[kcameron]

I've been wondering lately if there might be excessive restriction in the induction system of my RV-4's IO-360-A1B. I get about 1.1" drop at WOT when taking off from a near sea-level airport. To test it, I note the manifold pressure reading before starting the engine. I note it again during the takeoff roll. For instance, if the pre-start value is 29.4" then the value during takeoff will be about 28.3". Anyone here know if that's normal?

 

[elippse]

I've been wondering lately if there might be excessive restriction in the induction system of my RV-4's IO-360-A1B. I get about 1.1" drop at WOT when taking off from a near sea-level airport. To test it, I note the manifold pressure reading before starting the engine. I note it again during the takeoff roll. For instance, if the pre-start value is 29.4" then the value during takeoff will be about 28.3". Anyone here know if that's normal?

I wouldn't think so. I have a carb and according to Lycoming this carb has a 1.5" drop at WOT near sea-level. When I do a static run-up I see exactly that, 1.5" relative to engine off MAP and I have a 40 sq. in. K&N filter in the induction tract of my O-235. Unless there's a problem in your instrumentation, I'd say that 0.9" is excessive. Usually in my evaluation equations for injected engines I use 0.5" drop maximum. Do you have a filter, and if so, what is the drop with the filter element removed?

 

[kcameron]

Thanks for the input, Elippse.

I do have a filter: K&N RC-2600. My computations give ~47 sq in which is almost exactly the correct size for a 360 cubic inch engine at 2700 RPM.

The K&N site isn't 100% clear what the pressure drop is for their recommended filter size. Most of their writings indicate they use 1.5" of H20 which would be a little more than 0.1" of Hg. If that's true then I wouldn't expect much more than 0.2" Hg drop even if the filter is dirty. I guess I'll try running without it one of these days just to be sure.

I don't think it's the instrumentation. I've got an AFS 3400 which gives steady MP readings. I suppose there could be some errors due an interference between the pressure pulses and the EFIS's sampling frequency but I don't think so. For one thing, I installed one of Van's restrictor fittings and there's about 2' of hose between it and the MP sensor which should provide a decent amount of smoothing.

I'm thinking it might have something to do with the induction setup the builder chose. He used an updraft sump and a 90 degree adapter between it and the injection servo. Here are some pics:
http://www.cameron.com/~kcameron/rv-4/Pics/2010-07-31--13.19.00/DSCN0032.JPG
http://www.cameron.com/~kcameron/rv-4/Pics/2010-07-31--13.19.00/DSCN0033.JPG

Does anyone else here have some data for comparison?

 

[elippse]

That 90 looks like it has a very sharp bend at the top of the input side where it could cause the flow to separate with some pressure loss. One book on manifold design shows a fish-hook shape where it is desired to have a short 90, that is, it goes a little past, bends somewhat back toward the origin, then completes the 90. Is that clear? I dont know if you have enough space to put something like that in it.
Look at the flat input end of the filter. Can you picture a hole cut in it and a circular rotary valve put in it which could be operated to bypass the filter straight through the center? That blunt end of the filter, if the duct in front is very short, could also cause some flow separation. You might consider making a bullet-shaped input attached to the front of the filter made out of foam covered with a couple of ply of fiberglass. For a test you could carve something out of wood or balsa and glue it on to see if that reduces the loss at WOT. Air really loves to have someone take it by the hand, as it were, and lead it smoothly around curves. At least the throttle valve rotates in a direction to point the air up into the 90 at part throttle! That should be a help in guiding the air around.

 

[Kevin Horton]

For comparison purposes, I looked at recorded data from two flights on my RV-8. The engine is a IO-360-A1B6 which has the stock Bendix servo mounted on the front of the oil sump. I've got Van's snorkel, which takes the air from a flat K&N filter mounted horizontally on the bottom face of the left cowl air inlet.

I looked at the MP before start, during the initial take-off and during touch and goes at the end of the flight. The MP increased slightly during the take-off roll, and was equal to the pre-start MP when the airspeed was about 60 KIAS. The MP was about 0.1" higher than the pre-start value when the airspeed was about 90 KIAS.

I also looked at the MP during a long full throttle run at 2200 ft. The MP was 1.3" higher than the ambient pressure for this pressure altitude, so the induction system seems to be recovering about 85% of the available ram pressure rise, which is quite impressive. This is much better than I had expected.

