At The Glen I discovered that my re-engineered fuel system wasn't worth a dog's rear end. When I went to a fuel cell, I had to ditch the factory fuel pump, as it was located on the crossmember back by the port heat exchanger, and wouldn't pull fuel out of the top-loading cell.
As you can see from the picture, I mounted the pump in the smuggler's box, which I thought was a good idea from a safety standpoint, and assumed that the capacity of the pump would be sufficient to pull fuel out the top of the cell and back to the engine, even though there was an approximately 3' run of hose from the pickup to the pump inlet. The trouble is that the pump's application notes call for it to be no further than 12" from the tank, and below the level of the fuel. Looking back (and looking at the website of Aircooled Racing
www.aircooledracing.com the guys who invented the 911 cell for ATL) I should have put the pump low on the starboard side with the supply line running forward from the fill plate.
http://www.aircooledracing.com/atl-3.jpg
The fuel cell has a three-door tank with a pickup in the bottom, which I thought would be adequate to keep the fuel pressure up under high g conditions at the Glen.
But it wasn't, and as a result, in the middle of the esses at the glen, the fuel pressure dropped from 2.5 bar (I had it cranked up a little bit) to ZERO, and the engine began to stutter. The only way I was able to finish the enduro was by lifting off at the end of the straights and letting the plenum refill itself, allowing me enough power to make it through the corners.
So I now have sitting here another ATL three-door tank with an integral Walbro GSS340 gerotor style in-tank pump, and a spare pump just in case.
The application notes for this pump:
PSI l/Hr
0 295
10 283
20 264
30 244
40 229
50 209
60 194
70 174
80 143
90 105
100 43
110 4
OK, and the system pressure is between 1 bar (15 psi) and 2 bar (30 psi). According to ze little white book, the "E" motor wants 125 liters per hour. At a testing pressure of .8 bar +/- .2 bar, the pump should flow a minimum of 900-1000 cc with the power on for 30 seconds. This we will test. But at an operating pressure of 1 bar, the pump should flow 274 liters/ hour (interpolated), or 220% of capacity. THAT should do it. (But that's what I said BEFORE!
http://forums.pelicanparts.com/showthread.php?threadid=61592&highlight=fuel
As a final check what does the motor want at WOT?
HP = pounds per hour/ BSFC
assumed BSFC =.50 (rule of thumb)
155 = pounds per hour /.50
= 77.5 pounds per hour or 13 gallons/ hour or 50 liters/hour.
Does this look right to you guys? If we go back to this thread
http://forums.pelicanparts.com/showthread.php?threadid=113702&highlight=bsfc and pull the BSFC of 290g/HPh. . . (which is .64 pounds/hour per horsepower)
Premium gas = .780 kg/l density (from BOSCH Auto. Hndbk. 5th ed p. 244) or 780g/l. 290g/780g = .37l/HPh .37 x 155 = 57.35l/hour.
Okay, so the factory used about a 100% margin in sizing their fuel pump.
I'll let you know how it works out. On paper it looks fine but the real test will be at Gingerman. . .