[Capflyer]

I have embarked on a project that I hope will yield some lost speed. I read through just about every post related to cooling and cooling drag issues along with extensive web searches including links others have posted here.

A little history:
Currently there is about 260 hrs on my fuel injected, dual EI ignition plane from the last 20 months. Initially I had terrible but typical RV7A cooling issues. Last Spring I retrofitted a SJ plenum along with completely rebuilding the front of my stock Van's cowling for the 5.25" round inlets hoping this would help with lowering my CHT's. It ended up lowering my CHT's an average of 15deg. After Oshkosh I replaced my stock oil cooler with a SW which lowered my oil temps 40 deg. All the while I had tinkered with all the other pre-requisite fixes like double checking for leaks and sealing them up, cleaning up any debris between the cylinder fins, etc. Toward the end of last Summer I added louvers to the bottom cowl. Overall I saw my climb and cruise CHT's drop but still running in the 375 deg range in cruise. While helping a friend install an engine monitoring system in his Tiger I noticed that on his lower baffles wrapped around the bottom of the cylinders the sides were sealed with RTV to help with the Tiger's cooling problem. What the heck, I had tried everything else so why not this. Spent a good 10 minutes and a $5 tube of high temp Copper RTV and sealed the sides of mine. CHT's dropped an average of 25 deg!

Fast forward a few months. It's cold and I do not fly any XC's, that is until April when I flew to GA and SC. I found that in cruise flight my IAS and TAS were lower than expected and compared to XC's from last year. My compression check in December for my condition inspection yielded 78's and 79's, fuel and oil consumption all normal, etc. The plane is now polished and painted so a much smoother surface than last year.

Earlier this week I did some speed tests and compared them to ones done last year. The ones from last year were prior to the SJ retrofit and without the intersection fairings on the plane and the ones from this week were done with the louvers covered up.

Tests were done at 8K DA at Peak EGT and WOT and 2600 rpm, speed is in Kts.
Last Year: 149 IAS / 168 TAS / 22.4 MAP
This Year: 143 IAS / 160 TAS / 23.1 MAP

Interestingly, my manifold pressure increased and still yielded slower speed.

The only thing I can now attribute it to is excessive cooling drag. Here are some measurements:
My buddies stock RV8 has a 1:1.1 ratio inlet to outlet size
Mine with louvers closed off has a 1:1.4 inlet to outlet ratio, actual measurements are inlet 35.4"sq and outlet (exhaust pipes deducted) 51.7"sq.

For those not familiar with the nose gear jungle located at the firewall, between the engine mount, nose gear brace and mount, and bottom cowling brace, there is a lot of items internally that I believe create turbulence from the exiting air flow. This combined with tests others have done showing the airflow actually going back into the exit area bring me to my non-engineering degree conclusion that this could be another part of the problem.

I am going to try blocking off some of the exit area opening around the brace between the exhaust pipes and see if both reducing the exit area more along with cleaning up the exiting air will help.

All suggestions, comments, ideas, etc on this will be graciously accepted. My goal is to get something accomplished before flying to Oshkosh in a few weeks.

Here is a picture of the initial stages. Using cardboard to make templates for sheet aluminum to cover up the jungle and close off the center exit area.

 

[DanH]

<<lower baffles wrapped around the bottom of the cylinders the sides were sealed with RTV........... Spent a good 10 minutes and a $5 tube of high temp Copper RTV and sealed the sides of mine. CHT's dropped an average of 25 deg!>>

Sounds familiar.

<<...a lot of items internally that I believe create turbulence from the exiting air flow. >>

Loss of momentum, ie exit velocity.

<<templates for sheet aluminum to cover up the jungle and close off the center exit area.>>

Do some reading on form drag, specifically the drag coefficients for various afterbody shapes. A wedge fairing around the nose gear tubes won't do any good without an afterbody.

 

[David-aviator]

Well, I have not changed my inlet to exit ratio of some 220% one bit and I am still getting Vans numbers in speed and cooling is excellent. I just don't buy into this mantra that exit has to be a close to inlet as possible.