Notes:
The data has a resolution of 0.1".
There is a fairly small restrictor in the MP line. I'm not sure how much of the slow MP increase seen during the initial take-off is due to the restrictor, and how much is due to increasing ram air pressure as the speed increases.

 

[kcameron]

Impressive

Thanks for the numbers, Kevin. You seem to have a very efficient intake system. I guess I'll have to fix mine. Jealousy is a string motivator.

Maybe I'll try and find some local performance engine shop with a flow bench and see if I can isolate the problem.

I'm thinking of going with the Sky Dynamics cold air intake, sump, and 4-1 exhaust. One of these days.

 

[Alan Carroll]

I also looked at the MP during a long full throttle run at 2200 ft. The MP was 1.3" higher than the ambient pressure for this pressure altitude, so the induction system seems to be recovering about 85% of the available ram pressure rise, which is quite impressive. This is much better than I had expected.

Kevin,

The above analysis suggests that there is essentially no pressure drop attributable to the injector servo or intake tubes - is this correct?

 

[Kevin Horton]

Kevin,

The above analysis suggests that there is essentially no pressure drop attributable to the injector servo or intake tubes - is this correct?

That is my interpretation of the data. I am open to other explanations if anyone has one. I must admit that I expected there to be some loss in MP due to the servo and intake tubes.

As near as I can tell by comparing MP with engine off to the pressure altitude, my MP indication is between 0.05" and 0.1" too high at about 30". I don't have any similar comparisons in the range of 29" MP, which is where the MP was sitting during the full throttle run at 2200 ft. But, I think it is reasonable to assume that the error in the MP indication at 29" is similar to the error in the MP at 30".

Note: there has to be some pressure drop, as the pressure differential is what makes the air flow from the outside world into the engine. But the pressure drop appears to be very small.

 

[hevansrv7a]

Kevin's numbers are likely pretty good

I posted this info once before and got a lot of denials, so it's good to see that Kevin's seeing something like what I see. I will first note that my MP was checked by the same instruments that were used to certify my airplane for IFR.

At around 8000' density altitude and around 200 mph I get well over an inch of "boost". By this I mean that my MP is greater than ambient by more than an inch, perhaps as much as 1.2". As I level off and accelerate to top speed, WOT, I can watch the MP increase.

I am using the SJ cowl and intake system including a K&N from a Ford Explorer. The FI is vertical, like a carburetor; the air makes the 90 degree bend in a box designed by Will James. Will says my numbers are in general agreement with his testing. The air intake is forward facing in a snout so it directly faces the oncoming air plus an additional velocity imparted as the prop blades come around (which would average out).

There is also a slow-down before the air gets to the prop, so we have to be talking about net deltaV vs. the free stream.

I don't know the "delta V" that the prop imparts, but I can approximate the thrust as 212.66 pounds if anyone wants to do the rest of the math. (T=Mass x DeltaV).

So, bottom line, it is a complex analysis, but, if there is a pressure drop between the intake mouth and the location of the MP sensor, it is not much!

 

[elippse]

I don't know the "delta V" that the prop imparts, but I can approximate the thrust as 212.66 pounds if anyone wants to do the rest of the math. (T=Mass x DeltaV).

The deltaV you calculate from that formula is the final result far downstream from the plane due to the "vena contracta", and will be half as much at the prop disc. With apologies to LoPresti and his little addition to the blade TE, it's probably only 5 ft/sec or less immediately behind the blade which passes by the inlet so fast that the effect is probably immeasurable. At 8000' palt on a standard day and 200 mph TAS the maximum dynamic pressure, corrected for compressibility, is 1.15", so you're getting excellent recovery!

 

[Alan Carroll]

Note: there has to be some pressure drop, as the pressure differential is what makes the air flow from the outside world into the engine. But the pressure drop appears to be very small.

Kevin,

What RPM did you use for your test? I did some tests a while back at 8500', WOT, and RPM ranging between 2000 and 2700 (I have an O-360 with carb). I found about 0.5" Hg lower MP at 2700 than at 2000. Put differently, a greater pressure difference was needed to make air flow at a faster volumetric rate, as you'd expect. I imagine a similar relationship must exist for an injected engine, even though the total pressure drop is less than with a carb?