What's new
Van's Air Force

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

Oil Cooler outlet duct?

claycookiemonster

Well Known Member
I see many mounts and placements for oil coolers, but they all end at the cooler. Air is blown on it, and then...? The interior of a cowling, particularly below the engine has got to be a very turbulent place, and we're just dumping more turbulent warm air into it at the cooler.

Has anyone installed any kind of ductwork downstream of the oil cooler? It occurs to me that some sort of plenum full of warm air could extend to near the cowl exit where we could depend on a low pressure area to pull the air out of the cooler chamber. (P-51 afficionadoes might even achieve the purported thrust that aircraft was said to achieve by it's radiator)

It might only take 12 - 18" of ducting to lead the oil cooler exit flow near the cowl exit to assist the flow through the cooler as well as keeping that part of the heat somewhat contained?

Or, am I (ah-hem) full of hot air?
 
I have seen a few examples of this here on VAF.

I believe Dan Horton has one.

There was a gentleman named Bob Axsom with a RV6 that did a lot of racing that did a duct.

Alan Judy also, if my memory is correct.
 
Last edited:
A cooler exit duct is useless if it does not extend all the way to a low pressure area. In fact, it's probably worse, as it adds skin friction.

As our own Steve Smith explained to me, with a standard cowl the pressure drop (inside the lower cowl vs freestream) is right at the plane of the exit. Thus an oil cooler duct must reach all the way to freestream.

Mine is neither fish nor fowl. The long converging exit bell drops pressure below lower cowl pressure at a point prior to the exit plane, but not quite as low as freestream. I wanted to run rather high lower cowl pressures.
 
assist the flow through the cooler

Not sure I wanna do that.. like many others O-360 users I struggle to reach ideal oil temps, even with no air to the cooler, as in shutter fully closed.

On parallel thinking, one thing I wonder about is how to replace that Vernatherm automatic valve with a manual control...
 
This is a very interesting question and one that I plan to explore with my biplane build. Through the modify/fly/test route. My oil cooler is FW mounted up high and will receive dedicated ram air pressure for cooling but the outlet will initially be the large area where the cylinder cooing air exits the cowl. All mixed together. Plans are to add an exit plenum that dumps into the low pressure area of the exhaust tunnel and also a shutter valve at the cooler. All done incrementally with test flights in between. One of the smartest engine guys I know - Tom Wilson - suggests that this is the way to go. We'll see.
 
To the OP, I'm going to respectfully suggest holding off on this decision and save yourself some possible unneeded work.

There are a lot of variables that will determine how effective your oil cooling circuit will be: cooler size, things that will affect airflow conditions (engine plenum sealing, your flying environment, position of the baffling cooling air take-off), real oil flow rate, etc. There's a pretty decent mix of boundary and process conditions and their respective influencers.

Based on these and the oil cooler exit air temps reported here, increased airflow through it will have lees effect than one would think unless there is a pretty gross error elsewhere. I'd stick to the plans unless you encounter an issue. My dos centavos.
 
What is your goal?

These guys have addressed the basics and DanH quite well as usual. As an observation and pitfall, poor oil cooling has typically resulted from either, (1) speed pursuits where mass flow (and delta P) across the engine was reduced, or (2) excess bypass around the engine (again low delta P) resulting in restricted cowling inlet or exit of the standard cowl.

Either of these will reduce mass flow across the cooler, then the attention of the oil temps begin.

Which direction are you attempting to go, reduced exit, (speed) or something else?

Alan Judy made a beautiful separate circuit for his oil cooler. And a clever variable inlet opening for low speeds. It was needed because he restricted the inlet and exit air areas gain speed. His speed gain was quite successful in the small inlet, small exit category.
 
Good question. "My Goal" isn't clearly defined yet, as I'm still building.
I've just noticed the great care that goes into many of the "upstream" areas (plenums, oil cooler ducting, Scat tubing, etc) and then the relatively unscripted and chaotic "downstream" areas.
I'm looking intensely at all the contingencies my build may face in all areas, and trying to prepare for them now. I recall the posts trying to address whether the supposed low pressure area at the cowl exit really is low pressure, and how to address that. The Vetterman site has an interesting mod to the cowl exit to address this. So, I began to wonder about whether adding a few ounces of ducting after the oil cooler might make cooling more efficient? Maybe with better downstream management, I'd need less upstream input?

Just curious. I've been left alone in the house. It's either this, or I begin to chew my own sneakers.
 
Oil Cooler

Dan 57 mentioned manual control of oil flow to the cooler. Hardware store ball type valve modified for a manual push pull control cable. The vernatherm is removed. I believe the Lycoming parts manual shows the parts that replace the vernatherm.
Alan Judy had a inlet door for the cooler air in that could be completely closed. I don't remember what he had for cooler exit air.
 
Oil Cooler Fan

Has anyone tried using an electric fan for pulling/pushing air through their oil cooler? Like is done on automotive radiators. I did a search and didn't find anything. Seems like a good way to regulate oil temps.
 
The subject fan exists for when there's no velocity to motivate air through the cooler. Unless your engine oil temp is hitting redline prior to flight, what would be the benefit? There's better (lighter, cheaper, simpler) ways IF there's an issue to address.
 
