What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Tuft Testing RV-8 Sub-Cowl

Onewinglo

Active Member
My partner and I felt we were overcooling our O-360 XP so he designed and built a sub-cowl to reduce the exit opening and reduce cooling drag. We expect the reduction in exit opening will increase the CHT and oil temp. Testing has been limited due to weather and schedule but we have noticed a very modest increase in CHT and oil temp. I tufted the belly around the exit and recorded the flight to get an idea of how the sub-cowl disturbed the airflow. The test was at 3000' and 160 - 170 mph. A link to a short video is attached.
RV-8 with custom plenum and Sam James Cowl
Carbureted 360XP with Mags
FP Cato 3 bladed prop
https://youtu.be/qlX7UDfP8Ko
 
Neat, what camera was used to record this?

You should round the sharp edge aft of the exhaust pipes to >1" radius and add a little (0.25" - 0.5") spoiler on the opposite (fore) edge of the opening to separate the ambient stream so it rejoins the cowl exit air with minimal turbulence (can't tuft test this part, trial and error or CFD).
 
So Far, So Good

That didn't work so well. Darn shame...it's pretty glass work.

Its better than we expected so we are calling it a success, so far. I'm interested in CHTs and oil temp in the heat of the summer AND any noticeable change in performance flying high and fast.
This a lot of fun!
 
Neat, what camera was used to record this?

You should round the sharp edge aft of the exhaust pipes to >1" radius and add a little (0.25" - 0.5") spoiler on the opposite (fore) edge of the opening to separate the ambient stream so it rejoins the cowl exit air with minimal turbulence (can't tuft test this part, trial and error or CFD).

Thanks for the pointers Dan V.
The camera is a GoPro Hero 5. The mount was a Cessna strut mount attached to the gear leg with a double ball and socket arm. The ball and socket was fine at 160 mph but slipped at 200 mph.
 
The ramp surface of your sub cowl is separated because the angle is too steep, so there is too much adverse pressure gradient.

I trust that the leading edge of your sub cowl blends smoothly into the round lip at the front of the stock cooling channel? If not, then the flow is separating off of a sharp corner. If yes, then it is an adverse gradient problem as stated above.

This is just one of the reasons why it is far better to throttle the exit by making the actual cowl fair into to a smaller opening rather than add a bulge to the belly. The other reason of course is that the frontal area is reduced at the same time.
 
nice work

I agree with Dan, nice glass work.
I was tired after doing my James cowling....
Bravo to you guys continuing on...
Are the inlets stock? Interior baffeling Vans?
 
The ramp surface of your sub cowl is separated because the angle is too steep, so there is too much adverse pressure gradient.

I trust that the leading edge of your sub cowl blends smoothly into the round lip at the front of the stock cooling channel? If not, then the flow is separating off of a sharp corner. If yes, then it is an adverse gradient problem as stated above.

This is just one of the reasons why it is far better to throttle the exit by making the actual cowl fair into to a smaller opening rather than add a bulge to the belly. The other reason of course is that the frontal area is reduced at the same time.
Thanks for the pointers Steve. Yes the leading edge has a large radius that spans up to the ramp (cooling channel) above. Therefore, it must be the adverse gradient you pointed out. Thanks for the education, I need all I can get!
 
I agree with Dan, nice glass work.
I was tired after doing my James cowling....
Bravo to you guys continuing on...
Are the inlets stock? Interior baffeling Vans?
The inlets are the standard round inlets that come on the James cowling. We built a custom plenum that cools very well.
Thanks for the comments.
JP
 
Instead of after-fairing the exit chute, you can elect to remove it, and replace the whole thing with a flush panel, which reduces frontal area, and leaves an exit the width and height of the belly ramp.

I built progressively smaller exits to determine what exit area was really necessary. This one is the last in the series, extended four inches rearward to make the exit smaller due to ramp taper. I could keep the IO-390 under 380F CHT with this size, 1.625" high, but oil would push 215 if run hard. It would not have worked for slow flight and long taxi (like into HBC at OSH), as I would not have been able to keep oil under the limits. With a minimum flyable exit area established, I added a variable exit panel, the long term goal. Note the panel center-pivots, so half the additional exit adds no frontal area when open. It's good for about 16 sq in. Lower cowl internal pressure rises significantly when closed, meaning velocity rises in the primary exit.

The stuff in the exit photos is test gear, a coaxial pitot-static probe (left) and an exit air temperature probe (right) hidden from the pipe behind an insulated heat shield to block radiant energy from the hot pipe.

