage process may be preferred. As discussed; The .RLE data for a given layer consists of a set of start and stop transitions, with a site x associated with each transition. The data illustrated in Figure 23, for example, corresponds to the following start and stop sites and raster lines: raster line A = [(start 20), (stop 48), (start 60), (stop 89) ], indicated by the reference numbers 102, 104, 106 and 108 respectively and the frame line B = [(start 35), (stop 72)], indicated by the reference numbers 112 and 114. A method for storing these data consists of a linked list of start / stop transitions, as illustrated in the pseudocode of Figure 24. Compared to an array, a linked list is preferred because it easily allows for flexibility and variability in the number of transitions required by scanning line. The problem is that it results in the use of large numbers of small groups or memory groupings dynamically assigned which can significantly degrade performance for at least three different reasons. First, dynamic memory allocation takes a long time since it requires system calls. Second, each cluster or dynamic memory pool has a higher hidden memory storage associated with it which is used for accounting. Third, logically adjacent pieces of information are located in a non-contiguous memory that leads to a large number of cache misses. To overcome these problems, another form of data structure is preferred. At a resolution of 1200 DPI, a transition in a typical part can be represented by 15 bits. Thus, a 32-bit word (with two spare bits) can be used to represent a start / stop pair. This data structure is illustrated in the pseudocode of Figure 25. The "last" indicator is used to indicate whether the start / stop pair is the last one in the set for a particular scan line. If so, the last bit is set equal to a logical "1". If not, the bit is set to a logical "0". In this case, the next start / stop pair in the sequence is stored in the immediately adjacent memory site. This scheme allows large numbers of transition points to be stored in contiguous blocks of memory, with two bytes provided per transition. An example of this scheme is provided in Figure 26, where like elements are identified with similar numbers as used in Figure 23. As shown, line A consists of two transition pairs: [(start 20), (stop 48)] and [(start 60) (stop 89)], elements 102, 104, 106 and 108 respectively, which are stored in contiguous 32-bit words as shown. The "last" bit 122 in the first word resets or resets a logic "0" to indicate that additional data for the scan line is being followed, while the "last" bit 124 for the second word is set to a logical "1"to indicate that no additional data follows. Line B consists of only a couple of start / stop sites as indicated: [(start 37), (stop 78)] designated with numbers 112 and 114 respectively - and where the last bit 126 is set to the logic " 1"to indicate that no additional data for line B follows. Reference numbers 132, 134 and 136 refer to other used bits associated with each 32-bit word. The .RLE data is not initially created in the package format described above. Instead, as discussed in relation to Figures 13a-13c, it is initially created in an unpacked format and then converted to the packaged format. In summary, a block of memory is assigned to store transitions. Pointers or indicators are used to indicate where the data associated with each raster line starts ("present raster line" indicator or "present list" indicator) and a pointer indicating where unassigned memory begins (indicator of "next available site" or "next free site"). Each four byte (32 bit) word in this memory block is defined in such a way that the first 15 bits are used to store the x site of the transition and the second 15 bits are used to store the qv of the transition.

The 31st bit is used to define a "used" indicator which indicates whether the word has been assigned and used. The 32nd bit is used to define an end flag or indicator which indicates whether the entry in that word is or is not the last transition entry for the given scan line for which the word is associated. Initially, each frame line can be assigned to one or more words to store data. As transitions are introduced for each boundary segment in the memory block, they are added to the lists associated with the frame lines from which they are derived. By adding each new transition point to the raster line lists, several situations can be found. First, if there is no transition data in the memory block associated with a given frame line, the transition data is added to the word associated with the "present list indicator" for that frame line. Secondly, if there is transition data in the word associated with the present list pointer for the given frame line, the next word (this is "next word") to the last transition point recorded for that frame line (this is for that pointer or present list indicator) is inspected to see if it has been used. If not used, the new transition data is entered. Third, if the "next word" is occupied, then the word before the pointer or indicator of the present list (this is "previous word") is inspected to see if it is used. Otherwise, the list indicator present and all the registered transition data (for the frame line) are moved by a word and the new data of the transition point is added to the end of the displaced list. Fourth, if the "previous word" is occupied, all the transition data for the frame line (which includes the list indicator present for that line) are moved to the word marked by the "next available site" pointer. , the new transition data is added, the original word sites of the transitions are marked as available to add new data and the pointer or indicator of "next available site" is moved to the next site to the words just moved and to the word added Several modifications can be made to the procedure outlined above. For example, different word sizes can be used, bit assignments can be varied, initial allocation quantities can be varied for each frame line, initial assignments for each frame line and memory sites can be avoided Assigned as additional frame lines are needed to fully process the input segments, additional steps can be added for better use of control memory and the like. The process described above is exemplified in the description that follows and the associated figures. Figures 27a and b are based on the same data found in Figure 13 and as such reference is made to like elements with similar identification numbers to illustrate the process. A large memory area 93 is allocated to maintain the .RLE transitions and the pointer 101 is used to indicate the next available memory word (32 bits). In this example, the format of the word includes the following bit designation: the first 15 bits 142 register the value used to store the site x of the transition, the second 15 bits 144 register the value of the qv of the transition. The 31st bit 146 is the "used" indicator which indicates whether the word has been assigned and used. The 32nd bit 148 is the last indicator or indicate "end" which indicates whether or not the transition is the last recorded transition for the frame line. Figure 27a illustrates the situation before any transition data is added to memory 93. For procedural reasons, as will become clearer later in the present, the first word in area 93, as shown is marked as used. The next pointer 101 of "next free site" points to the second word in the area. Next, a pointer array 58 is set or adjusted with all pointers set to initial values with their "used" bits set to zero. As discussed above, each pointer is associated with a scan line and is used to allocate or locate the memory site for the first word (this is for the first transition) associated with that scan line. This pointer is called the "present list" pointer since it points to the first word in the list of transitions associated with the present scan line that is considered. To add a transition for a particular scan line to the array, if the pointer in the array is active a word with a "used" bit set to the logical zero, the pointer site is considered free and the transition is assigned to that word from memory. Figure 27b illustrates the situation where a first transition to memory has been introduced for 5 scan lines. The process of adding a transition for a scan line having a "used" non-zero indicator at the position of the "present list" pointer 94 is illustrated in Figures 28a and b. Figure 28a illustrates two words 150 and 160 which were already entered as belonging to the scan line associated with the pointer 94 of the present list. The word 