A Collaborative Study of Generative Design and Additive Manufacturing in Automobile Industry

 A Collaborative Study of Generative Design and Additive Manufacturing in Automobile Industry

Abstract—

A new emerging technology with great potential in the industry known as Metal additive manufacturing. Moreover, with the help of new and advanced technologies like generative design, we can maximize this potential for additive manufacturing solutions by computing complex optimized parts. recent studies on automotive design aim to cut vehicle and components weights, optimizing vehicle performances, and contribute to the challenge of reducing fuel consumption and operational costs. In this perspective, innovative materials and technologies are developed but also advances in design methods and tools. Generative Design is a different tactic to automatically augment the component design. The design process must be considered to accomplish the finest solution, about design parameters, requirements, and limits Additive manufacturing technologies can now be used in constructing metallic parts. This revolution in manufacturing technology makes way for the fabrication of new shapes and geometrical features

I. INTRODUCTION:

 Additive manufacturing (AM) methods have been usually used for rapid prototyping purposes for a great period of time in the last 30 years. They comprise of building an object “from scratch” or a semi-finished part acting as substrate. Thanks to many technological enhancements, these developments can now be used for swift manufacturing purposes [1]. Additive Manufacturing (AM) is an embryonic production technology in engineering every so often seen as the subsequent industrial revolution. Undeniably, AM is budding bigger every year and calculations on the related market see this trend growing even more, as a result of increased usage of this technology and expansion in the fields of application [2]. In contrast to orthodox manufacturing technologies, AM offers greater freedom of design and possibilities for mass customization [3]. Furthermore, this technology is capable of constructing prototypes or even finished parts in a short period without tooling or casts [4]. It can also yield intricate parts for both professional and personal uses. Additionally, evolving Computer-Aided Design (CAD) technologies like generative design can maximize the already huge potential of AM. The generative design tools help create optimized parts or assembly by employing computer power and optimization technologies [5]. Similar to the AM field, the generative design ecosystem keeps budding every year. There is the rising use of generative design resolutions in chief industries like the automotive and aerospace domains, as well as an increasing number of generative design tools.[2]

II. GENERATIVE DESIGN 

Generative Design (GD) is “a category of technologies that suggests design options, or optimizes an existing design, to meet criteria defended by the user” [5]. Indeed, designers lay down their part constraints and objectives in GD software. Then the software’s objectives are to propose an improved part design. The recommended options can be adjusted for weight, stiffness, frequency, etc. [6] In that way, GD has a substantial impact on the design process. Part of the design procedure is now automated by GD software, saving time for designers but also altering their orthodox working method. Their new role is now to make a comprehensive study of the part, so they can set up every applicable input (simulation parameters, criteria, and objectives) for the GD software. At the end of the GD process, it is also the designers’ role to evaluate the diverse design possibilities and choose amid them the best suited for their application,[7] Generative Design is a design method for apprehending the designer’s intent, engendering new solutions. Categorized by data-driven collaborative cloud-based technology, it depends on a highly automated activity. A set of parameters and rules (commands to the designer) is taken into consideration as the DNA of the design process; rules and parameters are considered as the genes. They are conjoined by evolutionary algorithms or even “brute force” calculations. The introduction of such trials into the design procedure allows the development of innovative design resolutions by adjusting the rules that outline a final design, difficult or impossible to attain via other methods. Grammar-based techniques exploit the principle of database strengthening the identification of rules, creating complex forms and patterns from simple stipulations. Generative design principles established particular attention in architecture; some good definitions coming from that field are reported here [8] Lars Hesselgren stated that Generative design isn't about designing the structure – It’s about designing the system that builds a structure. Paola Fontana states that it is the initial conditions of an object rather than modeling the final form in the Generative Design Process Kristina Shea appends that Generative Design Systems are made to create New Design Processes that generate spatially new yet effective and buildable designs through the exploitation of current computing and manufacturing capabilities.

