CFD: Computational Fluid Distraction?

HydroComp Web Log

[efficiency in motion for over 20 years]

CFD: Computational Fluid Distraction?
Miscellaneous Musings from the Technical Director

Monday, 02-OCT-2006 by Donald MacPherson - HydroComp Technical Director

I read a thread about "hull speed" last week on a boat design forum. The topic of hull speed is something that particularly bothers me for a number of reasons, some of which I put in a small report on hull speed that you can find on the Knowledge Library of our web site.

However, the thread actually took an interesting turn into the components of wave-making drag and CFD. The real gem in this thread is one sentence, which I hope the author will not mind my quoting here.

"So, after millions of dollars and many thousands of man-hours spent on the problem, we are still a long way from even a reasonable understanding of the physics involved. CFD is, of course, an enormous advance because we can now display our almost complete ignorance in full colour!"

CFD clearly has had some success in marine hydrodynamics, but success has come when the questions to be answered are small, such as for the design of waterjet inlet geometry, improving flow lines at the bow, or identifying the wake distribution around a stern. The big questions of full ship resistance, wave fields, propulsor thrust and torque have not been so well served.

So, is CFD a viable tool for hull form or propulsor design? This is a provocative question, I know, but at this point in time the answer is unclear. Even the ITTC has something to say about the use of CFD for hull form design (24th ITTC 2005).

"Today, numerical shape optimisation is still at his infancy in ship hydrodynamics and ship designers are still very sceptical about its usefulness. It can be only registered that there is a constant growth of papers dealing with this topic together with some real industrial applications."

The growth of papers may be the real milestone for CFD to date. A quick look at our technical library (of more than 3000 articles and references) tells me that some 15% of them make reference to CFD. This figure was probably about 1% only ten years ago. To put the papers in our library into further perspective, 60% refer to propellers, 20% mention Froude number, 10% model testing, and about 5% each to systematic series and sea trials.

The empirical design tools - model testing, systematic series, and sea trials - appear to be off of almost everyone's R&D radar, yet these are still the real workhorses of hydrodynamics today. This may change in the future, I grant you, but not yet.

Makes you go "hmmmm......", doesn't it.