What do you see are the biggest challenges facing CFD in the next 3 years?

In my point one of the main problems at the moment is a variety of models and methods. Which one to choose in this case to get as close as possible to reality solutions in minimum time is a very difficult question. This problem is similar to an optimization problem. Now it all rests on the experience and intuition of the researcher.  Nothing is so expensive in CFD as an error in the choice of model.

John: When you say methods I assume you are meaning numerical methods in the solver. And when you say models I assume you are meaning turbulence, chemical reaction and similar models. Is that correct? If so, I suspect there is more variability in models than methods. Would you agree?

Sergey: You’re right, it is more concerned with models (turbulence, chemical reaction, etc.) and to a lesser degree of numerical methods.

What are you currently working on?

It’s quite a wide range of tasks related with dynamics of disperse systems, multiphase combustion, multiphase supersonic jets and their interaction with the target.  Also it is the problem of unsteady heat transfer between supersonic jet and the target. Accordingly, solving these problems helps us improve our CFD code.

Streamlines over a re-entry vehicle as computed in HyperFlow.

John: Your CFD code is HyperFlow3D, right? You just released a 2D version as open source, OpenHyperFlow2D. What led you to that decision? What has been the response from users?

Sergey: Yes, HyperFLOW3D is our current working tool by which we try to solve our problems. In its present form, it grew out of an earlier version of the 2D solver – HyperFLOW2D which more than 20 years. First version HyperFLOW2D  (though then it had no name) was written in 1995, and a small piece of code is used since that time. In HyperFLOW3D used completely different preprocessor but the solver largely repeats the design, which was used in HyperFLOW2D.  Here is exactly this older version (after considerable cleaning code), and was released as open source version – OpenHyperFLOW2D (http://openhyperflow2d.googlecode.com).

Why Open Source? This is a very good question. I saw how good codes (not necessarily CFD) died without community when they became not interesting to their authors. It does not matter whether it is a large commercial project or simply author’s code. Although some of our postgraduate students used HyperFLOW2D as a tool, it was rather odd jobs. And of course interesting feedback from other members of CFD community.

How did you get to be where you are today?

I graduated from the Bauman MSTU and have worked there for over 20 years as researcher.

Who or what inspired you to get started in your career?

At school I was interested in physics and mathematics. Aerospace topics also was interested me. It was interesting to know how it works inside.

What advice do you have for young people entering the field today?

The insides of CFD are much more interesting than what you see in colored pictures. It is very cool if you have a chance to make the results of your simulation in the real prototype.

How do you know Pointwise?

I never used your software but enjoy reading your blog.

Can you share with us your favorite tools and resources that help you get your job done?

• CAELinux  http://www.caelinux.com/CMS/ (with Salome, gmsh, OpenFOAM)
• Tecplot http://www.tecplot.com/
• Intel Cluster Studio http://software.intel.com/en-us/intel-cluster-studio-xe
• NASA Technical Report Server (a lot of data for V&V your CFD code) http://ntrs.nasa.gov/search.jsp

If we were to come visit you where’s a good place to go out for dinner?

I’m always interested trying something new. So I like to visit new places, no matter in Moscow, London, Berlin or what that other city. Therefore, I am afraid that with the “favorite” place I have problems, I rarely where there was more than 2 times (on my own initiative).

John: You are the opposite of me. I like to find a favorite place and go back many times.

Thanks for sharing your work with us.