CN106903889A

CN106903889A - A kind of rapid section algorithm of STL models

Info

Publication number: CN106903889A
Application number: CN201510979383.6A
Authority: CN
Inventors: 姜海涛; 贺兵; 周盛宗
Original assignee: Individual
Current assignee: Individual
Priority date: 2015-12-21
Filing date: 2015-12-21
Publication date: 2017-06-30

Abstract

A kind of rapid section algorithm of STL models, according to the geological information and slice thickness of STL model intermediate cam dough sheets, by tri patch packet sequencing, reducing the number of times of tri patch traversal；By building the local dynamic station tri patch table between tri patch, reduce slice plane and tri patch seeks friendship calculation times；And according to the topological relation of most of tri patch in slicing processes, it is proposed that the algorithm of dynamic tri patch table is updated, the lookup number of times of slicing processes intermediate cam dough sheet neighbouring relationship is reduced, so as to improve the whole efficiency of Slicing Algorithm.Topology reconstruction and section work to STL models can be rapidly completed by more new algorithm, analyzed from time complexity：Best-case, i.e. i+1 layer are added without new tri patch, and all that topology reconstruction need not be carried out to tri patch, the required time is the contour loop according to topological relationship calculation current layer；So saving the substantial amounts of time.

Description

A kind of rapid section algorithm of STL models

First, technical field

STL model files are the normative documents in 3D printing system.Inventive algorithm is applied to 3D printing industry, further promotes the development of 3D printing technique, has both improve the precision of processing model, and the process time of model is shortened again.

2nd, background technology

The essence of 3D printing is Layered manufacturing, so in the design of Software of RP System system, slicing treatment is one of key content, traditional Slicing Algorithm is typically equal using lift height, thus can produce the contradiction between model accuracy and time-write interval：Lift height is small, and model accuracy is guaranteed, but the time of printing is relatively long；Conversely, the time-write interval shortens, but cause the precision of model not high big with model ladder application condition.In mechanical rapid shaping field, many scholars did in-depth study and improvement to section computational algorithm, if dividing by research object, section computational methods can be divided into：(1) grid section is calculated：Because the grid model of STL forms is normative document type used in 3D printing industry, therefore many section calculating objects are main based on the trellis-type model of STL forms；(2) direct slicing is calculated：Because the original 3D models built by modeling software are when STL form model datas are converted into, transformed error can be produced, therefore also many research methods are to consider that section is directly performed in original 3D model datas calculates.The purpose of hierarchy slicing is exactly that complicated threedimensional model is converted into simple two dimension synusia, and the 3-D geometric model in computer is represented with contour line by the treatment cut into slices.Contour line illustrates the border of hierarchy slicing, and being linked in sequence by the intersection section of layered plane and stl file model obtains the loop of closing and constitute, and each layer of two-dimentional synusia is made up of the totally different profile of shape one by one.The conventional method of STL hierarchy slicings algorithm is：By analyzing the position relationship between section and triangular mesh,Intersection section is obtained if section is intersected with triangular mesh,After obtaining the section and all of intersection section of STL grid models,All intersections section in the layering face is coupled together in an orderly manner,So as to obtain the contour line in the hierarchy slicing face,When obtaining contour line by this method,It has to be ensured that traveling through all of triangular mesh dough sheet,But when contour line is sought,Possible major part dough sheet does not intersect with section,So result in that data-handling efficiency is low and operand is than larger,In order to improve layering efficiency,Many researchers are being studied both at home and abroad,Propose some hierarchical algorithms,First it is exactly that original STL model datas are pre-processed,Then hierarchy slicing is carried out again,Find that main slicing delamination algorithm can be divided into following three class by analysis：Hierarchy slicing algorithm based on geometric topology information,Hierarchy slicing algorithm based on triangle surface positional information,Hierarchy slicing algorithm based on STL grid model geometric continuities,But slice engine classic in the world is also useless in this at present does well,Therefore for this demixing technology,There is very big development and optimization space for existing slicing delamination algorithm,And present delamination software also there are problems that it is many,Present delamination software formed precision and speed are also not quite similar,STL formatted files are converted into by CAD model,And the error produced by subsequent slicing treatment is by the formed precision for influenceing model last,And because the algorithm of model layers has a lot,Using different algorithms,The speed of model layers is also differed,Even differ tens times,In the market RP softwares are incompatible,It is extremely difficult data to be exchanged with each other between different RP systems,It is all that to set up a set of general hierarchical algorithm system be necessary.

