Simulation-Driven Continuous Engineering: Enabling Innovation Throughout the Product Lifecycle
CDFAM NYC Speaker Interview: Neel Kumar of Intact Solutions
Can you start by giving us some background info on Intact Solutions and the software solutions you offer?
Intact synonyms: entire, whole, complete; antonyms: broken, shattered, fragmented
Intact Solutions was founded in 1999 as a spinoff from the University of Wisconsin-Madison to commercialize meshfree analysis and simulation tools directly from native solid modeling representations, without meshing, conversion, or preprocessing.
The original technology relied on B-splines and other basis functions allocated on the non-conforming mesh of space and deployed novel methods for numerical integration and enforcing boundary conditions. All of this happened way before IGA, immersed FEA, finite cells, and other “meshless” methods in the same category. It was developed with funding from the National Science Foundations and became widely known after a press release on meshfree structural analysis of Michelangelo’s David Tell Them Where It Hurts.
Fast forward to about 2016, after several significant R&D projects and commercial successes (e.g. Scan&Solve for Rhino, DM Live Parts and Live Sinter, Intact.design, etc.). We realized that the real challenge and opportunity is not to keep developing new simulation codes, but to solve a much larger class of interoperability problems: any geometry, material, or FE solver should work in plug-and-play fashion supporting unprecedented complexity, heterogeneity, and automation.
Thus was born Intact.Simulation is a unique component simulation product that can be deployed anywhere in record time. The ability to handle complex simulation workflows and digital threads in additive manufacturing led to Intact.Additive, and because we always work with native representations and have access to all sensitivities, Intact.Generative encompasses a rich variety of shape, topology, and material optimization tools.
In addition to our product portfolio, the company is still a major player in advanced R&D projects, as witnessed by recent awards from NIST, DARPA, and ONR.
Your presentation at CDFAM NYC covers “Simulation-Driven Continuous Engineering.” What does this concept involve, how does it differ from other engineering methods and what are the benefits?
Let’s use an example to understand Simulation-Driven Continuous Engineering. Consider the additive manufacturing process: it starts with designers creating lightweight components using, say, lattices, often represented as implicit models.
During process planning, decisions such as build direction and scan strategy significantly affect the part’s performance. In manufacturing, data from in-situ monitoring and scanning can predict warping and defects that degrade the final part’s performance.
Performance prediction at all these steps can help optimize the design and process plan, as well as pre-qualify the printed part. Currently, there is no straightforward way to tackle all these performance prediction problems. Different geometric and material representations must be converted to boundary representations and then meshed. Additionally, there is no way to incorporate process and in-situ data to study their effect on the printed part’s performance.
This method breaks down traditional silos, enabling continuous improvement and optimization of products, which is even more important given advancements like computational design, additive manufacturing, digital twins, and the Internet of Things (IoT).
A Intact.Simulation workflow where simulation-driven design and validation are done in the same loop.
We achieve Continuous Engineering through a Common Neutral Simulation Format and Space and employing the Immersed Method of Moments.
This method works with native geometry representations without any conversion or meshing, irrespective of complexity. This, in addition to the ability to handle hybrid data types, allows continuity across all stages.
For example, it enables the exploration of larger parameter spaces in design, planning, or diagnostics. It also allows for the pre-qualification of metal additive parts, a significant focus in aerospace, and supports generative applications with Multiphysics objectives where designs are always validated.
On-going work that combines Generative capabilities with process-depenedent material properties.
We have several projects and workflows that demonstrate critical pieces of Simulation-Driven Continuous Engineering, which we will share in our presentation at CDFAM. This is the goal we are striving towards, and we believe it will significantly de-risk product development, enhance efficiency and innovation in engineering workflows.
Current simulation tools sometimes face challenges with complex geometry representations. Could you explain these challenges and how your approach addresses them?
Current simulation tools often require what’s known as “body-fitted” or “conformal” meshing, where the finite element mesh must resolve down to the smallest feature in the geometry. This approach poses significant challenges not only for complex geometries such as lattices with hundreds or thousands of cells, but also for traditional geometries with intricate details like fillets, small holes, embosses, and other small features. As a result, the meshing process often dominates the simulation time and routinely requires extensive manual tuning and pre-processing, such as removing small features and cleaning up the geometry to ensure it is “meshable”.