150 includes the bit allocations 150, 154, 156 and 158 that have the same definitions associated with the bits 142, 144, 146 and 148 of Figure 27b. Similarly, the word 160 includes bit assignments 162, 164, 166 and 168. Elements 156 and 166 give the value of the "used" flag. Elements 158 and 168 indicate whether or not the word (this is the transition) is the last transition so far recorded in the present list. As can be seen, item 158 indicates that the word 150 is not the last word, while 168 indicates that 160 is the last word used in the current list. First, the indicator "used" in the next word 170 after the end of the present transition list, which indicator is identified with the number 96 in Figure 28a, is inspected to see if the word is available. If the "used" indicator is set to logical 0, the word is available to store new transition details. If it is set to logic 1, the word is not available. If it is available, as shown in Figure 28a, then the new transition details can be placed in this word. The current or present list, as modified by the addition of a new transition, is illustrated in Figures 28b. In Figure 28b, the new transition details 97 are added to the word 170, the value of the "final" indicator element 168 is changed from "1" to "0" and the element 178 is given final indicator of the word 170 the value of "1" since 170 is now the final word of the present list. If the next word after the end of the present transition list is not available, then the availability of the immediately preceding word is inspected before the start of the present transition list. This inspection is presented when evaluating the value of the "used" indicator of this immediately preceding word. If it is available (as indicated by a value of "0"), then the entire list is shifted back by a word and the new transition is placed in the word which has recently been cleared. This process is illustrated in Figures 29a-29b in which in comparison with Figures 28a-28b reference is made to like elements with similar identification numbers. As shown in Figure 19a, the indicator or pointer of "present list" is associated with the word 150, the list ends with the word 160 and the next word after the end of the present list identified in the figure with the number 170, is not available (due to the value "1" in item 176), whereas the word just before the beginning of the list, identified with the n-number 180 is available (due to the value "0" in the element 186). The consequences of these evaluations are shown in Figure 29b, where the transition values previously associated with the words 150 and 160 are shifted to be associated with the words 180 and 150 respectively. The "present list" pointer is also moved to the word 180 and the new transition information is added to the word now available 160. As a result. The "end" indicator is still associated with the word 160 although it is no longer associated with the transition at the value x 60 (previous element 162, new element 152) but instead is associated with the transition at the value x 12 ( previous element 172, new element 162). In other words, the entire present list is shifted back by a word and the new transition 97 is stored in the cleared place. If there is no space in front of or behind the present transition list (that is, the word immediately preceding the pointer of the present list and the word immediately following the word containing the true end of the list indicator), all the The present list is copied to the space starting with the word indicated by the pointer "next available site" and the new transition is added to the end of this copied list. The "used" indicators of the original memory words in which the list was stored are then readjusted or reset to indicate that these original memory words are now available for use by immediately preceding and immediately following lists of the scan line. original sites. This process is illustrated in Figures 30a-30b, in which, with reference to Figures 28a-28b, 29a-29b, like elements are referred to with like reference numbers. Figure 30a illustrates that the word 170, after the end 160 of the present list, also as the word 180 preceding the word 150 containing the pointer of the present list, are now unavailable due to the "used" indicators 176 and 186 that fit "1". Figure 30a further illustrates that the word 200, where the "next available site" pointer is located. The word 200 follows the transition points already entered for all scan lines. Consequently, no new transition for the present scan line can be introduced in consecutive memory sites in those sites 150 and 160 that already contain transitions associated with the scan lines. As illustrated in Figure 30b the entire present list (transitions originally located at word 150 and 160) is copied to the area starting with the word 200 indicated by the pointer 101 of the next free site. The "used" indicators in the old memory, identified with the number 100 in Figure 30b, are reset or readjusted to indicate that this memory is now available. The pointer 94 of the present list is updated to point to the word 200, the new transition 97 is added to the end of the list in the word 220. The pointer of "next available site" identified with the number 101 is then updated to point to word 230 immediately following word 220 that contains the last entered transition 97 (this is the end of the list). Of course, if desired, one or more empty words can be left between the last entered transition 97 in the word 220 and the word pointed by the "next available site" pointer. This scheme is particularly efficient due to the nature of the .RLE data. Because the data is used to describe solid geometric objects, the number of transitions on a particular scan line is usually the same as the number of transitions on a neighboring line. This property is illustrated in Figure 31. The cross section of an object is illustrated from the top where spaced scan or scan lines are shown. To the right of each scan line, the number of transitions associated with that of scan is shown. Thus, if you want to add a transition to a particular scan line, it is likely that a transition will be added to a neighboring scan line. When a memory area is released, as described in Figures 30a-30b and the accompanying text, it is likely that the neighbor list has transitions that can be stored in this area, as illustrated in Figure 28a-28b and 29a- 29b and the attached text. Thus, large memory arrays develop fewer spaces than would be presented with random or disordered data. Also, there will be less loss of data stored in cache memory. When all the segments have been processed, the resulting lists are then stored in the x direction. The lines correctly submitted to Boolean operations are then extracted in the manner described above and the extracted lines are stored in the packaged format previously described. This mode operates directly on an .STL file without requiring the rounding of the vertices to cut planes and thus avoids at least some quantization error. However, some error of vertical and horizontal quantification is introduced, by means of the generation of the .RLE data since the cutting planes will only be located at discrete levels in the vertical direction and since the horizontal transitions will be limited to the borders of the pixels. An example of these issues is illustrated in Figure 32, which represents the quantization decisions associated with the representation of the on / off points 322, 324, 326, 328, 330, 332, and 334 for the frame lines 302 , 304, 306, 308, 310, 312, and 314. The center line of each frame line is illustrated by respective dashed lines associated with the boundary segment 300 which crosses through a plurality of pixels. In the figure, the region to the right of line is considered to be inside the object and it is considered that the region on the left is outside the object. For each frame line only one transition pixel can be selected to represent the edge of the object, regardless of how many pixels on that line intersect the border. Although there are many ways to determine which pixels form the boundary of the object, the illustrated procedure selects the boundary pixel for a given raster line as the pixel which contains the line segment and the center line of the raster line. In the case where the center line of the frame line is exactly the boundary between the two pixels, a decision is made as to whether or not to emphasize the object (this is solid) or the non-object (this is hollow) . As illustrated by the frame lines 302, 306, 310 and 314, the decision is made to emphasize the gap.