A. Topology Optimization: There are different tools accessible for GD, for instance, lattice infill or meta-structure blending. But the most widespread tool, booming automated optimization, is topology optimization.[5] Topology Optimization (TO) is the technology optimizing the material layout within a specified design space.[9] Topology Optimization goals are to augment a part property (weight, stiffness, frequency …) while valuing a definite set of limitations. To do so, the Topology Optimization process uses various mathematical algorithms and methods. Each Topology Optimization method has several versions aiming at faster optimization or addressing characteristic optimization’s issues one of the most popular Topology Optimization methods is the Solid Isotropic Material with Penalization (SIMP) method. This distinct method idea is to give an element a continuous virtual density between 0 and 1 and steer the result to 0 or 1 after each iteration with a penalization factor. Another noteworthy distinct method is the Evolutionary Structural Optimization (ESO) method which is based on biomimicry. This discrete method uses finite component breakdown to determine and remove the inefficient components and make the structure evolve into its optima counterpart. Indeed, after each iteration, the components with the lowest stress density are removed until all the residual components have an equal stress density [10]. One direct improvement of this method, even closer to mimicking nature growth, is to also add components near the highest stress density components as well as eliminating the lowest stress density one.[11] The discrete element methods face some disputes of their own, the two most common are the mesh dependency issues, i.e., optimization results vary depending on the mesh, and the checkerboard issue, i.e., stiffness is virtually high due to a checkboard pattern of the elements. Addressing these issues, a recent Topology Optimization method emerged: The Movable Morphable Component (MMC) method. In opposition to the two previous methods, MMC is not distinct. The core idea is to find the optimal structure topology by optimizing the thickness, shape, orientation, and layout of a set of morphable components, i.e., building blocks [12] Topology Optimization: on the left SIMP method iterations [9], on the upper right ESO method iterations [10], on the bottom right MMC method iterations [12] B. Synthesis of Generative Design The study of GD is essential to comprehend the GD tools optimizing operations. It is also acceptable to review the various tools possibilities, similarities, and differences. Multiple Design Options of the same General Motors Seat Bracket proposed by Autodesk Generative Design tool [15]

III. GENERATIVE DESIGN WITHIN THE DESIGN FOR ADDITIVE MANUFACTURING APPROACH

 Nowadays it is commonly accepted that in automotive and industrial applications, we can attain the key advantages from the introduction of advanced design procedures if we associate them with Additive Manufacturing (AM) processes and techniques. Therefore, we must contemplate also the coupling of evolutionary algorithms with innovative manufacturing processes, like Additive Manufacturing, and new materials. This combination introduces more degrees of freedom in the final design concept: for example, the mixing of materials with different properties allows having different properties distributed in different zones of the same part, leading to multifunctional concepts. The opportunities offered by AM are not constrained to multi-functional concepts. Deliberate that, over the last years, AM’s implementation has increased across industries, with the aerospace industry contributing about 10.2% of AM’s global revenues in 2012. AM provides the flexibility to create complex part geometries that are problematic to build using traditional manufacturing, such as internal cavities or lattice structures that help reduce parts’ weight without compromising their mechanical performance.[13] Additionally, AM’s impact on economies of scale and scope makes it a natural fit for automotive, which is mainly geared toward customized production. The new system, once manufactured thanks to AM, should fulfil the practical requirements in an innovative and more efficient way, also targeting a humbler design and a substantial cost reduction. Novel structural materials and advanced AM techniques make these technologies ready to be presented within the generative design process also for safety critical contexts, such as the aeronautical. Even if the advantages from the introduction of the couple Generative Design and Additive Manufacturing have been widely considered.

IV. GENERATIVE DESIGN APPROACH IN PRODUCT DEVELOPMENT AND AUTOMOTIVE INDUSTRY 

Typical automotive design practices consider the design process spread into three main phases: the conceptual design, the preliminary design, and the detailed design. Conceptual Design is an early phase of the design process. Preliminary Design means that part of the Development Phase where all of the geometric design elements, Detailed design is the phase where the design is refined and plans, specifications, and estimates are created The design solution is proposed throughout the typical diverging – converging process, in relation to design necessities and limits. Multidisciplinary optimization processes are currently being developed to support the designer in assessing the optimal solution, in relation to all design features and constraints. Generative Design is a novel form-finding process that considers structural performances, material properties, and ergonomic demand, throughout an automatic iterative holistic approach for component topology optimization.[14]