3rd, the content of the invention

The problem to be solved：

Complicated threedimensional model is converted to by slicing treatment by face profile information, the degree of accuracy of the profile information of acquisition directly affects the effect of shaping.And, as 3D printing develops, the target object of printing is increasing, become increasingly complex, this often leads to slicing treatment and takes greatly, and if one kind balance between the precision of shaping and time-write interval can not be found using traditional hierarchy slicing algorithm, so having to the optimization by algorithm to reduce the time of the time shortening printing shaping of the consumption in stl file pretreatment, the degree of accuracy for obtaining contour line is also improved, and then the effect of printing can be improved.

Technical scheme：

(1) packet sequencing

In order to reduce the number of times that tri patch judges with slice plane position relationship, packet sequencing is carried out to all of tri patch first, because the tri patch of STL models there are two features in slicing processes：1. the low tri patch in position is first cut, high rear cut in position.2. minimum constructive height falls the triangular plate in same slicing layer, and they intersect with slice plane, and slice height that they are hited first is identical, referred to as they have identical order, as shown in the Fig. 1 in Figure of description, Fig. 1 is tri patch and slice plane location diagram, triangular plate B and C in figure, its minimum constructive height appears in i-th layer, is hited first when i ＆ lt is cut into slices, according to two above feature, how all of tri patch implements Fast Groupinc And Sorting Method：Assuming that STL models have been layered n-layer by all of slice plane, then all of tri patch can be at most divided into n groups, and all tri patch with identical order are placed in same group, tri patch in i.e. one group corresponds to the set along the new intersecting tri patch of each slice plane of fabrication orientation, according to each layer of tri patch set Ti of formation, the corresponding matrix of each layer is set up, this completes the packet sequencing of all tri patch, so as to improve efficiency.

(2) local topology relation is built

After to all of tri patch packet sequencing,The matrix that can be just formed according to packet sequencing builds local topology relation,The all tri patch in T1 are all added in dynamic tri patch table first,Dynamic tri patch table is exactly the table that all tri patch intersected with current slice plane are constituted,And determine the syntople between each tri patch,Then according to the STL section intersection algorithms based on geometrical model topology information,Obtain the contour loop of ground floor section,In follow-up every layer slicing processes,Because the tri patch intersected with slice plane is change,So dynamic tri patch table also and then changes,That is when slice plane highly increases to Zi+1 (Zi+1+dz,Dz is slice thickness) when,Then need to be updated this active tri patch table,If the Zmax of some tri patch is less than Zi+1,I.e. the tri patch is completely in the lower section of current slice plane,No longer intersect with it,Then it is deleted from active tri patch table；The triangular plate ti+1j (j=1,2,3 ..., mi+1) included in Ti+1 is taken out from packet matrix, i.e., the triangular plate for newly intersecting with slice plane is added in active tri patch table；For every addition or a tri patch is deleted, will relative adjacency information in time update activity tri patch table.

(3) intergrowth into contour line is sought

In intersection process, for a slice plane Zi, first tri patch T intersected with the slice plane is calculated first, and obtain intersecting point coordinate, then adjacent tri patch is obtained according to local topology adjacency information, obtains intersecting point coordinate, followed the trail of successively, know and return to tri patch T, and obtain a contour loop for oriented closing；Said process is repeated, is finished until all contour loops are calculated, and finally give this layer of complete cross section profile.