And the beat goes on......what am I doing wrong for a change? True, I won't win a Reno air race, but it sure is ok otherwise.

 

[Rick6a]

Cool Beans

I don't know much about cooling drag and frankly, I am not all that interested in focusing my limited brain power to a study of it either. I just want my plane to perform the way I think it should perform.

A recent and simple modification seems to have significantly improved CHT's on climbout. I installed a pair of Avery cooling louvers. Prior to the mod, CHT on the hottest cylinder could quickly reach 435? in a 95-100 MPH climbout. If I did nothing about it, CHT's would continue to climb to a value I am not interested in discovering. Because of that, I would either level off momentarily or limit most departures to 120 MPH simply to keep the CHT's in the 400? ballpark. Leaned out and in cruise, CHT's have remained unchanged as before and routinely settle to around 380?-390?.

Now, I am not going to offer detailed, specific numbers comparing before and after cruise speeds using various power settings in my O-320 F/P equipped RV-6A because I'm not inclined to document such minutae. If I lost a bit of speed on top end....oh well. But I can say that after the mod, a recent 100 MPH extended climbout to 3500' on a 91? day produced a maximum 418? CHT on the hottest cylinder which in my airplane and experience is a great improvement. Interestingly...at any given time the hottest cylinder may or may not be #3. By the way, high oil temps have never been a problem in my airplane. Indeed, most of the time I have to block off a section of the cooler to warm the oil to acceptable levels. Per Van's recommendation, I installed the baffle mounted oil cooler as high as possible. As hot as it was the other day, I could not get the oil temp past 170? and we know 180? is preferred. Gotta love the precision and accuracy of that VM-1000 engine monitor. Finally, I thank Avery Tools for providing a fix to address an operational need. Apparently many RV's could use it to some advantage. Mine did.

 

[DanH]

David, do you happen to have a good photo of your under cowl exit area? I recall a photo of the side louvers.

 

[allbee]

I don't know much about cooling drag and frankly, I am not all that interested in focusing my limited brain power to a study of it either. I just want my plane to perform the way I think it should perform.

A recent and simple modification seems to have significantly improved CHT's on climbout. I installed a pair of Avery cooling louvers. Prior to the mod, CHT on the hottest cylinder could quickly reach 435? in a 95-100 MPH climbout. If I did nothing about it, CHT's would continue to climb to a value I am not interested in discovering. Because of that, I would either level off momentarily or limit most departures to 120 MPH simply to keep the CHT's in the 400? ballpark. Leaned out and in cruise, CHT's have remained unchanged as before and routinely settle to around 380?-390?.

Now, I am not going to offer detailed, specific numbers comparing before and after cruise speeds using various power settings in my O-320 F/P equipped RV-6A because I'm not inclined to document such minutae. If I lost a bit of speed on top end....oh well. But I can say that after the mod, a recent 100 MPH extended climbout to 3500' on a 91? day produced a maximum 418? CHT on the hottest cylinder which in my airplane and experience is a great improvement. Interestingly...at any given time the hottest cylinder may or may not be #3. By the way, high oil temps have never been a problem in my airplane. Indeed, most of the time I have to block off a section of the cooler to warm the oil to acceptable levels. Per Van's recommendation, I installed the baffle mounted oil cooler as high as possible. As hot as it was the other day, I could not get the oil temp past 170? and we know 180? is preferred. Gotta love the precision and accuracy of that VM-1000 engine monitor. Finally, I thank Avery Tools for providing a fix to address an operational need. Apparently many RV's could use it to some advantage. Mine did.

Thanks Rick for posting pics of your lower louvers. I'm concidering doing the same. currently I'm running 400 to 420 in CHT on take off and cruise. The 420 is coming from cylinders 1 and 3 the 400 is 2 and 4. I'm thinking something is up on the right side of the engine in the baffles. I've gone through and sealed up everything with silicon but the higher heat on the right side persist. I now have 65hrs on the engine and the oil burn is stabalized. The oil temps run right at 180. I to put the cooler as high as possible on the baffle. could the mineral oil be causing higher heat, be interesting to hear if possible. I'm getting ready to change to a standard blend.

