As high-performance computing (HPC) becomes ubiquitous, experts are saying that computational fluid dynamics (CFD) applications will soon replace many — but not all — uses of tow tank testing commonly deployed today in offshore oil platform development and design. An engineering executive from an international oil company (IOC) recently told me: “This is a game-changing technology and we want to see routine use of it by knowledgeable contractors that have the right hardware and validated, benchmarked software.” It’s a fair assumption that he’ll get — if not today, then tomorrow — what he wants. Jim Jeans, manager of HPC business development, Hewlett-Packard, says his company has seen “a ten-fold increase in demand for HPC in computer-aided engineering [CAE] in the Houston area.” Moreover, adds Jeans, by the early part of next year, new Intel and AMD microprocessors are due to deliver yet another quantum leap in computational performance. There’s nothing quite like the atmosphere of a room full of engineers, quietly excited about the prospects for increased computing power and more powerful applications: interrupting themselves with brief digressions and self-deprecating remarks—but working together in deep, unremarked collaboration. That was the scene at “ACUSIM Energy Day: Offshore Platfom and Wind Turbine Simulation,” held recently in Houston, and co-sponsored by HP, Intel Corp., and Microsoft Corp. ACUSIM’s AcuSolve is a general-purpose finite-element flow solver capable of extremely fast computations running on “distributed parallel machines,” says Dr. Farzin Shakib, founder and president. Most other commercial flow solvers, he added, are “finite volume.” And for over a year now, the company’s solutions have been able to run simultaneously with solid-structure analysis tools, without intervening middleware. The president of an engineering firm pointed out that thirty years ago automotive companies made extensive use of “scale modeling for automobile structural testing. It was all experiments. They still do scale model tests, but all the initial testing is by computer simulations. The oil industry is on the cusp of making a similar transition to CFD.” For an offshore platform design, a common use of CFD might be analyzing vortex-induced motion (VIM) associated with a spar; vortex-induced motion (VIM) associated with a riser; or other types of offshore structure complexities. Even with the very best CFD solvers, challenges remain for offshore applications. These include size (the GOM is much larger than a car, for example), geometries (such as dealing with the appendages fitted to platform structures), and surface roughness. Yet strategies for dealing with these complexities are being developed all the time. More generally, large IOCs are taking steps such as the following · Moving from simulations of one- to two-million cells to ones approaching 100-million cell simulations; · Moving to clusters that involve hundreds of cpus; · Developing benchmark data—a real key to CFD acceptance; · Bringing together AcuSolve and Abaqus for direct-coupled CFD and finite-element analysis; and · Participating in academic and OMEA conference activities. Dependent upon scale and instrumentation, tow-tank testing can cost anywhere from $150 to more than $3,000 per test. Performing CFD analyses can cost $2,000 per test. But the main benefit of CFD isn’t necessarily cost or replacing tow tank testing. Rather, it’s a precursor to model testing and a complement to it.