(19)
(11)EP 3 617 913 A1

(12)EUROPEAN PATENT APPLICATION

(43)Date of publication:
04.03.2020 Bulletin 2020/10

(21)Application number: 18192292.3

(22)Date of filing:  03.09.2018
(51)International Patent Classification (IPC): 
G06F 17/50(2006.01)
G06T 19/00(2011.01)
(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR
Designated Extension States:
BA ME
Designated Validation States:
KH MA MD TN

(71)Applicant: Bricsys NV
9050 Gent (Ledeberg) (BE)

(72)Inventors:
  • Ulianenko, Aleksei
    630133 Novosibirsk (RU)
  • Shilovskii, Maksim
    630096 Novosibirsk (RU)
  • Rukoleev, Evgeny
    630117 Novosibirsk (RU)
  • Ermolin, Egor
    633011 Berdsk (RU)

(74)Representative: Brantsandpatents bvba 
Pauline Van Pottelsberghelaan 24
9051 Ghent
9051 Ghent (BE)

  


(54)IMPROVED CREATION OF AN EXPLODED VIEW OF AN ASSEMBLY IN CAD


(57) The current invention concerns a computer-implemented method, a computer system, and a computer program product for creating an exploded view of an assembly comprising multiple parts in a computer-aided design model. For a test part and a test direction, a candidate set of candidate parts of the assembly is obtained. For the test part and each candidate part, a polygonal surface approximation comprising faces, edges and vertices is obtained. For a test vertex of the polygonal surface approximation of the test part, an intersection subset of faces of the polygonal surface approximations of one or more candidate parts of the candidate set is determined. The faces of the intersection subset intersect with a ray from said test vertex in the test direction. A blocking subset of parts is determined for the test part and test direction based on the intersection subset of faces.




Description

Technical field



[0001] The invention pertains to the technical field of computer-aided design (CAD), and in particular to exploded views of assemblies.

Background



[0002] An exploded view drawing of an assembly is a diagram, picture, schematic or technical drawing of an object, that shows the relationship or order of assembly of various parts. It shows the parts of the assembly slightly separated by distance.

[0003] J. Yu and J. Zhang, "Hierarchical exploded view generation based on recursive assembly sequence planning", International Journal of Advanced Manufacturing Technology 93(1-4), pages 1207-1228 (2017), doi: 10.1007/s00170-017-0414-y, discloses a method for automatic exploded view generation.

[0004] Yu (2017) illustrates that the function AUTO EXPLODE COMPONENTS of UG NX fails to explode most of the parts to the right position and direction. Often occurring problems, disclosed in the introduction section of Yu (2017), include:
  • too much information should be manually input;
  • automatically obtained information is not well used;
  • determination of feasible explosion directions is inexact, time-consuming, and limited to the axial directions of the global coordinate system; and
  • the influence of assembly sequence planning on the layout and visual effect of the exploded view is seldom considered.


[0005] Autodesk Inventor 2016 provides for an AUTO EXPLODE functionality. This functionality has been deprecated in Autodesk Inventor 2017. Similarly, SolidWorks used to provide for an AUTO EXPLODE functionality, which has been deprecated in recent versions. Reasons for deprecation overlap with the above-stated problems.

[0006] There is a need in the art for an improved automatic generation of an exploded view of an assembly, with an improved determination of suitable locations and/or explosion directions for the parts of the assembly.

[0007] Yu (2017) discloses that the key elements for generating exploded views are explosion directions, explosion sequences, and displacement vectors. The document discloses hierarchical assembly sequence planning (ASP) based on an extended interference matrix (EIM) and a hierarchical assembly tree (HAT).

[0008] The EIM contains information on interference of parts in both the global coordinate system (GBS) as well as a part-specific local coordinate system (LCS). To construct the EIM, Yu (2017) proposes a sequential interference detection algorithm between pairs of parts from coarse to fine, based on a sweeping bounding box method, an extruding bounding face method, an extruding feature face method, and a stepping precise detection method, to sequentially determine whether possible interferences according to a coarse analysis persist according to a finer analysis. For n parts in the assembly, the computational time required for the construction of the EIM scales as O(n2).

[0009] The HAT comprises an assembly root node, subassembly intermediate nodes, and part leaf nodes. The HAT is utilized to group EIM information bottom-up, whereby component (part or subassembly) EIM information of components of a subassembly with the remainder of the assembly (assembly modulo the subassembly) is combined into subassembly EIM information. Disassembly directions are determined top-down, whereby disassembly directions for a subassembly are determined first, based on subassembly EIM information, before determining disassembly directions for its components. The disassembly sequence (order of components and choice of directions) is determined based on a heuristic multi-rule screening (MRS) method taking parallelism, continuity, stability, directionality and manipulability into account. An exploded view is generated bottom-up, thereby adjusting the explosion displacement of a subassembly so that a bounding box of the exploded subassembly does not interfere with the remainder.

[0010] Constructing the EIM is the most time-consuming part in the method of Yu (2017), eventually being prohibitively limiting for assemblies with a large number of parts. Positive interferences are ultimately detected via the stepping precise detection method, requiring time-consuming iterative discrete moves. For each active part, nine directions are scanned.

[0011] The present invention aims to resolve at least some of the problems mentioned above.

Summary of the invention



[0012] In a first aspect, the present invention provides a computer-implemented method (CIM) for creating an exploded view of an assembly comprising multiple parts in a computer-aided design (CAD) model, according to claim 1.

