Method for generating a digital model of a part of a lighting device in a motor vehicle

Abstract

The invention relates to a method for generating a digital model of a component of a lighting device for a motor vehicle, implemented by a computer system, comprising: (E0) a step of providing an original digital model (OM) of a component of the lighting device; (E1) a step of generating an optimal design digital space (ODS) in which said original model can be extended; (E2) a step of generating, by a generative design algorithm, new digital models (NMs), each new digital model extending within said optimal design space and having a set of predetermined design and / or production parameter values; (E4) a step of selecting a digital model (NMs) from said plurality of new digital models according to one or more criteria applied to the sets of parameters associated with said new digital models. Figure to be published with the abstract: Fig. 1

Description

Title of the invention: Method for generating a digital model of a part of a lighting device of a motor vehicle

[0001] The invention relates to the field of automotive lighting devices. More specifically, the invention relates to a method for generating a digital model of a part of a lighting device, in particular for lighting, signaling or interior lighting, of a motor vehicle.

[0002] The lighting systems of motor vehicles include massive parts generally produced by injection molding of plastic or metal, such as optical components, mounting plates for optical modules, aesthetic components, and heat sinks. Due to the materials used and the injection molding process by which they are manufactured, these parts cannot be optimized, particularly with regard to their weight, density, geometric configuration, constituent materials, or carbon footprint.

[0003] However, in a context of reducing the weight, cost, and carbon footprint of lighting devices, it might be desirable to design some of their components so that they can be produced using additive or subtractive manufacturing processes, or 3D printing, from bio-based materials. Nevertheless, given the multitude of parameters likely to influence these design and / or production constraints, it is complex to define the optimal configuration of a given component.

[0004] There is therefore a need for a design process for a part of a lighting device that allows for a rapid achievement of an optimal configuration with regard to a set of design and / or production constraints.

[0005] The invention thus falls within this context and aims to meet this need.

[0006] To this end, the invention relates to a method for generating a digital model of a part of a lighting device of a motor vehicle, implemented by a computer system, comprising: a. a step of providing an original digital model of at least one part of the lighting device; b. a generation step, from said original model, of an optimal digital design space in which said original model of the part can extend; c. a generation step, using at least one generative design algorithm, of a plurality of new digital models of said part, each new digital model extending within said design space optimal and presenting a set of predetermined design and / or production parameter values ​​for said part according to this new digital model. d. a step of selecting a numerical model from said plurality of new numerical models according to one or more criteria applied to the sets of parameters associated with said new numerical models.

[0007] The invention thus proposes, starting from an existing model of a part, to automatically and digitally design, using one or more generative design algorithms, different configurations, or models, of parts functionally similar to the original part. Each configuration, however, varies in its design with respect to one or more design and / or production parameters likely to influence the weight, cost, mechanical or thermal characteristics, and carbon footprint of said part. It is thus possible to select, according to said design and / or production parameters, the optimal configuration with respect to a given design and / or production constraint. This optimal configuration could, for example, be the one that, with regard to the weight and materials from which it is designed, minimizes the carbon footprint of the part or that minimizes the cost of the part.

[0008] In the context of the present invention, and by way of non-limiting example, "computer system" means a desktop computer, a laptop computer, a tablet computer, test equipment, a computer server, a controller, a processor, a computing engine, and / or any combination of such equipment, appropriate to the method according to the invention. Furthermore, the steps of the method according to the invention may all be performed on the same device. Alternatively, the steps of the method according to the invention may be performed on several devices, connected to each other by a wired connection and / or by a wireless connection.

[0009] It may in particular be conceived that the method according to the invention is a software executed alone by the computer system or is executed as a module in a computer-aided design software.

[0010] In the context of the present invention, and by way of non-limiting example, "part of a lighting device" means an optical part such as a reflector or a lens, a plate intended to support an optical module, an aesthetic part, a heat sink.

[0011] In the context of the present invention, and by way of non-limiting example, a "digital model of a part" means a set of elementary digital objects representing said part. The digital model may, for example, be a 3D model, namely one or more structured meshes, each comprising a plurality of vertices connected to each other to define a plurality of faces or volumes. representing approximately the surfaces and / or volumes of the room. The 3D model can be stored in a digital file as a list of vertices, each associated with spatial coordinates and connections to other vertices, it being understood that any other suitable form of storage may be considered. Alternatively, the digital model could be an unstructured set of points, namely a point cloud devoid of any information regarding the connections between the points.