There is a variety of alternatives for selecting the transition. For example, you can choose to emphasize the solid by selecting the transition to occur, so that any pixel through which the line passes is counted as part of the object. Conversely, it may be selected to emphasize the gap by selecting that the transition occurs in such a way that only those pixels which are completely within the boundary of the object are included as part of the solid region. As an intermediate alternative, you can take an average of the transitions of the two previous alternatives. Other schemes for the determination of the transition sites may involve determinations of the percentages of the area of the solid or hole for the pixels of the border region and the like. The implementation of some of these techniques can be assisted by using the techniques described in the patents and patent applications referred to above, especially those involving cutting techniques. As a final example, an alternative could involve the subdivision of a pixel and make a decision as to whether the segment intersects with one or more of the subpixels. Whatever procedure is used, however, consistency is desired in the procedure used in relation to the part and the supports. Data Compensation Technique: Compensation is easily accomplished by moving the end points of the transitions in or out, keeping in mind that the end points of the adjacent segments should not be crossed. To prevent a support from contacting a part, for example, the .RLE data for the part could be expanded and then subjected to a Boolean subtraction operation of the total data present to obtain the .RLE data describing the the support region. Alternatively, the total data present could be expanded and the support data calculated as the Boolean difference between the total present data expanded and the data of the part. Support data could be calculated as the Boolean difference between the total data present and the data of the part. Then, the support data expands. Then, the actual support data is calculated as the Boolean difference between the expanded support data and the original data of the part. The compensation for adjusting the drop size along the scanning direction is easily carried out as long as the DPIs are at a resolution higher than the diameter of the drop. The compensation in the direction and is more difficult, but can also be carried out by staggering or recording in increments smaller or smaller than 300 DPI. It is useful to have the ability to convert .RLE data into vector data. As shown in Figure 33, the technique involves the connection of two consecutive "on" or "active" points or consecutive "off or inactive" points to form vectors, unless there is an intermediate point between the two, in which case the connection is not permissible. In Figure 33, for example, it is permissible to join point a and point a ', but it is not permissible to join point a to point c. The reason is that point b is between the two. Generation of Support Data Now we will describe a preferred process for the creation of data for the support structures. The process begins with the data provided from the data manipulation techniques described above. As described above, the Data Manipulation Subsystem provides object data (this is the part) and "total" data for each layer. Part data for a given layer consists of a series of start and stop points in adjacent frame lines, which define the XY sites of the part in that layer. The "total" data for a given layer consists of a series of start and stop points in adjacent frame lines, which define the Boolean union between the XY sites of the part in that layer and any desired support in that layer. Such data is illustrated in Figures 34a-34c. Figure 34a illustrates the data from part P [1] to P [10] for each layer (i.e., cross sections, sheet) 1 to 10, respectively, for a "peanut" shaped part, shown in FIG. floating way in the zx plane. In Figure 34a only one RLE line is shown for each of the cross sections P [l] to P [10]. The starting transitions are identified by the symbol "" 'while the stop transitions are identified by the symbol "-j". As you can see, the part data follows or tracks the boundary (that is, the extent) of the part. Figure 34b illustrates the "total" data T [l] to T [10] for each layer 1 to 10, respectively, for the part. They are also defined in terms of start and stop transitions. However, unlike the part's data, they do not necessarily trace the boundary of the part. As discussed above, the "total" data for a given layer consists of the Boolean union of the part data for all the layers above the given layer. Figure 34c illustrates a cross-sectional view (in the X-Y plane) of the data of the part and of the total data for a given layer. These data, identified as P [i] and T [i], respectively, comprise a plurality of start and stop transitions which are arranged along dashed lines H [i] in the X-Y plane. In a preferred embodiment, the dashed lines would be oriented parallel to the x axis. However, as indicated by other orientations of dashed lines are possible. In a preferred embodiment, the object data and the combined total data are used to determine the start and stop transitions for the supports, one layer at a time. If only one type of support is to be used in all regions that require support, you can define a single support style, which can be applied to each layer in the region defined as the difference between the total data for a layer and the data of the part for that layer. On the other hand, as discussed in US Patent Application No. 08 / 534,813 it may be advantageous to use different types of support structures for different sites, depending on how close or how far apart is any surface facing upwards and / or facing down the object. In addition, it may be advantageous to use different support styles depending on how far apart the region of the borders of the object is in the same layer. The techniques for carrying out horizontal comparisons are described in the North American Patent Application referred to above No. 08 / 428,951 which are applicable to the present invention to help define the regions of the support. For example, it may be advantageous to use two different support styles, one for use when a region is a few layers below a surface facing down and one for use anywhere. Alternatively, two physical support styles can be used in combination with a third style of "non-support" where the non-support style could be applied to the region that is within one or two pixels of the regions bordering the part or where the surface of the part above the object as a normal to the vertical which is greater than some critical angle. Many additional modalities using multiple support styles are possible and can be easily implemented by the teachings herein and those incorporated for reference (particularly, US Patent Applications Nos. 08 / 475,730; 08 / 480,670; 08 / 428,951 and 08 / 428,950). Additionally, the teachings herein may be applied to what might be termed interior object supports, wherein one or multiple support styles may be used in the process to form the interior portions of the objects. Examples of such techniques, as applied in stereolithography for the purpose of manufacturing investment casting configurations, are described in US Patent Application No. 08 / 428,950 previously incorporated herein. To further explain how data could be defined for different support regions, the following example is given, which corresponds to the hybrid support example described in US Patent Application No. 08 / 534,813. In terms of this example, 3 categories of supports are recognized: (1) thin, fiber-like columns, spaced in a chessboard configuration; (2) column supports of 3x3 pixels more substantial; and (3) intermediate or transition layers. Assuming that the "n" layer is about to be constructed or integrated, the technique involves determining how close each portion of the "n" layer is to an upwardly facing and / or downward facing surface of the "n" layer. object .. In the present embodiment, if a portion of the layer "n" lies within "r" layers (e.g. 5-10 layers) of a surface facing down or within "u" layers (e.g. -10 layers) of a face-up surface, the category of chessboard stands will be integrated or constructed for that portion; if between "s" (s = r + l) and "t" layers of a face-down surface (6-10 or 11-15) and more than "u" layers (for example 5-10 layers) of a surface facing upwards, the intermediate or junction support category will be integrated; and if more than "u" layers (5-10 layers) of a face-up surface and more than "t" layers (eg, 10-15 layers) of a face-down surface are to be integrated into the 3 x 3 column support. The above example is illustrated in Figures 46a and 46b, which illustrate identical side views of an object, with a space between an upwardly facing surface and a downwardly facing surface of an object. . Figure 46a illustrates the side view together with levels and hypothetical regions on which the formation of different support structures will be based. Figure 46b illustrates the side view where the space is filled with various types of support structures, according to the physical layout of the hypothetical levels and regions of Figure 46a. More specifically, Figure 46a illustrates the surface 402 of the object facing downward and a surface 400 of the object facing upwards, which are separated by a spacing or separation comprising the regions 404, 410, 408 and 406. The region 404 is located within the "u" layers of the surface 400 facing upwards and region 406 is located with "r" layers of surface 402 facing downwards. The region 408 is located "r" and "t" layers of the surface 402 face down and is located simultaneously to more than "u" layers of the surface 400 face up. The region 410 is located simultaneously to more than "u" layers of the surface 400 facing upwards and more than "t" layers of the surface 402 facing downwards. Regions 404 and 406 will be formed with chessboard type supports, region 408 will be formed with transition type supports (for example completely solidified) and region 410 will be formed with 3-by-3 column supports. Layers 414, 412, 424 and 416 are shown completely within the regions 404, 406, 408 and 410 respects. vamente. Therefore, these layers will be formed with a single type of support structure in its entire area. On the other hand, layers 418, 420 and 422 are shown