[0013] In a second aspect, the present invention provides a computer system for creating an exploded view of an assembly comprising multiple parts in a CAD model, whereby the computer system is configured for performing the CIM according to the first aspect.

[0014] In a third aspect, the present invention provides a computer program product (CPP) for creating an exploded view of an assembly comprising multiple parts in a CAD model, whereby the CPP comprises instructions which, when the CPP is executed by a computer, cause the computer to carry out the CIM according to the first aspect. The present invention may further provide a tangible non-transitory computer-readable data carrier comprising the CPP.

[0015] The present invention is advantageous as determining a blocking subset of parts for a test part and a test direction is performed quickly and analytically via ray tracing and polygonal surface approximations, thereby avoiding the need for multiple iterative discrete moves.

Description of figures



[0016] 

Figures 1 and 2 illustrate embodiments of algorithms according to the present invention.

Figures 3, 4, 5 and 6 show examples of assemblies and corresponding exploded views.


Detailed description of the invention



[0017] The present invention concerns a computer-implemented method (CIM), a computer system, and a computer program product (CPP) for creating an exploded view of an assembly comprising multiple parts in a computer-aided design (CAD) model. The present invention has been summarized in the corresponding section above. In what follows, the present invention is described in detail, preferred embodiments are discussed, and the invention is illustrated by means of non-limiting examples.

[0018] Unless otherwise defined, all terms used in disclosing the invention, including technical and scientific terms, have the meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. By means of further guidance, term definitions are included to better appreciate the teaching of the present invention.

[0019] "A", "an", and "the" as used herein refers to both singular and plural referents unless the context clearly dictates otherwise. By way of example, "a compartment" refers to one or more than one compartment.

[0020] "Comprise", "comprising", and "comprises" and "comprised of" as used herein are synonymous with "include", "including", "includes" or "contain", "containing", "contains" and are inclusive or open-ended terms that specify the presence of what follows, e.g. component, and do not exclude or preclude the presence of additional, non-recited components, features, elements, members, steps, known in the art or disclosed therein. Similarly, "based on" as used herein is synonymous with "based at least in part on" and is an inclusive or open-ended term, which does not exclude or preclude additional non-recited components, features, elements, members, steps, and the like.

[0021] A "set/list comprising an element" as used herein comprises computer-processable data enabling to obtain the element, and may refer to a set/list comprising a pointer to the element.

[0022] A "computer-aided design model" (CAD model), as used herein, comprises computer-processable data, preferably digital data, about one or more objects, such as an assembly or a part of an assembly. In the context of the present invention, a CAD model comprises multiple parts forming an assembly. A part is preferably a three-dimensional solid. The computer-processable data allows to derive a geometric property of a part and a relative geometric property between two parts. A non-limiting list of "geometric properties" of a part comprises a position; an orientation; a dimension; an inclination; and a radius of curvature. A non-limiting list of "relative geometric properties" of two parts comprises a relative distance; a relative orientation; a relative position; a relative angle; and a relative size.

[0023] A CAD model may be edited via a corresponding CPP, so-called CAD software. A non-limiting list of CAD software comprises 123D, ACIS, Advance Concrete, Advance Design, Advance Steel, AllyCAD, ArchiCAD, AutoCAD, BricsCAD, BRL-CAD, C3D, Caddie, Cadwork, CATIA, Chief Architect, Cobalt, Creo, DataCAD, DesignSpark Mechanical, Digital Project, Drawing Express, FINE MEP, form•Z, FreeCAD, HiCAD, IDEA Architectural, Inventor, IRONCAD, ItelliCAD, KeyCreator, LibreCAD, MEDUSA, MicroStation, Modelur, NanoCAD, NX, OpenCASCADE, OpenSCAD, Parasolid, PTC Creo, PowerCADD, progeCAD, PunchCAD, QCad, Revit Architecture, Revit MEP, Revit Structure, Rhinoceros 3D, RoutCad, SALOME, ShapeManager, SketchUp, Solid Edge, SolidWorks, SolveSpace, SpaceClaim, SpaceClaim Engineer, Tekla Structures, TopSolid, TransMagic, TurboCAD, VariCAD, VectorWorks, and VisualARQ.

[0024] In a first aspect, the present invention provides a CIM for creating an exploded view of an assembly comprising multiple parts in a CAD model. In a second aspect, the present invention provides a computer system for creating an exploded view of an assembly comprising multiple parts in a CAD model. The computer system is configured for performing the CIM according to the first aspect. In a third aspect, the present invention provides a CPP for creating an exploded view of an assembly comprising multiple parts in a CAD model. The CPP comprises instructions for performing the CIM according to the first aspect. The CPP comprises in particular instructions which, when the CPP is executed by a computer, such as a computer system according to the second aspect, cause the computer to carry out the CIM according to the first aspect. The present invention may further also provide a tangible non-transitory computer-readable data carrier comprising the CPP. The computer system according to the second aspect may comprise a tangible non-transitory computer-readable storage medium comprising the CPP. The three aspects of the present invention are hence interrelated. Therefore, any feature disclosed in this document may relate to each of the aspects of the present invention, even if it has been disclosed in conjunction with a particular aspect.

