Adam wrote: 

"Common wisdom says that a ram air duct facing into the slipstream

boosts MP, but we already have two big scoops, namely the cooling inlets!

Pressure rise from a ram air duct should be nearly identical to the air

pressure inside the cowl. Anyone have any thoughts?"

 

Indeed I do.  You are correct.  A bit of care is needed, however.

 

My Lancair IV mounts an IO-550 and so has the same induction air problems/opportunities as the Legacy.  I did not much care for the factory Legacy/ES induction "ram air" system because of the flow blockage, small tube size, and poor inlet flow coefficient arising from the sharp-edged angle-cut aluminium tube.  And it costs a lot. 

 

But the concept of collecting air from inside the plenum is sound - if one makes sure the air inducted is cool and the pressure recovery into the cowl is acceptably high, giving a bit of free horsepower from the recovery of ram pressure.

 

Many measurements show about 20F rise in air temperature in the upper cowl which can cut HP is inducted into the engine.  It is incorrect to attribute this to ram air heating.  The same ram air and friction effects create a similar temperature rise on the OAT sensor.  The temperature rise inside the top of the cowl is in addition to the ram air compression effect. 

 

So what is heating the air?

 

It is the air whipping through the inlets, then passing across the tops of the fins of the front cylinder heads without going downward past the rest of the head.  A portion of the cooling air passes over the top of the front heads, is heated modestly, and then flows back and mixes with the balance of airflow.  This then finds it way down through fins (and leaks) into the lower cowl driven by the pressure difference between the top and bottom of the engine.  One has to be careful to avoid prematurely heating the air, and then sucking it into the engine.  Twenty degrees temperature rise causes about 3.5-4% loss in horsepower, about the same losing an inch of manifold pressure.

 

I built my own induction system out of fibreglass patterned on the factory layout, but including some improvements.  Features include:

 

1) Air filter for ground and low level operation.

2) A large flapper valve that directs air either from the air filter or the unfiltered inlet.  No screws are used to hold the flapper onto the shaft.  Rather it is bonded in place and wrapped with 3 BID.  A micro switch indicates if the valve is cracked open bypassing the air filter.

3) A larger, shorter unfiltered air inlet.  The cross sectional area is about 1.5 times that of the throttle body so velocity is lower and friction losses cut by more than half. The inlet tapers and includes a small bell mouth opening (not much room in there) to improve the inlet flow coefficient.  I set the inlet off to the side of the cowl inlet stream so as to not block the flow from the cowl inlet and to minimize water, bug, and bird entry to the engine.

 

Referring to Chris Z's excellent article in Sport Aviation referenced by Scott, I subscribe to the external deceleration theory of cowl inlet design with most of the flow slowing occurring in front of the inlet where the deceleration process is frictionless.  I have sized the inlets (6 inch diameter) so that at high cruise the inlet velocity through the cowl opening is 40% of the free stream velocity.  This means that 86% of the free stream ram pressure is present at the inlet plane, and it is up to me to get as much of the remaining 14% as I can after the flow enters the plenum.

 

 Photos are shown below.  The part was made from a foam master which was then covered with fibreglass, slit in half horizontally, foam chewed out, cleaned up, and then bonded together with the flapper valve captured inside.  The flapper valve shaft is made from 1/4 inch hydraulic tubing.  It took a lot of fussing to get the flapper valve to seal without leakage (no obvious daylight when the flapper is seated).

 

Fred Moreno