William Humber, Senior CFD Engineer, ADSCFD.

I was born in New Jersey in 1984.  I did both my bachelors and masters work in Aerospace Engineering at Penn State University. While I was there I received the University Turbine Systems Research fellowship and spent my summers during grad school working at Pratt & Whitney in the Turbine Durability Group.  It was during a chance encounter there in 2008 that I met Bob Ni, the former head of CFD development for Pratt and Whitney.  My boss at the time, asked me if I could do some meshing work on the side for Bob.  The meshing work ended up with me joining Bob’s new company, AeroDynamic Solutions (ADSCFD), once I finished my graduate work.  I’ve been working with ADSCFD, primarily involved in meshing and post-processing technology development since then.

• Location: Rutland, VT
• Current position: Senior CFD Engineer, ADSCFD
• Current computer: Hodgepodge desktop i7 with 32 GB RAM (Windows 7)
• One word that best describes how you work: Multitasking

What software or tools do you use every day?

Some of the basic software that I use every day includes Visual Studio, Textpad, Intel Fortran Compiler, ParaView, and Excel 2003.  I’m mostly up to date with the rest of the Office suite but I still find the 2003 version of Excel to be superior to the modern versions.

What does your workspace look like?

Our company is virtual so everyone maintains an office in their house.  I have a main workstation in my office and then run my own mini cluster on site.

What do you see are the biggest challenges facing CFD in the next 5 years?

One of the main challenges we see is getting the man out of the loop whenever possible.  We did a webinar with Pointwise a couple years ago where we looked at different meshing strategies (multi-block structured, hybrid, fully unstructured) for a simple volute.  While the multi-block structured case gave the best results for a given mesh count, and ran the fastest, we found that the mesh generation time required meant that the potential speedup via parallel processing was much lower than the fully unstructured mesh.

The ability to remove the man from the loop is critical to implementing optimization schemes as well as when creating a design system, both of which are highly dependent on reliability, consistency, and quick turnaround time.

What are you currently working on?

We are currently spending much of our time on fluid-structure interaction development.  High cycle fatigue prediction is something we’ve been working on for a few years now in conjunction with the Air Force Research Laboratory and we’re finally getting to the point where we’re simply refining the process for commercialization.  In addition to development, a large amount of my time is spent on consulting work.

What would you say is your meshing specialty?

I’m not sure I would call it a specialty but a lot of my meshing ‘portfolio’ involves conjugate heat transfer meshes of engine hardware.

Any tips for our users?

The layers tab is definitely your friend as well as the ability to label the layers, especially for large complicated meshes.  Also, if you run many different software packages that require manipulating objects in 3D, a 3D mouse is an excellent investment that can significantly cut down the time required for certain operations.

What project are you most proud of and why?

We did a Phase II SBIR with the Air Force Research Lab for the development of a conjugate heat transfer capability to our solver that involved simulating a cooled turbine vane with over 600 film cooling holes.  Our initial results weren’t great, but over a year of root cause analysis we were able to resolve almost all the issues we were seeing.  In the end we were able to publish the most detailed conjugate film cooling simulation I had seen in the open literature up to that point.

What CFD solver and postprocessor do you use most often?

I use our solver, ADS Code LEO, exclusively as it handles both structured and unstructured meshes.  For post-processing I primarily use ParaView for visualization, but we’re pretty heavily dependent on Tecplot as well.  For lower level post-processing tasks I use a combination of internal utilities, Excel 2003, and Matlab.

Are you reading any interesting technical papers we should know about?

There’s an interesting paper that should be coming out soon that looks at a full wheel unsteady turbomachinery simulation that accounts for all the manufacturing variation in the individual airfoils.

Do you plan on attending any conferences or workshops this year?

ADSCFD will have a booth at TurboExpo this year.  So I’ll be in Charlotte in June.

What do you do outside the world of CFD?

I have many hobbies, but primarily I’m a ski instructor in the winter, ride road bikes in the summer, and play table tennis all year round.

What is some of the best CFD advice you’ve ever received?

One of the biggest challenges with CFD is how to reconcile differences with experimental data.  When you run a new case and see that first comparison it’s easy to get lost in all the possibilities of where things could be going wrong.  One of the greatest difficulties, and one I’m guilty of from time to time, is getting tunnel vision and assuming that the problem has to be something within my sphere of influence.  What’s important is to keep an open mind, and be methodical about identifying, and eliminating, all the possible contributors to the problem.

If you had to pick a place to have dinner, where would you go?

If I could go anywhere it would be Rodolfo’s Pizzeria in NJ from back in the late 90s (the original owners sold it sometime ago and it’s just not the same now).  If I was going somewhere tonight, probably The Prince and the Pauper in Woodstock, VT.