CN102279898A - Material having expected characteristics designed and manufactured with base material having specific characteristics - Google Patents

Abstract

The invention provides a material having expected characteristics and designed and manufactured with base material having specific characteristics. Usable basic materials can be showed in a object generating system. The basic materials having the characteristics can be used in the object generating system. Through a computer or a computation device, user input of the expected object and a set of characteristics of the expected object are processed, so as to obtain a reflection of positions of the basic materials. Thus, after the reflection is provided to an object generating device, the generated object is similar to the expected object and the user input of the set of characteristics. The characteristics of the basic material can include the elasticity of the basic material and the user input can be an expected shape and the elasticity. The object generating device can be a 3D multi-material printer and can be made of the multi-materials used in the printer. The generated object can be at lease similar to the expected shape and the expected elasticity.

Description

Design and make material by basic material with desired characteristic with definite characteristic

The cross reference of related application

This non-provisional application requires to enjoy the exercise question of submitting on February 24th, 2010 and is No. 61/307833 U.S. Provisional Application No. under " design and manufacturing with material of expectation deformational behavior " 35USC § 119 (e), quotes for any destination at this to comprise its full content.

Technical field

The present invention relates generally to a kind of object manufacturing system, and especially, relates to the object manufacturing of using multiple basic material, and the equipment that can place basic material according to the order of computing machine or electronic equipment.

Background technology

Many objects have the attribute that depends on the material of constructing it.For example, marble statue is hard, and the feather pillow is soft.Usually, the attribute of object is assumed that to be fixed, and by its structure decision.For example, the color of expection a loaf of bread is white, brown or brown.Its shape is also determined by method for making usually---bread normally circular or strip, the outside is hard and the inside is soft, this all is owing to contains the needs of plate of bread and the baking oven heating outside surface fact more than heating inside.

Known already in area of computer graphics, computer model and image composer can make the user break away from physical constraint---be easy to generate the video of the animal of speaking, floating bread, impossible physical phenomenon etc.Therefore, by suitable instruction is provided, animation system can be exported the video sequence of being created by instruction fully, and therefore is not subjected to the constraint of physique method.

Computer graphics is done film, and the 3D printer has also been done object.Though some constraints are arranged, might send the object of a set of computer instructions to the 3D printer, even print and use some objects that to make such as the object manufacturing technology of craft such as casting, engraving, milling with the printing arbitrary shape.

The 3D printer has been used to generate model machine, model and antique, but it is restricted to, although the net shape of object can at random be specified by computing machine (or using to the user of the computer interface of 3D printer), but object will be uniformly, because it is to be made by the employed any output material of specific printer (resin, plastics etc.).Certainly, do not exist completely freely, because if there are not enough supports, some instructions may cause the caving in of part of object.

Some objects are flexibly deformable, and they are that as the part that induces one of object they have those attributes.For example, the Association football of elastic deformation is not attractive itself with regard to Association football.Other examples comprise clothes and shoes, furniture, plant or even tissue or animal tissue.At the animation environment with in the real world objects manufacturing environment, the correct object characteristic of acquisition, definite properties of materials and other characteristics that is adopted are understood in important being.In animation, problem is easier to, because the behavior of dummy object does not need to meet physical law, yet the behavior of physical objects need meet physical law, and its behavior is subjected to very much, and its structure restriction---virtual bounder can be made by stone, but the physical objects ball of being made by stone can not bounce fully.

Anamorphic effect can be modeled with very various yardstick of the response function from interaction of molecules to overall situation support, and is modeled [Zohdi and Wriggers 2004] by continuous elastic method or lumped parameter model.By studies show that of numerical value alligatoring (coarsen) and homogenising, can be similar to the behavior [Kharevych et al.2009] of heteropical material as recently by the homogeneous material of mesoscale with microscale.

In some 3D printers, there are a kind of Available Material (rubber, plastics, resin etc.) and output object 100% to make by the sort of Available Material.Yet nearest development comprises for example OBJECT of many materials 3D printer ^TMConnex ^TMSeries 3D printer.These print west can " print " object, and this object comprises assembly of being made by first material and the assembly of being made by second material, and wherein two kinds of (or more) materials are in some parameters aspect differences such as density, elasticity, hardness for example.

Area of computer graphics has been contributed the system that is used for design and makes virtual costume [Okabe et al.1992], lint object [Mori and Igarashi 2007], paper skill object [Mitani andSuzuki 2004] or surface microscopic geography [Weyrich et at.2009].

What need is to create deformable object (or the object with other characteristic groups) to mate the characteristic of some desired user's appointment by computer instruction.Certainly, if desired characteristic is the characteristic that is provided by Available Material, problem is exactly simple, but the user has more complicated demand usually.

In the computer graphical background, known use finite element model emulation parenchima is as the data-driven modeling of the virtual material of deformable.The nearest research of distorted pattern in computer graphics shown in [Nealen 2006].

Be used for exactly the common scheme of modeling deformable material in the computer graphical and use continuous elastic method and finite element model.This scheme can the large-scale material of modeling, comprises the material with non-linear inhomogeneous deformation behavior.Usually, must selection can the cladding material behavior this structure (constructive) material model [Ogden 1997] of scope, then, under the situation of given object and constitutive model, adjust material parameter with the match empirical data.This scheme is introduced into biology-mechanical modeling in the computer graphics by [Terzopoulos et al.1987], and is applied to body part such as face [Koch et al.1996 subsequently; Magnenat-Thalmannet al.2002; Terzopoulus and Waters 1993; Sifakis et al.2005], hand [Suedaet al.2008], neck [Lee and Terzopoulos 2006], trunk [Zordan et al.2004; Teran et al.2005; DiLorenzo et al.2008] or the whole upper part of the body [(Lee et al.2009], the emulation of the effect that is applied to rupture [O ' Brien and Hodgins 1999]; Or even the interactive simulation after the model reduction step [ And James 2005].

In order to reach high fidelity, the elastic method dependence comprises the complex process of how much accurate modeling and parameter fine setting continuously.

Some researchists have designed the method for discerning the parameter of constitutive model according to the measurement of real-world object automatically.The estimation that these have covered parameter based on the modeling scheme of measuring is such as Young modulus and Poisson ratio [Bickel et al.2009], plasticity parameter [Kajberg and Lindkvist 2004] or the nonlinear viscoelasticity parameter [Kauer et al.2002] of Young modulus [Schnur and Zabaras 1992], Young modulus and Poisson ratio [(Becker andTeschner 2007], non-linear interpolation.

The function that the direct match overall situation of other the modeling scheme based on measuring is supported is described as material, rather than estimates local parameter.Pai et al.[2001] introduced the system of the shape that is used in united frame gathering object, elasticity, roughness features.They adopt Green function matrix representation [James and Pai 1999] to describe distorted pattern.After this, other people have expanded Pai and other people research, to increase robustness and to handle viscoelastic effect [Lang et al.2002; Schoner et al.2004].

The research in computer picture recently is intended to, even if the resolution of discretize is very poor, and the also high-resolution heterogeneous body of modeling [Kharevych et al.2009; Nesme et al.2009].Be known as this process of homogenising, attempt to find out parameter value, so that homogeneous material is mated in the behavior of object best with the constitutive model of low resolution sampling.Yet, only create that visual object is the one thing on screen, creating physical objects is a different matter.

List of references

[?

and?James?2005]BARBI?

J.，AND?JAMES，D.2005.Real-time?subspace?integrationfor?St.Venant-Kirchhoff?deformable?models.ACM?Trans.Graph.24，3(Aug.)，982-990.

[Bathe?1995]BATHE，K.J.1995.Finite?Element?Procedures.Prentice?Hall.

[Becker?and?Teschner?2007]BECKER，M.，AND?TESCHNER，M.2007.Robust?and?efficient?estimation?of?elasticity?parameters?using?thelinear?finite?element?method.In?Sim?Vis，15-28.

[Bickel?et?al.2009]BICKEL，B.，B?

CHER，M.，OTADUY，M.A.，MATUSIK，W.，PFISTER，H.，AND?GROSS，M.2009.Capture?andmodeling?of?non-linear?heterogeneous?soft?tissue.ACM?Trans.Graph.28，3(July)，89：1-89：9.

[DiLorenzo?et?al.2008]DILORENZO，P.c.，ZORDAN，V.B.，ANDSANDERS，B.L.2008.Laughing?out?loud：Control?for?modelinganatomically?inspired?laughter?using?audio.ACM?Trans.Graph.27，5(Dec.)，125：1-125：8.

[James?and?Pai?1999]JAMES，D.L.，AND?PAI，D.K.1999.Artdefo-accurate?real?time?deformable?objects.In?Proc.of?SIGGRAPH99，Computer?Graphics?Proceedings，65-72.