 

[Kevin Horton]

Tests were done at 8K DA at Peak EGT and WOT and 2600 rpm, speed is in Kts.
Last Year: 149 IAS / 168 TAS / 22.4 MAP
This Year: 143 IAS / 160 TAS / 23.1 MAP

If I understand correctly, the TAS values you reported were calculated from the IAS, altitude and temperature. Is there any chance that the errors in the IAS have changed between last year and now? For example, if you had a static system leak last year, that would make the IAS read high. Maybe that leak has been fixed, and now your IAS is reading more correctly.

Perhaps your aircraft hasn't slowed down in the last year. You may just have a change in the errors in your indicated airspeed.

 

[Capflyer]

If I understand correctly, the TAS values you reported were calculated from the IAS, altitude and temperature. Is there any chance that the errors in the IAS have changed between last year and now? For example, if you had a static system leak last year, that would make the IAS read high. Maybe that leak has been fixed, and now your IAS is reading more correctly.

Perhaps your aircraft hasn't slowed down in the last year. You may just have a change in the errors in your indicated airspeed.

Thanks Kevin for your responses. I have not changed anything in my pitot/static system so that has stayed consistent. My TAS is calculated for me in my GRT, I keep both on my flight instrument screen all the time. For my future tests I'll switch to using a three direction gps GS which will should be more accurate.

Doing some searches on drag coefficients on different shaped bodies as you suggested. Looks like instead of making a wedge shaped air flow diverter I'll make it more airfoil shaped which may actually be easier to do and bring it out the bottom between the stacks as opposed to finishing off the exit flat at the firewall. The real test for me though is closing off more of the exit surface area and see if that will increase the airflow out.

The Cirrus uses separate exit areas, one for each stack. I do not have the measurements but a friend of mine is a Cirrus salesrep and he talked about the engineering work that went into cutting down the cooling drag and that the outlet area is smaller than the inlet area although he didn't know off hand the measurements or ratio. I'll have to measure one at the airport later today.

 

[airbal]

rv-7A cooling

I have the same engine on the same plane,7A, and i was running cht of 410 at top of climb out and 230 oil temp during first 100 hrs. i installed avery louvers and lost 5kts but cooling was good 380cht and 200 oil temp.

now i have 250 hrs and i decided to cover the louvers and see if the engine could live with the reduced air flow. Good results so far. 380 at top of climb out and 210 oil on a 90 degree day. Also, i regained the lost speed.

I'm building a 10 and i have built and aluminum plenum with f'glass inlets to fit the stock vans cowl. Pray for good results.

 

[Kevin Horton]

Thanks Kevin for your responses. I have not changed anything in my pitot/static system so that has stayed consistent. My TAS is calculated for me in my GRT, I keep both on my flight instrument screen all the time.

Is it possible you had a static system leak from a loose fitting before and didn't realize it? Then maybe you removed and replaced something (ASI, altimeter, VSI, etc) and the leak went away when you tightened the fitting and the leak went away? That would explain why the IAS would decrease. The next big candidate would be the louvers, as suggested by airbal.

I think the biggest factor in cooling drag is the amount of air that comes in the inlets. Considering the aircraft as our frame of reference, all the air that comes in the inlet has to be decelerated. That requires a force from the aircraft pushing on the air, and this force increases the drag. If you reduce the amount of air coming in the inlets, the cooling drag will be reduced. The inlet area is the biggest driver in how much air comes in the inlet. Changing the outlet area will have a much smaller effect on the amount of cooling air that is used.

Cleaning up around the outlet may help, as it could allow the air to be going a bit faster as it comes out. The closer the exit air velocity is to the air velocity on the bottom of the cowling, the less turbulence you should have around the exit, I think. So reducing exit area may help some too.

Good luck.

 

[flickroll]

currently I'm running 400 to 420 in CHT on take off and cruise

What are your sea level take off EGT's? If they're north of 1300 by much at all, you do not have enough fuel flow.

 

[David-aviator]

David, do you happen to have a good photo of your under cowl exit area? I recall a photo of the side louvers.