[0025] The CIM comprises several steps. Reference is made to Figure 1. For a test part of the assembly and a test direction, a candidate set of candidate parts of the assembly is obtained (101). For the test part and for each candidate part of the candidate set, a polygonal surface approximation (PSA) is obtained (101). A PSA comprises faces, edges and vertices. Preferably, a PSA is a triangulation. Preferably, the faces are triangular faces. For a test vertex of the PSA of the test part, an intersection subset of faces of the PSAs of one or more candidate parts of the candidate set is determined (102). The faces of the intersection subset intersect with a ray from said test vertex in the test direction (102). Thereto a ray may be constructed from the test vertex in the test direction. A blocking subset of parts is determined, for the test part and test direction, based on the intersection subset of faces (103).

[0026] This is advantageous as ray tracing and PSAs allow for quick and analytical determination whether the test part is blocked in the test direction by a candidate part of the assembly.

[0027] In a preferred embodiment, a blocking subset of parts may initially be obtained. The blocking subset may be empty, or may comprise parts added to the blocking subset based on another test vertex or multiple other test vertices of the PSA of the test part. Determining the intersection subset of faces for a particular test vertex of the PSA of the test part may comprise determining the intersection subset of faces of the PSAs of the candidate parts of the candidate set which are not in the blocking subset. The one or more candidate parts for which the intersection subset of faces is determined may hence be the candidate parts which are not (already) in the blocking subset.

[0028] This is advantageous, as it avoids identifying a blocking candidate part twice, and thereby reduces the computational cost further.

[0029] In a preferred embodiment, a ray subset of test vertices of the PSA of the test part is determined. An empty blocking subset of parts is obtained. For each test vertex of the ray subset, an intersection subset of faces of the PSA of one or more candidate parts of the candidate set is determined, whereby the faces of the intersection subset intersect with a ray from the test vertex in the test direction, and one or more candidate parts of the candidate set are added to the blocking subset based on said intersection subset. Preferably, an intersection subset of faces of the PSA of the candidate parts of the candidate set which are not in the blocking subset is determined.

[0030] This is advantageous, as it allows for a stepwise build-up of a blocking subset, while avoiding double identification of a blocking candidate part.

[0031] In a preferred embodiment, obtaining a candidate set comprises the following steps. An index subset of index parts of the assembly is obtained. A view box for the test part encompassing the test part is obtained. The view box is bounded in directions perpendicular to the test direction and unbounded in the test direction. For each index part of the index subset a bounding box is obtained. The candidate set is determined, whereby the candidate set comprises the index parts of the index subset for which the bounding box intersects with the view box.

[0032] A hierarchical spatial index of the parts of the assembly may be obtained from the CAD model. Preferably, the hierarchical spatial index is an assembly R-tree. For each of one or more parts of the assembly, preferably each part of the assembly, a bounding box may be added to the hierarchical spatial index. Alternatively, the CAD model may already comprise for each of one or more parts, preferably each part, a bounding box. In yet another alternative embodiment, a bounding box may be determined for a part when needed, i.e. so-called on-the-fly generation of a bounding box. Determining the candidate set may involve querying the hierarchical spatial index, preferably the assembly R-tree, for parts comprising a bounding box which intersects the view box.

[0033] This is advantageous, as a prescreening based on the view box of the test part and bounding boxes of the index parts is performed to exclude a priori uninteresting parts from the candidate set.

[0034] In a preferred embodiment, obtaining an index subset comprises determining an index subset of index parts based on hierarchical proximity to the test part in the hierarchical spatial index. Determining the index subset may involve querying the hierarchical spatial index, preferably the assembly R-tree, for parts comprising a predefined hierarchical proximity to the test part.

[0035] This is advantageous, as for n parts in the assembly and an appropriately structured hierarchical spatial index, the computational time required for determining the blocking subsets for all test parts scales as O(n log(n)), rendering the automatic creation of an exploded view of an assembly with a previously prohibitively large number of parts possible via the present invention. Preferably, the hierarchical spatial index is structured based on subassembly relationships, spatial proximity and/or mutual constraints between parts.

[0036] In a preferred embodiment, a test vertex, and preferably all test vertices of the ray subset, may be identified via the following steps, which are performed for a vertex, and preferably each vertex, of the PSA of the test part. A tangent set of faces of the PSA of the test part is determined, whereby the faces of the tangent set comprise the vertex. It is determined whether the sign of the dot product of the outward or the inward normal direction of the faces of the tangent set with the test direction is uniform over the tangent set. In case the sign of the dot product is not uniform over the tangent set, the vertex is identified as a test vertex. Hereby, the sign of the dot product may be positive, negative or undetermined, the latter being the case when the dot product is zero. The ray subset comprises hereby vertices which may be interpreted as border points of the PSA of the test part.

[0037] In a preferred embodiment, for the faces of the intersection set, associated candidate parts are identified, corresponding surfaces of the associated candidate parts are obtained, and intersection of the ray from the test vertex in the test direction with the corresponding surfaces is checked. In this embodiment, the blocking subset is determined based on associated candidate parts for which the ray intersects with a corresponding surface of the associated candidate part. Preferably, intersection of the ray from the test vertex in the test direction with the corresponding surfaces is checked for corresponding surfaces of associated candidate parts of the candidate set which are not (already) in the blocking subset.

[0038] This is advantageous, as it allows for a coarse PSA of a candidate part comprising a limited amount of surfaces, edges and vertices, and which encompasses the candidate part, for prescreening purposes. Furthermore, ray tracing from a test vertex of the PSA of the test part can still be performed quickly and analytically, and mitigates the need of multiple iterative discrete moves of the test part or a surface of the test part.