[Kajberg?and?Lindkvist?2004]KAJBERG，J.，AND?LINDKVIST，G.2004.Characterisation?of?materials?subjected?to?large?strains?by?inversemodellng?based?on?in-plane?displacement?fields.Inter?national?Journal?ofSolids?and?Structures?41，13，3439-3459.

[Kauer?et?al.2002]KAUER，M.，VUSKOVIC，V.，DUAL，J.，SZEKELY，G.，AND?BAJKA，M.2002.Inverse?finite?elementcharacterization?of?soft?tissues.Medical?lmage?Analysis?6，3，257-287.

[Kharevych?et?al.2009]KHAREVYCH，L，MULLEN，P.，OWHADI，H.，AND?DESBRUN，M.2009.Numerical?coarsening?of?inhomogeneouselastic?materials.ACM?Trans.Graph.28，3(July)，51：I-51：8.

[Koch?et?al.1996]KOCH，R.M.，GROSS，M.H.，CARLS，F.R.，VON?BUREN，D.F.，FANKHAUSER，G.，AND?PARISH，Y.1996.Simulating?facial?surgery?using?finite?element?methods.In?Proc.ofSIGGRAPH?96，Computer?Graphics?Proc.，421-428.

[Land?and?Doig?1960]LAND，A.H.，AND?DOIG，A.G.1960.An?automatic?method?of?solving?discrete?programming?problems.Econometrica?28，3，497-520.

[Lang?et?al.2002]LANG，J.，PAI，D.K.，AND?WOODHAM，R.J.2002.Acquisition?of?elastic?models?for?interactive?simulation.InternationalJournal?of?Robotics?Research?21，8，713-733.

[Lee?and?Terzopoulos?2006]LEE，S.-H.，AND?TERZOPOULOS，D.2006.Heads?up！：biomechanical?modeling?and?neuromuscular?control?of?theneck.ACM?Trans.Graph?25，3(July)，1188-1198.

[Lee?et?al.2009]LEE，S.-H.，SIFAKIS，E.，ANDTERZOPOULOS，D.2009.Comprehensive?biomechanical?modeling?andsimulation?of?the?upper?body.ACM?Trans.Graph.28，4(Aug.)，?99：1-99：17.

[Lund?and?Stegmann?2005]LUND，E.，AND?STEGMANN，J.2005.On?structural?optimization?of?composite?shell?structures?using?a?discreteconstitutive?parameterization.Wind?Energy?8，109-124.

[Magnenat-Thalmann?et?al.2002]MAGNENAT-THALMANN，N.，KALRA，P.，L?

QUE，1.L.，BAZIN，R.，BATISSE，D.，AND?QUELEUX，B.2002.A?computational?skin?model：fold?and?wrinkle?formation.IEEETrans.on?Information?Technology?in?Biomedicine?6，4，317-323.

[Mikolajczyk?and?Schmid?2004]MIKOLAJCZYK，K.，ANDSCHMID，C.2004.Scale?&?affine?invariant?interest?point?detectors.International?Journal?of?Computer?Vision?60，1，63-86.

[Mitani?and?Suzuki?2004]MITANI，J.，AND?SUZUKI，H.2004.Making?papercraft?toys?from?meshes?using?strip-based?approximateunfolding.ACM?Trans.Graph.23，3(Aug.)，259-263.

[Mori?and?Igarashi?2007]MORI，Y.，AND?IGARASHI，T.2007.Plushie：An?interactive?design?system?for?plush?toys.ACM?Trans.Graph.26，3(July)，45：1-45：8.

[Müller?and?Gross?2004]M? LLER，M.，AND?GROSS，M.H.2004.Interactive?virtual?materials.In?Graphics?Interface?2OO4，239-246.

[Nealen?2006]NEALEN，A.，MLLER，M.，KEISER，R.，BOXERMAN，E.，AND?CARLSON，M.2006.Physically?based?deformablemodels?in?computer?graphics.Computer?Graphics?Forum?25，4(Dec.)，809-836.

[Nesme?et?al.2009]NESME，M.，KRY，P.G.，? ，L.，AND?FAURE，F.2009.Preserving?topology?and?elasticity?for?embeddeddeformable?models.ACM?Trans.Graph.28，3(July)，52：1-52：9.

[Neumaier?and?Pownuk?2007]NEUMAIER，A.，AND?POWNUK，A.2007.Linear?systems?with?large?uncertainties?with?applications?to?truss?structures.Reliable?Computing?13，149-172.

[OBJET]OBJET.Connex500?Multi-Material?3D?Printing?System.http://ww.objet.com/3D-Printer/Connex5001.

[O’Brien?and?Hodgins?1999]O’BRIEN，J.F.，AND?HODGINS，1.K.1999.Graphical?modeling?and?animation?of?brittle?fracture.In?Proc.ofSIGGRAPH99，Computer?Graphics?Proc.，137-146.

[Ogden?1997]OGDEN，R.W.1997.Non-Linear?ElasticDeformations.Courier?Dover?Publications.

[Okabe?et?al.1992]OKABE，H.，IMAOKA，H.，TOMIHA，T.，ANDNIWAYA，H.1992.Three?dimensional?apparel?cad?system.In?ComputerGraphics(Proc.of?SIGGRAPH?92)，105-110.

[Pai?et?al.2001]PAI，D.K.，VAN?DEN?DOEL，K.，JAMES，D.L.，LANG，J.，LLOYD，J.E.，RICHMOND，J.L，AND?YAU，S.H.2001.Scanning?physical?interaction?behavior?of?3d?objects.In?Proc.of?ACMSIGGRAPH?20O1，Computer?Graphics?Proc.，87-96.

[Rebonato?and?

1999]REBONATO，R.，AND?

CKEL，P.1999.The?most?general?methodology?to?create?a?valid?correlation?matrix?for?riskmanagement?and?option?pricing?purposes.Tech.rep.，Quantitative?ResearchCentre，NatWest?Group.

[Schnur?and?Zabaras?1992]SCHNUR，D.S.，AND?ZABARAS，N.1992.An?inverse?method?for?determining?elastic?material?properties?and?amaterial?interface.International?Journal?for?Numerical?Methods?inEngineering?33，10，2039-2057.

[Schoner?et?al.2004]SCHONER，J.L.，LANG，J.，AND?SEIDEL，H.-P.2004.Measurement-based?interactive?simulation?of?viscoelastic?sollids.Computer?Graphics?Forum?23，3(Sept.)，547-556.

[Sifakis?et?al.2005]SIFAKIS，E.，NEVEROV，I.，AND?FEDKIW，R.2005.Automatic?determination?of?facial?muscle?activations?from?sparsemotion?capture?marker?data.ACM?Trans.Graph?24，3(Aug.)，417-425.

[Sueda?et?al.2008]SUEDA，S.，KAUFMAN，A.，AND?PAI，D.K.2008.Musculotendon?simulation?for?hand?animation.ACM?Trans.Graph.27，3(Aug.)，83：1-83：8.

[Svoboda?et?al.2005]SVOBODA，T.，MARTINEC，D.，ANDPAJDLA，T.2005.A?convenient?multi-camera?self-calibration?for?virtualenvironments.PRESENCE：Teleoperators?and?Virtual?Environments?14，4(August)，407-422.

[Teran?et?al.2005]TERAN，J.，SIFAKIS，E.，BLEMKER，S.S.，NG-THOW-HING，V.，LAU，C.，AND?FEDKIW，R.2005.Creating?andsimulating?skeletal?muscle?from?the?visible?human?data?set.IEEE?Trans.on?Visualization?and?Computer?Graphics?11，3(May/June)，317-328.

[Terzopoulos?et?al.1987]TERZOPOULOS，D.，PLATT，J.，BARR，A.，AND?FLEISCHER，K.1987.Elastically?deformable?models.InComputer?Graphics?(Proc.of?SIGGRAPH?87)，205-214.

[Terzopoulus?and?Waters?1993]TERZOPOULUS，D.，ANDWATERS，K.1993.Analysis?and?synthesis?of?facial?image?sequences?usingphysical?and?anatomical?models.IEEE?Trans.on?Pattern?Analysis?andMachine?Intellgence(PAMI)14，569-579.

[Zohdi?and?Wriggers?2004]ZOHDI，T.I.，AND?WR1GGERS，P.2004.Introduction?to?Computational?Micromechanics.Springer-Verlag?New?York，Inc.

[Zordan?et?al.2004]ZORDAN，V.B.，CELLY，B.，CHIU，B.，ANDDILORENZO，P.C.2004.Breathe?easy：model?and?control?of?simulatedrespiration?for?animation.In?2OO4?ACM?SIGGRAPH/EurographicsSymposium?on?Computer?Animation，29-37.