Early test flight. The internal pressures are high enough that the first iteration would not fully close the door due to panel bulging. However, watch the tufts when the door closes at 120 knots, both behind the door and behind the primary exit.

https://youtu.be/nA5PY7PYBsU

Adding some structural ribs stiffened the panel enough to allow full door closure: https://youtu.be/aIBXAE2Ezn4
 

Attachments

  • 2012-10-27_18-30-09_773.jpg
    2012-10-27_18-30-09_773.jpg
    356 KB · Views: 13
  • 2012-12-15_16-20-09_817.jpg
    2012-12-15_16-20-09_817.jpg
    476.5 KB · Views: 18
  • Variable Exit Open.jpg
    Variable Exit Open.jpg
    322.6 KB · Views: 19
  • Variable Exit Closed.jpg
    Variable Exit Closed.jpg
    326.4 KB · Views: 12
Last edited:
Go to the vetterman website and look at Larrys work on his afterbody with twin exhaust pipes. The got 4 mph or so IIRC. RV7A, but should apply.

My exit velocity on my SJ (7) is so low that the lower cowl pressure is nearly equal to static pressure up to 170 kts. I just have other issues now to complete phase I, so appreciate your project and sharing.

Note the air wants to come back to the centerline, so a narrowing of the afterbody should help.

edit:
7Zy4U_20uo1zCRHShjb3a6JSKPddDgljl8jGIwMhRYaEavkU0wQzMbMyCnge3-kx2yC2xwOcM6yp4QKThPPTiuXKvH2REKz8foZ33qg357v3a7QWVpoFIFg0G-oPE6QcWyVGVK0DEsd8ndj2aJdMUfX8BGhvNCTBk7GdYreAd4bWcz-mzDpBjWqdTU0BTe6y6O8zbZcQOZlKUOY1OAHTv6K7xTlqgpNZj0us60fsQiNLTvKABxpzITyMdYX1WnL9DSDJ0ts22JoQZws2h8mAStEybneugV51QbAa-yXOwY7_bzbcdiigveeM0aDIRNAC5wGfBkgY9rldUB_fdJehCOVFIxb-yZe2zRsV6PxoRiLf7A7ctCMNjCCcawWeyEq5V81Vq2l0rrmmqxLHcaMUM1RYlzZFpOzR_34kJFuR2NkbcoRxjPvoZjtwB4qinhwYEwSsaNjEkkpAatNQlCsmrkEwItoFAEuyiU-l_cFkbMssG-B4rJ3oGjVXwGCcIdWkvCtaVegiaKB82zkt0rcrlxoYaS5iBM36a_QhaVRpEDfLGZysp6FgY2umwj8DE043AcnAvS4hSiIJZ_SDIbYFm7k4shTd1pFbf3gk0xeFXP7L3ytvcXIzcr4vdJ67uHa5Gc1VnvyTueU5JMyxfs-QlivqPuRa87an=w800-no
 
Last edited:
I built progressively smaller exits to determine what exit area was really necessary. This one is the last in the series, extended four inches rearward to make the exit smaller due to ramp taper. I could keep the IO-390 under 380F CHT with this size, 1.625" high, but oil would push 215 if run hard. It would not have worked for slow flight and long taxi (like into HBC at OSH), as I would not have been able to keep oil under the limits. With a minimum flyable exit area established, I added a variable exit panel, the long term goal. Note the panel center-pivots, so half the additional exit adds no frontal area when open. It's good for about 16 sq in. Lower cowl internal pressure rises significantly when closed, meaning velocity rises in the primary exit.

Dan, have you tested the 'variable exit panel' (aka cowl flap?) during hot weather and/or long ground runs like at OSH to confirm that it provides adequate cooling under those conditions? Or is that testing part of what you refer to by 'long term goal'? I've considered reducing the exit area on my RV-14 to try to pick up some speed (by all accounts it has more cooling than needed), but I'd rather not mess with a cowl flap and prefer to find a fixed exit area that provides an acceptable compromise in all operating conditions...
 
My partner and I felt we were overcooling our O-360 XP so he designed and built a sub-cowl to reduce the exit opening and reduce cooling drag. We expect the reduction in exit opening will increase the CHT and oil temp. Testing has been limited due to weather and schedule but we have noticed a very modest increase in CHT and oil temp. I tufted the belly around the exit and recorded the flight to get an idea of how the sub-cowl disturbed the airflow. The test was at 3000' and 160 - 170 mph. A link to a short video is attached.
RV-8 with custom plenum and Sam James Cowl
Carbureted 360XP with Mags
FP Cato 3 bladed prop
https://youtu.be/qlX7UDfP8Ko

I too am an aero engineer and I agree with Steve Smith's take that your aft-facing surface is too steep to keep the flow attached. However, based on my experience, you have done a pretty good job of keeping the flow-separation damage to a minimum. The separation bubble is not very big and is not very strong. You could easily do much worse!