partially located within regions 404 and 410, 410 and 408 and 408 and 406 respectively. Accordingly, these layers will be formed with different types of support structure, depending on the XY location of each portion of the layers. Figure 46b illustrates the regions 432 and 430 of the solid object, above and below the downface surface 402 and the upward facing surface 400 respectively. Regions 404 and 406 are indicated filled with chessboard supports (an active pixel, an inactive pixel). Region 410 is indicated filled by 3-by-3 column supports (3 active pixels, one inactive pixel). Region 408 is indicated as filled in a solid region of supports. This modality can be presented in the form of an equation. To present these equations, the following terminology is used: Cn (D): the area elements of layer n on which the "chessboard" category of the supports must be integrated as determined from the face surfaces toward aba j o. Cn (U): the area elements of layer n on which the category of "chessboard" of supports must be integrated as determined from the surfaces facing upwards. Bn (D): the area elements of layer n on which the category of "chessboard" of supports must be integrated as determined from the faces facing downwards. Sn: the area elements of layer n on which the category of 3x3 pixel column supports must be integrated. Pi: the area elements of the part in the cross section "I". Pn: the area elements of the part in the cross section "n". Tn: the area elements of the total data in the "n" cross section. ?: the Boolean sum of the area + elements: Boolean union of the area elements. -: Boolean difference of the elements of area D: Boolean intersection of the area elements. r: the number of layers below a face down feature which are formed with chessboard supports. u: the number of layers above an up-face feature which are formed with chessboard supports. s: r + l = the number of layers below a surface face down at the end or end of the transition type supports. t: the number of layers below a face-down surface on which the transition-type supports begin. With this terminology in mind, the following equations define the preferred method for determining the supports for layer "n" according to this modality: (I) c, (> > - £ P ~ F. ••• t (2) cw t (- ^) f | r. (J) *. (ß) -? 'Lc, (z » - », - c, (t < 4 > *. T. -. - c. (0 > - <:" (£ /) - *. (?) Equation (1) indicates that the area of the "n" layer over which the chessboard category of the supports must be constructed or integrated, as determined from the surfaces facing down, is calculated by taking the Boolean union of the data of the part of the layers above the layer n and then calculate the Boolean difference between the data representing this bound area and the data of the part for the layer "n". Equation (2) indicates that the layer area "n" on which to build or integrate the category of chessboard supports, as determined from the surfaces facing up, is calculated by taking the union Boolean of the data of the part of the "u" layer below the layer "n", calculate the Boolean difference between the data representing this bound area and the data of the part for the layer "n" and then calculate the intersection between this data and the total data for layer "n". The purpose of this last calculation is to avoid integrating or constructing supports when in fact there are no layers of the part above layer "n". Equation (3) indicates that the area on the layer n on which the joint supports must be constructed or integrated, as determined from the surfaces facing down, is calculated by 1) taking the Boolean sum of the Part data from layers "s" to "t" above layer "n" and 2) then differentiate from the data added from stage 1, t the representative data of the areas on which they will be integrated or will build the chessboard supports on the n-layer (below the surfaces of ca, a down and above the surfaces facing upwards) and the data representative of the areas on which the part itself is located. will integrate over layer "n". Essentially, this equation establishes a priority between the binding and chessboard supports. It requires that in the areas which are inside the "u" layers of an upside-facing surface and within the "s" to "t" layers of a face-down surface (such as an area --- below of a curved surface continuously), that priority will be given to the construction of the edrez board supports. Finally, equation (4) stipulates that the area on the "n" layer on which the three x three pixel column supports will be integrated or constructed, is determined by taking the total data for layer "n" and determine the Boolean difference between these data and 1) the data of the part for layer "n", 2) the data representative of the area or areas of layer "n" on which the chessboard supports will be integrated and 3) representative data of the area or areas of layer "n" on which the joining supports are to be built or integrated. As it is evident from the previous discussion, equations can be defined for several regions where it is desired to form different types of support structures. Figure 37 illustrates an arc-like support structure that requires a different construction or integration configuration as it approaches more closely to a downward facing surface 24. As indicated, the arc-type support starts at the surface -23, which may be the surface of a construction or integration platform, an up-face surface of the object or a surface associated with the previously formed supports. As such, this support structure is a hybrid support with many different support styles (for example 10 or more) required for its formation. Of course, it would be possible to add a number of layers of chessboard supports between the tops or tops of the arches and the face-down surface that is supported or supported. Once these data have been determined, the next step in the process is to format the data for its output or emission to the control computer. As discussed, the control computer will also load this data as the object data in the bitmap to control the print head, as well as the X-, Y- and Z- steps. Style files are used for this purpose, one for each category of object structure and support structure. A style file for a given type of object or support consists of the core or central configuration which is repeated throughout the area in which the category of the object or support is to be integrated. Style files are used to modulate the configuration or geometric shape of construction or integration associated with a given region. This data modeling technique simplifies data handling and storage requirements. For example, the style file associated with the category of "chessboard" supports in the present embodiment is the 2x2 pixel configuration shown in Figure 38a. As a second example, the style file associated with the 3x3 pixel column supports in the most preferred mode is the 4x5 pixel configuration shown in Figure 38b. Of course, many other style configurations are possible. These configurations or geometric style figures are repeated one after another, usually starting at the site (x, y) (0,0) to define a repetitive configuration in the XY space. This global configuration is associated with the start and corresponding transition data for the object and support regions. The combination of style file information and object information can be presented before the transfer of data to the control computer or can be presented after the transfer. Typically, the object and style information are combined into a single data set after they are both transferred to the control computer. In the present, the preferred style file associated with the part consists simply of a solid pixel configuration of 1 x 1, which indicates that the interior of the part is always solid. In the present, the most preferred replica of configurations is fixed in the X-Y plane. With respect to the most preferred 3x3 support configurations, the result is that some of the 3x3 pixel columns are decreased at the border of the part. This effect is illustrated in Figure 39a. As shown, portions 30 and 31 of the 3x3 pixel columns are not constructed or integrated due to their proximity to the border 32 of the part. The result is that these two supports have diminished surface areas. If the columns are not retracted or removed from the part boundary, this presents little problem, since the formation of the part will form the other portion of each column formed partially. However, the construction or integration supports in contact with the part tend to damage the surface finish of the object, to thereby result in another problem. In the case that the supports are retracted or removed from the part, one solution to this problem is to allow the replication configuration to vary, to allow the 3x3 supports to follow the part boundary. This procedure is illustrated in Figure 39b. Gradual changes in the position of the support column can be obtained by using offset pixel configurations as described in US Patent Application No. 08 / 534,813. As mentioned above, another problem that sometimes arises is that the 3x3 support columns are sometimes built or integrated in direct contact with the part. This problem is illustrated in Figure 39c. As shown, the supports 33 have been constructed or integrated in direct contact with the part 33 (the supports 34 shown in dashed lines are below the part and are illustrated only for purposes of clarity). One solution to this problem is to remove these supports by a pixel or more to separate the supports from the part. This can be done simply by adjusting the start and stop transition data for the media. In the present mode this adjustment is optional due to the transaction or exchange involved: when the support is pushed back by a pixel, the surface area of the columns will be decreased, possibly to cause an accumulation problem. A few warnings about the preferred method to perform Boolean calculations. As discussed, the data involved in these calculations is formatted as a series of start and stop transitions. It has been found that this format facilitates Boolean calculations by allowing them to function as a series of arithmetic calculations. For example, to carry out a Boolean difference operation between two sets of transition data, it is only necessary to subtract arithmetically the corresponding start and stop transitions from each other. The result is a significant improvement in the calculation speed.