My cooler is mounted almost flush with the top skin of the cowl. My plan was to use a scoop and force air down through the cooler and then duct it out the exhaust tunnel. Yesterday when I was looking at it I thought...what about sending air the other direction. Out the top skin using a fan and creating a low pressure area for the exit air. Yes, it would add the complexity of a fan but saves a fair amount of ducting work and would seem to give me greater control over oil temps. Especially with the already-planned shuttle valve. Just spit balling ideas here.
 
Actually, the more I think about it, the more inclined I am to think this is the way to go. Use the slipstream to "pull" air through the cooler. Fan or no fan.

I have a "door" on the pilot's side window of my Bonanza that is about the same size as an oil cooler. There is so much low pressure at that opening in flight that anything above 130 mph makes it almost impossible to open the door. If I don't have a firm hold on it until it's about 45 degrees open it will slam shut. Also, any skydiver knows that the open door of a jump plane is a tremendous hazard for anything loose in the cockpit. Like an open parachute!:eek: Lots of negative pressure at that opening from the slipstream. Besides, I get free windshield defrosting this way.:)

What do you guys think?
 
really

My cooler is mounted almost flush with the top skin of the cowl. My plan was to use a scoop and force air down through the cooler and then duct it out the exhaust tunnel. Yesterday when I was looking at it I thought...what about sending air the other direction. Out the top skin using a fan and creating a low pressure area for the exit air. Yes, it would add the complexity of a fan but saves a fair amount of ducting work and would seem to give me greater control over oil temps. Especially with the already-planned shuttle valve. Just spit balling ideas here.

I think this would be ugly, and also the air pressure might be higher in front of the windscreen. Seems tried and true is the simplest. But to each there own. Not to mention all the rain that soaks everything when sitting on the ground....
 
CHT not OT

Seems like just about every RV6 and 7 that I know has issues with CHT (which can be addressed by sealing the dozens of leak paths), while concurrently having no Oil Temp issue. Now if you have an engine with piston squirters, and intend on a lot of Palm Springs / Yuma summer flying, maybe you can get into an oil temp issue. Somewhere before or after melting the heads. :-|

Personally, I installed a larger oil cooler than standard, mounted on the firewall, fed with a large duct, and dumps into the lower cowl. No oil temp issue despite flying in the SW US. Always hit a CHT issue first.

My only interest in oil cooler exits is to reduce cooling drag.

See the exit air ducts at https://www.kitplanes.com/so-youd-like-to-go-faster/
 
My only interest in oil cooler exits is to reduce cooling drag.

Which is unlikely.

Remember, cooling drag= mass x loss of velocity. To cut cooling drag, design to reduce the quantity of air flowing through the system, or increase its exit velocity, or both.

Reality...The oil cooler accounts for only a small percentage of the mass flow, and we really can't reduce it anyway, not without also reducing cooling capacity. Increasing the velocity of a small mass would offer only a tiny drag reduction, but it's a moot point, because ducting will probably slow the flow, not increase it, due to skin friction.

As before, the only good reason to duct the cooler exit is to reach a low pressure area, because cooling was inadequate when dumping into high pressure. More deltaP means more mass flow, increasing heat rejection, but also increasing drag, not decreasing it.
 
Pulling air through OC into slipstream

I think that's what I'm going to do, let the slipstream pull air through the cooler and dump it overboard through the top skin. And use a shutter valve to regulate temps.

As mentioned in my earlier post this is a biplane build and I don't have many of the restrictions we find in RVs. I have plenty of space to work with, not concerned about rain, the canopy is ~ 3ft aft of the cowling. If I need additional flow I can duct air to the bottom of the cooler from the upper deck high pressure area. I am most concerned about rejecting as much heat as possible from the cooler but am also concerned about too much cooling at cruise settings. I want as much control over the cooling process as possible. From what I can tell nobody as tried this yet so what the heck. I can always go back to installing a scoop in the top skin and forcing air through the cooler if I need to.
 

Attachments

  • Oil cooler installation.jpg
    Oil cooler installation.jpg
    462.9 KB · Views: 112
Which is unlikely.

Remember, cooling drag= mass x loss of velocity. To cut cooling drag, design to reduce the quantity of air flowing through the system, or increase its exit velocity, or both.

Reality...The oil cooler accounts for only a small percentage of the mass flow, and we really can't reduce it anyway, not without also reducing cooling capacity. Increasing the velocity of a small mass would offer only a tiny drag reduction, but it's a moot point, because ducting will probably slow the flow, not increase it, due to skin friction.

As before, the only good reason to duct the cooler exit is to reach a low pressure area, because cooling was inadequate when dumping into high pressure. More deltaP means more mass flow, increasing heat rejection, but also increasing drag, not decreasing it.

Interesting point. I would have thought that reducing the turbulent uncoordinated flow in the lower cowl (even if just the oil cooler contribution) would reduce cooling drag and/or airframe drag.
 
Debate= lots of words with no quantifiable result.

Which is unlikely.

Remember, cooling drag= mass x loss of velocity. To cut cooling drag, design to reduce the quantity of air flowing through the system, or increase its exit velocity, or both.