Unfortunately, anything you do to add frontal area requires a lot of length to smooth it out downstream to avoid flow separation. The fuselage is a good example. You rapidly increase frontal area, then you spend about 10 ft or more fairing it back to a small cone to keep drag low.

Did you substantially increase frontal area in the forward cowl? I am a bit uncertain on how extensive your modification is to the outer moldline of the RV. It seems like you modified two cowl parts ... is that correct?
 
No frontal area increase

I too am an aero engineer and I agree with Steve Smith's take that your aft-facing surface is too steep to keep the flow attached. However, based on my experience, you have done a pretty good job of keeping the flow-separation damage to a minimum. The separation bubble is not very big and is not very strong. You could easily do much worse!

Unfortunately, anything you do to add frontal area requires a lot of length to smooth it out downstream to avoid flow separation. The fuselage is a good example. You rapidly increase frontal area, then you spend about 10 ft or more fairing it back to a small cone to keep drag low.

Did you substantially increase frontal area in the forward cowl? I am a bit uncertain on how extensive your modification is to the outer moldline of the RV. It seems like you modified two cowl parts ... is that correct?
No we did not change the frontal area of the lower cowl. We simply added the sub cowl and changed/reduced the exit in an effort or reduce cooling drag. The link below has a couple images that will help in my explanation.
Thanks for the comments!

https://photos.app.goo.gl/wUgNCDJEEkjXS5A32
https://photos.app.goo.gl/GRXanzkgtVp85Y223
https://photos.app.goo.gl/PiLQ3QxUtDN8iQIz2
 
Kyle, I hate to display my ignorance, but is the inflection point on the following edge of the exit opening?
Thanks!

I may have used an inappropriate term, but what I was referring to is the sharp radius I *think* I see where your new exit ramp turns parallel with the direction of flight.
 
Dan, have you tested the 'variable exit panel' (aka cowl flap?) during hot weather and/or long ground runs like at OSH to confirm that it provides adequate cooling under those conditions

Yes, it has seen all conditions, including OSH.

We can talk after S&F if you like.
 
No we did not change the frontal area of the lower cowl. We simply added the sub cowl and changed/reduced the exit in an effort or reduce cooling drag. The link below has a couple images that will help in my explanation.
Thanks for the comments!

https://photos.app.goo.gl/wUgNCDJEEkjXS5A32
https://photos.app.goo.gl/GRXanzkgtVp85Y223
https://photos.app.goo.gl/PiLQ3QxUtDN8iQIz2


Personally, I think you did fantastic work. I would wager that if you did a tuft test of the original design, you would find significant improvement. The part you added is probably light, and you may well have improved performance (drag reduction v. weight addition) in addition to getting to a better temperature for operation of the engine. Hard to tell without understanding the "baseline." If it were me, I would repeat the tuft test with the original configuration. Thanks for posting this experiment!
 
Baseline

Personally, I think you did fantastic work. I would wager that if you did a tuft test of the original design, you would find significant improvement. The part you added is probably light, and you may well have improved performance (drag reduction v. weight addition) in addition to getting to a better temperature for operation of the engine. Hard to tell without understanding the "baseline." If it were me, I would repeat the tuft test with the original configuration. Thanks for posting this experiment!

Brian,
Baseline is a good idea. I had not considered that. Our sub-cowl is easy to remove so I may take it off, tuft and record a flight.
Thanks!
 
jLLf_dLUoKuG6IVQutEsPKJORcQv957c6-Mm4jCT1T6drFGBZiltl8FHshowppftsuyzOwxLdl947Xm0lKDqeRbhRDDrWXhRE8Eh7vkHOG123COKXzbl1Plz_zId-uYbdBAKE--O2tAdPlBwkWdeDCkyxWzHv0EQbnoiVn-EL8kSBdbS1DyBoGTnZGeh8gN9E8TRMSKQNdBCNbeOWJVjueyBPsNH33VhlMq4Rp4jqjQFMk0M7xcajYMaFSDvqWS-1dWI11uwhF1K1djirifj5gzokE75yWwndoCFLghvwtd1Hlhbqs0brzZrSjhQ0YlSiMMxJOPibNrlaO8yN6_OJ6JV64rN-oIYunlgZGZaoUsBSTXB6y6JfAin78LsfPP78AzOcS9bNlVHcXLv1hvS4JdSbYfaGjxTNOrSWjNYr3f6HAvg1WWo9vnupVmJCj4Db02whva_dpDAZXQ-VPIvw_TF23NSQaTYEsLDg73kmsEVYhAlw0cDamLalDMElqaYD3Z7IoM7r0-pyvUJ002dOmzxh_b70dr97x9n4TLlLl12RsO_HB9PN2OywaszHJCNKis19y-o1xL4b0M7HZ42LX-aahJnb2ZcadcnOeQ-5Y26UKtYgYvErXj4vW_shr7rYjwBHrltE89zsw83TPSgvZWmFmV6Mbc=w1179-h884-no


