Thanks guys. Bunch of experiments going on here. The primary is the large, external diffusion low velocity inlets; I'm shooting for maximum pressure recovery and hope to maintain maximum exit velocity.
Tweaks include relocating the inlet centers to about 1.5" above the crankshaft plane. They also move outboard, a natural consequence of switching to round inlets. The relocation improves the duct shapes in the vertical; more of a straight shot rather than the typical steep ramps. It should also reduce flow reversion because external pressure at the inboard sides of the inlets should be increased compared to stock slot inlets.
Although low velocity (the system is exit throttled) there is certainly some intake velocity, so internal divergent angles remain important. In this case it mostly drives the shape of the ducts and forward plenum wall nearest and above the crankshaft. The shapes there are not perfect, merely reasonable and better than box shapes. The left inlet had to accomodate the prop governor. NASA 3405 suggests some recovery loss because of a governor in the intake. I faired this one into the duct wall with a urethane collar gasket, but the head still protrudes into the intake flow. Again, not perfect but maybe better.
You can probably start an argument in an aero geeks bar (anybody having visions of a Gary Larson cartoon?) regarding plenum volume. I vote for large volume, and this one more or less mimics the inside of the upper cowl. However, it doesn't look large (and it's not) because this is all the available space when you shove a big angle valve motor into an RV-8 cowl. The front and rear upper corners on the left side are really tight; note the one-piece bent valve cover baffles on that side. I even rounded the mounting lugs on the cylinders in order to lower that side away from the inner cowl surface.
What looks like duct tape on the cylinder face is actually fiberglass cloth impregnated with Loctite 598. All the cylinder baffles are lined with this "gasket" and every other gap is sealed. The only air path out of this plenum is between fins.
The remote oil cooler mount is much like one I did on a previous project for a friend. Worked very well. 4" duct, not 3", and again some care in smooth transition between the round duct and the rectangular oil cooler face. The black transition is glass; I outlined how to make duct parts in another thread (
http://www.vansairforce.com/community/showthread.php?t=44856). This one is ducted all the way to the cowl exit.
I am not suggesting you do any of these things, and you certainly don't
need to do any of these things. They are experiments to suit my personal fancy. Hopefully they will add both speed and operational advantage....but they may be just a waste of time. We'll see