Summary of the invention

In the embodiment of object generation system and/or its ingredient, available basic material is characterized in characterization, and wherein the object generation system can use multiple Foundation of Representation material like this.New characterization has been described.

Use new characterization, and/or traditional characterization, use a computer or calculation element, handle user's input of one group of desired characteristic of the desired object of expression and this desired object in new ways, mapping with the position of the placement of the part that derives multiple basic material, so that when mapping when being provided for the object maker, the object of generation is similar to user's input of the desired object of described expression and desired one group of characteristic.

In a certain embodiments, the characteristic of basic material comprises the elasticity of basic material, user input shows the desired shape of object and all or part of desired elasticity of this object, the object maker is many file printings of 3D machine, and the object that is generated is owing to formed and approximate at least desired shape and elasticity by the employed multiple basic material structure of many file printings of 3D machine.

Here the method and apparatus of Miao Shuing can be the part of many file printings of 3D machine or may only be used to generate the instruction or the prescription to be used to generate physical objects with one group of desired characteristic.The desired one group of characteristic of object to be generated and the characteristic of basic material may be included in one or more in density, elasticity, heat conduction, thermal capacity, hardness, molecular behavior and/or the electromagnetic radiation response.Prescription may be the form of understanding with one or more modules that can be easy to import and be easy to many file printings of 3D machine, for example is used for the different locational instruction that different points (blob) or layer (layer) with different basic materials are placed on the operated 3d space of many file printings of 3D machine.

The position of calculating multiple basic material may comprise that the search expression has the search volume of combination of the basic material of different qualities, and the position in definite search volume, certain standard of this position optimization, for example threshold value of the mean difference of characteristic on object or the body surface.Search may be an iteration, and may comprise and deleting to simplify search procedure.Delete and to comprise and deleting so that in the zone of the zone of " impossible " of search volume the space segment of the material that uses the negative amount (for example, corresponding to) or the operation that need can not realize, not search for by specific printing mechanism.

In other respects, the embodiment according to the characterization system of many aspects of the present invention can be used to characterize basic material.When all changes of Fundamentals of Measurement material actual time not, can measure according to other of the subclass of material and not the carrying out of from the measurement of having carried out, inferring with those different attributes, characterize different attributes.For example, measure in the time of can being T1, T2, T3 at the thickness of material M, and when thickness T 4 and T5, do not measure, but calculate or measurement when inferring at thickness T 4 and T5 from the measurement when thickness T 1, T2 and/or the T3.

The operation that is used for determining prescription or is used to construct desired physical objects can be implemented by hardware and/or software, and function described here may be implemented with the computer-readable instruction that is stored on the tangible medium.

Detailed description and accompanying drawing will provide the better understanding to performance of the present invention and advantage below.

Description of drawings

Fig. 1 shows the example of real-world object and duplicate.

Fig. 2 shows the example of the step in the process that designs the deformable material with desired deformational behavior.

Fig. 3 shows the example of the branch and bound method with cluster.

Fig. 4 shows the assembly of measuring system of the deformational behavior of the deformational behavior of the deformational behavior that may be used to measure object to be duplicated (target object), multiple basic material to be used and/or the object that checking is printed.

Fig. 5 shows the figure of force-displacement curve of the subclass of measured material, and this figure shows its non-linear behavior.

Fig. 6 shows the basic material under 15 Newton force.

Fig. 7 shows the comparison side by side of the material of real and emulation.

Fig. 8 shows the result of the affirmation of compound substance.

Fig. 9 shows the figure of affirmation of the algorithm for design of based target.

Figure 10 shows manufacturing example and the spatial array that is obtained and the error evaluation of based target design.

Figure 11 shows the force-displacement curve of the corresponding point on the duplicate and the original.

Figure 12 shows the error amount table of various models.

Figure 13 shows transversely isotropic basic material of being made by the hollow tubular of different size and distribution.

Figure 14 shows the parameter of skin stool, and the force-displacement curve of the corresponding point on the duplicate and the original.

Figure 15 shows the example hardware system that may be used to computing system and/or be used to implement technology as described herein.

Embodiment

The various embodiment of new departure have been described, this new departure be used for measuring current material characteristic, be used to measure the non-existing but properties of materials that can infer from the measurement of the characteristic of current material and determine one group of object and generate instruction for given one group of desired object characteristic.Measurable characteristic of material also may comprise the measurement of appearance characteristic such as color, transparency, quality, reflectivity, refractivity or the like.For example, as shown in Figure 1, the folder flip-flop has multiple color and reflectivity, for example, and the black bright surface.Other measurable characteristics may comprise taste, smell and/or the quality of object.For example apple is flavoursome and smell.

Use these new processes and/or device and/or method, can be by determining that material is described and parameter design object, so that object looks like desired mode and shows in desired mode.As in this explanation, by make a plurality of different compound substances, deformation when measuring it and being subjected to various power and these measurement results are compared with simulation result, implement test to confirm (Validation) emulation and material model.After the explanation of the operation of each assembly, these simulation results can be described.

Fig. 1 shows two examples of real and the object that duplicates, and wherein the object that is duplicated can be used to measure by use, emulation and the data-driven process of combination that obtains the nonlinear foundation material of the desired deformational behavior of coupling generate; Go out as shown, those objects can be by many materials 3D printer " printing ", and this many material 3D printer uses two kinds of materials with different internal microstructure and character of surface.Fig. 1 (a) is the image of a real physical objects, here is the rubber sandals.Fig. 1 (b) is the data representation of rubber sandals, is actually the model of the deformational behavior of rubber sandals.Fig. 1 (c) is corresponding object, and its 3D printer (or equivalent apparatus) that is had the multiple material of different internal microstructure by use generates, so that the deformational behavior of the approximate former rubber sandals of the deformational behavior of the sandals of being printed.Similarly, Fig. 1 (d) is the image of slippers, and Fig. 1 (e) is the data representation of these slippers, and Fig. 1 (f) is corresponding object, it generates by having different internal microstructure, thus the deformational behavior of the approximate former slippers of the deformational behavior of the slippers of being printed.In addition, can select this two kinds of materials, so that desired " outward appearance and the impression " of object to be provided.For example, the slippers in Fig. 1 (f) are formed by the material of black, and this material has the corresponding bright appearance that is caused by the light absorption of material and scattering power.In other embodiment, for the object such as the apple of printing, material can be configured to provide the smell that has as apple to apple, maybe can provide different smell or tastes, for example the aroma and flavor of strawberry.

Details as Follows, the system that is used to create these objects can comprise the measuring system of the characteristic of for example distortion that is used to measure object to be duplicated, ductility, elasticity, smell, taste etc., and can be used to measure the surface properties such as reflectivity, color, surface texturisation, transparency.For example, be used to measure the measuring system of the deformational behavior of object to be duplicated, (this can be the system identical with the system of Measuring Object to be used to measure the measuring system of multiple distortion of materials behavior to be used, and/or may be provided in advance) about the part or all of information of material, comprise and be used for implement calculating (here with other local descriptions) pattern (pattern) with the multiple material of determining to be used for to duplicate (or make be not the thing of having a mind to duplicate) measured object, the result's of the control of the details of perhaps measured object the hardware and/or the computing system of software, and 3D printer or equivalent apparatus, described 3D printer or equivalent apparatus are connected to computing system so that can provide about what material to the instruction of where going (or which what material moves to) to the 3D printer.

Therefore, be in operation, the deformational behavior of user's (may be terminal user, deviser, artist or the like) acquisition target object, appearance, smell, taste etc., and obtain the deformational behavior of expression object and the data set of other characteristics.User's (people or computing machine) can revise data set according to expectation then, such as so that to make not be the object that is intended to accurately duplicate the original, but has the object of the characteristic different with the original.Then, this data set can be output to printer, to generate the object that is generated.Can use a computer or electronics or some steps of mechanical system robotization.

Can be by approximate behavior, appearance, the composition of mesoscale homogeneous material with heteropical material of microscale, but interesting here one side is this process and can be regarded as the inverse process of homogenising, that is, reaching desired large scale behavior than the suitable combination on the small scale by material.This can comprise the design of the based target of deformable model, and solves the difficult problem of anisotropy, non-homogeneous and non-linear behavior.This can exceed the computer animation of such object, and feasible physics with deformable bodies of desired vision, sense of touch, sense of smell, the sense of taste and/or behavior creates possibility.

Use the technology of describing here, the process and the expression that are used for the deformable bodies that nonuniform line that alligatoring has microscale is are provided, it uses the data-driven scheme and as the expansion of linear numerical alligatoring to the nonlinear material behavior, it can be gathered non-linear behavior and keep efficient working time simultaneously.The material design proposal of based target also is provided, and this scheme can be used in computing machine, 3D printer or other systems based on electronics, and this scheme is similar to desired mesoscale deformation characteristic by the microscale material via Combinatorial Optimization.This is to have reversed the reverse modeling scheme of notion of homogenising.The result of such design process can be as the input of many file printings of 3D machine, to be used for the actual physical manufacturing of deformable bodies.

