Another Fine Mesh

Chavaly Venkatesh, Partner, KFour Metrics.

An impartial absence of malice towards any subject that didn’t involve formalin meant I could pretty much play the field on leaving school. The dice were rolled, and I found myself working towards a Bachelor’s degree in Mechanical Engineering at the Indian Institute of Technology, Madras. The first few semesters were an extension of the anodyne academics at school till I sat through a course on Dynamics of Machinery taught by Dr. V. Ramamurti. Boy, could he teach. I spent the remaining semesters at college doing every course he offered, and developing an appreciation of both engineering mechanics and engineering judgment.

The next few years were equally formative and shaping. I spent two of these working for a Master’s Degree under the stellar guidance of Dr. Eric Becker at the University of Texas at Austin. In between discussing motorcycles, the appalling state of some leading textbooks, the discipline of computer programming, the art of literature surveys and the difference between a college-degree and education, he taught me most of what I now know about the Finite Element Method. Degree in hand, it was off to the Tata Research, Development and Design Center at Pune where Dr. E. C. Subbarao was all you could ask for in a boss. The lessons I learnt from him on managing people have stayed with me since.

While my own job involved applied research using finite element and finite difference methods, it was educative to watch as TCS, TRDDC’s parent company, took up the distribution of Unigraphics in India. The lure of computer graphics was strong, and I moved to OMC Computers, a Hyderabad based company that distributed Silicon Graphics computers together with a wide range of scientific software.  The market for FEA was challenging, so the obvious thing to do was take that challenge head on by starting a company to focus on CAE and CAD. SDRC’s I-Deas on Windows NT was my platform of choice, and as the company grew we took up the distribution of I-Deas, provided design services to the automotive, power and defense sectors, and distributed HyperWorks. Growth often brings home the lesson that small can be beautiful too. After some introspection and evaluation of our priorities for life, Shaila and I sold the business to start KFour Metrics in November 2006.

In 2014 I chanced upon an interview with John Chawner of Pointwise which made us sit up and think. Experience with computational mechanics had shown that preprocessing is inelegant, and sometimes downright ugly. Could this product that had seen decades of work on a subset of preprocessing for CAE – grid generation specifically for CFD – be worth looking at? It helped that in the interview John criticized one of my pet hates – the “democratization” of CAE. We called Richard Matus to find out what Pointwise was doing in India, and here we are: distributing Pointwise in India.

• Location: Hyderabad, India
• Current position: Partner, KFour Metrics
• Current computer: HP Laptop with Windows 8.1 and Dell Workstation with Linux Mint.
• One word that best describes how you work: Patient

What software or tools do you use every day?

Firefox, Windows Live Mail, Libre Office, K-9 and Signal. And, of course, Pointwise.

What does your workspace look like?

Venkatesh’s current workspace.

My dislike of air conditioning means I prefer wide open windows and a slowly swirling ceiling fan. This doesn’t always sit well with Hyderabad weather, where summers sometimes cross 45 degrees Celsius, but I continue to hold out. I rely on SpeedFan to let me know when the machines need a break.

My main workspace is a large, flat table on which I pile things as they come in. I often have multiple tasks going on at the same time, which makes for multiple piles, each of which I tend to leave undisturbed till the task is complete. Since some projects take quite a while before I finish, it can get messy at times.

The walls are lined with cupboards double-stacked with books. Many of these are indispensable technical references but some are there just for the joy of possession. The pick of the latter is a 1910 edition of “Theoretical Mechanics” by Percey F.Smith and William Raymond Longley. I come from a line of reading-addicts, so the bookshelves also house a collection of fiction and non-fiction going back to the early 1930s.

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

Of the technology challenges, dealing with CAD data is easily at the top of the list. A better concealed, and therefore more pernicious, problem is the tendency of improvements in user-interfaces to lull engineers into complacency. The first page of printed output from any FE application consists of a here-be-dragons warning. How many users even see printed output today?

What are you currently working on?

We’re looking forward to hosting Pointwise in India later this year, and are spending our energy working on making sure Pointwise users here can get the most out of this.

What would you say is your meshing specialty?

I can’t pick out any single area as a specialty. I’ve logged enough time in this field to find all types of meshes equally engrossing, but coming from long experience with FE meshing, T-Rex is fascinating. Local stoppage and prism recombination are both remarkably obvious and remarkably impressive once you see them in action.

Any tips for our users?

That’s easy: Glyph. Invest the upfront time in learning Tcl and you will find the API is a pretty logical extension of interactive Pointwise usage.

What project are you most proud of and why?

Once case that really stands out is when Pointwise delivered a grid that had about half the cells the analyst had braced for, even while maintaining resolution in areas of interest and with a significantly better grid quality. Can’t say more than that without crossing the confidentiality-limit!

What CFD solver and postprocessor do you use most often?

I don’t do much with CFD solvers beyond checking grids. OpenFOAM, SU2 and Fluent are the ones I use from time to time. For post-processing, ParaView rules.

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

My immediate interest centers around overset methods. “A Direct Cut Approach for Overset Hole Cutting” (Ralph W. Noack, 18th AIAA Computational Fluid Dynamics Conference, June 25-28 2007, Miami, FL) and “Best Practices in Overset Grid Generation” (William Chan, Reynaldo Gomez, Stuart Rogers, Pieter Buning, AIAA 2002-3191) head the list of papers I’m currently studying.

Do you plan on attending any upcoming conferences or workshops?

I will be at the 2016 Pointwise User Group Meeting in late September. Schedule permitting, the ICFMFA 2017 and the Aeronautical Society of India’s Annual CFD Symposium are on my wish-list for 2017.

What do you do outside the world of CFD?

I find it very hard to understand art. It doesn’t seem to yield to hard work. The nut I’ve been trying to crack for the past 11 years and counting, is music. I play in a small band that meets as often as it can, given our different schedules. Sometimes it feels like the band sessions are more about arguing how to play what and why, but we do make progress.

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

To exercise extreme skepticism about the output of any computer simulation.

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

If you’re a vegetarian, as I am, the thali at The Spicy Venue in Jubilee Hills is hard to beat.