I do but I can't find it right now.

The opening is like Vans except larger, 5x15 vrs 4x14. Also I have the bottom of the opening cut forward about 2".

Another thing to note here, the 4 exhaust pipes do take up some of the area. I need to subtract their area from the exit total to come up with a more accurate ratio.

 

[DanH]

Just for fun consider the case of the cowled radial. Rather lopsided inlet-exit ratio....but all the air in front of the huge inlet does not pass through the system. It is exit throttled.

Now switch to a flat engine and try this thought. You can throttle the inlet or the exit. Either will reduce cooling mass flow.....so create two systems with the same mass flow, one inlet-throttled, one exit-throttled, by adjusting the appropriate orifice as necessary. Now compare them.

If you throttle the inlet, exit velocity will be low. If you throttle the exit, velocity will be increased. Same mass flow, same cooling capacity, but which is less drag?

 

[N91CZ]

.... I just don't buy into this mantra that exit has to be a close to inlet as possible.

.

David,
You are correct. The parameters that determine inlet and exit areas are totally unrelated to each other. Therefore the inlet/exit ratio simply falls out. It doesn't mean much as a design parameter. I have seen good systems that range from 80% to 250%

 

[johnny stick]

What is really important

Hi all,
I did a bunch of research on this for formula series race cars as it applies to radiator entrance and exits. On the race cars, one car used a vortex to extract the hot air from the radiator, this is probably not appropriate here, but the others were similar to the cowl exits of RVs

From this experience, IMHO the most important thing for reduced drag is how the outside air behaves near the exit. I think the exit area needs to be set to make the exit airflow match the velocity of the free stream airflow. If this is followed to the end, then the cowl that has a lot of internal turbulence would need to have a smaller exit area to match free stream flow. Perhaps the internal cowl airflow also needs to be smoothed to create the highest mass flow for proper cooling, as mention earlier, and the exit area set to make a good match between free stream velocity and exit flow velocity. I think it is a two part solution. JMHO.

 

[allbee]

What are your sea level take off EGT's? If they're north of 1300 by much at all, you do not have enough fuel flow.

Can't tell ya, our altitude for station is 2000ft. I suppose I can fly into a trench and get down to 1000ft. Just a joke there. anyway, I'll be going over to arlington next week, I'll see what I get. I'm normally runing about 1350EGT in cruise. those temps on the cht are the worse I get.

 

[Bill Wightman]

Cooling drag analysis - methodology

Cooling drag analysis on air cooled engines has been studied in depth since the 30's. Some basic relationships I've used in the past to analyze duct flow are:

1 - Cooling drag is a function of momentum losses in the flow thru the engine compartment. The baseline analysis method, then, is momentum analysis. This requires knowing the inlet (V_inf) and exit (V_0) velocities, and the total mass flow through the system. Cooling drag is obtained by dividing the increment in the momentum energy by a velocity that is the average between V_0 and V_inf.
2 - For fixed-baffle geometry, the pressure drop thru the baffle will be proportional to the dynamic pressure ahead of the baffle.
3 - Throttling the exit has been proven the most effective way to control flow through the system. Reference NASA contractor report 3405 done by Miss State and Texas A&M.
4 - Moving air through the engine is totally dependent on the pressure differential between the inlet and exit orifices. The larger the delta-P, the more air you will move. To that end, throttling the exit has another major advantage: exit area coefficient of pressure (C_p) can be dramatically lowered by use of a cowl flap, or a backward-facing step (such as the RVs use). Exit pressures can be practically reduced by about 50%. The P-51 achieved this with a simple radiator door.

The difficulty in putting numbers to this and sizing the inlet/exit areas is that its hard to obtain the engine baffle pressure drop. I have a source on it, but its generic "piston engine" data. The basic relationship is the more airfow you put through the engine, the larger the baffle pressure drop will be. Its really non-linear too. So, if I were doing this, I'd instrument the plane to take this data and do some test points. Then the analysis would be pretty straight-forward.