Hard to be sure, but it looks like the bottom surface just to the right of the exhaust stacks is curving up a bit where the air exits. This would tend to force the exit airflow back up toward the pipes, and create a bit of turbulence.

I think a small filler piece inside the lower surface with a smooth airfoil shape on its upper surface would direct the airflow downwards and smooth out the overall airflow. See the right spoon drawing below, imagine it is rotated clockwise 90*, and the inside of your lower cowl fairing.

Coanda effect. A world of info written by the late Paul Lipps.

You can look up his posts here. http://www.vansairforce.com/community/member.php?u=5053

coanda_effect.gif
 
Last edited:
What is the basic size (height width) of your cowl exit, both counting and not counting the chute under the fuse? My 7 SJ exit height (IO cowl) is 1" lower than the stock Vans and the SJ carb cowl. I was worried it might be too much, but now having measured the upper and lower cowl pressures, found the the lower was about equal to static pressure. The oil streaks (repaired now) on the fuse are like your tufting indicates, some retro flow just around the left/right edges.

If you measure pressures, upper/lower, we can share. ;)

Thanks for posting your results!
 
jLLf_dLUoKuG6IVQutEsPKJORcQv957c6-Mm4jCT1T6drFGBZiltl8FHshowppftsuyzOwxLdl947Xm0lKDqeRbhRDDrWXhRE8Eh7vkHOG123COKXzbl1Plz_zId-uYbdBAKE--O2tAdPlBwkWdeDCkyxWzHv0EQbnoiVn-EL8kSBdbS1DyBoGTnZGeh8gN9E8TRMSKQNdBCNbeOWJVjueyBPsNH33VhlMq4Rp4jqjQFMk0M7xcajYMaFSDvqWS-1dWI11uwhF1K1djirifj5gzokE75yWwndoCFLghvwtd1Hlhbqs0brzZrSjhQ0YlSiMMxJOPibNrlaO8yN6_OJ6JV64rN-oIYunlgZGZaoUsBSTXB6y6JfAin78LsfPP78AzOcS9bNlVHcXLv1hvS4JdSbYfaGjxTNOrSWjNYr3f6HAvg1WWo9vnupVmJCj4Db02whva_dpDAZXQ-VPIvw_TF23NSQaTYEsLDg73kmsEVYhAlw0cDamLalDMElqaYD3Z7IoM7r0-pyvUJ002dOmzxh_b70dr97x9n4TLlLl12RsO_HB9PN2OywaszHJCNKis19y-o1xL4b0M7HZ42LX-aahJnb2ZcadcnOeQ-5Y26UKtYgYvErXj4vW_shr7rYjwBHrltE89zsw83TPSgvZWmFmV6Mbc=w1179-h884-no


Hard to be sure, but it looks like the bottom surface just to the right of the exhaust stacks is curving up a bit where the air exits. This would tend to force the exit airflow back up toward the pipes, and create a bit of turbulence.

I think a small filler piece inside the lower surface with a smooth airfoil shape on its upper surface would direct the airflow downwards and smooth out the overall airflow. See the right spoon drawing below, imagine it is rotated clockwise 90*, and the inside of your lower cowl fairing.

Coanda effect. A world of info written by the late Paul Lipps.

You can look up his posts here. http://www.vansairforce.com/community/member.php?u=5053

coanda_effect.gif
Mike, you are correct in that the bottom surface does curve up a bit. I see how that could disturb air flow. Thanks for the Coanda effect concept. We may incorporate the shape if we modify this thing.
JP
 
Since one of your goals is to reduce cooling drag, has there been any performance change with and without it?

Dave


Dave, we have not flown enough to get any real performance data yet. I'll call it "In Progress" and I'll report back when we have something.
We have noticed a modest increase in CHT and maybe oil temp, but we have lots of margin with our temperatures.
 
Over the last 20 years or so I have spent a lot of time on cooling exits, always trying for a balance between cooling and speed.