As summing up among Fig. 2, the scheme that is used to measure, design and make the material with desired deformational behavior here has a plurality of visibly different steps.As this illustrate, the basic material that printer/maker must be used is measured and characterized.This can be done in advance and store in the storer of electronically readable as material depot.Then, measure object to be duplicated, and pass through the iterative search and the simulation process in materials used storehouse, obtain separating of some phantom types that can be sent to printer/maker.Then, by various processes, for example manual evaluation or machine assessment can contrast the original and confirm the result.Affirmation can be the part of bigger backfeed loop.

In the result that shown search is conciliate, the phantom type comprises the layer of different basic materials.Do not need to limit search so that the cross section is identical on whole phantom type, the side line that is parallel to the object surfaces of being exported may run into different basic materials.This may be to need such variation so that the part of the feasible object of being exported feels to be different from other parts of the object of being exported, but also may be used to find out than (for example better separating under the situation that has constraint, more approaching phantom type), described constraint refers to separating of the fixedly number of plies that only allows to have basic material.

In test, considered multiple basic material.By printer for example being set to print the cube of the handled basic material of only a kind of printer, some basic materials can be shaped as the output object of printer.Exemplary printer is Connex material more than 500 3D printer.Except with the output of printer as basic material, can also obtain other basic materials, for example various standard foam that are easy to obtain, gelinite and rubber, its cross over different distortion from very soft to stone and wide region rigidity.

As being the part that given target object calculates phantom type to be generated, coarsening process be used to use Finite Element Method (FEM) data-driven, nonlinear strain-stress relation comes the modeling basic material.This simplifies expression makes computing system to determine what kind of the distortion thicker or thin version of basic material sampling may be according to the distortion of basic material sampling.This makes and need not to carry out the distortion that the combination in any of the basic material that piles up just can be accurately predicted in too much measurement.

But simplify the expression time spent when this, computing system can design the compound substance that mates desired deformational behavior best, and described desired deformational behavior can be drawn by the expectation of the virtual variation of mating target object or coupling target object.For example, target may be the phantom type that mates the target object of being measured as described above under the situation of given one group of basic material exactly, perhaps target may be the phantom type of output object, this output object what position in office all be the twice of target object big and be 1.5 times of target object firmly.

In order to simplify the material design process, in certain embodiments, computing system can be programmed to follow the optimization of based target.In such optimization, the user specifies desired material by exemplary distortion and its corresponding power is provided, and the compound substance of the automatic calculating optimum coupling of computing system.Because configuration space be combination and the index rank big, as a rule, computing system can be configured to operation truncation search procedure, it deletes the state that desired material specification is produced the coupling of difference.

Fig. 3 shows the example of the branch and bound method with cluster.Here, tree root shows the two kinds of materials A and the B in the first bag chamber in three bags chambeies (cell).Each layer of tree comprises the possible option in subsequent packet chamber.The branch of the suboptimum of tree can be picked out, and similar distortion can be by cluster.

Common use meeting is more complicated, but in this simple example, has three layers (1,2 and 3) and two kinds of basic materials to use.Though gross space have eight kinds may, in this example, determine: if on two-layer is material B, then desired result is provided is impossible in any combination of Xia Mian extra play.Therefore, " BB* " branch of search tree is picked out.In addition, can generally obtain under the situation of identical effect at definite " AB* " and " BA* " structure, these two branches can be by " cluster ", and therefore has only a branch to need to continue.In this example, have only four branches need continue rather than eight.

Existence is automatically discerned traditional technology of the parameter of constitutive model by the measurement of real-world object, and these traditional technology can be used or be used at this.By type of material is restricted to uniformly, the model that computing system adopted needs the degree of freedom of much less.The homogeneity of basic material is not a restriction, because the phantom type can be by making up the heterogeneity in the object that various homogeneous materials represent finally to be exported.If desired, can be by simultaneously single nonlinear model being fitted to the robustness that whole input examples increases fit procedure.In some cases, may not need these features.

The first step of design and manufacture process can be regarded as the modification of homogenising, and wherein the heterogeneous material of accurate yardstick is the actual physical material.Yet, in second step, take the scheme opposite with homogenising, generate heterogeneous object of the rough power-deformation data of match by small scale material with known behavior.

Measuring system

Fig. 4 shows the assembly of a measuring system, and this measuring system may be used to measure the characteristic of object to be duplicated (target object), the deformational behavior of multiple basic material for example to be used, and/or be used to verify the deformational behavior of the object of being printed.

Figure 4 illustrates the system of the robotization that is used to measure material deformation, comprise that robotic arm 402, power-torque sensor 404 and vision subsystem 406 are with the surface displacement on the tracking and testing material 410.In specific embodiment, robotic arm 402 may be the arm of 4 degree of freedom being provided by little prototyping system (MicroProto System), power-torque sensor 404 may be the 25 6 power-torque sensors of nanometer that provided by ATI, and vision subsystem 406 may comprise that seven cameras are with the tracked surface displacement.For the accessorial visual collection, basic material may be painted or mark on locate pattern (pattern).In specific embodiment, the resolution of robotic arm 402 is 0.003mm and its repeatable 0.01mm of being, the maximum magnitude of power-torque sensor 404 is 125N and the resolution with 1/24N, and vision subsystem 406 comprises seven high-resolution Basels navigation (Basler Pilot) cameras, and it is with the resolution work of 1600x1200 pixel.Adopt the instruction of [Svoboda et al.2005], camera may be oriented on the semicircle of robotic arm top, covers (occlusions) to minimize and uses additional scattered light.

The nonlinear material model

In certain embodiments, basic material may demonstrate as by shown nonlinear, the hyperelastic stress-strain behavior of the measured force-displacement curve among Fig. 5.Most of adopt basic material on the microscale aspect, comprise complicated structure, it influences deformational behavior significantly.As shown in Figure 6, deformational behavior can change, and Fig. 6 shows the different material under fixing power (15 newton's power).Can use other power, but this example shows variability well.

In the scheme of data-driving, use the non-linear interpolation of the local linear material properties that the exemplary variations by the basic material of mesoscale draws to gather non-linear behavior by measuring system.Conceptive, this can be regarded as the alligatoring algorithm of data-driven, the deformational behavior that expression is caused by the microstructure on the mesoscale aspect.

In order to realize this, measuring system is sampled to the stress-strain function of material under one group of little strain value, and represent each stress-strain point by linear combined rotational model (co-rotaional model), as shown in [Bickel et al.2009].Under the deformation state of the set point in having the material of any strain value, in data structure, material is represented by the non-linear interpolation of the linear model of being sampled partly.

The linear FEM here can use Cauchy's linear strain tensor.Given displacement fields u, tensor can be defined as shown in the equation 1.

$\mathrm{\ϵ}\left(u\right)=1/2(\▿u+{(\▿u)}^T)$ (equation 1)

By measuring system by the utmost point decompose to extract deformation gradient rotating part, and subsequently the conversion stiffness matrix, partly use in the content shown in [M ü ller and Gross 2004], the unchangeability of the strain under rotating can obtain.

Because the symmetry of strain and stress tensor, they can both be represented as 6 dimensional vectors in data structure.Given strain tensor, 6 dimensional vectors of its equivalence can be constructed to shown in equation 2, and are as the same for stress.

ε=(ε _Xxε _Yyε _Zzε _Xyε _Xzε _Yz) ^T(equation 2)

Subsequently, linear material can be represented as 6 * 6 matrix, E, and it makes that the stress and strain vector is interrelated, shown in equation 3.

σ (u)=E ε (u) (equation 3)

The non-linear behavior of basic material can be represented by the function that strain-stress relation E is defined as local train ε (u).Matrix E usually can be by one group of less parameter p parametrization.With such parameter-definition is the nonlinear function of strain, thereby matrix E can be represented as nonlinear function E (p (u)).Although this scheme is applicable to general anisotropy behavior, will be described in the parameter p of two types in fact very relevant material below: the isotropy and the parameter p of isotropic material transversely.Transversely isotropic material is described below.

The parametrization of isotropic material

For isotropic material of uniform linearity, matrix E can be represented by two Lame parameter lambda and μ.In other words, parameter vector be p=(λ, μ).Use the Lame parameter, the stress and strain tensor is interrelated shown in equation 4, and E can be derived by this equation matrix, partly may use in the content shown in [Bathe 1995].

σ (u)=2 μ ε (u)+λ tr (ε (u)) I (equation 4)

In isotropic material of linearity, strain can be by three constant I of the strain tensor of symmetry ₁(ε), I ₂(ε), I ₃(ε) gather well.As equation 5 expressions, use constant to represent strain, nonlinear model can be counted as the function of 2 values in the three-dimensional domain.