I used to have a decent spread sheet worked up for this, but have lost it since. Dan, I wanted to send you that spread sheet... its gone it seems

 

[DanH]

<<The opening is like Vans except larger, 5x15 vrs 4x14. Also I have the bottom of the opening cut forward about 2".>>

Got it David, thanks. I sure would be curious to see what would happen if you closed off the louvers....and equally curious about leaving the louvers open and closing off some lower cowl exit area with an afterbody.

Lemme throw out a thought for discussion. Seems like the side-by-side airplanes show up in "poor cooling" reports more often than the tandems. I suspect (no data) the wide belly located just forward of the wing junction results in higher localized pressure at the cowl exit at high AOA, as compared to the longer nosed 8 or the narrow 4. I also suspect moving cowl exit area outboard may locate it in an area of lower pressure, examples being some (but not all) arrangements of louvers, and the dual outlets seen on Cirrus, Lancair, and some of Larry Vetterman's experiments. Comments?

<<Dan, I wanted to send you that spread sheet... its gone it seems.>>

Bill, thanks for looking. I can always use the education.

 

[L.Adamson]

My 6A cools just fine; as does another one I'm familiar with. But then I'm aware of 9's that do not cool as well. Differences in cowling, outlet????

L.Adamson --- RV6A

 

[David-aviator]

<<The opening is like Vans except larger, 5x15 vrs 4x14. Also I have the bottom of the opening cut forward about 2".>>

Got it David, thanks. I sure would be curious to see what would happen if you closed off the louvers....and equally curious about leaving the louvers open and closing off some lower cowl exit area with an afterbody.

I debated removing the side Bonanza vents before freezing the changes and proceeding with the finish and paint work. I taped the vents closed on the outside with duct tape and took off. In flight, no apparent change was noted in cooling and I decided before landing they would go. But then I noticed the reason there was no change, the internal pressure of the lower cowl area had blown the duct tape off on both sides. There obviously is much air exiting the cowl through those vents. I debated some more about taping the vents shut on the inside but decided enough is enough, the airplane was ground for the refinish and paint work, the 10 day job that took 5 weeks.

(Man, is it raining in this part of Missouri today. 2.5" and it's still coming down.)

 

[Capflyer]

Closed Exit Area & Opened Louvers

Over the weekend I fabricated a plate that covers the exit area with holes to fit the exhaust pipes through. My initial tests were just around the pattern because of time constraints. What I found was by covering the exit area and leaving the louvers open, my CHT's in the climb were just slightly higher than with both open. Max CHT was 425 for a 1000' climb at 1200 ft/min. OAT was 80 deg. If I have a chance today and the WX cooperates, I will try some cruise/speed tests. I'll try and post a picture of it later along with some results.

 

[N91CZ]

Single vs Dual Exit Ramps

Lemme throw out a thought for discussion. Seems like the side-by-side airplanes show up in "poor cooling" reports more often than the tandems. I suspect (no data) the wide belly located just forward of the wing junction results in higher localized pressure at the cowl exit at high AOA, as compared to the longer nosed 8 or the narrow 4. I also suspect moving cowl exit area outboard may locate it in an area of lower pressure, examples being some (but not all) arrangements of louvers, and the dual outlets seen on Cirrus, Lancair, and some of Larry Vetterman's experiments. Comments?

.

By splitting and moving the exits outboard, you avoid the nose gear and can build real exit ramps into the fuselage. In a taildragger, you don't have a gear leg in the way, so a single exit ramp works just fine

 

[Capflyer]

Inlet Size

Chris,

I have been using a lot of the information from your excellent website and some of the links you provide. Maybe I missed it but you did a lot of experimentation with inlet sizes. Curious what size you started with and ended up with. I am currently using the 5 1/4" rings from Sam James and would also like to try some different sizes, probably going down. One thing I also need to do is clean up the inside of the inlets on the plenum. In front of cylinder 1 there is a large step at about an 80deg angle up about 1.5" high.

I also read through your report on induction inlet sizes. Any thoughts on the vans stock oval inlet and do you think that changing this to a round inlet would make a difference on the ram air?