Lessons learned

1. Not all engines, even those with the same designations, cool the same
2. Until you get the cowling inlets flowing properly, gradually smoothing the inlet area of the plenum and plugging all holes, there is not much use playing around with the outlet
3. Spinner gap seals help the system work properly, and yes the cooling system is just that, from inlets to outlet, a system.
4. the goal at the outlet is to get the air exiting the cowling to align with the outside air.

Number 4 is where I see a problem with the sub cowling outlet in this thread.
the outlet actually forces the air down at an angle to the relative airflow surrounding the aircraft. This will add drag and likely negate any benefits of the part. I would suggest cutting the "bump" out of the part and glassing in the outlet sides so that they are parallel to the airflow. This will get your outlet air travelling in the same direction as the relative airstream. If this works you can then reduce the outlet size to get the outlet required for hot day climbs, or to add a cowl flap of some kind so that you can manually adjust cooling in the air.
You will know if you have things right if the oil from the breather runs straight down the belly and remains attached to the fuselage the entire way.
I have found that extending the bottom of the cowling exit at least two inches aft of the firewall greatly aids in getting the outlet air flowing in the proper direction.
The exit air is pushed out of the cowling due to the differential pressures within the cowling. I have no proof, but I feel that if you get your outlet air flow in the right direction, the outside air may even help to pull the air out.

I have a cowl flap on my aircraft but seldom use it now that have the system flowing properly. If you live in the deep south then you will likely need some adjustment if you are looking for maximum speed/cooling.
 
Last edited:
Tom said: " I have no proof, but I feel that if you get your outlet air flow in the right direction, the outside air may even help to pull the air out."

It doesn't, as there's nothing between the molecules to "pull" the exit air out.

Try John Thorp's article from the December 1963 Sport Aviation article on "Cowling and Cooling" will help! (I'll have to learn how to add attachments!)

Till then, the EAA S/A archives will have to suffice!
 
If I may rephrase; a vacuum could be created with the passage of the faster moving airstream, relative to the slower air mass exiting from the cowling which could be at a lower pressure then the passing airstream.
When I did my differential air pressure tests I did not test the relative pressure exiting the cowling relative to the airstream around the bottom of the fuselage. I will refrain, in the future, from using speculation.
 
Last edited:
Over the last 20 years or so I have spent a lot of time on cooling exits, always trying for a balance between cooling and speed.

Lessons learned

1. Not all engines, even those with the same designations, cool the same
2. Until you get the cowling inlets flowing properly, gradually smoothing the inlet area of the plenum and plugging all holes, there is not much use playing around with the outlet
3. Spinner gap seals help the system work properly, and yes the cooling system is just that, from inlets to outlet, a system.
4. the goal at the outlet is to get the air exiting the cowling to align with the outside air.

Number 4 is where I see a problem with the sub cowling outlet in this thread.
the outlet actually forces the air down at an angle to the relative airflow surrounding the aircraft. This will add drag and likely negate any benefits of the part. I would suggest cutting the "bump" out of the part and glassing in the outlet sides so that they are parallel to the airflow. This will get your outlet air travelling in the same direction as the relative airstream. If this works you can then reduce the outlet size to get the outlet required for hot day climbs, or to add a cowl flap of some kind so that you can manually adjust cooling in the air.
You will know if you have things right if the oil from the breather runs straight down the belly and remains attached to the fuselage the entire way.
I have found that extending the bottom of the cowling exit at least two inches aft of the firewall greatly aids in getting the outlet air flowing in the proper direction.
The exit air is pushed out of the cowling due to the differential pressures within the cowling. I have no proof, but I feel that if you get your outlet air flow in the right direction, the outside air may even help to pull the air out.

I have a cowl flap on my aircraft but seldom use it now that have the system flowing properly. If you live in the deep south then you will likely need some adjustment if you are looking for maximum speed/cooling.

Tom,
Thanks for sharing your experience on this issue. I'm sure you are correct about the exit air should be aligned with the slipstream. After tuft testing and pondering everyone's suggestions the concept is getting much clearer. Our current plans are to fly for a while and evaluate a modification later.
 
Revised RV-8 Sub-Cowl

I tufted our RV-8 and flew with the modified sub cowl a few days ago. The point of the sub cowl is to reduce the exit opening to raise CHT and oil temperature, which it has. (Our XP360 was running a bit too cool). The below video indicates the exit flow is not perfect but better than the first attempt.
https://vimeo.com/308621727

Here is a video of the first attempt. Note the large lip on the trailing edge of the opening which caused flow separation.
https://vimeo.com/308619285
 
really good flow

I tufted our RV-8 and flew with the modified sub cowl a few days ago. The point of the sub cowl is to reduce the exit opening to raise CHT and oil temperature, which it has. (Our XP360 was running a bit too cool). The below video indicates the exit flow is not perfect but better than the first attempt.
https://vimeo.com/308621727
That looks like some really good flow. Any feeling for impact on speed?
 