[0039] Figure 2 illustrates an exemplary embodiment of an algorithm (CIM) according to the present invention. The algorithm comprises several steps to determine a blocking subset of parts of the assembly for a test part of the assembly and a test direction. A view box is obtained for the test part and the test direction (201). The view box encompasses the test part, is unbounded in the test direction, and is bounded in other directions, i.e. directions perpendicular to the test direction. The view box may be a bounding box of the test part, which is extended in the test direction to infinity. An index subset of index parts is obtained for the test part based on hierarchical proximity to the test part in an assembly R-tree (202). Note that step (202) may precede step (201), as there is no dependency between them. A candidate set of candidate parts is determined (203). The candidate set comprises the index parts of the index subset for which a bounding box of the index part intersects the view box. For the test part and for each candidate part, a triangulation comprising triangular faces, edges and vertices is obtained (204). Note that for all parts of the assembly, a triangulation may be determined beforehand. Step (204) may therefore precede any one, multiple or all of steps (201), (202) and (203). A ray subset of test vertices of the triangulation of the test part is obtained (205). The ray subset is determined by checking for each vertex of the triangulation of the test part, whether the vertex is a test vertex. Each vertex is tangent to a tangent set of multiple triangular faces sharing the vertex. In case the sign of the dot product of the test direction and one of the outward or the inward normal direction of the triangular faces is not uniform over the multiple triangular faces of the tangent set, the vertex is identified as a test vertex, and added to the ray subset. Subsequently, for each test vertex of the ray subset, potentially blocking parts for the test part in the test direction are identified via a ray from the test vertex in the test direction (206). First, the triangular faces of the triangulations of one or more candidate parts which intersect the ray is determined, the so-called intersection set of triangular faces. Associated candidate parts are identified for the triangular faces of the intersection set. Corresponding surfaces of the associated candidate parts are obtained. Then it is checked with which corresponding surfaces the ray intersects. A blocking subset of parts is determined, based on associated candidate parts for which a ray from a test vertex of the ray subset in the test direction intersects with a corresponding surface of the associated candidate part (207). Preferably, the blocking subset comprises the associated candidate parts for which a ray from a test vertex of the ray subset in the test direction intersects with a corresponding surface of the associated candidate part.

[0040] In a preferred embodiment, a disassembling direction and a disassembling order may be determined based on determined disassembling directions of blocking parts. A disassembling sequence comprising zero, one or more scheduled parts of the assembly may be obtained. Each scheduled part is associated with a disassembling direction. The assembly further comprises unscheduled parts, which the disassembling sequence does not comprise. A removal set of unscheduled removal parts may be obtained. Each removal part comprises one or more removal directions which are not blocked by unscheduled parts of the assembly. For each removal part of the removal set and for each scheduled blocking part of the blocking set of said removal part, presence of the disassembling direction of said scheduled blocking part in the one or more removal directions of the removal part may be verified. In case at least one presence is verified over the removal set, a disassembling direction associated with a maximum number of removal parts for which presence is verified may be identified as the next removal direction. Otherwise, a removal direction associated with a maximum number of occurrences over the removal parts of the removal set may be identified as the next removal direction. One or more removal parts of the removal set comprising the next removal direction may be added to the disassembling sequence in association with the next removal direction.

[0041] The inventors have found that a technically useful exploded view may be obtained by determining the disassembling sequence based on previously determined disassembling directions of blocking parts of removable parts.

[0042] The disassembling sequence may comprise a sequence of zero, one or more steps, each step comprising one or more scheduled parts of the assembly associated with a common disassembling direction. A new step associated with the next removal direction may be added to the disassembling sequence, whereby the new step comprises said one or more removal parts of the removal set comprising said next removal direction.

[0043] This is advantageous as it allows for treating the parts of a step simultaneously. A common displacement along the disassembling direction may be determined for all parts of a step of the disassembling sequence. When, for example, two parts are joined by multiple bolts and nuts, a common disassembling step for all nuts may be determined and a common disassembling step for all bolts may be determined.

[0044] In a preferred embodiment, adding one or more removal parts of the removal set comprising the next removal direction to the disassembling sequence in association with the next removal direction may involve a determination of an addition set of removal parts comprising the next removal direction, whereby the removal parts of the addition set have mutually non-interfering bounding boxes. An empty addition set may be obtained. An aligned removal set comprising the removal parts of the removal set which comprise the next removal direction may be obtained. For each pair of removal parts of the aligned removal set is determined whether the removal parts of the pair are deemed adjacent. Deemed adjacent removal parts comprise bounding boxes which touch or intersect. While the aligned removal set comprises unprocessed parts: an unprocessed removal part with a maximal number of deemed adjacent unprocessed removal parts is selected from the aligned removal set; the selected removal part is added to the addition set; and the selected removal part and the deemed adjacent unprocessed removal parts are identified as processed. The removal parts of the addition set are added to the disassembling sequence in association with the next removal direction. Preferably, a connection graph comprising nodes and edges is obtained. Each node of the connection graph represents a removal part of the aligned removal set. Each edge of the connection graph between two nodes of the connection graph represents deemed adjacent removal parts. While the connection graph comprises nodes: a node of the connection graph which comprises a maximal number of edges is selected; the removal part represented by the selected node is added to the addition set; and all nodes connected to the selected node, the selected node, and all corresponding edges connecting at least one of these nodes are removed from the connection graph.

