Ken Wurtzler is Director of Operations for Cobalt Solutions, LLC (cobaltcfd.com).
What do you see are the biggest challenges facing CFD in the next 3 years?
One of the challenges is the notion that CFD can be reduced to a “push-button technology”. I guess a corollary to this is a “one-size-fits-all” view. The desire is to use CFD for more advanced analyses and in a mutli-disciplinary environment. The complexity of the desired solution has increased while at the same time, it is expected to be able to automate everything to get an answer – using just one code. To me, that seems to be pulling in two different directions. I worked on the AIAA Drag Prediction Study II and I remember how different grids gave much different results using one flow solver. There is a lot of variability in grids for complex configurations (as you well know) and the user has to be able to document that variability. No one expects a novice to put a model in a wind tunnel, turn it on, wait a few minutes, and then get an answer.
John: I agree, but the key word in your last statement is “novice.” An engineering with an understanding of wind tunnel testing should be able to get answers in short order. The key for CFD is training these same engineers on how to run the virtual wind tunnel. Would you agree?
Ken: Somewhat agree. The key word in your statement is “answers”. Too often the result is treated as the answer but is actually information/data that has to be interpreted to arrive at an answer. That interpretation depends on the understanding of the grid, problem set-up, and input parameters to the CFD code – which come from using the code and experience. I guess it comes down to what kind of training occurs.
What are you currently working on?
Right now, working on getting Cobalt Version 7 out the door. Cobalt V7.0 introduces a new Overset module which provides improved hole-cuts and improved efficiency. It was developed along the coding guidelines as Cobalt – robust, accurate, and user-friendly. The idea is to have the solver do the work and to minimize user up-front work. The hole-cut interface is calculated to lie midway between non-cuttable boundaries of overset grids. The result is a very smooth hole-cut which can easily handle vastly different scales such as a full scale aircraft with tiny gaps in control surfaces. It is also much faster than the previous Overset module – both in non-motion and motion cases.
John: You’re aware of the addition of overset grid assembly capabilities to Pointwise. Would that help your users or have you bypassed that altogether?
Ken: Pretty much bypassed that. Bob Tomaro wrote the grid assembly code for the first Overset version in Cobalt about 10 years ago. There was an attempt to incorporate a grid assembler written by a third-party but license other issues arose. We have kept everything in-house since then.
For another example of mesh generation with Pointwise’s software by Cobalt Solutions, see this case study of paddle deployment for an Argus missile.
How did you get to be where you are today?
I graduated from the University of Michigan in 1986 with a BS in Aero Engineering. Started working in the Aero-Methods Group of the Flight Dynamics Lab in 1987. This group was directed to develop computational methods for aerodynamic analysis. We worked with all the defense contractors at the time to see what they were working on. We began in-house efforts in flow solvers and grid generators and began supporting in-house analysis work for other groups in the Lab. Started with structured grids and flow solvers and then moved into the unstructured arena. I spent a lot of time applying CFD methods to in-house and contract development projects where I quickly learned what worked and what didn’t.
A few of us in the group decided in 1999 to try to continue what we were doing but in a small company. We left the Lab in 2000 and started Cobalt Solutions, LLC. We have continued the development and support of Cobalt for 15 years.
Who or what inspired you to get started in your career?
I had a fascination with airplanes since a young age. Always wanted to be a fighter pilot – but bad eyes kept me from doing that in the military. I knew I wanted to study aerospace engineering in college before I started high school.
John: Any thoughts on the need to attract more young people to STEM careers?
Ken: Bring in real-world experience into the classroom. GE Aviation in Cincinnati does a good job in local high schools working with kids who are interested in engineering – offering technical projects for classes to work.
What advice do you have for young people entering the field today?
Remember why you started in engineering. You can easily get dead-ended in engineering so keep your options – education, career broadening – open if you want to stay in engineering. The field can always be a springboard to something else if that is desired.
John: What exactly do you mean by “dead-ended”?
Ken: Maybe pigeonholed in a certain area of engineering is a better phrase. You have to keep ahead of changes – which happen quickly now.
How do you know Pointwise?
I worked at Wright-Patterson AFB in what was then called the Flight Dynamics Lab in 1988. My office put out a request for contracted work for grid generation capability. The winner was from a group at General Dynamics (RIP) that ended up creating Gridgen V6. In that group were a few young, energetic engineers – John Chawner, John Steinbrenner and Chris Fouts. I helped manage the government distribution of the software to DoD contractors all the way up to Gridgen V9. I still use a version of Gridgen.
John: What is it you still like about Gridgen?
Ken: Its familiarity and its robustness. In our development of our new Overset module, I typically work with one model at a time and make constant changes to the grid and then see how Cobalt performs on it. I can quickly make the needed changes and know the grid will “work”.
John: There’s something here about old dogs and new tricks but I’ll let sleeping dogs lie.
Can you share with us your favorite tools and resources that help you get your job done
Cobalt for flow solver of course. Gridgen (now onto Pointwise) has always been a favorite for grid generation. I use EnSight (www.ceisoftware.com) for flow visualization. I run my solutions on the DoD HPC computational resources in support of projects.
Your blog is very useful to check on what is happening in the computational world. The DoD HPC website www.hpc.mil is also useful to me.
If we were to come visit you where’s a good place to go out for dinner?
The Precinct – eccentric owner Jeff Ruby really goes out of his way to provide impeccable service along with great steaks. http://www.jeffruby.com/precinct
John: The Pine Club is still open, right? We always enjoyed going there for steaks back in the day (when we jokingly called it the Pine Box.)
Ken: The Pine Club is in Dayton (near Univ of Dayton campus). Yes, still open and only takes cash. And they haven’t expanded – they shoehorn you into your table.