Computational Mapping of Biomimetic Structures in Response to External Loading Conditions
CDFAM Expert Speaker Series Interview with Matt Shomper
With a background in both mechanical and biomedical engineering, Matt Shomper has designed some of the most advanced medical devices produced using 3D printing. Possessing deep expertise in engineering metamaterials through computational design, as well as experience in navigating complex designs through the FDA approval process, Matt has pushed the boundaries of what is possible with computational design and additive manufacturing.
With his newly launched consultancy, ‘Not A Robot,’ Matt offers advanced engineering services in the design of not only products but also the systems necessary to build computational workflows for medical devices and beyond.
In the interview, Matt discusses his presentation at the CDFAM Symposium on the topic of Computational Mapping of Biomimetic Structures in Response to External Loading Conditions.
He talks about the benefits of using algorithms that mimic natural responses to external loading conditions, his experience in creating and validating bio-inspired structures for medical applications, and how these structures could be utilized in other engineering fields.
Shomper also delves into the challenges of selecting the right natural structure for a specific engineering problem and the difficulty of modeling and optimizing these structures. He briefly discusses the consulting services provided by his company, Not A Robot Engineering, and how his expertise can benefit clients in various industries.
Can you start by giving us an overview of the topics and key points you will be discussing in your presentation at the CDFAM Symposium, which is entitled ‘Computational Mapping of Biomimetic Structures in Response to External Loading Conditions‘?
Absolutely! So, most people that follow my work know that I’m a little obsessed with all things biomimicry – nature as applied to human-use products.
Most natural structures are optimized over time as a response to their environment, whether that’s forces, temperatures, sunlight, fluid flow, etc. This is directly opposite to how most engineers and designers create products – entirely static and rigid. There’s an attempt in design to make these structures more organic, but ultimately I’ve found most of the attempts is just slapping radii on already pre-formatted geometry.
One of the things I’d like to show is my recent attempts at creating algorithms that naturally respond to their input conditions. What they create sometimes doesn’t jive with an engineering understanding of the problem, but I think this is why it’s so fascinating – it forces us out of our rigid box of the “known” and into this space where we kind of let the solution manifest of itself. Which feels scary since we don’t have fundamental control over every “voxel” of the solution.
Natural Response to Loading Condition
I’ll be showing some load-bearing and temperature-response scenarios with varying input geometry. Essentially showing that once you build the algorithm the designer is given the ability to specify any input field and tweak the loading scenario and then it’s off to the races!
In simulation-driven design, where an algorithm generates or modifies geometry in response to applied loads, forces, and constraints, how do you determine the necessary load requirements for a specific medical implant?
Oh boy, this is hotly debated (at least amongst the nerds that want to join in the discussion, haha).
Since implants reside in a highly regulated industry there are strength requirements predicated on some already existing devices. Like, devices that “were legally marketed prior to May 28, 1976 (random right?), or has been found SE (substantially equivelant) through the 510(k) process” Here’s a super fun read if anyone is interested. /s
I hear what you’re thinking… “Wait, our standards for these implants are based on almost 50-year-old devices?” The answer? Yeah, basically.
Now some modernization has been done, but ultimately we still have to compare current devices to hard chunks of metal that had way too much stiffness.
So, now for my take. There have been a ton of studies showing what the loads are in various joints in the body. And there are publicly-available databases where one can investigate loading conditions in humans across a range of body types and activities (standing, walking, lifting, etc.). So we know what kind of loading the joints see on a day-to-day basis. And it is much less than what the FDA requires when you submit a device for approval!
So, when I’m looking at a specific algorithm that modifies a structure based upon inputs, my desire is to use the anatomical loading scenarios. I typically utilize the bevy of research that’s out there, depending on which “joint” I’m designing for.
This is even more important if your desire is for the implant to flex or move!
How did your interest in investigating natural structures for their possible application in medical device design first develop, and what approach do you take to select the proper mathematical models for representing these structures?
Back when additive in orthopedic devices was becoming a thing (In 2010, an early adopter of electron beam based powder bed fusion technology from Arcam received one the very first FDA approvals for an orthopedic implant made via additive manufacturing), I had the privilege of working on a front-end research project. The owner of the company I was at was a bit of a mad scientist and was interested in geometrical complexity not offered by any softwares at the time.
So back around 2014-ish I attempted to use parametric modeling in Solidworks to create “lattice” structures. It was frustrating and time-consuming. In 2016, I downloaded nTopology Element and was introduced into a world where the unit cells were already available and I just had to give them a space to live.
Because I was so ingrained in a world where biology was king (shape of bone, surrounding anatomy, stiffness, structures), I started comparing what structures were out there to the natural world. And unfortunately, noticed where things came up short. You see, as cool as a Voronoi algorithm is there’s no reason to call it trabecular bone.