[0045] This is advantageous as it enables quick determination of a minimal amount of successive disassembling steps, whereby each disassembling step comprises parts with non-interfering bounding boxes. Certain parts may comprise a joint possible removal direction, but joint removal over the same displacement does not provide a technically useful exploded view. Consider, for example, an assembly comprising two tangent cogwheels, for which the exploded view preferably comprises a distance in between the two cogwheels.

[0046] In a preferred embodiment, for a test direction a directed blocking graph comprising nodes and directed edges is constructed based on said blocking subset. A node of the directed blocking graph represents a part blocked by or blocking a part in the test direction. A directed edge from a first node to a second node represents one of the part associated with the first node blocked by or blocking the part associated with the second node. Preferably, for each test direction of multiple test directions a corresponding directed blocking graph is constructed. Determining the removal set may be based on the directed blocking graphs. The removal set may comprise all parts of the assembly for which an associated node in at least one of the directed blocking graphs does not comprise one of incoming or outgoing directed edges. In case a directed edge from a first node to a second node represents the part associated with the first node being blocked by the part associated with the second node, the removal set may comprise all parts of the assembly for which an associated node in at least one of the directed blocking graphs does not comprise an outgoing directed edge, i.e. the part is not blocked in the test direction associated with the particular directed blocking graph. Upon adding one or more removal parts to the disassembling sequence, each node corresponding to the one or more removal parts may be removed from the directed blocking graphs.

[0047] This is advantageous as it allows for quick querying of graphs to obtain the removal set and for simple bookkeeping of which parts are unscheduled.

[0048] In a preferred embodiment, for a part of the assembly three global and six local test directions are determined and for another part of the assembly a reduced number of, preferably at most eight, more preferably at most five, even more preferably at most two, most preferably one, test direction(s) is determined. For said another part, preferably each test direction is parallel to a symmetry axis. For example, one test direction parallel to the main axis of symmetry may be determined for a standard fastener such as a bolt, a nut or a threaded rod. Determination of a reduced number of test directions may be based on geometric properties, relative geometric properties and/or semantic properties. Preferably, determination of a reduced number of test directions for a part is based on a semantic property of the part. A semantic property of a part may comprise an identification of the function of the part.

[0049] This is advantageous, as non-useful directions are not even considered as test directions for determining a disassembly sequence. For example, a nut will always comprise a disassembling direction parallel to the length direction of the bolt on which it is fastened.

[0050] In a preferred embodiment, a displacement for a part of the assembly along a disassembling direction is determined based on a viewing point and a viewing direction.

[0051] This is advantageous, as a technically useful exploded view of an assembly should visibly display all parts of the assembly while maintaining a compact spatial extent. Adjusting the displacements depending on the point of view, solves this problem.

[0052] The invention is further described by the following non-limiting examples which further illustrate the invention, and are not intended to, nor should they be interpreted to, limit the scope of the invention.

Examples


Example 1: Embodiment of an algorithm according to the present invention



[0053] The present example relates to a particularly preferred embodiment of an algorithm (CIM) according to the present invention.

[0054] A spatial assembly R-tree (assembly tree) is obtained. The CAD model may comprise the assembly tree. The assembly tree may be determined based on the CAD model. The assembly tree comprises leaf nodes, each corresponding to a part. A leaf node thereby does not correspond to a subassembly comprising multiple parts. A leaf node also does not correspond to a part containing a subassembly. The assembly tree may comprise a root node, subassembly intermediate nodes, and part leaf nodes. Preferably, a part is a three-dimensional solid.

I. Initialization



[0055] In an initialization step, for each part:
  1. a. the associated three-dimensional geometry is triangulated;
  2. b. a bounding box is computed; and
  3. c. the bounding box associated with this part is added to the assembly tree.
Pre-computation of the triangulations and bounding boxes is advantageous as they can be re-used when necessary at negligible computational cost.

II. Determination of the disassembling sequence



[0056] In a following step, the disassembling sequence is determined. Hereby, determination of test directions for a part, detection of blocking parts for a part in a test direction, and addition of parts to the disassembling sequence is performed bottom-up, starting at the leafs and gradually working up the assembly R-tree. In particular, the algorithm may allow for addition of a part to the disassembling sequence, before the test directions of another part (higher up the assembly R-tree) are determined. A schematic overview of the algorithm is provided. Where indicated, certain aspects are detailed further below.

II.1 Schematic overview



[0057] While there is any leaf node in the assembly tree:
  • For each leaf node in the assembly tree (corresponding to a test part):
    • Determine the test directions along which the test part can be moved (either three global and six local ones or a reduced number as discussed) and add them to a list of unchecked test directions for the test part;
    • While there are parts comprising an unchecked test direction:

      ∘ For each test part comprising an unchecked test direction:

      ▪ For each unchecked test direction of the test part:

      ❖ Determine the blocking subset of parts for the test part and test direction (further detailed below, see II.2);

      ❖ Add the information about the parts of the blocking subset in a directed blocking graph corresponding to the test direction, whereby a directed edge (i, j) means that part i is blocked by part j;

      ❖ If the test direction is a global direction or the opposite direction is a local direction for some parts in the blocking subset, add related information corresponding to the opposite direction to the graph as well

      ❖ Remove the test direction from the list of unchecked test directions;

    • Find the base part, i.e. the part that has the most outgoing edges in total;
  • While at least two parts remain which have not been moved for the exploded view (unscheduled parts):
    • For each test direction:

      ∘ Find all parts associated with nodes of the corresponding directed blocking graph which do not have outgoing directed edges;

      ∘ If at least one such part is found, add the test direction in conjunction with the corresponding parts to a map of candidate directions (the map of candidate directions thereby comprises the removal set of unscheduled removal parts);

    • If the map of candidate directions is empty, break (no parts can be moved);
    • Obtain a next removal direction based on the map of candidate directions (further detailed below, see II.3);
    • Obtain an addition set of parts that can be moved in the next removal direction (further detailed below, see II.4) and add a corresponding step comprising the addition set and the next removal direction to the disassembling sequence;
    • Remove all edges from all directed blocking graphs that are incoming to the parts of the addition set and remove all nodes corresponding to those parts;
    • Remove the corresponding leaf nodes from the assembly tree and replace the corresponding subassembly with the base part (note that hereby the subassembly node may be converted into a leaf node).