The conformal meshing process is also particularly sensitive to small defects in the geometry, such as slivers or gaps, and requires boundary representations (B-reps) or STL formats to be flawless. Additionally, conformal meshing is limited to working with B-reps and necessitates converting other geometry types, such as implicit models, G-code, or CT scans, into B-reps. This conversion process can be time-consuming and may introduce additional errors or inaccuracies, further complicating the simulation workflow.
IMM works with any geometric representation by decoupling the geometry from the analysis space. This also allows simulation resolution to be controlled independently of the geometry resolution.
This technique allows us to work directly with native geometry representations without the need for conformal meshing. Consequently, we can handle complex geometries, including lattices and intricate details, without any pre-processing or manual adjustments.
Our methods accommodate various geometry types, including implicit models, G-code, and CT scans directly. This flexibility not only reduces the time and effort required for meshing but also ensures more accurate and reliable simulations.
Our approach results in a simulation that is robust, fully automated, and seamlessly integrated, which can significantly de-risk product development.
Designers can explore larger design spaces and optimize complex designs more effectively, enabling them to innovate faster and with greater confidence. Since mesh resolution is independent of geometric resolution, you can easily perform fast coarse simulations during concept design and accurate refined simulations during validation. This consistency also allows the same simulation setup to be used for concept design, design validation, and part qualification, regardless of the geometry format at different stages.
What exactly is ‘meshless’ simulation? How do you represent material and forces in space if not via a mesh?
You can think of it as a background 3D “graph paper” that allows you to perform computations.
The mesh defines the basis (functions) for the simulation space and determines the quality of the analysis/simulations. So the real goal is to eliminate not the mesh, but meshing.
How? This is what is unique about Intact’s technology. Materials, boundary conditions, parameters, and sensitivities are intrinsic to the model and are always defined on native representations of geometry. The innovation happens at run time when all of this information is immersed into a mesh of space, to approximate solutions of the underlying differential and integral equations. As far as we can tell, Intact’s proprietary method of moments is superior in combination with accuracy, performance, and generality to just about any other FE technology today on the market, including other meshfree methods.
High-level overview of Immersed Method of MomentsTM
Rapid Design Study using Intact.Simulation for Synera
We also offer Intact.Simulation for Automation, which includes our Python API, PyIntact, as well as a Command Line Interface for headless operation. In addition to all the previously mentioned geometry formats, Intact.Simulation for Automation supports “slice” or “surface” models for analyzing composite plies or thin structures, as well as “Lines” and “Curves” for simulating G-code models.
This versatility allows our solution to be integrated with a wide range of design tools. For example, our partners have used our API to integrate with Houdini to create custom computational design pipelines.
Our goal is to provide a robust, automated, and flexible simulation solution that enables advanced custom design workflows fully integrated with the design tools that users already use.
What is the key takeaway you would like the attendees at CDFAM to take away from your presentation?
The key takeaway I want attendees at CDFAM to understand is that traditional finite element analysis (FEA) is not well-suited for computational design and exploring large parameter spaces, whether in design, process planning, or other applications.
In addition, computational design often employs a variety of tools, including non-traditional ones like Houdini and Blender. All these tools must be able to seamlessly integrate and work together to create an efficient and effective workflow.
Intact is a leader in meshfree simulation, providing a robust, automated, and easy-to-integrate simulation solution.
Our technology has been fully benchmarked and is already integrated with the most popular computational design tools, including Grasshopper, nTop, and Synera. Additionally, our Python API and headless mode support custom workflows, offering unparalleled flexibility and efficiency for users. This integration and automation enable designers to explore large design spaces more effectively, de-risk product development, and innovate with confidence.
Finally, What do you hope to gain from attending CDFAM NYC?
Attending CDFAM NYC presents an invaluable opportunity to learn about the latest trends in the adoption of computational design, including the various application areas and types of workflows being utilized.
Meeting and networking with professionals in this field will provide us insights into how different industries are leveraging computational design and help us identify how Intact’s solutions can be effectively integrated into their processes, as we believe we are still scratching the surface of what is possible. Additionally, we aim to better understand the simulation needs of the community.
Finally, we also want to spread the word about Intact’s unique simulation capabilities and the advancements it can unlock. As computational design is still emerging, users may be cautious about integrating new tools into their workflows. However, we believe that automated and integrated physics simulation is a key foundational element for the future of computational design.
By sharing our vision and capabilities, we hope to demonstrate how our technologies can streamline and enhance design processes, ultimately fostering greater innovation and efficiency within the community.