P (I ₁, I ₂, I ₃): IR ³→ IR ²(equation 5)

The parametrization of nonlinear material attribute

Given basic material, computing system can use by one group of little P corresponding to different strain value { ε _iParameter vector { p _iRepresent the data structure of its nonlinear strain-stress relation.Subsequently, use (parameter, strain) to as the RBF center, computing system or measuring system can be defined as whole material behaviors shown in the equation 6 by RBF interpolation (seeing as [Bickel et al.2009]).

(equation 6)

And, should be appreciated that to use a computer handles from data internal memory or its input, and suitable result is outputed to internal memory or its output, finishing this description of this relation, and can be in the such step that has or do not have to implement under the situation of manual intervention.

For uniform basic material, single group parameter vector is enough to describe the behavior of the arbitrary objects of being made up of single basic material.This decreased number with the basic material parameter arrives | p|P, wherein | P| is the radix (for isotropic material is 2, is 5 for isotropic material transversely) of parameter vector.In the example here, the number at RBF center is usually at the P=6 of isotropic foam with have between the P=12 of printed material of complicated internal microstructure.Based on local train, calculate the RBF interpolation in the mode of spatial variations, the different non-linear behavior of the zones of different that the system that makes can the emulation object.

A kind of scheme that is used for the behavior of the composite body that emulation made by basic material is the FEM scheme of the class static state of [Bickel et al.2009].A given simulation status, computing system can calculate the strain of all elements and implement every element (per-element) calculating of parameter vector according to equation 6.Computing system can recomputate every element stiffness matrix subsequently, and implements the new step of FEM emulation.

The match basic material

The attribute of our match basic material mates one group of input example best with the distortion of toilet emulation.In fit procedure, need RBF center { ε _i(that is) and its corresponding weights as the strain value of data point, w _jThe value of (for example, being used for equation 6).At first suppose known P RBF center.Provide and have measured displacement

With corresponding power

One group of exemplary distortion, according to equation 7, computing system calculates RBF weight w by the error that minimizes in the displacement.

$\hat{w}=\arg \min \left\{\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\left|\right|x_i(p,F_i))-{\stackrel{\‾}{x}}_i||}^2\right\}$ (equation 7)

In order to define the RBF center, computing system at first fits to material the material of uniform linearity, and obtains one group of constant material parameter.Use these parameters, computing system is all measurement operation FEM emulation, and record strain value and select the RBF center by with the P point of the scope that covers measured value strain space being sampled.Use these RBF centers, computing system can the match material parameter, obtains strain space and better covers but move repeatedly iteration.

Free optimization problem can cause physically incorrect material parameter.Under the situation of isotropic material, be easy to come gauge Lame parameter by calculating Young modulus and Poisson ratio and it being projected to physically effective value.Under the situation of isotropic material, can there be such constraint in the horizontal, promptly by using [Rebonato and

1999] technology, stiffness matrix are positive definite.

Measured power

Perpendicular to the surface.Yet, during measuring, the contact area under force probe also stands little tangential force, and we find that the power of those disappearances produces little error of fitting.Can increase the match quality by the tangential force of calculating disappearance, this tangential force can produce coupling fully on the surface point of being surveyed, and introduces this tangential force subsequently again as known power in optimizing process.

Test result is as follows, has the assessment of match quality, reported the basic material of institute's match and measured exemplary deformation ratio than the time error amount.

The material design of based target

Provide the scheme of the material design of using based target at this, it is similar to the deformational behavior of desired microscale by the material via the microscale of Combinatorial Optimization.Therefore, the combination of measuring system and computing system can export indication to many materials 3D printer and generate the output object of being printed with desired attribute.A scheme that obtains this output is that operation is automatic or automanual, the material design process of based target.

The optimizing process that is used to obtain the composite structure that is made of one group of basic material is described below.This process can use a computer and realize and receive the right example of description to body surface, desired power-displacement and one group and have basic material with the known distortion attribute of the nonlinear material model representation here as input.

The formulism that problem is conciliate

Some factors when building or constructing the embodiment of measurement/calculating/tectonic system, have been considered.Hereinafter will describe in detail now it.

Design process can be such optimization problem, and this optimization problem is chosen in the distribution of the possible basic material of M kind in the object of manufacturing, so that power-displacement example that its coupling is imported.Can be by representing that with the bag chamber of the individual rule of one group of N desired body form makes this problem discretize, the bag chamber of each rule is made by identical basic material.Suitable distribution by basic material realizes the desired unevenness of final object and possible anisotropy.In each bag chamber, can from the possible basic material of M kind, select single material.At this, in the text, " design " refers to the set to certain selection of basic material and its distribution, usually is stored in the data structure of electronically readable.Design can be expressed as vectorial m=(m ₁, m ₂..., m _N), wherein mi is an integer, its will the type of the basic material of i bag in the chamber from

Point out in the possible basic material.

In order to test each design, the material that we will be specific selects to be dispensed to the bag chamber of object, uses behavioral illustrations (profile) the emulation object of the power of user's appointment and the error of measurement in surface displacement with the FEM scheme of class static state described herein.The surface displacement of all input examples is collected in the big vector x.Given displacement of targets

The displacement error of design is exactly

The design that searching has the optimization of minimum displacement error is exponential problem, needs the possible design of test MN kind.

Use the branch and bound method of cluster

The problem that occurs when finding the solution such design optimization problem is the non-convexity of design space, and the risk that obtains locally optimal solution when only observing local adjacent domain [Lundand Stegmann 2005] that causes thus.

In order to solve this discrete optimization problem, computing system can use decision tree, so that comprise the option in a bag chamber in the design on each layer of tree.The root of tree has M child, and wherein each child represents the material in the first bag chamber in selecting, and other N-1 bag chambeies still do not have to determine.As mentioned above, Fig. 3 illustrates object and the two kinds of decision trees that possible material is selected that are used to have three bag chambeies.

Can use the branch and bound algorithm to pick out whole branch of decision tree [Land and Doig1960].During traversal of tree, the error dmin of the minimum of the design that our storage is tested up to now.When the new node of of access tree, that is, improve a new bag chamber, we use this minimum error to pick out (if possible) with the subtree of this node as root.

Under the situation of the width of given tree, picking out branch, still can to cause the quantity of possible design to be tested be reluctant.Yet several designs usually produce similar deformation result.We can limit the width of decision tree on each layer with those node clusters to coming together, and therefore limit the quantity of design to be tested.

Gauge is estimated

We are with { m} _a=(m ₁, m ₂..., m _l, x...x) being defined as with node a is root and the design that is positioned at layer l.Preceding l bag chamber is to determine along this branch, and remaining bag chamber is still debatable (representing with x).We are by considering that debatable bag chamber is that identical situation is estimated by { m} _aThe gauge of the distortion that produces.That is to say that in order to estimate gauge, we use basic material

In each fill debatable bag chamber, be all input example x _jCalculate the distortion that is caused, and the result of gauge M kind situation is [x subsequently _j].We use a bounding box of alignment as gauge in high-dimensional, that is, and and the maximal value of each dimension of resulting motion vector and minimum value.If

Then we to pick out with node a be the branch of root.When testing new design, then we suitably upgrade d _Min

There are some non-monotonic quantitys owing to the non-linear of material with along simulation process, gauge estimation scheme right and wrong sane (conservative).Recently, proposed to be used for being provided with the effective ways [Neumaier and Pownuk 2007] of gauge displacement, but the gauge that is used for the practicality of nonlinear setting is still a difficult problem at linear FEM.Yet identical piece can be considered as the behavior (from very soft extremely stone) of the limit, therefore according to those instructions, can expect that the combination of these materials can produce the distortion between the centre, and in these distortion, this gauge is estimated can not pick out optimal design.

The cluster strategy

We travel through decision tree in the mode of breadth-first, so therefore the father's layer with n node produces the layer that another has nM node.The gauge that is evaluated on this new layer need be calculated nM ²Individual design.In order to limit the width of tree, and limit total quantity of design to be tested thus, before operation splitting, we are the cluster node on each layer.

We use the K-means clustering procedure, use for the identical in twos exemplary displacement difference that the descendant assessed square and as between two nodes apart from dist (a, b), with the n on one deck node cluster to K bunch.In form, distance matrix is:

$\mathrm{dist}(a,b)=\underset{j}{\overset{M}{\mathrm{\Σ}}}\left|\right|x(m_{1_a},...m_{1_a},{\stackrel{\‾}{m}}_j,...,{\stackrel{\‾}{m}}_j)-$

$x(m_{1b},...m_{1b},{\stackrel{\‾}{m}}_j,...{\stackrel{\‾}{m}}_j){\left|\right|}^2$ (equation 8)

Cluster representative is the node of the centre of moment of close cluster.When layer decomposes, only need tested K M ²Individual design.Because the height of tree equals to wrap the quantity N in chamber, this cluster strategy, this cluster strategy approximately is limited to O (KM with total number of design evaluation ²N).Notice, because based on the picking out of gauge, the actual quantity of the design of being tested can be less.Some preferred embodiment in, use the cluster of K=20.This cluster scheme is a cost with the disappearance globally optimal solution.