Beneficial effect：

The characteristics of algorithm synthesis consider and combine existing Slicing Algorithm, the advantage of existing existing Slicing Algorithm, their limitation is overcome again, the packet sequencing of tri patch reduces the operand to all tri patch name placements, by the dynamic table of the crucial tri patch for building of local topology, ensure that in slicing processing, only the tri patch in active tri patch dynamic table is participated in and the friendship of asking of slice plane is calculated, and reducing carries out the number of times that tri patch judges with slice plane position relationship.Using topological relation, what the intersection point set for obtaining section was ordered into, without being ranked up again, end to end oriented closed outline can be directly obtained, simplify the process for setting up slicing profile；And when tri patch and slice plane ask friendship, the intersection point on side need to be only calculated certain tri patch, by the syntople in face, an intersection point of adjacent dough sheet can be inherited, but it is fairly time consuming to set up the process of complete STL data topology information, and all algorithms establish each layer of topological relation of part, and plain comparison range is searched in diminution, the workload of building topology relation is effectively reduced, the substantial amounts of time has been saved；Oriented closed outline ring can be directly obtained in slicing processes, it is to avoid the search of intersection section annexation judges.

Specific embodiment

(1) height and slice thickness of the STL threedimensional models according to loading on slice direction, calculate the maximum hierarchy number n of whole tri patch packet sequencings, the value of the Zmin and Zmax of each tri patch is calculated, the number of plies i where determining the tri patch sets up packet matrix Ti (i=1,2,3 ... ..., n), as shown in the Fig. 2 in Figure of description, Fig. 2 is the tri patch packet matrix diagram of STL models.

(2) since ground floor, i=1 during for ground floor generates the dynamic tri patch table of the first slicing layer according to T1 for section, determines the topological relation between each tri patch in table.

(3) carry out asking friendship to calculate and generating slicing profile according to the section intersection algorithm based on geometric topo-relationship.

(4) slice of data is analyzed, and removes the data point of redundancy and preserved.

(5) at this moment number of plies increase is i=i+1, if i ＞ n, turns to step (8)

(6) as the number of plies increases, while to update the tri patch in dynamic tri patch table (according to the method that as above technical scheme is introduced).

(7) step (3) is turned to

(8) section is fully completed, and terminates.

As shown in the Fig. 3 in Figure of description, Fig. 3 is the whole process figure of STL model slices layering to whole implementation process.

Claims

1. the characteristics of one kind is based on existing STL (stereolithography) model layers slicing treatment algorithm, on this basis Proposing carries out packet sequencing and dynamic tri patch table to tri patch and sets up data structure according to the topological structure of levels Slicing Algorithm, it is characterized in that：The algorithm is carried out to institute according to the minimum and maximum Z coordinate of tri patch and the thickness of layering Some tri patch carry out packet sequencing and dynamic tri patch table, and set up the topological structure of levels, which reduce and are cutting Friendship amount is sought to the traversal number of times of tri patch, sequence number of times and when intersection profile is sought during piece, slicing profile is simplified Construction process, improve the whole efficiency of Slicing Algorithm.

2. Slicing Algorithm according to claim 1, it is characterized in that：Tri patch according to STL models in slicing processes two Individual feature：1. the minimum constructive height Zmin of each tri patch is different, then also different from the order that its intersecting slice plane occurs, The tri patch for setting low of ascending the throne is first cut, high rear cut in position, as shown in the Fig. 1 in Figure of description.Design is all Tri patch packet sequencing is, in order to reduce the number of times that tri patch judges with slice plane position relationship, to improve layered velocity.

3. Slicing Algorithm according to claim 1, it is characterized in that：After the packet matrix for establishing STL models, it is possible to root Dynamic tri patch table is generated according to the matrix, first by T₁In tri patch be all added in dynamic tri patch table, then Dynamic tri patch table is constantly updated further according to the topological structure between the topological relation and levels of same layer, until having cut into slices Into.