 

[N91CZ]

Inlet Sizing

Chris,

I have been using a lot of the information from your excellent website and some of the links you provide. Maybe I missed it but you did a lot of experimentation with inlet sizes. Curious what size you started with and ended up with. I am currently using the 5 1/4" rings from Sam James and would also like to try some different sizes, probably going down. One thing I also need to do is clean up the inside of the inlets on the plenum. In front of cylinder 1 there is a large step at about an 80deg angle up about 1.5" high.

I also read through your report on induction inlet sizes. Any thoughts on the vans stock oval inlet and do you think that changing this to a round inlet would make a difference on the ram air?

Mike,
I really didn't experiment with different sizes. I just made the originally intended size for the Lancair 360 actually work well. It is very, very small and even Lancair never again went that small (in^2/hp) on any later model. The only place you'll routinely see inlets this small is at the Reno Air Races. Smaller inlets shift the burden of pressure recover from external to internal. External recovery is easy to get, but you pay a drag penalty. Internal recovery requires better internal flow control. One needs to decide up front how much work one is willing to invest in the interior before getting to small on the inlets. Typical literature discusses inlet velocity ratios between 0.3 and 0.7. Even at 0.7 you should be looking at a good diffuser. Any higher and it becomes a necessity. The distance available between the prop and engine factor into this. Without an extended hub of some sort options are more limited. The large 80 deg step you describe is deadly. Flow separates at that point and you are left with the equivalent of a sudden expansion with its associated losses. Many a cooling system has been killed by such sudden steps in an area of high velocity flow. Your inlet diameter is quite generous however, so this will have mush less impact. What is critical is the velocity of the air when it hits that expansion. If you've already recovered most of your pressure then it is no problem. Keep in mind that we cannot get ideal internal expansion given the geometric constraints of the typical engine installation. The goal is to recover as much as possible.
I'm not too familiar with the oval inlet. Round has many desirable characteristics that make it very attractive. In general, if you wish to capture ram pressure in the induction system, make sure you maintain a positive seal at every connection.
good luck
Chris

 

[N91CZ]

Inlet Sizing

Chris,

I have been using a lot of the information from your excellent website and some of the links you provide. Maybe I missed it but you did a lot of experimentation with inlet sizes. Curious what size you started with and ended up with. I am currently using the 5 1/4" rings from Sam James and would also like to try some different sizes, probably going down. One thing I also need to do is clean up the inside of the inlets on the plenum. In front of cylinder 1 there is a large step at about an 80deg angle up about 1.5" high.

I also read through your report on induction inlet sizes. Any thoughts on the vans stock oval inlet and do you think that changing this to a round inlet would make a difference on the ram air?

Mike,
I really didn't experiment with different sizes. I just made the originally intended size for the Lancair 360 actually work well. It is very, very small and even Lancair never again went that small (in^2/hp) on any later model. The only place you'll routinely see inlets this small is at the Reno Air Races. Smaller inlets shift the burden of pressure recover from external to internal. External recovery is easy to get, but you pay a drag penalty. Internal recovery requires better internal flow control. One needs to decide up front how much work one is willing to invest in the interior before getting to small on the inlets. Typical literature discusses inlet velocity ratios between 0.3 and 0.7. Even at 0.7 you should be looking at a good diffuser. Any higher and it becomes a necessity. The distance available between the prop and engine factor into this. Without an extended hub of some sort options are more limited. The large 80 deg step you describe is deadly. Flow separates at that point and you are left with the equivalent of a sudden expansion with its associated losses. Many a cooling system has been killed by such sudden steps in an area of high velocity flow. Your inlet diameter is quite generous however, so this will have mush less impact. What is critical is the velocity of the air when it hits that expansion. If you've already recovered most of your pressure then it is no problem. Keep in mind that we cannot get ideal internal expansion given the geometric constraints of the typical engine installation. The goal is to recover as much as possible.
I'm not too familiar with the oval inlet. Round has many desirable characteristics that make it very attractive. In general, if you wish to capture ram pressure in the induction system, make sure you maintain a positive seal at every connection.
good luck
Chris