Measure and basic material

In order to obtain object surfaces distortion, can use automatic measuring system to obtain the details of the material of many different geometry and surface properties with variation with material properties widely.One example system is above being described.This system can be used in sampling, the combination that is used for the basic material of model validation, complex object to be duplicated and its copy of printing that is used to confirm purpose of surveying basic material.

In certain embodiments, on the surface of basic material, drawn gauge point.In an example, those are in the grid of the rule of spacing 3mm.Can use scale and affine invariant point detecting device (scale and affine invariant blob detector) from image, extract mark position ([Mikolajczyk and Schmid 2004]) and can be in all frames the trace labelling position.According to the rigidity of material, the quantity of the deforming step that system implements can be from 30 to 200.The maximal value of power can be in 35 to 50N scope.Mark of being followed the tracks of and corresponding power can be recorded to surface mesh.

In having the example of isotropic basic material, OBJET ^TMConnex ^TMThe printer of many materials is used in a series of in service.In service at each, printer can use two kinds of different materials at the most, for example, and Vero White (firmly) and Tango Black Plus (soft).Printer can also mix these two kinds of materials with predefined ratio and produce the isotropic material with moderate rigidity.Specific in service, we mainly use Tango Black Plus (TBP) and are called as the composite material of the digital material (DM50) with shore 50.Except these two isotropic basic materials, we additionally measure 8 kinds of isotropic materials in the McMaster-Carr line directory, comprise rubber and foam.Fig. 5 shows the curve of surface displacement, and wherein surface displacement depends on the power of the subclass that puts on measured material.

For modeling with make and to have even the material (especially, softer material) of bigger deformation range, may need transversely isotropic basic material.This can be by solving the hollow space of an inside option as the object of being printed.In some printers, only when hollow space when the z axle is crossed over whole object, just allow hollow space.Under these circumstances, as shown in figure 13, can use the hollow tubular of different size and distribution.At this, perhaps only used the size of two kinds of hollow tubulars, but the multiple variation of hollow tubular also is possible.These objects are isotropic on the surface level perpendicular to hollow tubular.This material can be considered as transversely isotropy.When using other not have the 3D printer of these restrictions, can obtain more flexibility.

For such material, matrix E can be shown in equation 9 and is represented to have 5 degree of freedom, (E11, E12, E13, E33, E44}.Subsequently, the nonlinear material model can be considered five value functions in the sextuple strain zone, shown in equation 10.

$E=\left[\begin{array}{cccccc}E11& E12& \mathrm{<figure-callout\; id="13"\; label="E"\; filenames="HSA00000441784700031.png,HSA00000441784700071.png"\; state="\{\{state\}\}">E</figure-callout>}13& 0& 0& 0\\ E12& E11& \mathrm{<figure-callout\; id="13"\; label="E"\; filenames="HSA00000441784700031.png,HSA00000441784700071.png"\; state="\{\{state\}\}">E</figure-callout>}13& 0& 0& 0\\ E13& E13& E33& 0& 0& 0\\ 0& 0& 0& E44& 0& 0\\ 0& 0& 0& 0& E44& 0\\ 0& 0& 0& 0& 0& \frac{(E11-E12)}{2},\end{array}\right]$ (equation 9)

P (ε): IR ⁶→ IR ⁵(equation 10)

All basic materials are printed the piece into 4cm (wide) * 5cm (length) * 2.5cm (height).Distortion (side view) when some in these materials shown in Figure 6 are subjected to 15 Newton force.

Confirm and the result

Fig. 7 illustrates really and comparing side by side between the material of institute's emulation.Material model has very well been represented the elastic behavior of the basic material on the mesoscale rank.In Fig. 7, (first trip) that distortion isotropic material (left column) and transversely isotropic material (right row) will be obtained from actual object compared with emulation (middle row) and displacement error (tail is capable) based on FEM.Notice, assessment errors on the surface indicia position only, and subsequently interpolated error to be used for visual purpose.In addition, error evaluation depends on the accuracy of the measuring system in scope＜1mm.Very little extensional vibration effect (pitching effects) may not be traceable on the material of microscale, and thus may be in error visual disappearance.If it is important, then can use other system.

For isotropic basic material, we use six RBF centers, and for transversely isotropic basic material, we use 12 RBF centers, have caused every kind of basic material that 12 and 60 parameters are arranged separately.Match to material model on average needs two hours, but only needs to implement once.Figure 12 is illustrated in mean value, standard deviation and the maximal value of the error that applies the material under the different loads.

The affirmation of piling up.Then, we will illustrate, and we can predict the behavior of the compound substance of being made by the combination arbitrarily of basic material exactly.For different compound substances moves a plurality of emulation, and use Connex 500 printers to make those compound substances.Then, the behavior of using native system to measure these compound substances, and it is compared with emulation accordingly.

Fig. 8 illustrates the result of the affirmation of compound substance, wherein be assigned to and comprise two-layer compound substance and be printed from two of basic material (DM501 and the DM502) material properties that independently obtain the match, be under 8 and 21 newton's the situation being subjected to load subsequently, the distortion (up) of real-world object is compared with the distortion of emulation (middle row).Descending is the emulation of error.In the example of compound substance shown in Figure 8, be under the situation of 10N and 20N being subjected to load, average error is 1.98mm and 2.16mm.

The affirmation of the design of based target.Then, we confirm the design process of the based target here.As first step, we test it on known reproducible material.We select the given layer and the combination of its thickness.Subsequently, our this compound substance of emulation, and use its distortion as input to searching algorithm.

Fig. 9 illustrates the affirmation of the design process of based target.Based on the one group of material design that generates at random, those designs of emulation, and the distortion of its emulation is the input to the design search procedure of based target.Fig. 9 illustrates the design that is obtained of searching algorithm and the comparison on known basis.The histogram of Fig. 9 top illustrates the RMS error.In case make a kind of and corresponding search output in these designs that produce at random, then compare its force-displacement curve.

Go up this strategy of test in the 20 kinds of different material of selecting at random designs (5 layers, every layer has 9 kinds of different materials and selects).Though search does not guarantee to find out global optimum, very approaching separating (average RMS error is 0.067mm) usually found out in search.For further implementation should confirm that we had made those compound substances and have remeasured its attribute.We also show those results in Fig. 9.

Figure 10 illustrates by the manufacturing example of the design of a resulting based target in the approximate foam of combination trial of the material that uses Connex 500 to print and the combination and the error evaluation in the space that is obtained.As input, three exemplary distortion (top keeps left) of foam have been specified to the material design process of based target.By obtain by Combinatorial Optimization and use the material of the more accurate yardstick that the 3D printer makes to be similar to desired deformational behavior (keeping right in the top) subsequently.The curve map of below illustrates the relation of the power-displacement of approximate thing employed basic material foam and manufacturing.

Duplicate object

For most of complicated results, we try to represent whole process by duplicating the object with complicated deformational behavior.Especially, we duplicate sandy beach folder flip-flop and soft group footwear.We use the Cyberware scanner to scan the 3D geometric configuration of each object.Subsequently, we use our measuring system to obtain the deformational behavior of each object.Then, we use the design process of based target to use our basic material to find out the best approximate of distortion of materials attribute.At last, we use many file printings machine to print the duplicate of those objects.As seen from Figure 11, duplicate has and very similarly behavior of the original.In order further to confirm this scheme, we are illustrated in the force-displacement curve of the corresponding point on the duplicate and the original in Figure 11.

By the process shown in Figure 11 is to duplicate sandy beach folder flip-flop (up) and hair soft group footwear (descending).Left column illustrates original image, and middle column illustrates the object that is duplicated, the spatial array that its optimizing process by based target with basic material obtains.As seen from the figure, by power displacement curve (right row) indication of corresponding point on the original and duplicate, duplicate has very well mated the deformational behavior of the original.Dashed curve has characterized basic material.

Figure 12 is the table of error amount.These data by various isotropy (soft/moderate/hard foam) and transversely the match of the parameter of the material of isotropy (the material of being printed) with cylindrical hole structure draw.Show the surface displacement error under the loading condition of little, medium and big power, and this surface displacement error by measure and relatively the distortion of the material block (material of being printed) of the distortion of the material block (isotropy) of 5 * 5 * 2.5cm or 5 * 4 * 2.5cm realize.