V. GENERIC GENERATIVE DESIGN WORKFLOW 

Before starting the GD process, an initial study relating to the part Specification is required. This study is decisive since all the final GD options projected in the end rely on it. Meaning, the designer needs to have a seamless understanding of all part interactions with its environment before starting the GD process entire volume available in which GD software can operate, every space outside the design space cannot be used. Thus, according to its former study of the part, the designer’s role is to determine the utmost space the part could ft in, so the results at the end of the process are the most augmented ones. To complete the design space, the designer also needs to specify the conserved geometries and voids within the design space, i.e., the functional volumes. Then, the designer applies the set of loads and boundary conditions, i.e., the load cases, on this design space. Generic Generative Design Workflow [16] A. Translation Phase The first phase in the GD process, based on the outcomes of the past functional analysis, is translation. In this phase, the part specifications are “translated” into inputs for the GD software. The designer must define the basis of his optimization: the design space. The design space is the entire volume available in which GD software can function, every space outside the design space cannot be utilized. Thus, according to its preceding study of the part, the designer’s role is to govern the maximal space the part could fit in, so the results at the end of the process are the most optimized ones. To complete the design space, the designer also needs to mention the preserved geometries and voids within the design space, i.e., the functional volumes. Then, the designer applies the set of loads and boundary conditions, i.e., the load cases, on this design space B. Processing Phase To begin with the second phase, the designer needs to arrange his/ her optimization, i.e., to point out the optimization objectives and restraints. Some usual parameters are the material, the percentage of the design space volume for the solution, the minimal element size, etc. The designer can now run the optimization process. The second step of the GD process is the computing, this step usually does not include the designer except for software proposing live feedback and live alteration on the optimization runs. With all the previous step information, the software finds solutions attaining the specified objectives and constraints. In the third step of the GD process, the designer must evaluate the options proposed by the GD software to find the one best suited to its requirements. The designer’s proficiency is once again decisive as the choice is based on the part context in terms of cost, manufacturing, production, quality, etc. If the designer is not satisfied by the GD proposed options or if he wants to enhance the optimization results, he can loop back to the first step to adapt the optimization parameters and produce new options. In the fourth step, the designer can refine even more the solution subsequently by blending complex meta-structure in the optimized part. For example, some software proposes to blend complex meta-structures like lattices, gyroids, bone-infill-like structures, etc. Once this augmentation is done, the designer should finally run an analysis of his/her solution to determine its performance and certify the validity of the part specifications. If the specifications are not met, the designer should loop back to the first set-up step and modify the initial parameters.