[0012] In the context of the present invention, and by way of non-limiting example, "generative design algorithm" means an algorithm capable of autonomously or semi-autonomously generating, from a digital model of a part, another digital model of that part which satisfies one or more design and / or production constraints.

[0013] In one embodiment of the invention, the step of generating said optimal design digital space involves identifying an envelope of the maximum space of the lighting device that can be occupied by said part. For example, said maximum space could be the space of coincidence, or exclusive AND, of, on the one hand, the maximum envelope containing said part and, on the other hand, the lighting device. In this embodiment, the optimal design digital space thus corresponds to the maximum volume that any variation of the original part is likely to occupy with respect to the other parts of the lighting device, or even the environment of the lighting device in the motor vehicle.

[0014] Advantageously, the generation step of said optimal digital design space includes a substep of extracting, from said original digital model, critical areas of said part with respect to the lighting device and a substep of estimating load constraints of said part from said critical areas, the generative design algorithm receiving as input said critical areas and said load constraints.

[0015] The term "critical zone" includes, in particular, attachment points of said part to another part; load zones that may be subjected to compression, tension, shear, torsion, and / or bending; vibration zones, particularly resonance or damping zones; heat zones, particularly dissipation or concentration zones; and functional zones that may support another part. It may be provided that said critical zones are labeled in the original digital model, the extraction of said critical zones thus corresponding to a segmentation of said labeled zones in the original digital model. Alternatively, it may be provided that said critical zones are segmented manually by a user through a visualization interface. and manipulation of the computer system on which said original digital model is displayed.

[0016] Advantageously, said load constraints can be estimated, automatically or semi-manually, by simulation, in particular by a finite element analysis, also called FEA (from the English "Finite Element Analysis").

[0017] In one embodiment of the invention, said generative design algorithm is a topological optimization algorithm arranged to determine said plurality of new numerical models, each extending within the optimal design numerical space and each exhibiting a configuration that optimizes a cost function determined from one or more of said predetermined design and / or production parameters and said load constraints. This topological optimization algorithm thus determines, iteratively or non-iteratively, the value or values ​​of said design parameter(s) that optimize said cost function.

[0018] In one example, the algorithm iteratively modifies the values ​​of each of the design and / or production parameters until a stopping condition is reached, such as a predetermined number of iterations and / or the crossing by the cost function of a minimum or maximum threshold value.

[0019] Said design and / or production parameters may, for example, be the materials composing the part, the density and / or distribution of these materials, or even the geometric configuration of the part. Said cost function may, for example, be determined from the mass of the part, the cost of the part, the carbon footprint of the part, the natural frequencies of the part, the mechanical strength of the part, the heat dissipation of the part, or any other design and / or production constraint likely to be influenced by the design and / or production parameters.

[0020] In one embodiment of the invention, the topological optimization algorithm is a SIMP-type algorithm. This SIMP algorithm (from the English "Solid Isotropy Material with Penalization") defines an initial mesh of finite elements filling the optimal design digital space and iteratively adjusts the distribution of the elements and / or the density of each finite element to maximize or minimize said cost function.

[0021] In another embodiment of the invention, said generative design algorithm is an evolutionary algorithm arranged to iteratively determine said plurality of new digital models from a plurality of initial digital models generated from the original digital model, the algorithm being arranged to select, at each iteration, at least two digital models resulting from the previous iteration by evaluating a cost function determined from one or more of said design and / or production parameters predetermined and / or said load constraints, to cross the selected numerical models and to iteratively modify in the optimal design digital space, by variation of the value of one or more of said predetermined design and / or production parameters, said crossed numerical models.

[0022] This type of algorithm, also called a genetic algorithm, may, for example, include a first step of segmenting the original digital model into a plurality of zones, a second step of generating said plurality of initial digital models by random and / or heuristic and / or statistical variation of the design and / or production parameters in one or more of said zones of the original digital model. These initial digital models thus form a first generation of models on which a first iteration is applied. The resulting crossed digital models then form a second generation on which a second iteration can be applied, with the selection, crossover, and modification steps being iterated until a stopping condition is reached.

[0023] In yet another embodiment, said generative design algorithm may be a machine learning algorithm, for example of the neural network, support vector machine or random forest type.

[0024] It may be conceived indifferently to employ only one of the algorithms of the embodiments which have been described to obtain said plurality of new numerical models, or to employ several of these algorithms, for example in parallel, said plurality of new numerical models being formed by the sum of the new numerical models resulting from each of these algorithms, or in a sequential manner, the models from one algorithm being provided as input to another algorithm.

