This Website is not fully compatible with Internet Explorer.
For a more complete and secure browsing experience please consider using Microsoft Edge, Firefox, or Chrome

The CFD interview

The CFD interview

by The CAE Guy


Now that I have a few months to further settle in to the aero-thermal department manager role, I have had the chance to think more about expanding the role of Computational Fluid Dynamics within the department and potentially within the rest of the company as well. Our department certainly has an overall CFD strategy, but how that CFD strategy relates to the testing strategy within the entire development schedule could easily be re-thought. There is also the possibility of evaluating the always-developing software tools that we in the CAE community are well aware of. 

However, from my experience implementation of new (for the company, not newly developed) software techniques are closely tied with the availability of computational resources in the form of faster computers or solving on more CPUs. These are certainly ways to expand Computational Fluid Dynamics. Nevertheless, those aspects – better software and more application of Computational Fluid Dynamics – both really only address efficiency. At some point additional manpower must be addressed to realize any significant gains in output. That means hiring. And that “buck” stops at the manager – me[1]. However,while I was the CFD Group Leader I could completely advocate for additional CFD manpower whole heartily (and I did), but I must now take the interests of the entire department into account. That means balancing testing and computational resources based on a number of factors, such as immediate need or future need. First and foremost, from the company perspective, manpower is simply another resource that all departments are vying for (along with test facilities, equipment, test vehicles, computers, etc.).

Nevertheless, once authorization for additional manpower is agreed to, that means hiring, which in turn, means interviewing. I have been in a number of interviews – on both sides – and feel like I have a good idea of what I am looking for and what would work for our company. First and foremost, I am looking for someone that is interested in and enthusiastic about working in the automotive industry. Someone that has done their homework on our company. Someone that knows the fundamentals of their specialty. For CFD specifically, that means not only the rough/general equational development behind CFD, but also knowing and being able to explain aerodynamics and/or heat transfer. In my opinion, if you do not have some understanding of the physics or mathematics behind Computational Fluid Dynamics, then you will be less successful. I do not expect undergraduates to know the advantages/disadvantages of the various predictor-corrector numerical schemes or the advantages of finite volume techniques over finite difference techniques. In addition to any relevant work-related experience, I would (and do), however, ask the following question, as we interview a number of aerospace students:

“From an aerodynamic point of view, what is the difference between a streamline body like and airplane and blunt body like a car”.

You see, I want to know that they’ve considered the difference between the relatively balanced viscous and pressure forces on an airplane that they studied in college and the pressure-dominated forces on a road vehicle. I also expect them to understand problem solving as vehicle development is primarily about problem solving. To that end, I typically ask one seemingly innocuous question to see how they think. A couple of my favorites are:

“If you are in a car with a helium balloon in the middle of the back seat and you turn left, which way does the balloon move and why?”

“You have two completely separate rooms with no way to see from one to the other; one room has three light switches and the other has three incandescent light bulbs;If each of the light switches is connected directly to one of the light bulbs, then what is the minimum number of trips to determine which light switch goes to each light bulb?”

These are not intended as “trick” questions as the interviewees are certainly welcome to ask clarifying questions, but I want to see them go through the process of solving the problem. I find this is invaluable to helping determining how someone will perform. Finally, specifically for CFD analysts, I ask if they are OK with sitting in at a computer workstation for 90-95% of the time. This is actually a deal breaker for a CFD position as I have found no matter how smart and/or capable a candidate is, if they cannot “sit still” or really want to “work with their hands”,they will not be successful at CFD.

What are your thoughts on this? How can we best combine testing and modeling functions? Or anything else for that matter?

-The CAE Guy

[1]    [1] While the phrase “the buck stops here” was popularized by US President Harry S. Truman, he actually received sign that popularized the phrase as a gift.