VI. CASE STUDY on How Generative Design Enhances Autonomous-

Vehicle Development [19] The technological ramp to completely independent vehicles present striking challenges for the company’s imperative independent-vehicle (AV) programs. Advanced detector technology, high-speed and highbandwidth data networks, and slice-edge artificial intelligence are all decisive to the functional and marketable success of AVs. In addition, utmost approximations prognosticate that AVs will bear billions of country miles worth of testing to certify their safety. Manufacturers will need to integrate the assignments learned through dissembled and real-world testing into their AV designs to remain competitive. Sensors Field of Perception in Autonomous Vehicles [17] 1. Autonomous- vehicle platforms must connect an array of advanced detectors and computers through high-speed data networks to perceive, assess, and act on environmental stimulants. The Society of Automotive Engineers (SAE) states six situations of complication for independent vehicles, from zero to five. An auto with position-two autonomy may feature active voyage control, a lane-departure warning system, lane keep help, and parking backing. In total, this auto requires about 17 detectors to enable its motorist-backing systems. The calculations performed by such an auto’s automated systems are fairly primitive. The lane- keep- help system, for case, is only assigned with covering the vehicle’s position relative to the lines of the road. Should the motorist begin to transgress, the system will notify the motorist or take corrective action, but the ultimate responsibility for control of the vehicle lies with the motorist. A position-five AV will have complete control over the driving task, taking no mortal input. As a result, a position-five auto is projected to have further than 30 fresh detectors of an important wider variety to cover the immense number of tasks an independent vehicle will need to attain.On top of the ultrasonic, compass camera, and long-and short-range radar detectors of a position-two auto, position five will bear long-range and stereo cameras, LiDAR, and dead-reckoning detectors. The increase in detectors will increase the quantum of wiring demanded in the harness and the necessary computational coffers to handle the gigabits of data being produced by the detectors. 2. A completely independent vehicle will bear numerous types of detectors to directly perceive dynamic driving surroundings. While designing a position-five platform, masterminds will need to perform armature and concession analyses to examine architectural proffers, similar to a centralized vs. sphere vs. distributed armature. These analyses will need to regard for hundreds of factors and millions of signals while enhancing function locales, network quiescence, error rates, and more. Despite these challenges, the independent drive is a raising request. At least 144 companies have blazoned AV programs. Some of these are major automotive manufacturers seeking to stay ahead of the coming assiduity interruption, but utmost are startups or companies from other diligence seeking to enter a traditionally impenetrable request. These companies warrant assiduity-specific experience and the engineering coffers to spontaneously force their way through the complications of independent vehicle design. Indeed the major automotive OEMs will face problems that their heritage design overflows are ill-equipped to handle. To contend, these companies will need a new design methodology that permits youthful masterminds to design accurate and optimized systems, which can only be done by landing the experience and knowledge of expert masterminds. They will need generative design. Generative Design and Engineering Generative design takes system delineations and conditions as input and generates architectural flings for the sense, software, tackle, and networks of the electrical and electronic systems using rules-grounded robotization. These rules capture the knowledge and experience of the expert masterminds to guide youngish masterminds throughout the design. Landing this IP helps companies to advance both vehicle infrastructures and new generations of masterminds as they learn and apply being company knowledge. 3. Generative design uses rules-grounded robotization to induce proffers for the sense, software, tackle, and networks of the electrical and electronic (E/ E) system. The adding electrical and electronic content of ultramodern vehicles is formerly conclusive current design styles to their limits, yet the complexity of automotive systems will only continue to grow in the future. Autonomous buses will contain the most complicated electrical and electronic systems yet seen in the automotive assiduity. For case, further, than 30 detectors, country miles of wiring, and hundreds of ECUs will be needed to gather, move, and process the data necessary for independent driving. The data networks will need to be extremely fast to support real-time perceptivity, decisiontimber, and action to help collisions and detriment to mortal passengers or climbers. Masterminds developing these vehicles will also need to balance performance conditions against power consumption, physical space constraints, weight, and thermal considerations. Generative design authorizes automotive masterminds to attack the challenges of electrical and electronic systems design for independent vehicles. It employs rulesgrounded robotization for rapid-fire design admixture, enables masterminds to design in the environment of a full vehicle platform, and tightly integrates colorful design disciplines to ensure data abidance. Originally, employing robotization throughout the process will help design brigades manage design complexity without adding time-torequest. Robotization helps masterminds concentrate on the most dangerous aspects of the design and verification of the functionality of the E/ E system’s functionality and reduces crimes from homemade data entry. This empowers masterminds to concentrate further of their time on applying their creativity and inventiveness to creating the coming generation of automotive technology improvements. Robotization also applies company IP to the generated proffers through design rules, adding the delicacy and quality of the designs. Next, designing in the full platform environment helps masterminds to understand the way signals, cables, and other factors are enforced across the entire vehicle platform, thereby reducing crimes at interfaces or due to the complexity of the harness. This design inflow also enables brigades to exercise undisciplined data across vehicle platforms to ameliorate quality and reduce development costs. Eventually, a tightly unified terrain enables the electrical masterminds to partake data with masterminds and tools in other disciplines, similar to mechanical or PCB design. The relations between the electrical, mechanical, and software factors of a vehicle are adding. Flawless synchronization of data between these disciplines improves the integration of them into a single system. Generative design also creates a nonstop thread of data from the primary system description and conditions to full-scale product and service. The same data forages each stage of the generative design inflow so that nothing is lost between design stages or design disciplines. This nonstop thread of data keeps all engineering platoon members up to date and working with the most current data while also icing that designs are meeting colorful conditions for functionality, safety, weight, and so forth. 4. Generative design ensures data continuity from initial system definitions through production and after-sales for full traceability and acquiescence with requirements. Data continuity ensures that projects have a single data source, providing a clear picture of the innumerable inter-domain and inter-system interactions. Designs can be automatically checked against design rules to ensure their functionality, accuracy, and quality. As changes are made to the design, they can be examined with detailed impact analysis that will inform the engineer of issues the change may cause in other domains. For instance, moving or removing an ECU could be evaluated for its impact on network timing, signal integrity, or physical clearance and collision issues. As a result, changes are made knowing their full impact on the system. Generative design will be a key enabler for new and established automotive companies in their quest of developing fully autonomous vehicles. The ability to generate electrical system architectures automatically enables early exploration and optimization of designs while entrenching company IP into the design flow. In addition, a singular source of data promotes consistency between domains, design reuse, and enhances the analysis of change impact. Finally, tight integrations between the electrical domains and with mechanical and product lifecycle management tools streamline the entire design flow from the outset through production. The massive intricacy intrinsic in AV design will continue to push the tools and methodologies used by automotive engineers. This is especially true in the electrical and electronic systems domains as they come to dominate the operation of a vehicle’s safetycritical systems and conveniences. The winners in this disruptive technology will be those companies that can most effectively integrate the advanced technologies required for an autonomous drive into a package that’s reliable, safe, and attractive to consumers, and then get those technologies to market quickly and with a high level of quality.