The reason is that Boolean operations involving N data points based on polygonal data are essentially N2 operations ~ while arithmetic operations using start and stop transition data are essentially proportional to N. Another point is that the data Intermediate boolean junctions calculated for the "n" layer, that is, the Boolean union of the data layers u above and below the n 'layer between the "s" and "t" layers above the "n" layer ", can not be used in any subsequent processing. The reason is the lack of "memory" associated with the Boolean union operation as illustrated by the following equations: (5) '¿, .1. at¿ (arithmetic) (6)? * | (Boolean) As indicated, with the arithmetic operation, the last item in the sum has an effect on the final sum which can be subtracted when the calculations are carried out for the next layer. On the other hand, with the Boolean operation, the n-th item does not necessarily have any impact. Thus, the effect of this item can not necessarily be subtracted when the calculations are carried out for the next layer. Although equations (1) to (4) above produce accurate results, they can lead to an excessive calculation time. As such, in some circumstances, it may be desirable to use equations that can give approximate results but that involve fewer calculations. Excessive calculations can be avoided by making assumptions that the slope of a part surface does not change sign in a given number of layers (for example 10 layers, approximately 10-20 mils) or that any change in direction represents a negligible variation in the position of the cross section. In other words, the assumption is that the surface of the part does not change quickly or drastically. This point is illustrated in Figures 35a-35b. Figure 35a illustrates a part which is consistent with the assumption. As can be seen, the slope of the surface of the part, identified as S in the figure, does not change sign or direction over a given number of layers, for example 10 layers. On the other hand »- Figure 35b shows a part which is inconsistent with the assumption that the direction of the slope of the surface does not change sign. However, depending on the amount of variance at the XY position of the surface, the change in direction can result in a negligible variation in the position of the cross section. As can be seen, the slope of the surface of the part, identified as S 'in the figure, changes sign in, for example 10 layers. For a given number of layers, the thinner the layers, the more likely the assumption is to be maintained. If these previous assumptions are made, the following formulas can be used to reduce the mathematical calculations: (7) Cn = (Pn-1 + Pn.u-Pn) ntn (8) Bn = Pn-t + Pn-s-Cn- Pn (9) Sn = Tn-Pn-C "-Bn instead of being based on the Boolean sum of the area of each cross section within a region, like the original equations (1) to (4), these equations use the information of the cross section of only the upper and lower cross sections of the region. If the assumptions are always true, these formulas produce accurate results. In any case, in practice, they have shown that they are very good approximations. It should be noted that in order to carry out the calculations mentioned above, it is necessary to have simultaneously the data of the layers (t + u + 1) (for example for t = 10, u = 5, data for 16 layers is required). This is because the support data for the "n" layer is dependent on the data of the part and the total data for the layers "n + l" to "n + t", the layers "nl" to "nu" "and of course for layer" n ". In order to keep such data in an immediately accessible form, it is advantageous to use a temporary or intermediate ring memory. As shown in Figure 36, a temporary or intermediate ring memory is a circular memory in which the data of the part and the total data are stored for the n + u + 1 layers (for example 16 layers). Figure 36a illustrates the state of the buffer or time in terms of a 16-layer example (t = 10, u = 5) when the calculations for layer n are to be carried out. A pointer, identified as PTR in the figure s-e used to point to the layer present under consideration. As indicated, the data for the layers "n + l" to "n + 10", "n" and "n-l" to "n-5" are stored in the buffer or temporary. A second pointer, identified as LAST in the figure, is used to point to the last entry in the temporary or intermediate memory, in this case, the entry is for layer n-5. After the calculations for the "n" layer have been completed, it is necessary to update the buffer or temporary in the preparation to carry out the calculations for the layer "n + l". To accomplish this, the PTR pointer is first updated in such a way that it points to the data for layer "n + l". Then, the data indicated by LAST is overwritten by the data for the next layer to be added to the buffer, in this case the "n + 11" layer. Finally, LAST is updated to point to the data which is now the last entry in the buffer, in this case, the data is for layer "n-4". The result of these three calculations is illustrated in Figure 36b. Figure 36c illustrates the state of the buffer at the point when the calculations for the layer n + 2 are going to be carried out. Then this process is repeated until the calculations for all the layers have been completed. A number of alternative modes are possible to manipulate the data of the 3d object into data useful for controlling an SDM apparatus. For example, in an alternative mode, the calculations mentioned above are carried out using Boolean operations on polygonal data instead of transition overlays. In another, the data for all the layers of the part are stored simultaneously in a memory instead of in a temporary ring memory. In yet another embodiment, it is possible to equalize the accumulation rates of the fiber-like, thin, and the part supports by employing multiple steps of the print head. It should also be appreciated that it is possible to calculate the junction data or the transition support data from the surfaces facing upwards, that is, Bn (U). This data could be used to form transition supports between the supports of thin fiber-like columns, starting at an upward facing surface of the object and the 3x3 column supports seated thereon. In addition, it should also be appreciated that it is not necessary to calculate the data of Cn (U) separately from the data of Cn (D), if the style file for both is the same. Of course, if it is proposed that the two style files be different, then both categories of data must be maintained. It should also be appreciated that it is possible to construct or integrate an arbitrary number of support types or categories on a given layer, by using the present invention instead of the three that have been discussed. This can be done simply by adding additional style files and equations to determine the areas in which the new media categories will be integrated. Building styles and support styles: For optimal data manipulation it is advantageous not to embed the formations from the regular configuration to the RLE data because this would make the RLE files excessively large and would make data manipulation impractical in a timely or synchronized manner.

As such, it is advantageous to keep the cross-object information of and the support independent of the exact exposure settings (ie, deposition settings) until the printing of the layer is to occur. As mentioned above, at an appropriate time, the cross section data (eg, in the form of information RLE) is subjected to a Boolean intersection operation with the appropriate construction style settings to define the exact configuration that will be used to define the details of the deposition. For example, this can be used to create chessboard configurations on a fast basis. An example of this is illustrated in Figures 40a-40c, in which reference is made to like elements with similar identification numbers. Figure 40a illustrates the desired image 28 to be printed. As shown, the desired image consists of two components. The first component, identified with the number 29, is a solid. The second component, identified with the number 30, which one wishes to form with an on-off chessboard configuration. For the reasons discussed above, it can be prohibitively slow and of extensive memory usage, to convert the image 30 to a honeycomb configuration on a pixel basis per pixel basis. Further manipulations of the data for the image 30 may be unduly complicated and slow when placed in a honeycomb configuration too prematurely. The transfer of data to a storage device (that is, a hard drive or filter driver) can also be unduly troublesome to maintain in a detailed format. Thus, as shown in Figure 40b, the data for both configurations is maintained or converted into solid form (minimum transitions) for further manipulation after which they are transmitted to a digital signal processor which is responsive to control the emission of jets and movement in X, Y, Z. Then, as shown in Figure 40c, the data 31, associated with component 30, which are in solid form, are "Y ed" (i.e. to a Boolean intersection operation) logically with the honeycomb / chessboard configuration 32 in order to change the solid data to the desired modulated shape representative of the modulated cross section configuration to be ejected in jets. Once it is in its final modulated form, it is preferred that no additional storage of the data is presented, but instead they are used to control the firing of the jets with or without additional manipulation. In this example, the data for component 29 and 30 must now be submitted to an "O ed" operation together to produce the single bitmap that contains the entire desired set of data. Then these combined data are used to control the trigger or drive of the print head. The data provided with the RLE file to the modulator includes information on the construction style / configuration of the support for use as discussed above. As discussed above, the association of the RLE data with the modulation data is carried out through the use of Style files, each of which stores a "style" or particular construction configuration. Examples of construction or interaction configurations are shown in Figures 41a, 41b and 41c. Figure 41a illustrates a chessboard construction configuration suitable for use in the construction or integration of a category of supports as described in US Patent Application 08 / 534,813. Figure 41b illustrates an appropriate configuration for use in the integration or construction of a second category of supports, as also described in US Patent Application 08/534, 813. Figure 41c illustrates a configuration which specifies that the solid has been constructed or integrated. Many other styles of integration or construction are possible, in which multiple styles of construction or interaction by exposure are included. Such examples are shown in Figure 41d in which the alternative spaced scan lines are solidified in successive steps. In this example, the configuration 56 is exposed during a first step and the configuration 57 is exposed in a second step. Another example is shown in Figure 41e, in which the alternative spaced columns are solidified in successive steps. In this example, the configuration 58 is exposed during a first step and the configuration 59 is exposed during a second step. A third example is illustrated in Figure 41f in which the non-overlapping chessboard configurations are solidified in successive steps. The configuration 60 is exposed in a first step and the configuration 61 is exposed in a second step. To associate the different style files with the different regions of the object and the support, the RLE format is made to include a designation of the construction configuration for each different set of frame line transition information that is passed to the modulator. The conceptual format of the .RLE file is illustrated in Figure 47. Through this file format, a user can specify virtually any construction configuration for a given pair- or pairs of transition points. Data Fake In addition to providing a bitmap containing the correct pixel information to control the firing of the jets, the data must be easily extractable from the bitmap and provided to the firing mechanism in the correct order. This in need to place the data in a removable form brings us to the next stage in the process of data manipulation. This next stage is called falsely. For example, the data may be processed in such a way that the necessary information is available to allow the jets to fire simultaneously although the adjacent jets may not be located on adjacent or still located frame lines on their respective frame lines and, above the same X coordinate. As such, the falsification refers to a process of realignment of the data which is required, for example, when the scanning head is placed at an angle to the scanning direction (as illustrated in FIG. Figure 2b) when multiple heads are used and will be fired simultaneously or in sequence or simply because the jets are not spaced over adjacent raster lines. In Figure 2b, for example, the holes (3) and 10 (4) which are aligned in Figure 2a, move in the scanning direction by a distance d ", as shown in Figure 2b, when the scanning head is angular in relation to the direction of exploration.