Figure 14 illustrates the parameter of skin stool, comprises the force-displacement curve that is illustrated in corresponding point on the original and the duplicate.Left column illustrates the original, and middle column illustrates the cross section of the spatial array of the object that duplicated and basic material.In this example, stool is segmented into two zones, territory, hard area under clasp and remaining softer zone.Confirm this deformational behavior by the power displacement curve (right row) that compares in the clasp zone and in softer zone.

The example hardware that is used for computing system

Figure 15 illustrates the example system 100 that is used to be implemented in calculating described herein.After having read the present invention, those skilled in the art should be able to dispose and/or computing machine programmed and implement various calculating and the operations of describing herein, and need not too much experiment.

In an example shown, computer system 100 generally includes watch-dog 110, computing machine 120, keyboard 130, user input device 140, computer interface 150 or the like.In some cases, operation is automatically fully, and does not need some I/O equipment.In certain embodiments, computation process is the same with measuring process to be structured in the identical computing machine, can be via the wired or wireless mechanical organ that is connected directly to such as robotic arm, this mechanical organ is the part of measuring process, and can close printer via wired or wireless directly coupling, so that by calculating the output print object.

In various embodiments, display/watch-dog 110 (if use) can be implemented with the form of CRT, LCD display, plasma display, direct projection or rear projection DLP, micro-display etc.In various embodiments, watch-dog 110 can be used for vision ground display image etc., and as the part of interactive environment.

In the present embodiment, user input device 140 is implemented with the form of computer mouse, tracking ball, tracking plate, operating rod, wireless remote control, plotting sheet, voice command system, eye tracking system etc. usually.The embodiment of network interface 150 comprise Ethernet card, modulator-demodular unit (phone, satellite, cable, ISDN), (asynchronous) Digital Subscriber Line (DSL) unit, FireWire interface, USB interface and or the like.For example, network interface 150 can be connected to computer network, to the FireWire bus or or the like.In other embodiment, network interface 150 can physically be integrated on the mainboard of computing machine 120 and/or comprise software driver, or or the like.

In various embodiments, computing machine 120 generally includes familiar machine element, such as processor 160, and such as the memory storage equipment of random-access memory (ram) 170, disc driver 180, and and the interconnected system bus 190 of above-mentioned parts.RAM 170 or other storeies can be stored and treat the computer instruction carried out by one or more processors, with the mechanism of some functions that realize in software of describing herein as influence.In one embodiment, computing machine 120 comprises the Core of one or more Intel ^TMMicroprocessor.In addition, in the present embodiment, computing machine 120 generally includes the operating system based on UNIX.

RAM 170 and disc driver 180 are examples of computer-readable tangible medium, it is configured to the embodiments of the invention that are used to store, comprise the computer-executable code, data such as image file, object model, the expression data that realize technology described here, from model to printer the instruction, process prescription, executable computer code of conversion and/or or the like.The other types of tangible medium can comprise the hard disk of magnetic storage medium such as floppy disk, networking or portable hard drive, optical storage media, holographic memory and/or bar code such as CD-ROM, DVD, such as the memory storage of flash memory, the semiconductor memory of ROM (read-only memory) (ROMS), battery-backed volatile memory, networking and or the like.

In various embodiments, computer system 100 can also comprise software, its make by on the network such as HTTP, TCP/IP, RTP/RTSP agreement with or the like communicate by letter and become possibility.In alternative embodiment of the present invention, can also use other communication softwares and host-host protocol, for example IPX, UDP or or the like.

Figure 15 can implement the representative of computer system of the present invention.Those skilled in the art of the present invention be it is evident that many other hardware and software configurations are fit to use with the present invention.For example, computing machine can be desk-top, portable, rack-mounted or the graphic tablet configuration.Additionally, computing machine can be the computing machine of a series of networkings.In addition, can consider to use other microprocessors, such as the Xeon of Intel ^TM, Pentium ^TMOr Itanium ^TMMicroprocessor; Or from Advanced Micro Devices, the Turion of Inc ^TMOr Opteron ^TMMicroprocessor; And analog.In addition, can consider the operating system of other types, such as Vista ^TMOr Windows XP ^TMOr the Solaris of the product of other Microsofts, Sun Microsystems ^TM, Linux, Unix or analog.

In other embodiments, above-mentioned technology can realize on chip or auxiliary process plate.Many Configuration Types of computing equipment can be used in and are implemented in the whole bag of tricks described herein.In addition, have the processing element of the computing power of different stage, for example, microprocessor, graphic process unit, risc processor, flush bonding processor or analog also can be used in realizes various embodiment.

The scheme of broad sense

A plurality of detailed examples more than are provided, but have should be appreciated that the present invention is not limited to specific example.Usually, given N kind has the material of measured attribute, and electronic/mechanical/computing system can design and make compound (layering) material with desired attribute.Desired attribute for example can be distortion, elasticity, ductility, sponginess, flexibility, taste, smell, surface color, surface reflection, surface transparent degree, surface texturisation, such as hair, skin or the like.

The first of system is a measuring system, and the automatic measurement that it can implement the surface deformation that causes being subjected to given power draws the expression of depending on data-driven, stress-strain.Subsequently, function with known stress-strain that given N kind is piled up and known thickness of material usually can derive the function of the stress-strain of combined material by using the search procedure that realizes branch-and-bound search.

Subsequently, the function of given desired stress-strain, computing system can export the material layer and the thickness thereof of approximate desired best function to printer.By discrete/Combinatorial Optimization, each material has M kind thickness (for example, 1,2,3 ..., M mm) in one, and problem is the combinatorial search problem (for example, attempting all possible combination) with effective search strategy (for example, deleting invalid separating apace).Thus, can describe possible output the space (for example, provide one group of material with function of known stress-strain, what is the function of the then all possible stress-strain that can obtain by piling up these materials?)

Material type as basic material can be uniform material (wherein, every layer is identical) or heterogeneous material (wherein, every layer of variation that can have the space).The function of stress-strain can spatially change (being different for each surface point for example).Can use the file printing machine of mixing to come manufactured materials.If can not duplicate the function of some stress-strains physically, then can use design output mapping process, this process remains on the relative mistake between the distortion attribute and does not keep its absolute value.

The exemplary application of such system can comprise that manufacturing learns the material of distortion fully as human skin or Electronic cartoon.This can also be used for personage's mask (old performer, different people).Another Application is to make still significantly cheap, the light soft-tissue material that is out of shape as more expensive material fully.

Further discuss

As indicated above, the desired characteristic of (for example, by some characteristics of measuring some basic materials and the characteristic of inferring remaining basic material) characteristic of determined basic material and physical objects to be created can be flexible.Under these circumstances, for example, an appointment is determined in the search of the search volume of different combination that can be by representing flexible basic material, its appointment will have different flexible basic materials and where be placed on, so that the physical objects of structure is had, at least approx has desired elastic behavior.Yet characteristic does not need to be restricted to elasticity, but can be relevant with weight distribution, thermal capacity and/or other characteristics of describing herein.

Herein in the description of embodiment, complete system is described to use the consumptive material of many materials 3D printer, the consumptive material that characterizes those many materials 3D printer, the desired characteristic of determining object to be generated, calculating where to use which consumptive material and uses those consumptive materials to generate subsequently in the position of being calculated, at least approx generate, have the physical objects of desired attribute.Should be appreciated that not all embodiment needs all aspects of the present invention.For example, system can only be used to generate the instruction that is used to make physical objects.Except being used to generate the device of physical objects, the present invention can also comprise the physical objects that new technology generated that use is described herein.System can generate physical objects after receiving result of calculation, and the consumable material that can not need separating and measuring to import just draws result of calculation (because these characteristics are measured and known, or because these characteristics by emulation or estimation, or the like)

At this, the new process that is used to the real physical material with desired scatter properties to calculate " prescription " has been described.In specific example, under the situation of the desired material (by the behavioral illustrations appointment) of given N kind basic material (every kind have measured characteristic explanation) and output, processor (for example, computing machine, computing equipment and/or electronic circuit) calculate " piling up " (layer quantity, every layer thickness and basic material of distributing to every layer) of those materials, its approximate best desired material (that is its behavioral illustrations).As explaining that in other places identical scheme can be used in other basic material and desired material behavior/attribute.

As explained, use to have the ability to come on the position, to place or the machine of immobilization material with the multiple material that can be placed or be fixed, by determining the attribute of every kind of material, the desired shape of physical objects, the desired attribute of the part on the surface of physical objects, by attribute based on determined material, the desired property calculation one set constructor instruction of the desired shape of physical objects and physical objects, and by generating physical objects or being used to generate the instruction of physical objects by output, can generate physical objects according to digitized description, thereby when generating physical objects, it will have, at least approx have, desired shape, and attribute, such as taste, smell, flexibility, elasticity, surface reflection, surface color, the surface transparent degree, or the like.