VII. CONCLUSION 

Generative Design is an innovative procedure to support the designer in widely exploring the design space. It is not only a topology optimization, nor an evolutionary algorithm, but it syndicates several optimization modules to topology definition within a CAD environment, according to design requirements, limits, and the bounding space. The output is not only the most appropriate solution, while it is a family of different results that the designer could properly select and modify. The solution space is generally established considering freeform shapes: it would not be possible to reach a better solution using a traditional design method. Additionally, the selected shape is designed to be manufactured by an Additive Manufacturing process. Even if some case studies and some tools have been developed, the potentials brought by Generative Design principles are not yet discovered enough. Some examples have been well-thought-out in small component design. Additive manufacturing is formerly being used at the sports auto manufacturer in prototype construction, manufacturing spare corridor for classic sports buses as well as in other areas. For the first time, the pistons for the high-performance machine of the 911 flagship model were designed using topology optimization, the GT2 RS, are now being produced with a 3D printer with the help of generative design. General Motors also used generative design before this time in an evidence-of-conception design to develop a featherlight seat-type prototype for its electric buses of the future. The technology is also proving its value for the future of space exploration.

ACKNOWLEDGMENT 

The work presented in this paper was to a very large extent based on Assignments for Bachelor’s Degree Conducted by Prof. Rajeshwar Deshmukh

REFERENCES [1] Vayre, B., Vignat, F., & Villeneuve, F. (2012). Metallic additive manufacturing: state-of-the-art review and prospects. Mechanics & Industry, 13(2), 89-96.

 [2] Wohlers, T. (2014). 3D printing and additive manufacturing state of the industry. Annual Worldwide Progress Report. Wohlers Associates.)

 [3] Frazier, W. E. (2014). Metal additive manufacturing: a review. Journal of Materials Engineering and Performance, 23(6), 1917-1928.

 [4] Gibson, I., Rosen, D. W., Stucker, B., & Khorasani, M. (2021). Additive manufacturing technologies (Vol. 17). Cham, Switzerland: Springer.

 [5] Schnitger, M. (2018). An introduction to generative design. Cadalyst. Longitude Media.

 [6] Khan, S., & Awan, M. J. (2018). A generative design technique for exploring shape variations. Advanced Engineering Informatics, 38, 712-724.

 [7] Krish, S. (2011). A practical generative design method. Computer-Aided Design, 43(1), 88-100.

 [8] https://generativedesign.wordpress.com/2011/01/29/ what-is-generative-designing/

 [9] Bendsoe, M. P., & Sigmund, O. (2003). Topology optimization: theory, methods, and applications. Springer Science & Business Media.

 [10] Liu, X., Yi, W. J., Li, Q. S., & Shen, P. S. (2008). Genetic evolutionary structural optimization. Journal of constructional steel research, 64(3), 305-311.

 [11] Zhao, F. (2014). A nodal variable ESO (BESO) method for structural topology optimization. Finite elements in Analysis and Design, 86, 34-40.

 [12] Guo, X., Zhang, W., Zhang, J., & Yuan, J. (2016). Explicit structural topology optimization based on moving morphable components (MMC) with curved skeletons. Computer methods in applied mechanics and engineering, 310, 711-748.

 [13] Lipson, H. (2012). Frontiers in additive manufacturing. Bridge, 42(1), 5-12.

 [14] Hemmerling, M., & Nether, U. (2014). Generico-A case study on performance-based design.

 [15] Danon, B. (2018). How GM and Autodesk are using generative design for vehicles of the future. Autodesk.

 [16] Briard, T., Segonds, F., & Zamariola, N. (2020). GDfAM: A methodological proposal of generative design for additive manufacturing in the automotive industry. International Journal on Interactive Design and Manufacturing (IJIDeM), 14(3), 875-886.

 [17] Burcicki, D. (2019, 01 25). How Generative Design Enhances Autonomous-Vehicle Development. Retrieved from ElectronicDesign:electronicdesign.com/markets/auto motive/article/21807500/how-generative-designenhances-autonomousvehicle-developmen