Reality...The oil cooler accounts for only a small percentage of the mass flow, and we really can't reduce it anyway, not without also reducing cooling capacity. Increasing the velocity of a small mass would offer only a tiny drag reduction, but it's a moot point, because ducting will probably slow the flow, not increase it, due to skin friction.

As before, the only good reason to duct the cooler exit is to reach a low pressure area, because cooling was inadequate when dumping into high pressure. More deltaP means more mass flow, increasing heat rejection, but also increasing drag, not decreasing it.

Interesting point. I would have thought that reducing the turbulent uncoordinated flow in the lower cowl (even if just the oil cooler contribution) would reduce cooling drag and/or airframe drag.

Probably a debatable situation, good for a beer at the OSH Pavilion - although not sure how it might shake out.

If a control volume is made around the cowl with the appropriate inputs, outputs then one can argue any internal drag reduction for the air flow from cooler to exit can organize the flow, help reduce the drag loss (aka pressure, enthalpy actually). But, reducing loss will unbalance the parallel flow through the engine, and increase total mass flow. Now, if one needed a tick more oil cooling to achieve balance, then there is a solid benefit, but the numbers would have to say whether the increased mass flow increased external drag over the reduction of the internal drag of the chaotic flow. Regardless of the small difference (either way), the measurement of external drag would be hard to measure. Oil cooling is (a little) easier to quantify. Maybe we should look at the increased engine power provided by lower friction losses of higher oil temperatures. :rolleyes:

I keep considering a duct for my cooler for increasing ambient capability, but my internal debate is not finished.

Either way, Dave Anders internal flow passages for cooling air is very clean.
 
Difference in loss through the forest of tubes, wires, and hoses, vs loss through a duct would be very difficult to quantify, given everyone's engine compartment is a unique mess.

What is not hard to quantify is use of a duct to modify deltaP across the cooler. From an earlier version of the exit, we have...

In round numbers, at 175 KTAS the reduced exit area on the modified cowl resulted in more than 6" H2O lower plenum pressure. Ballpark, I had something in excess of 15" in the upper plenum. So, 15 less 6 would be 9" deltaP available if the cooler exit simply dumped into the lower cowl volume. Per the SW charts, flow through a 10611 cooler would be 50 lbs/min at sea level density.

To drop oil temperature to where I like it (angle valve 390), I needed more deltaP, so the cooler outlet was ducted to the top of an exit bell of smoothly decreasing area, i.e. increasing velocity and decreasing pressure. Measured pressure at a point 3" forward of the plane of the exit, near the least area section, is only 1.5" H2O above pressure measured 2" aft of the exit, outside the cowl. Connecting the oil cooler duct to the top of the bell 3" forward of the exit meant the oil cooler deltaP became 15 less 1.5, or 13.5, less some additional unquantified duct loss. Let's call the result 13" for this example. Again per the SW charts, the additional 4" deltaP boosts cooler mass flow an additional 10 lbs min, which increases heat transfer by about 5%. Note the split between deltaP applied to CHT and deltaP applied to oil cooling. The cylinder fins are seeing 9", while the cooler is seeing 13"

Note this result requires more than just a duct. It also requires a connection to a low pressure location, internal as above, or external, like we've seem with some bluff body exits. The exit bell has some skin friction, so although internal, it's not a free lunch

It is possible to roughly quantify the relationship between cylinder mass flow and oil cooler mass flow. The Lycoming charts tell us maximum acceptable CHT typically requires 2 lbs/sec (ballpark), or 120 lbs/min, meaning cylinder mass flow is at least double oil cooler flow, even with a large cooler pushed with reduced deltaP. The sort of CHT we really like would require more like 2.75 lbs/sec, or 165 lbs/sec. A smaller cooler like the popular 8432 would flow 46 lbs/min, so the ratio becomes roughly 3 to 1. Point is, even if a cooler duct conserved some energy, it acts on a small percentage of the total flow. It makes more sense to work on accelerating all the flows.
.
 

Attachments

  • exit.jpg
    exit.jpg
    8.5 KB · Views: 92
  • PB010001.JPG
    PB010001.JPG
    227.9 KB · Views: 108
Last edited:
Ten years ago when I was active in the Sport Aircraft Racing League I was always making changes to my airplane. One change that I made was to exit the air from the oil cooler through the side of cowling just aft of the cooler. I fabricated a rough aluminum duct that flowed to the side of the cowling where the air exited.
This exit size was covered by a “bluff body duct” that was bonded on the side of the cowling with the opening to the aft. There were many versions of my outside duct and eventually I found the right shape and oil temps came down. The duct is about three inches deep, and about 8 to 10” long. It is now permanently bonded to the cowling just forward of the firewall. It works for cooling and I think that it is also esthetically pleasing. Testing on different versions were done at altitude but in the end the shape was determined by cooling and how it looked!
The air is exiting straight aft and there may be some associated drag but race speeds confirmed that it was not limiting speed and the cooling was adequate.
A better internal duct would undoubtedly do a better job.
The old Red Green saying is “ its temporary, unless of course it works”!
 
Back
Top