However, the data used in connection with the configuration of Figure 2a, would require that the jets 10 (3) and 10 (4) fire at the same time to strike at similar X sites. With the configuration of Figure 2b a distortion would be caused by the use of such data. Consequently, the data must be falsified, in this example, to correct this relative displacement. The problem is that the amount of data involved is relatively large and the distortion must be carried out in real time. For example, an ink jet in a typical configuration would need only 500nS to pass over a given pixel. Thus, any process of falsification that operates on individual pixels can no longer take this time per pixel (on average) in order to adjust to the speed of data consumption. A typical digital signal processor, for example a C31 processor, running at 40 MHz, has a cycle time of the order of 50nS. Accordingly, if the time over any pixel site is of the order of 500nS, there are only 10 cycles available to operate on a given pixel.

On the other hand, each processor instruction requires a minimum of one cycle. Frequently, several cycles are required to overcome conflicts of the main distribution line, end-of-term conflicts and memory wait states. Thus, each instruction can effectively require 2-4 cycles. Thus, only about 3 instructions can be really dedicated to each pixel. The problem is that to carry out a typical operation, such as adjusting an individual pixel to a logical "1" requires approximately 6 instructions. Thus, it is not feasible to carry out operations on a pixel by pixel basis. Instead, operations that operate on multiple pixels at a time, such as 32 pixels, are required. Some typical operations could include cleaning the image, moving the image, emitting the image, the "Y" operation of two images together or the "X" operation or the two images together. These types of instructions usually require fewer instructions (2 or 3 instead of 6) and operate at 32 pixels at a time. In general, they operate at approximately 100X faster than the operations on the individual pixels. As discussed above, a control computer performs the functions of cutting an .STL or CTL file and calculating the .RLE data for the various cross sections. A Digital Signal Processor (DSP) coupled to the printhead must take this .RLE data, decompress it, falsify the data according to the jet arrangement and then issue the data to the jets. As discussed, "distortion" refers to the process of manipulating the image data to compensate for the arrangement of jets or possibly other factors. Since the data, once decompressed, may not be able to be manipulated in a sufficiently fast manner, it is advantageous to have the ability to manipulate the data while they are still in compressed form (for example, while still they are in the .RLE format). Another critical preference that saves time is that the data is stored in memory in such a way that a word of 2 bytes or 4 bytes contains pixels that it is desired to emit each one at the same time.

Then the process of falsifying the data involves _- simply shifting the start and stop transitions in the scanning direction by an appropriate amount, while retaining the data associated with the pixels to be emitted at the same time in the same word . Then the data is decompressed and the individual words are sent to the printhead when the appropriate site is found in the X direction. The technique is illustrated in Figures 42a, 42b, 42c, 42d, and 42e, in which reference is made to like elements with similar identification numbers. Figure 42a illustrates the image transformed into pixels of the original cross section. Figure 42b illustrates this data in .RLE format. As shown, the data for the individual scan lines identified in the Figure with the numbers 25 (1), 25 (2), 25 (3), ... 25 (10), have been compressed into data representative of the start and stop transitions. Figure 42c illustrates the process of falsifying this data to adjust a print head which is angular in relation to the scanning direction. In this figure, it is assumed that the printhead has 5 jets and is angular such that the individual jets are relatively displaced-from the successive jets by a pixel. Thus, the data for the scanning line 25 (2) is displaced by a pixel in relation to the scanning line 25 (1); the data for the scanning line 25 (3) is offset by a pixel in relation to the scanning line 25 (2), etc. The process continues until it meets the last scan line 25 (6). Since that is the sixth line of exploration and will not be explored in the same step as the first 5 lines, that line is not displaced in relation to the others. Instead of this, the scanning line 25 (7) is displaced by a pixel in relation to the scanning line 25 (6). The scanning line 25 (8) is displaced by a pixel in relation to the scanning line 25 (7). The scanning line 25 (9) is displaced by a pixel in relation to the scanning line 25 (8) etc. During this process, the falsified data are "grouped in bands" in such a way that the data associated with the shots will be presented at the same time if they accumulate or collect in a single word. Then this data is decompressed successively from one band at a time. -The process is illustrated in Figure 42d. The data for the pixels in each of the columns 27 (1), 27 (2), 27 (3), ..., 27 (12), each represent data which will be fired at the same time. Thus, each of these columns of data is stored in individually accessible words and thus accessible simultaneously. A band-forming index 26 is also maintained to grade data from one column at a time. As it is found that one column is decompressed at a time (that is, each transition is converted to an on / off bit, for example 32 bits at a time) with reference to Figure 42d, for example the index of band formation is located in column 27 (8). Thus, as shown, the data in that column is uncompressed. The remaining data in columns 27 (9) to 27 (12) are still in compressed format. However, as discussed, those data will be decompressed as the band formation index is found. Next, the data is sent sequentially to the print head of one column at a time. The process is illustrated in Figure 42e. As shown, the band-forming index has been readjusted or reset and then used to successively graduate the columns 27 (1) -27 (12) a second time. As shown, the index is currently located in column 27 (5). Thus, the data in that column is emitted at the print head. The data in the remaining columns 27 (6) -27. { 12) will be issued in turn. Flight Time and Jet Shot: Before the previously generated data results in the deposition of drops of material at the desired sites, a critical function must be carried out. As the data is loaded into the ink jet head for firing, the system must end when the head of the ink jet has reached the proper place to drop or drip its material. The appropriate trigger time as discussed in the US Patent Application referenced previously 08 / 534,813, actually occurs a few times before the head is positioned on the appropriate deposition site. This premature trigger compensation is called the flight time correction. However, the system must still end when it is in the proper place to emit the premature trigger signal. The details of this determination process are given below. To allow construction with a desired scanning line resolution, it is important to have the ability to shoot the jets at any desired position along the scanning direction. This can be problematic when an encoder is used to indicate the actual position X, where the encoder may not have the guide actuators in the required positions. In effect, the encoder may be of lower resolution than that to which it is desired for printing. Since the resolution coders are more expensive later on and you want to keep equipment costs low and since it is a disadvantage to be limited to a single resolution or resolutions which are multiples of the guide spacing. Other means are desirable to determine the exact firing positions. The exact firing positions, as explained below, are determined by carrying out an interpolation of the distance between the guide lines based on a calculated average speed and a known elapsed time since the last guide was passed. Then the trigger sites are determined using the known trigger point known and the estimated interpolar estimated value of the actual position. The stage X-12 (see Figure 1) has associated with it an encoder for use in determining the position of the print head in the X direction, in such a way that the firing impulses for the print head can be initiated at the appropriate time. In a preferred embodiment for carrying out this function, a glass plate, identified with the number 34 in Figure 43, is employed, over which the lines 33 are recorded, which are separated from each other by microwaves A light and photodiode detector (not shown) is also used to determine when these lines pass and to interrupt DSP each time the print head passes through one of these lines. A pair of detectors (not shown) are also used to indicate whether the print head moves to the left or to the right. To avoid altering DSP with signals caused by vibration and the like, a digital hysteresis circuit (not shown) is used to protect or shield DSP from transient interruptions caused by vibrations and the like. From these circuits, it is possible for the DSP to determine the position of the print head in 10 microns, and also to determine the direction of movement. In order to print at a finer resolution of 10 microns, a counter is provided within the DSP to start counting whenever the DSP passes through one of the lines mentioned above. When the counter reaches a certain value the DSP causes a trigger signal to be generated to activate the print head. A second counter is also provided to deal with the situation illustrated in Figure 44. The signals of T0, Ti, T2, T3 and T4, identified with the number 35, represent the signals generated by the encoder of the print head pitch beyond the lines 33 illustrated in Figure 43. The lines identified by the number 36, in contrast, indicate the desired firing positions. For the signals To ', Ti', T2 ', T3', all these signals follow the corresponding signals To, Ti, T2, T3, respectively. Thus, a single counter can be used in the generation of these signals in the manner described. The problem that is presented is illustrated by the signals T and T4 '. Since T4 'actually precedes its corresponding signal T4, a second counter must be provided to generate this signal in response to the presence of the signal t3. An algorithm for generating the trigger signals is illustrated in Figures 