II.2 Detail: Determine for the test part and test direction a blocking subset



[0058] 
  • Obtain an empty blocking subset;
  • Extend the bounding box of the test part in the test direction to infinity;
  • Obtain an empty ray subset of test vertices (border points);
  • For each vertex from the triangulation of the test part:
    • Obtain a tangent set of all triangular faces that share the current vertex;
    • Obtain an uninitialized sign variable;
    • For each triangular face:

      ∘ Compute the dot product of the (outward or inward) triangular face normal direction and the given direction;

      ∘ If the sign variable is uninitialized, set it to the sign of the dot product;

      ∘ Otherwise (sign variable initialized), if the sign variable does not match the sign of the dot product, then:

      ▪ Add the point to the ray subset;

      ▪ Break.

  • Obtain the candidate set of candidate parts by querying parts with bounding box intersecting the extended box using the assembly tree;
  • Obtain an empty triangular R-tree and add for each candidate part from the query the triangulation of the candidate part to the triangular R-tree;
  • For each test vertex from the ray subset (border point):
    • Construct a ray along the test direction from the test vertex;
    • Obtain the intersection subset of triangular faces by querying all triangular faces which intersect this ray using the triangular R-tree comprising triangular faces;
    • For each triangular face of the intersection subset:

      ∘ Determine the associated part;

      ∘ If the associated part is in the blocking subset, continue;

      ∘ Get a corresponding true surface of the associated part;

      ∘ Determine an intersection point of the corresponding true surface with the ray using ray tracing;

      ∘ If an intersection point is determined, add the associated part to the blocking subset, and add the current test part to the blocking subset of the associated part for the opposite direction;

  • For each candidate part not in the blocking subset:
    • If the test direction is not in the list of unchecked test directions for the candidate part, add the opposite direction to the list of unchecked test directions of the candidate part.


[0059] This algorithm has several advantages:
  1. 1. Compared to algorithms using Boolean operations such as clashes, no iterative moves of the test part are required: the result is obtained using a few simple analytical calculations.
  2. 2. Compared to algorithms using projections, less expensive operations are required.
  3. 3. Compared to algorithms using rasterization, a smaller number of points is used.
  4. 4. By using R-trees, the computational complexity is O(n log(n)) for a representative assembly instead of O(n2), whereby n is the number of parts.

II.3 Detail: Select a disassembling direction



[0060] 
  • Obtain an empty map of previously used directions;
  • For each unscheduled removal part in the map of candidate directions:
    • Determine a set of possible directions for the unscheduled removal part, i.e. all test directions for the current part in the map of candidate directions;
    • Determine a set of previously (i.e. scheduled) blocking parts for the unscheduled removal part;
    • For each previously blocking part:

      ∘ Obtain the corresponding disassembling direction;

      ∘ If the disassembling direction is in the set of possible directions, add the disassembling direction in conjunction with the unscheduled removal part to the map of previously used directions;

    • If the map of previously used directions is not empty, return the direction from the map of previously used directions with the largest number of assigned parts as the next removal direction, otherwise, return the direction from the map of candidate directions with the largest number of assigned parts as the next removal direction.

II.4 Detail: Obtain an addition set of parts



[0061] 
  • Construct a connection graph, whereby each vertex corresponds to a part which can be moved in the next removal direction and each edge corresponds to touching or intersecting bounding boxes of the corresponding parts;
  • Obtain an empty addition set;
  • While the connection graph is not empty:
    • Determine a base vertex, i.e. the vertex with most edges in the connection graph;
    • Add the part associated to the base vertex to the addition set;
    • Remove all vertices connected to the base vertex, the base vertex, and all edges connecting at least one of the previous vertices from the connection graph;
  • Return the addition set.

III. Generating the exploded view



[0062] The disassembling sequence is reversed.

[0063] For each step of the disassembling sequence:
  • If it is the first step, continue;
  • Find the furthest point A of the current assembly based on the disassembling direction corrected for the desired 2D projection:
    • Determine the projected direction onto the screen plane of the disassembling direction, based on the position of the camera, the vector towards the 3D scene, and the disassembling direction;
    • If the projected direction is the zero vector, find the furthest point A based on the disassembling direction, otherwise, find the furthest point A based on the projected direction;
  • For each part in the step:
    • Find the furthest point B of the current part in the opposite direction;
    • Move the part along (A-B) + gap * direction vector, whereby the direction vector is based on the disassembling direction if the projected direction is the zero vector, and based on the projection direction otherwise.

IV. Discussion



[0064] Yu (2017) utilizes test directions of global and local coordinate systems to determine blocking parts. The present invention may also consider for standard fasteners (such as bolts, nuts, threaded rods, etc.) only the main axis of symmetry.