Bone on the left. Not bone on the right
So I started researching natural structures and quickly realized there was not a lot of “canned” geometry out there that did what I needed it to.
When I was given access to an early alpha of nTop Platform (what it was called at the time), the mathematical canvas was perfect for what I needed it for. I was able to build structures from their base equations, not relying on nodes and struts anymore.
Most all random natural structures are built on noise. But it’s a matter of tuning and taming those noise functions that make engineered structures something special.
So I usually start by envisioning the best way to get the structure on a canvas in math form, so it’s “infinitely available.” This can be using a range of tactics: base equations, mathematical remaps, tessellation using trig or modulo functions, etc.
Once I have the implicit function ready to go then I can build the structure and go from there, whether it’s evaluating with FEA, functionally grading, applying to lattice bodies, or more!
With your extensive experience in developing and validating bio-inspired structures for medical applications, how do you envision these or similar structures being utilized in other engineering fields?
I think most fields that utilize advanced designs are based on similar engineering principals to additive orthopedic devices.
How can we utilize the design space to the maximum but also minimize our “penalization” variables. Those penalties may be cost, weight, volume fraction, surface area, pore size, or whatever other quantifiable input.
As with most engineering problems, one must first understand the design inputs and produce a solution that meets these requirements. And I find that most computational structures are very good “chameleons.” In other words, they can modulate themselves to achieve many goals depending on the design inputs and required targets.
Even though my primary expertise is in orthopedic devices, I look forward to broadening how these custom structures can be used in other industries: aviation, space, automotive, consumer products, oil and gas, etc.
What signs or indicators could suggest that these types of bio-inspired structures might be effective in addressing a specific engineering problem? Additionally, how would an engineer or designer identify which particular natural structure could offer potential solutions, and determine the appropriate mathematical models needed to replicate or emulate those geometries? Should they just call you to help?
I think the “signs” are demystified by looking at the specific biological structure one is trying to mimic.
Does the structure exist due to surrounding forces? Heat transfer? Fluid flow? Something else. Sometimes it’s due to multiple influences. Take our buddy the mantis shrimp for example. It’s dactyl clubs aren’t really optimized for loading force but rather impact force, which is an entirely different engineering problem that must be solved.
But it’s also not easy to evaluate these structures. And even if you have one identified it’s even harder to model it. And then once you get it on a canvas you have to figure out how to apply it to your solution, and then optimize it.
So in short, yeah… just give me a call. 😀
Could you share some details about the consulting services you provide, the types of challenges you can help address, and how your expertise might benefit potential clients in various industries?
Canned answer incoming…
Not A Robot Engineering provides premiere engineering design consultation to all aspects of industry, with specialization in computational design software packages that set designs apart. From automotive brackets to orthopedic implants, Not A Robot excels at providing solutions where internal teams or other consultants have fallen short.
Combining field-based design, algorithmically-driven workflows, simulation feedback loops, advanced implicit modeling, and other next-gen techniques with traditional modeling to drive unique design outputs that exceed the requirements.
I find that most companies don’t lack the vision, but when it comes to advanced additive design they lack the resources to fully execute on the vision.
I’m willing to come alongside these companies and offer fully-built solutions down to basic training services on how to best leverage the tools and technology.
Additionally, due to my work in a highly-regulated industry I’m also pretty good at navigating qualifications, regulatory challenges, testing and validation, and other requirements inherent to good product development.
Aside from addressing specific engineering challenges, in what other ways can a company engage your services to enhance their processes, products, or overall business performance?
In my time in the engineering world, I’ve had the privilege of launching a myriad of products from concept to final solution – all of them in highly-regulated industries. So I can also provide development services, project planning, budget and scope analyses, regulatory consulting and gap analysis, and more.
I also have experience as an original member of a startup engaged in building a company around a new technology. I’d recommend visiting my business website and taking a look at my service offering to get some inspiration of how I can help!
What are the main points you intend to convey to the audience in your upcoming presentation, and what are your personal goals or desired outcomes for participating in the CDFAM Symposium?
I intend to show what’s actually possible with advanced computational design tools, not just something in a research lab or in a conceptual sketchbook.
My hope is to inspire a bevy of designers and engineers to look at what the physical world has already “engineered’ and take these ideas and turn them into a physical reality.
CDFAM is being billed as a thought leadership event, and based upon the list of speakers certainly seems poised to live up to the hype. I’m hoping to engage in real high-level conversations with other professionals in the computational design space about AI, generative design, algorithmic design, simulation-driven feedback, and more!
Finally, Did you make this?
As a human language model, I am unable to comment as to whether I was involved in the creation of this video. <empathetic human gesture> However, you may go to notarobot-eng.com for additional information about my creator and the services offered!
<human smile>
To hear more from Matt and network with other experts in computational design and additive manufacturing register to attend CDFAM in NYC June 14-15.