[0025] Advantageously, said predetermined design and / or production parameters include one or more of the following parameters: material type, weight, natural frequency, stiffness, thickness, deformation limit, cost, and carbon footprint. Materials that may be used in the generative design algorithm may include bioplastics, such as polyamides (PA11 or PA12), which may optionally be reinforced with carbon fibers (PA11CF) or glass beads (PA12GB). Deformation limits that may be used in the generative design algorithm may include, in particular, maximum tensile strength or yield strength. The carbon footprint may include the amount of carbon dioxide emitted during the manufacturing of the part.

[0026] Preferably, the selection step may be a manual, semi-automatic or automatic selection step of a numerical model from said plurality of New digital models can be defined according to one or more criteria applied to the sets of parameters associated with said new digital models. For example, the computer system could calculate, for each new digital model, an index using the same calculation method associated with a given design and / or production constraint, based on all or part of the predetermined design and / or production parameters, such as a weighted average. The selected model could then be the one with the highest or lowest index.

[0027] Alternatively, the selection may be implemented by a machine learning algorithm, in particular a classifier, trained to select a digital model from among a plurality of digital models based on the design and / or production parameters associated with these models.

[0028] In one embodiment, the generative design algorithm is configured to generate each new digital model as a point cloud, and the process includes a step of transforming the selected digital model into a mesh of elementary surfaces. In this embodiment, the selected digital model can thus be transformed into a file, for example of type STL, which can be used by an additive or subtractive manufacturing device.

[0029] Advantageously, the process comprises a step of supplying said mesh of elementary surfaces to an additive or subtractive manufacturing device and a step of manufacturing said part according to said digital model selected by said additive or subtractive manufacturing device from said mesh of elementary surfaces. The additive manufacturing device could, for example, be an MJF (Multi Fusion Jet) type 3D printer.

[0030] The invention also relates to a computer system capable of implementing the process according to the invention.

[0031] The invention also relates to a computer program product comprising instructions which, when the program is executed by a processor, lead the latter to implement the steps of the process according to the invention.

[0032] The invention further relates to a computer-readable storage medium comprising portions of code from a computer program intended to be executed by a processor to implement the steps of the process according to the invention.

[0033] The invention also relates to a part of a lighting device of a motor vehicle designed using the method according to the invention.

[0034] The present invention is now described by means of purely illustrative and in no way limiting examples of the scope of the invention, and from the accompanying drawings, in which the various figures represent:

[0035] [Fig.l] represents, schematically and partially, a method for generating a digital model of a part of a lighting device according to an embodiment of the invention;

[0036] [Fig.2] represents, schematically and partially, the implementation of a step of generating an optimal design digital space of the process of [Fig.1];

[0037] [Fig.3] represents, schematically and partially, the implementation of a generation step, by a generative design algorithm, of a digital model of said part of the process of [Fig.1];

[0038] [Fig.4] represents, schematically and partially, different models obtained at the end of the step of [Fig.3].

[0039] It should be noted that in these figures the structural and / or functional elements common to the different variants may have the same references.

[0040] Of course, various other modifications can be made to the invention within the scope of the annexed claims.

[0041] A method for generating a digital model of a part of a lighting device of a motor vehicle is shown in [Fig.1] according to an example of an embodiment of the invention.

[0042] In the example described, the process is implemented by a desktop or laptop computer on which computer-aided design software is run, including a software module whose code allows certain steps of this process to be implemented.

[0043] In a step E0, an original digital model MO of at least one part of the lighting device is provided. In this example, the original digital model MO may have been selected by a user from a set of part models representing the lighting device HL, using computer-aided design software. Alternatively, the original digital model may be a file, for example in XML or STEP format, uploaded to the computer-aided design software.

[0044] In the following description, said part is a plate intended to be fixed to a housing of the lighting device and to support an optical module of this lighting device. The method according to the invention can be used for other types of parts, and in particular an optical part such as a reflector or a lens, an aesthetic part, or a heat sink.

[0045] In a step El, an optimal design digital space EN is generated from the original digital model MO.

[0046] The different sub-steps of this step El, which generates the optimal design digital space EN, from the original digital model MO (represented in hatching) and the rest of the lighting device HL, are shown in [Fig.2].