[0065] Just defining likely directions is not sufficient, as in realistic assemblies, most parts are blocked by other parts. Such directions should be excluded from the list of possible removal directions on each iteration. To effectively find which parts are blocked by which, an algorithm based on spatial trees and ray tracing is used and allows to find blocking parts effectively. The inventors have found that the algorithm works substantially faster compared to the one disclosed in Yu (2017).

[0066] The choice of a possible disassembling direction greatly affects if a technically suitable exploded view is obtained or not. Technical suitability involves visibility and/or non-interference of the parts, compactness of the exploded view, and an easily understandable assembly order for constructing the assembly from its parts. In a first stage of the algorithm, a map between possible removal directions and parts that can be removed is constructed. Then, using this map, the most suitable disassembling direction is chosen using several heuristics. The inventors have developed two heuristics sufficient to provide a valid result. The first heuristic is trying to select the direction which has already been selected. To do that, it is checked, for each current part which can be removed and for each part that has blocked such a current part, whether the direction of removal is in the set of non-blocked directions for this part. If this is the case, the current part is added to a map between already used directions and parts non-blocked in that direction. The already used direction with the largest number of parts to move is selected, if available. If not available, the direction with the largest number of current parts to remove is selected. Thereby, as it would be unclear to see parts touching each other in an exploded view, a filtering of the set of parts to remove is performed. Only parts which are non-interfering can be moved simultaneously. Filtering is performed via the connection graph, thereby guaranteeing that only non-interfering parts will be moved simultaneously.

[0067] Exploded views are used as two-dimensional drawings for documentation. Exploded view algorithms work in three-dimensional space. Yu (2017) provides an algorithm whereby bounding boxes of subassemblies are moved to not interfere. However, non-interference in 3D does not imply non-interference in 2D. In the present algorithm, a displacement based on an extent of the current exploded view in the disassembling direction and further based on a projected direction is determined. This mitigates interference in 2D, while maintaining the minimal gap.

Example 2: Further implementation details



[0068] This second example pertains to an algorithm according to the first example, comprising further implementation details as specified below.

[0069] In the DWG-format, it is possible to define parts of geometry in the drawing as blocks, and then refer to those blocks using so-called block references. Block references allow to refer to both local and external (defined in another file) blocks. Such references are lightweight and contain the ID of the block, the transformation that should be applied to the content of the block, and some other properties. This feature allows to create an effective and lightweight representation of assemblies and their exploded views.

[0070] In the context of the present invention, each part may be represented as a block, either local or external. When a part is inserted into an assembly, it means that the block reference to this part is added to the block of the assembly geometry. The assembly structure is stored in the database using a hierarchy of dedicated objects containing IDs of blocks as well as other important data such as names and assembly structure. Each part has a GUID (unique ID) that cannot be changed during any operation. Therefore, each instance of any part can be described by a chain of GUIDs. When an exploded view is created, a dedicated block for this exploded view is created in the database, and then all block references representing parts are copied to that block recursively. Because block references are lightweight, the exploded view is lightweight as well. Copies of block references allow to move them in exploded view independently from the assembly itself. To identify the component corresponding to a given block reference in the exploded view, each block reference is assigned with data information containing the chain of GUIDs, the original transformation and the final transformation computed by the automated exploded view algorithm.

[0071] When the assembly is changed (for example, by adding new parts or by removing some parts), the exploded view can be automatically updated as well. The update procedure creates a copy of the current state of the assembly, but it also tries to match the parts with their previous copies. If a match is found, a modification transformation is computed for the previous copy in order to restore the modifications after the update. In case M is the modification transformation in matrix form, A is the transformation applied by the algorithm in matrix form, and F is the final (current) transformation in matrix form, their relations may be expressed as F = M*A and M = F*A-1. The modification transformation is stored in the map between new copies of the parts and modification transformations.

[0072] When all block references are copied, the algorithm is launched on that copy. After the algorithm finishes, modification transformations will be re-applied to corresponding parts. This is advantageous as it allows to keep user changes when the modifications of the geometry are not of a large scale.

Example 3: Example exploded views of embodiments



[0073] Figures 3, 4, 5 and 6 show several illustrative examples of assemblies (301, 401, 501, 601) and corresponding exploded views (302, 402, 502, 602).

[0074] Figures 3 and 4 demonstrate results generated via the algorithm disclosed in example 1, but without correction for projection. Figures 5 and 6 demonstrate results generated via the entire algorithm as disclosed in example 1.


Claims

1. Computer-implemented method for creating an exploded view of an assembly comprising multiple parts in a computer-aided design model, comprising the steps of:

- obtaining for a test part of the assembly and a test direction a candidate set of candidate parts of the assembly;

- obtaining for the test part and each candidate part of the candidate set a polygonal surface approximation comprising faces, edges and vertices;

- determining for a test vertex of the polygonal surface approximation of the test part an intersection subset of faces of the polygonal surface approximations of one or more candidate parts of the candidate set, whereby the faces of the intersection subset intersect with a ray from the test vertex in the test direction; and

- determining based on the intersection subset of faces a blocking subset of parts.


 
2. Computer-implemented method according to preceding claim 1, wherein the step of obtaining a candidate set comprises the steps of:

- obtaining an index subset of index parts of the assembly;

- obtaining for the test part a view box encompassing the test part, the view box bounded in directions perpendicular to the test direction and unbounded in the test direction;

- obtaining for each index part of said index subset a bounding box;

- determining a candidate set, comprising the index parts of the index subset for which the bounding box intersects with said view box.


