Dr. Lieven Vervecken is CEO and Co-Founder of Diabatix. The company specializes in generative thermal design. With a focus on heat transfer, the cooling components generated from Diabatix ColdStream software are made to cool the high-powered tech of today and tomorrow. We recently had the opportunity to learn more about Diabatix, ColdStream, and what we can look forward to in the future.
Thank you for taking a few minutes to share your thoughts on a few design and simulation-related topics with us. Can you tell us a little about your professional background and how you got to where you are today?
I studied mechanical engineering at the University of Leuven and did an additional master’s program in nuclear engineering. In 2015, I earned a PhD in the modeling of the dispersion of radioactive gasses. In early 2016, I combined my expertise in executing atmospheric simulations with that of my co-founder, Joris Coddé, in thermal modeling. And that was the beginning of Diabatix.
What were the circumstances behind launching Diabatix? What is the pressing market need that your customers are able to uniquely fill with your software?
We saw a need for improvement in the field of thermal design and wanted to contribute, but we didn’t really have dedicated software or something like that developed yet. We only had the idea, “that’s something that could be done better”. We started as consultants and in parallel developed our generative design technology. That way, we gradually evolved into an engineering company and used our software internally to help us with customer projects. Upon request of our customers, we released our ColdStream platform early 2021. As a result, we are now a software company rather than an engineering company, offering our customers access to the software that we have been developing in the background for years.
Your company recently launched ColdStream 2. Can you tell a bit more about that?
With last year’s initial launch of ColdStream we basically released a public front-end to our in-house software containing only the functionality that was most frequently needed in customer projects. Throughout the year, we gradually added a number of new features based on customer feedback. Small changes were also gradually introduced into the platform, but a number of big changes and additions were grouped into our ColdStream 2 release. In particular the addition of being able to select standardized heat sinks was a major game changer for ColdStream. This addition is broadening the scope and transforming an expert tool purely used for customized design into a general thermal engineering tool for any thermal design and any additional needs that thermal engineers might have.
So, it is adapted to the customer itself then?
Where can our readers learn more?
More information can be found on our website www.diabatix.com. You can also check our YouTube, LinkedIn or other channels.
What related services does Diabatix offer?
ColdStream is our central technology and our main platform, and we try to make it as self-supporting as possible. But if needed, we do provide software and engineering support. So, if you have new customers that are not very familiar with the technology yet, we help them get acquainted with it by teaching them how to use ColdStream in the most effective way and to interpret the results it produces. At the same time, we are always open to new projects that would help bring ColdStream to the next level. We always try to find a healthy balance between taking up new engineering projects that require additional developments into the front- or back-end of the software or implementing customer requests within the current capabilities of ColdStream.
Is Diabatix planning to broaden its generative design solutions beyond heat sink design?
Yes, that’s very likely, but the flow of physics and the thermal physics will always remain our core. We can’t exclude that at a certain moment we will introduce structural design and structural constraints to thermal design as well. But we won’t add it with the goal in mind of having a new structural design tool. It will be added to support the thermal design.
Rev-Sim’s theme and focus is bringing simulation to everyone who needs it – beyond just the experts. Could you comment on this aspect of your contributions? Do you consider your product to be a great example of democratizing design and simulation?
Yes, absolutely. I think the best proof is that when you talk about thermal design and thermal analysis, people immediately think of CFD experts, people with a PhD, with 10 years of research, studies and experience in the field. We want to change this and this ambition is even visible in our organizational chart. If you look at our Head of Product, Roxane, and our head of R&D, Ine, neither have a degree in thermal engineering (Roxane has a PhD in architectural engineering and Ine has a background in civil engineering). But it was a well-considered choice to make them responsible for the developments of both the back-end software and the front-end software. Their primary reasoning is always, “Do we understand the purpose of a button, field, value, etc.? If not, it needs to be changed.” Of course, it still contains features for the expert user, but they will never block the non-expert.
In your opinion, how accurate are the underlying simulations that drive the generative design? How can a designer get confidence in the accuracy of these simulations?
It is as accurate as the input. For example, if you want to simulate the temperature of a beer can when you put it in the fridge, but you forget to configure the environment temperature to 4°C, you cannot expect to get an accurate result. So, while the simulations are state-of-the-art concerning what is available from a physical modeling point of view, they are still only as accurate as the boundary conditions that are configured by the user. If you have a very accurate description of the environment in which a component, machine or product will be used, then you can make an almost perfect match between simulation and practice.
So, when there is a mistake, it is mostly due to wrongly entered data?
Well, the software itself doesn’t make mistakes. It solves the set of equations that the user wants it to solve. So, if the user makes an error in judgment or an input error, it will result in a certain deviation from reality. What we did try to do with ColdStream is to reduce or to minimize the room for error. For instance, before a user starts a simulation the whole setup will be checked automatically, and feedback is provided if anything out of the ordinary is detected. With every update the set of checks that are performed gradually increases to help the user as much as we can.
How essential are democratization and automation to the wider adoption of simulation – in your case, thermal and fluid simulations?
I think that the user base of the simulation at this moment is only a fraction of its full potential. The better the access to the simulation tools and the easier they are to use, the more widely adopted they will become. So, from that point of view, ease-of-use and automation are essential if you want to have a wide integration of the simulation into the product development cycle in the first place.
I understand that, especially in the early days when simulation tools were being developed, simulation tools were only for the happy few, the true experts. But I think we are now at a point in time where you can and must simplify those tools and bring the required level of expertise to a level that becomes accessible to a wide range of engineers.
Is Diabatix leveraging Artificial Intelligence, Machine Learning, and other similar technologies to speed up the underlying design evaluations?
Yes, a part of the design workflow is controlled by a module that autonomously makes decisions instead of the user. That part of the code falls perfectly within the category of AI, but, to be clear, it all boils down to a set of advanced algorithms and statistics. For the simulation itself, there is no AI involved.
Why do you think it is important for Diabatix to participate in the Rev-Sim online community, and how do you think an organization like ours provides value to the broader simulation community?
Especially for young companies like us, despite the fact that our technology is not available in any other tool, it is still a challenge to make yourself known to the broader engineering public. What’s interesting about Rev-Sim is that your mission is to bring together those who are ready to take the next step in engineering and in simulation. And that’s also what our mission is all about. We want to bring this new technology to the market, to help engineers take the next step in thermal design by leveraging the power of simulation. From that point of view, if you share a common mission, it is only logical that you start collaborating.
Thank you for taking some time to talk with us. Enough about work… tell us a little about yourself. What you enjoy doing in your spare time, interesting hobbies, and so on.
Well, I have 2 small kids, two sons, one of four years old and one who just had his first birthday, and they are quite a handful. So, they basically fill my evenings and my weekends, but it is the best time investment anyone can have I think, to spend time with your wife and kids.
Besides that, I enjoy the regular things like participating in sports from time to time, trying to open a book, typically it is business related that book, and I really enjoy cooking. I think that’s it. I hope, that in a few years when the kids are a little bit older, there will be some more room for other hobbies, but at this moment, my kids are my biggest hobby.