[0047] To this end, step El includes a substep of segmentation Eli, in the space in which the lighting device extends, of an envelope, such as a box BB, encompassing the part represented by the original model MO. This box BB can be automatically generated from the maximum dimensions of the original model MO along three orthogonal axes.

[0048] In a second step E12, the optimal design digital space EN is generated by estimating the coincidence space, obtained using an exclusive AND operator, between the box BB on the one hand and the rest of the lighting device HL on the other. The optimal design digital space EN thus corresponds to the maximum volume that any variation of the original part is likely to occupy with respect to the other parts of the lighting device HL, or even the environment of the lighting device HL in the motor vehicle.

[0049] Furthermore, in a third substep El3, critical zones Z are extracted from the original digital model MO. These critical zones Z can then be labeled in this model MO and can therefore be automatically segmented. Alternatively, it may be provided that said critical zones are segmented by a user, for example by manually defining, via the design software, regions in the model MO corresponding to these critical zones Z, the parts of the model MO corresponding to the selected regions being segmented to define said critical zones Z.

[0050] In the example of [Fig. 2], the critical zones Z correspond to load zones, namely points of attachment of said part to another part, and functional zones capable of supporting another part. These load zones are thus likely to be subjected to compression, tension, shear, torsion, and / or bending. Alternatively, the critical zones Z may correspond to vibration zones, in particular resonance or damping zones, or to heat zones, in particular dissipation or concentration zones.

[0051] Finally, in a fourth step E14, CC load constraints are estimated from the original MO model and the critical zones Z. The CC load constraints can, for example, be estimated by simulation through a finite element analysis, or FEA, implemented by a software module of the design software.

[0052] In a step E2, new digital models MNj of the part are generated by one or more generative design algorithms from the critical zones Z, the load constraints CC, and the optimal design digital space EN. The generative design algorithm(s) implemented in this step E2 are thus designed to vary, in order to arrive at a new digital model MNj, design and / or production parameters of the part so that this new digital model MNj extends within said optimal design digital space EN while satisfying a cost function determined from one or more of these predetermined design and / or production parameters and critical zones Z; and / or said load constraints CC.

[0053] It may be provided that this or these generative design algorithms are likely to vary one or more of the following design and / or production parameters: type of material, weight, natural frequency, stiffness, thickness, deformation limit, cost, carbon footprint.

[0054] We have thus represented in [Fig.3] an example of the generation of a new numerical model MNj by a topological optimization type algorithm SIMP.

[0055] In this example, in a first substep E21, the algorithm defines an initial mesh of finite elements, namely voxels, filling the optimal design numerical space EN and in which the critical zones Z; are positioned.

[0056] Then, in a second step under E22, starting from the critical areas Zi5, it adjusts the density of each finite element to minimize a cost function determined from the mass of the part, the cost of the part, the carbon footprint of the part, the natural frequencies of the part, the mechanical resistance of the part, the thermal dissipation of the part.

[0057] Substep E22 is thus iterated until a stopping condition is reached. This stopping condition corresponds to the cost function crossing a minimum threshold value.

[0058] At the end of the last iteration, in a substep E23, the point cloud thus obtained can be transformed into a mesh of elementary surfaces forming the new numerical model MNj.

[0059] Each implementation of the SIMP algorithm thus results in a new numerical model MNj, differing from the other MNj models and the original MO model by a set of design and / or production parameters. Figure 4 shows a table representing three distinct models MNb, MN2, and MN3, as well as design and / or production parameters such as manufacturing process, material, weight, natural frequency, cost, and carbon footprint.

[0060] It may be foreseen, in other sub-steps not shown, that a finite element analysis will again be implemented on this new numerical model MNj, in particular to allow a user to validate his design with regard to the CC load constraints.

[0061] In other unrepresented variants, the generative design algorithm may be an evolutionary or genetic algorithm arranged to determine iteratively generates said plurality of new numerical models MNj from a plurality of initial numerical models generated from the original numerical model MO, or from a machine learning algorithm, for example, a neural network, support vector machine, or random forest. Several distinct algorithms can be implemented in parallel, or sequentially, with the models produced by one algorithm being provided as input to another.

[0062] At the end of each of the implementations of the generative design algorithm(s), one of the numerical models among said plurality of new numerical models MNj is selected in a step E3.

[0063] For these purposes, the entire set of new MNj digital models can be presented to a user with their design and / or production parameters, via the design software, for example in the form of the table in [Fig. 4]. The user can then select one of these models, according to one or more criteria applied to these parameters.