 
3. Computer-implemented method according to preceding claim 2, wherein the step of obtaining an index subset comprises the steps of:

- obtaining a hierarchical spatial index of the parts of the assembly from the computer-aided design model; and

- determining an index subset of index parts based on hierarchical proximity to the test part in the hierarchical spatial index.


 
4. Computer-implemented method according to any one of the preceding claims, wherein a polygonal surface approximation is a triangulation, and wherein said faces are triangular faces.
 
5. Computer-implemented method according to any one of the preceding claims, comprising the steps of:

- determining a ray subset of test vertices of the polygonal surface approximation of the test part;

- obtaining an empty blocking subset of parts;

- for each test vertex of the ray subset:

• determining an intersection subset of faces of the polygonal surface approximations of one or more candidate parts of the candidate set, whereby the faces of the intersection subset intersect with a ray from said test vertex in the test direction;

• adding one or more candidate parts of the candidate set to the blocking subset based on said intersection subset.


 
6. Computer-implemented method according to any one of the preceding claims, comprising, for each vertex of the polygonal surface approximation of the test part, the steps of:

- determining a tangent set of faces of the polygonal surface approximation of the test part, which comprise said vertex;

- determining whether the sign of the dot product of the outward or the inward normal direction of the faces of the tangent set with the test direction is uniform over the tangent set; and

- in case said sign of said dot product is not uniform over the tangent set, identifying said vertex as a test vertex.


 
7. Computer-implemented method according to any one of the preceding claims, comprising the steps of:

- obtaining a disassembling sequence comprising zero or more scheduled parts of the assembly, each associated with a disassembling direction, the assembly further comprising unscheduled parts which the disassembling sequence does not comprise;

- obtaining a removal set of unscheduled removal parts, each removal part comprising one or more removal directions not blocked by unscheduled parts of the assembly;

- for each removal part of the removal set and for each scheduled blocking part of the blocking set of said removal part, verifying presence of the disassembling direction of said scheduled blocking part in the one or more removal directions of the removal part;

- in case of at least one presence verification, identifying a disassembling direction associated with a maximum number of removal parts for which presence is verified as the next removal direction, otherwise, identifying a removal direction associated with a maximum number of occurrences over the removal parts of the removal set as the next removal direction; and

- adding one or more removal parts of the removal set comprising the next removal direction to the disassembling sequence in association with the next removal direction.


 
8. Computer-implemented method according to preceding claim 7, wherein said step of adding one or more removal parts of the removal set comprising the next removal direction to the disassembling sequence in association with the next removal direction comprises the steps of:

- obtaining an empty addition set;

- obtaining an aligned removal set comprising the removal parts of the removal set comprising the next removal direction;

- determining for each pair of removal parts of the aligned removal set if the removal parts of the pair are deemed adjacent, deemed adjacent removal parts comprising bounding boxes which touch or intersect;

- while the aligned removal set comprises unprocessed removal parts,

• selecting an unprocessed removal part from the aligned removal set with a maximal number of deemed adjacent unprocessed removal parts;

• adding the selected removal part to the addition set; and

• identifying the selected removal part and the deemed adjacent unprocessed removal parts as processed,

- adding the removal parts of the addition set to the disassembling sequence in association with the next removal direction.


 
9. Computer-implemented method according to preceding claim 8, comprising the steps of:

- obtaining a connection graph comprising nodes and edges, each node representing a removal part of the aligned removal set, each edge between two nodes representing deemed adjacent removal parts;

- while the connection graph comprises nodes,

• selecting a node of the connection graph comprising a maximal number of edges;

• adding the removal part represented by said node to the addition set; and

• removing all nodes connected to said node, said node, and all corresponding edges connecting at least one of these nodes from the connection graph.


 
10. Computer-implemented method according to any one of the preceding claims, comprising the step of constructing for said test direction a directed blocking graph comprising nodes and directed edges based on said blocking subset, a node representing a part blocked by or blocking a part in the test direction, a directed edge from a first node to a second node representing one of the part associated with the first node blocked by or blocking the part associated with the second node.
 
11. Computer-implemented method according to preceding claims 7 and 10, wherein for each test direction of multiple test directions a corresponding directed blocking graph is constructed, wherein said removal set is determined based on the directed blocking graphs, and wherein, upon adding a removal part to the disassembling sequence, each node corresponding to the removal part is removed from the directed blocking graphs.
 
12. Computer-implemented method according to any one of the preceding claims, wherein for a part of the assembly three global and six local test directions are determined, and wherein for another part of the assembly a reduced number of test directions is determined, preferably parallel to a symmetry axis, such as one test direction parallel to the main axis of symmetry for a standard fastener such as a bolt, a nut or a threaded rod.
 
13. Computer-implemented method according to any one of the preceding claims, wherein a displacement for a part of the assembly along a disassembling direction is determined based on a viewing point and a viewing direction.
 
14. Computer system for creating an exploded view of an assembly comprising multiple parts in a computer-aided design model, the computer system configured for performing the computer-implemented method according to any one of the preceding claims 1 to 13.
 
15. Computer program product for creating an exploded view of an assembly comprising multiple parts in a computer-aided design model, the computer program product comprising instructions which, when the computer program product is executed by a computer, cause the computer to carry out the computer-implemented method according to any one of the preceding claims 1 to 13.
 




Drawing






















Search report









Search report




Cited references

REFERENCES CITED IN THE DESCRIPTION



This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Non-patent literature cited in the description