About 40-50 years ago a lot of experimental work was done in this area in the US. For the reasons enumerated here the research finally ended. I always thought that to be a mistake and am glad to see someone else take it up.

I wonder what would be the effect of putting a 50 micron hole in the middle of a dimple? Certainly someone needs to take a look at that in a windtunnel.

I can see why they gave up on this research some years ago. The amount of computing power required would be enormous. Even with the machines today which are several orders of magnitude faster than the ones I first used (I won't say how long ago that was but lets just say that I know about flipping bytes to do math on a PDP 8), you'd have to have something much larger than the desktop I'm on to do the job.

Patrick -

instationary CFD problems involving millions of nodes and sophisticated, highly nonlinear turbulence models (Reynolds stresses etc.) are still best solved on big iron supercomputers with very fast data interconnects, e.g. something with a ccNUMA architecture and hundreds of processors. CPU clock speed as such isn't everything - beyond a certain size, data transport architecture becomes the limiting performance factor for monolithic problems. Disclosure: I used to work for SGI, a supercomputer maker.

When Mssrs. Kloker and Messing say they don't have enough compute horsepower, I suspect they don't mean they need a fancier PC or commodity cluster.

For something like this, even more money has to go into the experimental validation. Scale models will only get you so far. Also, fluid dynamics calculations in particular are still as much of an art as a science; any given turbulence model only covers only certain classes of problems and certain parameter ranges. Pick the wrong one and you end up with garbage out.

Rafeal,

I wasn't referencing clock speed per se but actual computational capacity.

4 Cell processors properly addressed (each of the 8 cores on each processor similar to hyperthreading in a multi pipeline processor) are capable of turning out 1TFLOP of calculations per second.

Lucas,
Funny you mentioned the whole "hole in a dimple" thing. I have connections with Callaway golf. The recently started using small dimples within dimples. The result has been impressive. I cannot provide any imperical data, though I will try and obtain some.

Roger,
Interesting stuff, but....
How do you store the LH2? High pressure no doubt.
What about localized storage of LH2? Hydrogen has a nasty tendancy to leak.

John -

LH2 is liquid hydrogen at 20 Kelvin, at atmospheric pressure. In an aviation application, you would want to minimize the weight of the fuel tank (not as negligible as Roger suggests) by using a cylindrical tank capped by hemispheres. This sharply reduces the surface-to-volume ratio relative to in-wing tanks, which were found to be impractical for cryoapplications.

http://www.flug-revue.rotor.com/FRheft/FRH9809/FR9809k.htm

The optimum would be a single, large spherical tank with internal baffles to keep the liquid phase from sloshing around. This might be feasible in a blended wing-body design (cp. B2 bomber), if placed immediately behind the cockpit. BWB designs are also aerodynamically efficient.

As for boil-off, an airport needs a lot of electrical power so it could operate its own genset/SOFC. The fuel would be delivered frequently to limit the volume that must be stored on site. Similarly, all commercial aircraft feature APUs (currently small gas turbines) to generate power for on-board systems. An LH2-based aircraft with a fuel cell and electric motor would not need a separate APU, given high efficiency in part load. Any excess electricity produced while on the ground could be fed into the airport's grid.

The one snag is that you need very high power levels to take off in the first place. You don't want to overengineer your fuel cell stack if at all possible. In-wheel electric motors and perhaps, inductive couplers embedded in the runway for grid connectivity might do the trick.

Andrey,
Supercavitation. They are fitting them on autocannon shells as well (to kill inbound torpedos, naval mines and other underwater targets).

Hydrogen for planes are new to me. The idea of using a fuel cell and an electric engine is talented. What you need is to be able to boost energy at take of and landing. A ceramic condensator can contain a lot of energy and discharge it rapidly - and it is not heavy.

In the calculation you also have to take into account that both a fuel cell and an electric engine can deliver continous torque regardsless of how thin the air becomes whereas a jet looses torque rapidly with altitude. This allows a plane with fuel cell and electric engine to fly much higher where there are much lower air resistance which translates into higher speed and lower power consumption.

This combination may do the trick for the SAX plane concept as both noise and power could be reduced.

Regarding the hydrogen this can be produced very cheaply at location by low cost plants that tap into the electric grid outside peak hours or when windmills produce excess energy. This way the capacity factor of electricity plants can be kept high which will benefit their cost effectivenes and actually also their power efficiency because stop and go running of a power plants is very inefficient.

Also the air pollution in the form of unburnt fuel can be reduced.

I think Roger Pham should take this further.

Regarding dynamic fluid dynamics then please note that Moores law is still in function and that Intel has stated that we will be able to handle 100 times more cycles at the same cost for every ten years. It will be a while before dynamic fluid dynamics will be employed in standard planes but when it is time for it the computing power will be at hand.