[0064] Alternatively, an index may be calculated in a later step of the process, using the same calculation method associated with a given design and / or production constraint, with the model exhibiting the highest or lowest index, depending on the method used, being automatically selected. Alternatively, the selection may be carried out by a machine learning algorithm, in particular a classifier, trained to select a numerical model from among a plurality of numerical models based on the design and / or production parameters associated with those models.

[0065] In the example of [Fig.4], by applying a selection criterion aimed at selecting the model with the lowest weight and the lowest cost, the MN3 model will thus be selected.

[0066] In a step E4, the digital model MNj selected at the end of step E3 is transformed into a file, for example of type stl, and transmitted to an additive manufacturing device.

[0067] Finally, in a step E5, one or more parts can be produced by this manufacturing device from this stl file.

[0068] The preceding description clearly explains how the invention achieves its stated objectives, namely, to provide a method for designing a component of a lighting device that allows for the rapid achievement of an optimal configuration with respect to a set of design and / or production constraints. These objectives are achieved in particular through the use of one or more generative design algorithms that allow for the variation of one or more design and / or production parameters of the component that are likely to influence the weight, cost, mechanical or thermal characteristics, carbon footprint of said part, and thus to select, according to said design and / or production parameters, the optimal configuration with regard to a given design and / or production constraint.

[0069] In any event, the invention cannot be limited to the embodiments specifically described in this document, and extends in particular to all equivalent means and to any technically operative combination of these means.

Claims

Demands

1. A method for generating a digital model of a part of a lighting device of a motor vehicle, implemented by a computer system, comprising: a. (EO) a step of providing an original digital model (MO) of at least one part of the lighting device; b. (E1) a step of generating, from said original model, an optimal design digital space (EN) in which said original model of the part can extend; c. (E2) a step of generating, by at least one generative design algorithm, a plurality of new digital models (MN) of said part, each new digital model extending in said optimal design space and having a set of predetermined design and / or production parameter values ​​of said part according to this new digital model; d.(E4) a step of selecting a numerical model (MN j) from said plurality of new numerical models according to one or more criteria applied to the sets of parameters associated with said new numerical models.

2. A method according to the preceding claim, characterized in that the generation step (El) of said optimal design digital space comprises the identification of an envelope (BB) of the original digital model (MO) of said part, the identification of an envelope of a space of the lighting device (HL) that can be occupied by said part, the optimal design digital space (EN) being formed by the union of said envelopes.

3. A method according to any one of the preceding claims, characterized in that the generation step (El) of said optimal design digital space (EN) comprises a substep of extraction (El3), from said original digital model (MO), of critical zones (¾) of said part with respect to the lighting device and a substep of estimation (E14) of load constraints (CC) of said part from said critical zones, the generative design algorithm receiving as input the said critical zones and the said load constraints.

4. A method according to the preceding claim, characterized in that said generative design algorithm is a topological optimization algorithm arranged to determine said plurality of new numerical models (MN;) each extending in the optimal design numerical space (EN) and each having a configuration optimizing a cost function determined from one or more of said predetermined design and / or production parameters and said load constraints (CC).

5. A method according to the preceding claim, characterized in that the topological optimization algorithm is a SIMP type algorithm.

6. A method according to claim 3, characterized in that said generative design algorithm is an evolutionary algorithm arranged to iteratively determine said plurality of new digital models (MNi) from a plurality of initial digital models generated from the original digital model (MO), the algorithm being arranged to select, at each iteration, at least two digital models resulting from the previous iteration by evaluating a cost function determined from one or more of said predetermined design and / or production parameters and / or said load constraints (CC), to cross the selected digital models and to iteratively modify in the optimal design digital space (EN), by varying the value of one or more of said predetermined design and / or production parameters, said crossed digital models.

7. A method according to any one of the preceding claims, characterized in that said predetermined design and / or production parameters include one or more of the following parameters: material type, weight, natural frequency, stiffness, thickness, deformation limit, cost, carbon footprint.

8. A method according to any one of the preceding claims, characterized in that the generative design algorithm is configured to generate each new digital model (MNi) in the form of a point cloud, and in that it comprises a transformation step (E4). of the selected numerical model (MNj) into a mesh of elementary surfaces.

9. A method according to the preceding claim, characterized in that it comprises a step of supplying said mesh of elementary surfaces to an additive or subtractive manufacturing device and a manufacturing step (E5) of said part according to said digital model (MNj) selected by said additive or subtractive manufacturing device from said mesh of elementary surfaces.

Images