45a-45b. As shown, an interruption, identified in Figure 12a with the number 57, is generated as the print head passes through one of the lines of the encoder. Then, in step 38, a timer (not shown) of the encoder is read and associated with the position of the print head. This stage is carried out on several lines of the encoder. The resulting data is stored. In step 39, the average speed of the print head is calculated from the stored data by dividing the change in position by the change in time over the prescribed encoder lines. In step 40, the distance? D between the next trigger site and the last line of the encoder is determined. In step 41, this value is used to calculate the time differential? T (1) from the last encoder line to the next trigger site taking into account the left / right compensation i and the flight time compensation . Then, in step 42, this value is loaded into a first trigger timer, which, as discussed, initiates a trigger pulse when it has finished. In step 43 (Figure 45b), the time differential? T (2) for the next firing position is calculated in the manner described in relation to? T (l). In step 44, this value is inspected to see if the next trigger position is located beyond the next line of the encoder. If so, then that trigger pulse can be started offset from the next line of the encoder. If not, in step 45, that value is loaded in a second trigger timer. Then, in step 46, a return of the interrupt is initiated. Alternative modes can be used to link the position of the encoder to the output of the trip commands. An alternative as such, uses timing signals from the location of the multiple encoder guide to derive a more accurate representation of the average speed of the scanning head. In this preferred modality, the last 8 time signals of the encoder guidance sites are averaged to produce a time signal, which may be associated with the position of the fourth encoder guide backwards. The 8 time signals from the previous encoder guidance sites are averaged to produce a time signal which can be associated with the 12th backward encoder guide. These two averaged time signals are used to derive an average speed value for the print head scan. From a determination of the distance between the 4th return encoder and the next trigger site, the average speed, the time elapsed since the 4th encoder guide was crossed, a time when the jet will reach the correct firing site is estimated, a timer is started or activated using the estimated time and the jet triggers when the time interval has elapsed.

This completes a discussion of the basic firing position improvement algorithm. It should be noted that several improvements or modifications are available, which include compensation based on the acceleration of the print head or the use of more than one trigger counter to further increase the print resolution in relation to the increase that can be obtained by means of two counters. While the embodiments and applications of this invention have been shown and described, it will be apparent to those skilled in the art that many more modifications are possible without departing from the concepts of the invention herein. Accordingly, the invention is not restricted, except in the spirit of the appended claims.

Claims (101)

1. CLAIMS 1. A rapid scale modeling apparatus, characterized in that it comprises: a dispenser or dispenser to supply or controllably distribute a material that can flow which is soluable after being supplied or distributed; a platform for holding a cross section of a three-dimensional object and providing a work surface for integrating or constructing a next cross-section of the object; at least one grader coupled to the dispenser or distributor and to the platform, to relatively displace the dispenser or distributor and the work surface, in at least two dimensions, comprising a scanning direction and a graduation direction; and a controller coupled to the grader and spout to cause the material to be stocked or distributed on the work surface in accordance with a selected style. The apparatus according to claim 1, characterized in that the apparatus is a modeling apparatus by selective deposition, and the dispenser or dispenser is configured to selectively deliver or distribute the material according to the selected style. 3. The apparatus according to claim 2, characterized in that the controller is configured to provide the style which is a construction or integration style. 4. The apparatus according to claim 2, characterized in that the controller is configured to provide the style which is a style of support. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the overprint in the scanning direction. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style, which specifies a higher resolution in the scanning direction than in the graduation direction. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction or integration style which specifies a higher drop density ratio for the upper faces facing upwards than for the inner regions of the object. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the extension of a surface region face down to a plurality of layers above a surface face down . The apparatus according to claim 8, characterized in that the controller is configured to provide the construction style which specifies the extension of the surface region face down to 5 layers above the face down surface. The apparatus according to claim 8, characterized in that the controller is configured to provide the construction style which specifies the extension of the surface region facing down to 10 layers above the surface facing down. 11. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies a higher drip density ratio for the formation of upper faces facing upwards than for the inner portions of the object. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the extension of an upper face-up region to a plurality of layers below a face-up surface . The apparatus according to claim 12, characterized in that the controller is configured to provide the construction style which specifies the extension of the upper region face up to 5 layers below the surface facing upwards. 14. The apparatus according to claim 12, characterized in that the controller is configured to provide the construction style which specifies the extension of the face-up region to 10 layers below the face-up surface. 15. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the use of a higher drip density ratio for the formation of boundary regions of the object for the formation of inner regions of the object. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the extension of a border region to a plurality of drop widths within the object. 17. The apparatus according to the rei indication 16, characterized in that the controller is configured to provide the construction style which specifies the extension of the border region to at least two drop widths inside the object. 18. The apparatus according to claim 16, characterized in that the controller is configured to provide the construction style which specifies the extension of the border region to at least 4 drop widths within the object. 19. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the formation of an interior region of the object with chessboard supports. 20. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the formation of an interior region of the object with in-line supports. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the formation of an interior region of the object with column supports. 22. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style, which specifies the formation of an interior portion of the object with aerial supports. 23. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the change of the scanning direction between two layers. 24. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style, which specifies inverting the scanning direction between two layers. 25. The apparatus according to the rei indication 3, characterized in that the controller is configured to provide the construction style, which specifies the reversal of the graduation direction between two layers. 26. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the alteration of the scanning and graduation directions between two layers. 27. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the inversion of the scanning and graduation directions between two layers. 28. The apparatus according to claim 3, characterized in that the spout or distributor comprises at least one multi-jet ink jet distributing head. 29. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style, which specifies the formation of the object by means of frame scanning. 30. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the formation of the object by means of raster scanning having a length and width limited to an assortment region required by a layer that is formed. 31. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the formation of the object by means of vector scanning. 32. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the formation of an interior region of the object by means of frame scanning and a boundary region of the object by means of Vector exploration. 33. The apparatus according to claim 28, characterized in that the controller is configured to provide the construction style which specifies the randomization or reordering layer-by-layer of the jets which distribute over any XY site. 34. The apparatus according to claim 28, characterized in that the controller is configured to provide the construction style which specifies the printing using all the jets in a test configuration and detect from it the jets in which they are not firing properly. 35. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies that an object temperature must be maintained at a temperature greater than a minimum temperature to reduce the distortion by the corrugation. 36. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the formation of the different components of the object separately to allow the surfaces of the object to be reoriented as face surfaces. upwards during the construction of the parts and then combine the components formed separately. 37. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies supports of chess boards. 38. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies the use of a greater number of steps per layer to form supports than to form the object. 39. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies supports of active drip / idle drip chess boards. 40. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies in-line supports. 41. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies supports in a straight line. 42. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies curved line supports. 43. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies supports in dashed lines. 44. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style with shelves over at least part of a layer. 45. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies the movement of the supports on the layers above the shelf. 46. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies the shelves on less than 10 consecutive layers. 47. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support, which specifies shelves on less than 5 consecutive layers. 48. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies shelves over a whole first layer. 49. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies shelves in X-Y regions that are not on shelves in a previous layer. 50. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies complementary shelves on the subsequent layers. 51. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style, which specifies column supports. 52. The apparatus according to claim, characterized in that the controller is configured to provide the style of the support which specifies column supports with shelves. 53. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies displaced column supports. 54. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies supports of N-by-N columns defined in terms of the width of the drop. .55. The apparatus according to claim 4, characterized in that the controller is configured to provide the support style which specifies supports of N-by-N columns defined in terms of pixels. 56. The apparatus according to claim 55, characterized in that N is 2. 57. The apparatus according to claim 55, characterized in that N is 3. 58. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies the change of the configuration of the support in regions at least N layers below an underside surface. 59. The apparatus according to claim 58, characterized in that N is 4. 60. The apparatus according to claim 58, characterized in that N is 9. 61. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies the decrease in the ratio of the density of the drop in regions that approach a surface facing downwards. 62. The apparatus according to claim 61, characterized in that the style of the support which further specifies the use of at least one layer of shelves when making a transition of supports, of lower drop density ratio. 63. The apparatus according to claim 61, characterized in that the style of the support which further specifies the switching of column supports to checkerboard supports in regions when approaching a surface facing downward. 64. The apparatus according to claim 4, characterized in that the style of the support which further specifies the change of the support configuration in regions greater than a predetermined number of layers above an upward facing surface. 65. The apparatus according to claim 64, characterized in that the predetermined number is 4. 66. The apparatus according to claim 64, characterized in that the predetermined number is 9. 67. The apparatus according to claim 64, characterized because the style of the support which also specifies the decrease in the density ratio of the drop in regions after leaving a surface face up. 68. The apparatus according to claim 67, characterized in that the style of the support which further specifies the use of at least one layer of shelves in the transition from the higher density to the lower drop ratio supports. The apparatus according to claim 68, characterized in that the style of the support which further specifies the switching of the column supports to the chessboard supports after leaving a face-up surface. 70. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies arc supports. 71. The apparatus according to claim 4, characterized in that the controlled.r is configured to provide the style of the support which specifies supports by air pressure. 72. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies the differentiation of the support configuration in different areas of a cross section. 73. The apparatus in accordance with the claim, characterized in that the controller is configured to provide the style of the support which specifies the displacement of the support structure from an object boundary by a first "" number of predetermined pixels in the scanning direction and a second predetermined number of pixels in the direction of graduation. 74. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies the construction or integration of supports with a material different from that used of the surfaces and border regions of the object. 75. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies the bulk assortment of supports on each layer after the material used to form the object is selectively dispensed or dispensed. 76. The apparatus according to claim 4, characterized in that the controller is configured to provide the style of the support which specifies the use of a water soluble material to construct or integrate the supports. 77. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the construction or integration at a uniform temperature. 78. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction style which specifies the shooting of the subpixel. 79. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction or integration style which specifies the firing of the subframe line. 80. The apparatus according to claim 3, characterized in that the controller is configured to provide the construction or integration style which specifies the use of a material with a black body radiator added. 81. The apparatus according to claim 78, characterized in that the construction or integration style further specifies the formation of subpixels by means of displaced flight time data. 82. The apparatus according to claim 4, characterized in that the style of the support specifies larger drops for the supports than for the object. 83. The apparatus according to claim 3, characterized in that the construction or integration style also specifies a material having interlaced thermal conductors inserted. 84. The apparatus according to claim 3, characterized in that the construction or integration style further specifies object-sensitive interleaving. 85. The apparatus according to claim 3, characterized in that the construction or integration style further specifies the compensation of the drop width. 86. The apparatus according to claim 3, characterized in that the construction or integration style also specifies the compensation of the width by overwriting. 87. The apparatus according to claim 3, characterized in that the controller is configured to provide a construction style which specifies a spacing between the orifice of the spout or distributor and the work surface which is large enough such that the drops form hemispherical drops on the impact. 88. The apparatus according to claim 4, characterized in that the style of the support specifies delaying the construction or integration of support behind the construction or integration of the part by at least one layer to avoid the distortion of the supports caused by the Flattened 89. The apparatus according to claim 3, characterized in that the construction or integration style specifies the flattening by means of only the fusion. 90. The apparatus according to claim 3, characterized in that the construction or integration style specifies flattening by means of fusion in combination with scraping. 91. The apparatus according to claim 3, characterized in that the construction or integration style specifies the flattening by means of fusion in combination with scraping and rotation. 92. A method for rapid scale modeling, characterized in that it comprises the steps of: dispensing or controllably distributing a material that can flow, which is solidifiable after it is dispensed or distributed; supporting a cross section of a three-dimensional object and providing a work surface for integrating or constructing a next cross-section of the object; relatively displace the spout or distributor and the work surface in at least two dimensions, comprising a scanning direction and a graduation direction; and stocking or distributing the material on the work surface according to a selected style. 93. The method according to claim 92, characterized in that it also comprises selectively dispensing or distributing the material on the work surface according to the selected style. 94. The method according to claim 93, characterized in that the style is a style of construction or integration. 95. The method according to claim 93, characterized in that the style is a style of the support. 96. The apparatus according to claim 1, characterized in that it further comprises means for forming support structures which branch out beyond the material stocked onto an immediately preceding sheet, wherein the branching results in more support structures being laid in contact with a surface of the object facing downwards the number of structures from which the branching starts. 97. The method according to claim 92, characterized in that it additionally comprises the step of forming support structures which branch outwardly beyond the material stocked onto a immediately preceding sheet, wherein the branching results in further structures of support that are brought into contact with an object surface facing downwards that the support structure number from which the branching starts. 98. The apparatus according to claim 1, characterized in that it additionally comprises means directing the drops to a focal plane below a level of a real work surface in order to obtain a self-correcting accumulation in a direction perpendicular to a plane of the cross sections. 99. The method according to claim 92, characterized in that it additionally comprises the step of directing the drops in a focal plane below a level of a real work surface to obtain a self-correcting accumulation in a direction perpendicular to a plane of the cross sections. 100. The apparatus according to claim 1, characterized in that it additionally comprises means for directing a cooling gas on the surface of a partially formed object and means for withdrawing the cooling gas from the area above the surface. 101. The method according to claim 92, characterized in that it additionally comprises the step of directing a cooling gas on the surface of a partially formed object and the step of separating the cooling gas from the area above the surface.