Speed Increase? - Maybe

Mickey,
Nothing conclusive about speed yet. I've only made two runs at altitude due to weather. One flight due North and one South.
Normally our -8 yields 185 mph true, at 2500 rpm, at 8000 DA.
Flight with modified sub cowl yielded 187 mph true, at 2500 rpm, at 8000' DA.
More importantly our temps increaseed as desired. Oil temp was 172* and CHTs 295 - 320*. I forgot to record the OAT but it was a cool Louisiana day, 60* F at the airport.
 
I feel a bit silly chiming in, with actual aero engineers participating, but...

No one's mentioned the down-turned exhaust pipes' effect on flow out/around the opening. Is my failing memory deceiving, or hasn't there been testing that shows that and exhaust column exiting at an angle to the free-stream causes a significant increase in drag? (IIRC, the penalty for parallel exit flow is increased cockpit noise.)

Likely a minor detail, but what about filling in the sides of the afterbody? How is the air behaving in those 'corners'? Any chance that it flows back into the corners, then tries to follow the curve down and then horizontal toward the pipes, at an angle to the freestream?

Charlie
 
Cut the pipes?

Charlie, I agree the pipes in the slipstream likely cause drag. I sent Clint at Vetterman an email asking about cutting the tips at an angle parallel with the sub-cowl opening.
I want the exhaust pointed away from the belly to reduce noise, but the tips may be a bit long as they are. I look forward to hearing Clint's opinion.
 
I feel a bit silly chiming in, with actual aero engineers participating, but...

No one's mentioned the down-turned exhaust pipes' effect on flow out/around the opening. Is my failing memory deceiving, or hasn't there been testing that shows that and exhaust column exiting at an angle to the free-stream causes a significant increase in drag? (IIRC, the penalty for parallel exit flow is increased cockpit noise.)

Likely a minor detail, but what about filling in the sides of the afterbody? How is the air behaving in those 'corners'? Any chance that it flows back into the corners, then tries to follow the curve down and then horizontal toward the pipes, at an angle to the freestream? Charlie
Great question and you are right. The drag you talk of is called "interference drag". It is low speed air mixing with high speed air at angles. The mixing is drag. However as was said there are practical issues. If you have the pipe exiting parallel and near to the belly it can cause not only noise but structural damage.

1) The exhaust will cause noise a thumping into the cabin. People find it very objectionable.

2) Sonic and pressure pulse will vibrate and fatigue the structure, causing cracks and loose rivets.

3) On the Pro side, 4 into 1 with 2/125 collector exiting parallel ti free air-stream can produce thrust.

4) The down turn pipe minimizes or negates 1 through 3 (good and bad).

You can minimize 1 and 2 by dampening the belly with sound deadening liner in the cabin and placing a heavy plate on the belly over the skin. All this adds weight. Nothing is free.

Note Augmentor Tube - The augmentor tube uses the velocity of the exhaust gases to produce a low pressure on one side of the engine that helps pull cooling air through it. There is a large tube (augmentor) with curved bell mouth on inner cowl side. The exhaust pipe dumps into the augmentor tube entrance with cowl cooling air... The Piper Apache twin I owned had this. It clearly speeds up the cowl exit air, which induces better cooling. However I read it provides some thrust as well? The Pipe augmentor tube was buried in the nacelle and fairly aft facing. I rationalize air coming out as fast or faster than air stream and parallel to it, the better. The RV does not have room to develop an augnemtor tube, or does it?
 
Last edited:
Proper design of an Augmenter Tube - one factor is sufficient length that an exhaust pulse is still in the tube when the next cylinder's pulse enters the Augmenter Tube, denying a chance for reverse flow.

# of cylinders and operating RPM are relevant factors to length

FWIW
 
exit pressure must match local exterior flow at exit

From Tom Martin:
" I have no proof, but I feel that if you get your outlet air flow in the right direction, the outside air may even help to pull the air out."

The fundamental rule of subsonic exit flow is that the internal flow pressure at the exit must match the external flow pressure at the exit.

