I recently had the pleasure of visiting Dassault Systèmes’ Simulia headquarters for their first ever Simulia analyst conference. As Dassault Systèmes’ CAE (Computer Assisted Engineering) brand, Simulia encompasses all DS simulation solutions. The one-day event provided the Simulia team an opportunity to show off some of their latest technologies through a marathon of presentations and a handful of demos. Topics included Virtual Human modeling, Augmented and Virtual Reality, Conceptual Engineering, and wearable electronics, but the Additive Manufacturing presentation and the Systems Engineering presentation and demo stuck out to me as the most interesting of the day. (For a general overview of the event, check out Dick Slansky’s recent blog.)
It’s no secret that your typical CAD software falls short when it comes to designing for additive manufacturing (AM). The truth is, your typical design engineer falls short as well. These shortcomings are particularly evident when dealing with material properties, utilizing organic shapes, and designing for manufacturing. Simulia has developed a set of simulation solutions to alleviate these issues and furthermore, empower engineers to innovate in the field of additive manufacturing:
- In-Silico Materials Engineering: In conjunction with Biovia, which encompasses DS’s chemical, materials, and bioscience research software, Simulia continues to develop methods to predict macro-scale material properties based on microscale structure and chemistry. This will be an invaluable tool for companies to develop application specific materials for both polymer and metal additive manufacturing.
- Function Driven Generative Design: To capitalize on the ease by which AM can produce highly complex organic shapes, engineers need modeling tools that can generate these complex forms to fit specific criteria. Simulia’s generative design software takes a set of parameters applied to a basic design and optimizes that shape for reduced stress, decreased weight, and printability. The software can also generate non-parametric lattice structures for greater light-weighting capability. A limitation of this approach, which is inherent to most generative design tools I have seen, is the resulting element model must be manually reconstructed. Simulia’s solution is an intuitive manual tracing process. While it appears more elegant than other methods I have seen, I would love to see solution that requires little to no post processing. (Frustum, a small startup, claims to have solved this issue).
- Process Definition and Production Planning: Simulia’s approach to the modeling the AM process is composed of a five-part framework: heating inputs, cooling boundaries, time fidelity, part fidelity, and material evolution. The method allows simulation of support structures, path, process, & build information, localized heating, material phase changes & layup, the combination of which provides predictions of part distortions, build time, and other manufacturing metrics. Although it’s not as flashy as generative design, I consider it the most valuable tool of the set when it comes to the advancement of AM from a prototype to production.
Systems engineering and simulation is vital to the success of product development in industries like automotive and aerospace where projects take years and result in an intricate convergence of numerous engineering domains. As these systems grow in complexity (for instance, when a wire routing change in a car can affect the performance of nearby sensors, thus degrading transmission performance) simulating and validating designs earlier in the development phase becomes increasingly important. Success necessitates a comprehensive library of Multiscale, Multiphysics tools like the one Simulia has assembled through acquisitions and internal development.
Simulia’s philosophy is to push system simulation beyond a validation of function to a validation of experience. This undertaking was exhibited at the event with a live demonstration: visitors were invited to test out a driving simulator, complete with steering wheel, break and accelerator pedals, and a 120° three monitor display. The demo took driver inputs and fed them in real time through a pre-defined multi-body, Multiphysics simulation, which provided haptic and visual feedback to the user. At any time throughout the simulation, parameters like ride height, spring rate, steering ratio, and transmission control logic could be adjusted, and the new experience evaluated. The demo, as expected, lacked a level of immersion necessary to really experience the simulation, but it was an excellent showcase of the power of the software. I was told Dassault has a fully immersive simulation room at their Boston Campus, which I suspect is a much more powerful experience – I look forward to checking it out.