Wasn't sure where this belonged, so I put it here and in my gallery thread.
Here are some real world results of water injection at work. My car is a 2003 A4 1.8T with a medium sized turbo (GT2860R with a 68mm aluminum compressor, AKA GT2868). I'm running 27psi from about 4000 to 7200rpm. I have two different timing maps I can load, one for HFS-3 (labelled Race Gas or High Octane) and one for HFS-3 disabled (labelled Pump Gas or Low Octane). When switching between the high octane (with meth) and low octane (without meth) map, I gained about 5 degrees actual ignition timing. I believe there may be room for one or more degrees of timing advance on the High Octane map with meth, since my knock sensors don't detect any knock at these levels.
This graph is a scatter plot, with 6th order polynomial trend lines, grouped into three different classes...
1) High Octane map with Meth (3 runs)
2) Low Octane map with Meth (1 run)
3) Low Octane map without Meth (2 runs)
Raw Data:
http://spreadsheets.google.com/ccc?k...nc&hl=en#gid=0
Spoolin up video, for fun:
http://www.youtube.com/watch?v=6fY6no9rLVk