When you open a cowl flap or have a diverging exit angle, you are creating an area of lower pressure in the external flow, which yes indeed helps extract more flow through the internal flow path, because you are exiting into lower pressure. The penalty for this is the increase in frontal area needed to create the accelerated external flow, and the likelihood that exiting internal flow may be at a lower velocity than the external flow.

When you have an exit in an area where the external geometry is tapering or sloping back towards smaller cross sections as you go downstream, the other flow is slowing down, pressure increasing, and you are exhausting into an area of elevated pressure compared to the free stream. The internal flow at the exit is therefore at a higher pressure, lower velocity than it could have been. In small doses, this region of decelerating flow, with rising pressure, can be good - the pressure recovery on the aft-facing slope provides some thrust. But if the pressure gradient is too adverse (pressure rising too fast) the boundary layer will separate and you will have a separated flow region trapped in the area where the body is getting smaller, causing pressure drag - which is a way of saying that you are not getting the pressure recovery that you would have gotten had the flow stayed attached.

The general shape of the lower cowl is causing a flow acceleration (cross section area increasing as you go downstream) that peaks somewhere just ahead of the firewall, causing high velocity and low pressure. From that point, the flow starts decelerating to return to nearly free-stream velocity along the underside of the fuselage where the flow is in roughly the free-stream direction and no longer curving. This is one reason why the area near the flrewall is a particularly good place to have the cooling flow exit.

Note also that although the exit pressures must match, the exit velocities in general will not, because the internal flow has gone through other passages that have caused a loss of total pressure (pitot pressure) compared to the free stream flow. This is partially offset by an increase in temperature in the case of cooling flows. In a well-designed cooling flow process, it is possible to have the exiting cooling flow actually have higher velocity than the external flow at the exit, and can produce thrust.
 
So in laymans terms ?

Appologies for any drift here but hopefully my question will help others. I am currently cowling a -4 with 0360 and Vetterman 4 pipe downturn exhaust. Archives have suggested:
1-Provide nice curved surface to guide air around the engine mount tubes
2-Dimension the exit area to conform to the Horton Vi/Vo ratio
3- If you need climb cooling add cowl flap
4- Dont cut off the downturn tubes but ?maybe? trim parallel to free stream
Would Steve or Dan please suggest a ?prescription? for a layman to follow on this task ?
 
Appologies for any drift here but hopefully my question will help others. I am currently cowling a -4 with 0360 and Vetterman 4 pipe downturn exhaust. Archives have suggested:
1-Provide nice curved surface to guide air around the engine mount tubes
2-Dimension the exit area to conform to the Horton Vi/Vo ratio
3- If you need climb cooling add cowl flap
4- Dont cut off the downturn tubes but “maybe” trim parallel to free stream
Would Steve or Dan please suggest a “prescription” for a layman to follow on this task ?

Larry, first suggestion is to think in terms of building a system. Whatever individual modifications you might make should be considered in a system context. Bandaids only go so far, and real bolt-on speed is rare.

Next, remember the little things. I'm going to take one subject as an example, that 4-pipe exhaust. You're thinking about cutting off the downturns, but in the context of how to reduce cooling drag, the best bet would be to stack it under a workbench somewhere.

Why? First, they have a lot of surface area, a minor deal, but it all adds up. I don't know the tailpipe diameter, but I'll guess 1.25"D for illustration. Total exhaust area is 4.9 sq in for the four pipes, exactly the same as a single 2.5"D tailpipe on a 4-into-1. However, skin friction is roughly doubled, because total surface circumference is doubled, and one system goal was exit velocity.

Consider cowl exit shape. Four pipes don't package well, and they eat up a lot of exit area, thus requiring more frontal area. Our goals were to reduce exit area and frontal area. Look up Axel's RV-4, which is about as far as a fella can go eliminating frontal area a 4-pipe. Frontal area is reduced, but the pipes are outboard, near the gear leg fairings. That junction is not likely to be low drag, and the pair of down-pointing exhaust streams form an invisible wall. It's not just a matter of eliminating lengths of pipe hanging down in the freestream. It's also the plume of high velocity exhaust.

Axel.jpg


Let's look at Dave Anders' RV-4, or his friend Brian Schmidtbauer's Mustang II. Those are fixed exits with a single tailpipe. Overall cowling frontal area is about the same as Axel's, or maybe even a wee bit more. However, neither suffer from an exhaust plume wall, and in Dave's case, the outboard areas near the gear legs are slick. The exits are at the point of maximum fuselage cross section with little or nothing to trip the flow; consider that in the context of Steve's pressure tutorial. If desired, either could add variable area forward of the firewall plane, and shrink the fixed exit some additional amount, but instead they've gone with a degree of exhaust augmentation to supplement low speed cooling.

