Integrated or Specialized CFD?
Over on engineering.com, the guys at Tech4PD debate whether it’s better to have CFD tools that are fully-integrated (CAD, pre, solver, post) or specialized (i.e. best in class [my term, not theirs]).
- Jim Brown takes the side of specialized tools as it’s more important to get the right answers than to be able to compute many answers quickly.
- Chad Jackson thinks that for CFD to be applied practically it has to be part of a fully-integrated toolchain.
Of course, the correct answer is that they’re both right and they’re both wrong.
Jim is right because getting the right answer is most important. He’s also wrong because of his assumption that all CFD problems are sufficiently complex that the risk of wrong-ness is very high. There are classes of CFD problems that you could classify as simple, or perhaps workmanlike.
Chad is right because at the end of the day all that matters is the work you managed to accomplish during the day which is an issue of productivity and if all your tools were seamlessly integrated you’d be very productive. He’s also wrong because he assumes that specialized tools aren’t likewise, if not integrated, or at least compatible, with each other.
Interestingly, the question to be voted on by you and me doesn’t match the video’s title. Instead, the question is “Where is the biggest value in implementing CFD?” I voted and you should too if for no other reason than for either Chad or Jim to suffer the consequences of their position.
During the latter half of the video they look into their crystal ball and predict what’s ahead for CFD. I urge you to watch and hear their predictions.
- Thule and IMAGINiT applied CFD to the aerodynamics of rooftop cargo boxes to reduce drag by 35%.
- Tecplot presents a case study on the role of turbulence in the exchange of heat and gas between the atmosphere and the ocean.
- Emirates Team New Zealand will compete for the America’s Cup after their recent victory in the Louis Vitton Cup. ETNZ uses ANSYS and Pointwise software products as parts of their CFD arsenal.
- Symscape used Python scripting with Caedium CFD to automate an angle-of-attack study of one of the Drag Prediction Workshop cases.
Other News & Events
- The workshop Progress in Wall Turbulence: Understanding and Modeling will be held in Lille, France on 18-20 June 2014.
- In A Defense of Computational Physics Patrick Roach is “defending the Computational Physics enterprise against what amounts to an empty nihilism that seems to be popular among academics.” [Note: The quote is from the linked blog post and not Roach’s book.]
- What would make a superfluid (think of a inviscid fluid like liquid Helium) behave opposite how a normal fluid behaves? In a normal fluid, 2D turbulence evolves from small to large structures (think of a boundary layer transitioning) while 3D turbulence evolves from large eddies to smaller structures (think of the plume of smoke from a candle as it rises). But 2D turbulence in a superfluid devolves into smaller and smaller scales.
- The U.S. Association for Computational Mechanics presented its highest award, the John von Neumann Medal, to the University of Texas’ Mary Wheeler for her work on subsurface flow modeling.