Shown here, the scheme that the material with desired deformational behavior was measured, designed and made in a kind of complete being used to has been proposed.Our model can represent to have with emulation the nonlinear elastic deformation behavior of object of the micromechanism of complicated inside.In order to ensure the good coupling between the behavior of real distortion of materials and its emulation, the measuring process of use data-driving is estimated the model of the nonlinear stress-strain of every kind of material.In addition, show, the material design proposal of based target can be similar to desired overall deformational behavior by the material via the more accurate yardstick of Combinatorial Optimization.Material design by closing at measurement, emulation, based target and print between the loop, we have confirmed complete operating process, and show the approaching coupling between the object that can be implemented in institute's emulation object and real manufacturing.The design of our based target can be used in the 3D hard copy.

Modification can comprise the template of this process as system in a lot of future, those system extension institute's scope (such as dynamic distortion attribute or plasticity) with material properties manufacturing emulation.Use instruction herein, this model can extend to the dynamic and plastic deformational behavior and the measurement of improvement, so that its scope that can obtain wideer distortion attribute (for example, the measurement of material extensibility and dynamic deformation) maybe can guarantee and comprise previous physical knowledge, keep (volume preservation) such as capacity.

In other modification, strategy can be used for selecting best the number of the degree of freedom (RBF center) of this material model, be used to find the equilibrium point between accuracy and the overfitting.The speed of advance (emulation) step of material homogenising strategy to be used to improve nonlinear material can be provided.This improvement in company with more senior search strategy can transfer to accelerate design and the manufacturing that inverse step allows extremely complicated heterogeneous material.

In some extension examples, replace printing the level of different materials, it can extend to basic material spatial array (for example, voxel) arbitrarily in simple directly mode.Many file printings of 3D machine in the future may be able to use the basic material of the attribute that more has even more change, and those instructions can extend to printer in the future.

Should be appreciated that, move or move by " we " execution place in " we " execution that the present invention describes, can consider that performed action can be by another system's execution such as automatic design system, and/or as the part of the process in the physical world, in computer memory or in the part of electronic processes, the part of the system that independent existence or conduct are bigger.Certainly, some steps can be implemented by the people, but many can the realization with electric, electronics or computer installation, and can need not the people and intervene and realize.

After reading the present invention, those of ordinary skill in the art can associate more embodiment.In other embodiments, can advantageously carry out the combination or the son-combination of above invention disclosed.The exemplary arrangement of assembly only illustrates for the example purpose, and should be appreciated that will consider in an alternate embodiment of the invention in conjunction with, add, rearrange and similarly.Thereby although the present invention describes with respect to exemplary embodiment, those of ordinary skill in the art will recognize that numerous variation examples are possible.

For example, process described herein can be passed through nextport hardware component NextPort, component software and/or the enforcement of its combination in any.Correspondingly, instructions and accompanying drawing should be considered to illustrative rather than restrictive.Yet; this is obvious; promptly under the prerequisite that does not depart from the of the present invention wider spirit illustrated in the claim and protection domain, can make various changes and variation, and the present invention is intended to cover all changes that fall into claims protection domain examples and is equal to example.

Claims (15)

1. computed method is used to generate the method for the prescription that is used to generate physical objects, and described method comprises:

Determine the characteristic of multiple basic material;

Determine one group of desired characteristic of object to be generated;

Calculate the position of the part of described multiple basic material, one group of described desired characteristic of its approximate at least described object to be generated; And

Output is used for generating according to the part of being located of described multiple basic material the instruction of object.

2. method according to claim 1 is characterized in that, the position of calculating multiple basic material comprises:

A) generate data structure, it is corresponding to the search volume on the space of at least two kinds combination in the described multiple basic material that has different characteristics in expression;

B) expression of the initial position in the described search volume of storage, wherein said initial position is corresponding to the combination of described multiple basic material;

C) one group of characteristic of estimation expectation, its object of being made corresponding to the described combination of the described multiple basic material of described initial position by basis draws;

D) described one group of characteristic and the described one group of desired characteristic of estimating compared;

E) determine whether the difference between described one group of characteristic estimating and described desired one group of characteristic is lower than targets threshold;

F) if the described difference between described one group of characteristic estimating and described desired one group of characteristic is not less than described targets threshold, then discern follow-up location in described search volume, wherein said follow-up location is different from described initial position;

G) expression of the described follow-up location of storage;

H) regulate one group of characteristic of described expectation according to described follow-up location; And

I) repeating step c) to h), be lower than described targets threshold up to the described difference between described one group of characteristic estimating and described desired one group of characteristic.

3. method according to claim 2 is characterized in that, discerns described follow-up location and is included in before the identification, deletes described search volume, and discerns described follow-up location from the part of not deleting of described search volume.

4. method according to claim 2, it is characterized in that, discerning described follow-up location comprises and deletes described search volume, and realize deleting to be in order to meet the structure constraint, one of them structure constraint is added or by one or more the added constraint of physical attribute in the described multiple basic material by the physical machine that will generate described object.

5. method according to claim 4 is characterized in that, described structure constraint comprises a constraint, and promptly the position in described search volume can not be corresponding to the basic material with negative amount.

6. method according to claim 2, it is characterized in that, discerning described follow-up location comprises and deletes described search volume, and realize deleting by the redundancy section of removing described search volume, if wherein two or more parts of described search volume diverse location to basic material in the described object that generates provides same or similar one group of characteristic estimating, then it is redundant.

7. method according to claim 1 is characterized in that, one group of described desired characteristic of the described characteristic of described multiple basic material and described object to be generated comprises one or more in density, elasticity, heat conduction, thermal capacity and/or the hardness.

8. method according to claim 1 is characterized in that, one group of described desired characteristic of the described characteristic of described multiple basic material and described object to be generated comprises molecular behavior.

9. computed method is used for characterizing according to the subclass of the material of the attribute with variation the method for the material of the attribute with described variation, and described method comprises:

Measure first group of properties of materials with attribute of variation; And

Calculate the described characteristic of the material of second group of attribute with described variation, the value of the attribute of the described variation of at least a material in wherein said second group of material is not the value of attribute of the described variation of the material in described first group of material.

10. method according to claim 9 is characterized in that, the attribute of described variation comprises one or more in thickness and/or the elasticity.

11. be used for being generated by the combination of the multiple basic material that does not have the characteristic of mating one group of desired characteristic fully the device of the physical objects with described desired one group of characteristic, described device comprises:

Storer is used for the data of the characteristic of the described multiple basic material of storage representation;

Storer is used to store programming instruction;

Processor, it is connected to the described storer and the described storer that is used to store programming instruction of the data of the characteristic that is used for the described multiple basic material of storage representation;

Program code, it is stored in the described storer that is used for storing programming instruction, and described program code comprises being used for implementing in the search volume of the combination of described basic material deletes the program code of search;

The source of described basic material is used for the search volume and separates; And

Three-dimensional, be subjected to computer-controlled printer, be used for separating to generate described physical objects according to described search volume at least approx by placing described basic material.

12. device according to claim 11 is characterized in that, is used to implement the described program code of deleting search and comprises:

A) program code is used to generate data structure, and it is corresponding to the search volume on the space of at least two kinds combination in the described multiple basic material that has different characteristics in expression;

B) program code is used for storing the expression of the initial position of described search volume, and wherein said initial position is corresponding to the combination of described multiple basic material;

C) program code is used to estimate one group of characteristic estimating, and it is by drawing according to object of making corresponding to the described combination of the described multiple basic material of described initial position;

D) program code is used for described one group of characteristic and the described one group of desired characteristic of estimating compared;

E) program code is used to determine whether the difference between described one group of characteristic estimating and described desired one group of characteristic is lower than targets threshold;

F) program code, the described difference that is used between described one group of characteristic estimating and described desired one group of characteristic is not less than under the situation of described targets threshold, discern follow-up location in described search volume, wherein said follow-up location is different from described initial position;

G) program code, the expression that is used to store described follow-up location;

H) program code is used for regulating described one group of characteristic estimating according to described follow-up location; And

I) program code is used for repeating step c) to h), be lower than described targets threshold up to the described difference between described one group of characteristic estimating and described desired one group of characteristic.

13. device according to claim 11 is characterized in that, is used to implement the described described program code of search of deleting and is configured to illustrate physical constraint described three-dimensional, that be subjected to computer-controlled printer.

14. device according to claim 11, it is characterized in that one group of described desired characteristic of the described characteristic of described multiple basic material and described object to be generated comprises one or more in density, elasticity, heat conduction, thermal capacity and/or the hardness.

15. device according to claim 11 is characterized in that, one group of described desired characteristic of the described characteristic of described multiple basic material and described object to be generated comprises molecular behavior.

Images