Looks familiar to RV14 owners, yes?

Dave%20Anders.jpg


Brian%20Schmidtbauer.jpg


Augmentation comes with its own design issues, so I elected to work toward a good variable exit strategy for low speed cooling. The RV-8 already has clean areas outboard, toward the gear legs, so I chopped off a whole bunch of frontal area by eliminating the entire "backward coal shovel" from the bottom of the stock cowl, eventually reducing the fixed exit to a slot 1.625" high. It's supplemented with variable for climb or cool cruise, which only adds about 8 sq in of frontal area when fully open...not what most would consider a "cowl flap". The tailpipe is slash cut to point the plume rearward, although not directly aft where it would be entirely thrust. That's a compromise to keep the airplane practical; low noise and belly drumming, not so much floor heat. I visualize the plume as an ordinary tube extended into the airstream. Like that phantom tube, angling it back reduces drag. Exhaust smut on the belly tells me it reattaches a little past the wing spar, which tells me the compromise is working; recall that RV8s tend to crack the belly skin at the ends of the floor ribs, just in front of the spar...where I did not want it to drum. Point here is that (1) it's a single plume, not two wide plumes, and (2) sometimes compromises are necessary. If I was building an RV-8 for Reno, it would be different.

Cowl and exhaust as delivered. Targeted exit area sketched in, tailpipe as yet uncut.

Cowl%20Exit%20New.jpg


Finished exit, with variable area forward and cut pipe. The trash in the exit is measurement gear, all since removed.

2012-11-10_16-47-27_889.jpg


I can't offer a specific prescription, just ideas. You'll have to do like the rest of us...sit on a stool and stare at it a while, read some books, think about it in your spare time, then build, then measure if you really want to know how well you did.
 
Last edited:
Great post Dan. The fast Reno guys know the plume drag can be significant but they don't have to care much about noise. They go to great lengths to eliminate this from control surfaces too with exotic gap seals.
 
Last edited:
Thanks Dan !

As usual, the feedback is absolutely excellent. I can do this with minimal stress except for exhaust. Just FWIW, Clint and Mr. Vetterman both insist the four pipe yields better engine results than crossover or 4 into 1 and is NA for the -4. There are 4:1 available at $$$$$, but I get the feeling it might be a trade-off between drag and engine performance. As suggested, a chair in the shop with a beer and the airplane is next :rolleyes:
Thanks much Dan
 
...Clint and Mr. Vetterman both insist the four pipe yields better engine results than crossover or 4 into 1 and is NA for the -4. There are 4:1 available at $$$$$, but I get the feeling it might be a trade-off between drag and engine performance.

Yer' welcome, and I think your feeling hits squarely on the truth of the matter.

All good choices are based on reasoned compromise. My reasoning was that I could go faster, cheaper, via drag reduction than HP increase. I am not always right. Just ask Ms Patti....
 
Reno

Great post Dan. The fast Reno guys know the plume drag can be significant but they don't have to care much about noise. They go to great lengths to eliminate this from control surfaces to with exotic gap seals.

...and the fast glass forced induction guys have also been known to occasionally burn the belly composite on their fuselages when the plume gets too close too soon...

I think Andy Findlay had this happen in 2017.

Skylor
 
...and the fast glass forced induction guys have also been known to occasionally burn the belly composite on their fuselages when the plume gets too close too soon...

I think Andy Findlay had this happen in 2017.

Skylor

Modified for 2018 along with many other lessons learned in 2017 for the Gold win. Can't learn without pushing the limits or making some mistakes.





Augmentor detail. A large portion of the engine cooling air passes through these tubes aided by the tremendous amount of exhaust energy present here. It worked extremely well compared to non-augmented setups with similar power.
 
Last edited:
Tapped on photo and the link says is private. I would like to see more so as to make out the augmentor in the photo. Maybe could you make the photos public? Thanks
 
Tapped on photo and the link says is private. I would like to see more so as to make out the augmentor in the photo. Maybe could you make the photos public? Thanks

These photos share the Flickr account with some of my other public modeling photos which I don't want mixed together. You can just right click these here on VAF and save to your computer. I don't have any more showing the augmentor detail.
 
Last edited:
Exhaust plume refinement ?

Dan,
Something used on antiques to abate noise was converting a round pipe into a manifold with a slot for an exit ( reference some radials and early flat 4 cyls). This might have potential to consolidate 2 pipes into a flat slotted receptacle that might enhance airflow over exhaust exit, and allow a more parallel discharge to fuse without drumming ? Critique welcome from all.
 
Back
Top