What's new
Van's Air Force

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

Vans on alternative engines = ok

Lufthans

Well Known Member
Gents,

Just a little trivia - my recent communication with the mothership:

Our authorities allow us a fair bit of freedom in what we do with our aircraft, provided that they can find someone else to blame when things go wrong. (Sorry, I am being cynical here. They are in fact good people).

And so my planned “major modification” of an engine swap from Lycoming to Subaru on my RV-3B does require either a zillion pages of calculations and months of going back and forth with them, OR a “no objections” from the designer/ kit manufacturer. Easy choice which one to try first….

Having understood that our friends in Oregon are not the first ones to cheer up alternative engine projects, it was not without apprehension that I contacted them. And I must say, kudos to the mothership! After explaining my plans, and giving some resume of my previous Subaru builds, there was a very swift (and I am paraphrasing) “Yeah, sure. Looks good to us. Please understand that we cannot help you with parts, but by all means go ahead. Good luck with your build”.

That went smoothly. Tick off one hurdle. 999 more to go…. :)

Now if only I could figure out how to make room for a big enough radiator in a belly pod to cool 280 hp without having to pull the radiator up into the fuselage itself like the P51 did…. :D
 
......Now if only I could figure out how to make room for a big enough radiator in a belly pod to cool 280 hp without having to pull the radiator up into the fuselage itself like the P51 did…. :D

So...why not a P-51 scoop? Might have to get creative on dealing with the elevator pushrod, but otherwise looks doable without serious structural issues. Have a look at the P-51D+RV-3 overlay below. Could actually look pretty cool!

Capture.JPG
 
280 hp in an RV-3 should give you a fair ROC...:cool:

You'd need about 225 in2 of face area and 500-550 in3 of volume so pretty big scoop if external to the belly skin like my setup is.

Cutting into the belly skin to recess the rad introduces more structural questions and solutions. Leaving the skin intact should result in less trouble with the authorities- hopefully.
 
Last edited:
I keep waiting for someone to try the 300 HP Yamaha (that some of the STOL Drag guys are using) in an RV-3. Would make an interesting hot rod with all sorts of limitations on speed - but it would climb like the devil!
 
I keep waiting for someone to try the 300 HP Yamaha (that some of the STOL Drag guys are using) in an RV-3. Would make an interesting hot rod with all sorts of limitations on speed - but it would climb like the devil!

Teal Jenkins, owner of Skytrax and manuf. of the gearbox for the Yamaha that the STOL guys use, has one installed on the RV9 as a test bed.

https://www.youtube.com/watch?v=99zvYZzk16k&t=552s
 
Last edited:
Here are a couple photos for ideas. You might have room under the Subie for a decent duct if you can afford the weight FWF.
 

Attachments

  • geoff.jpg
    geoff.jpg
    95.4 KB · Views: 206
  • rv8scoop.jpg
    rv8scoop.jpg
    336.5 KB · Views: 202
280 hp in an RV-3 should give you a fair ROC...:cool:

You'd need about 225 in2 of face area and 500-550 in3 of volume so pretty big scoop if external to the belly skin like my setup is.

Cutting into the belly skin to recess the rad introduces more structural questions and solutions. Leaving the skin intact should result in less trouble with the authorities- hopefully.

These are EXACTLY my thoughts. I've had a good look at your (excellent) YouTube videos on this topic, and compared the numbers with my own observations (which matched).

And indeed struggling to find enough depth underneath the belly to make it work, but not willing to cut into the main structure.

Made plenty of P51 sketches though :)

Would look something like this: (And this obviously is not my aircraft, but gave me a nice side view)

RV3-51.jpg
 
Last edited:
I keep waiting for someone to try the 300 HP Yamaha (that some of the STOL Drag guys are using) in an RV-3. Would make an interesting hot rod with all sorts of limitations on speed - but it would climb like the devil!

My plan is to do this with a Subaru EJ-25 STi engine. Keep the Vne as it is and enjoy a 5000+ fpm climb rate.

95% of my flying is in short flights, divided between close formation flying and (mostly) quick dashes to play with cumulus clouds. My Jodel does 2500 fpm and I love it for that. If I could double the climb rate, I'm sure that would be fun ;)

And yes, there is no such thing as wide open throttle in level flight then. No sweat, I can live with that.
 
There was a clean RV-8 in Oz with an STi using wing rads. I'll see if I can dig up photos.

I've got a bunch of photos on this one, thanks. I've got wing rads on my Jodel. That wing is a lot better suited to this idea though - 6 ft chord, 9 inch depth, box spar only, so no rear spar. Fabric covered and no flaps. All of this made it easy to make BIG ducts inside the wing, rather than make some underslung arrangement.

I think on the RV, I prefer the P51 scoop. Forego on the oil/water heat exchanger (which would require even more from the water cooling setup) but do a regular air-cooled oil cooler. Ditto intercooler. And if all else fails, I could always add a cooler underneath the wing...
 
With the RV-3, you won't be running high power for very long in the climb before you're into forbidden airspace there if I recall so you might not need to size for extended climbs at 85 knots. Should be able to climb at Vne at at least 1000 fpm.
 
Last edited:
With the RV-3, you won't be running high power for very long in the climb before you're into forbidden airspace there if I recall so you might not need to size for extended climbs at 85 knots. Should be able to climb at Vne at at least 1000 fpm.

Good point. I particularly like the fact that I can throw anything at my Jodel and temps don't move a single degree. But indeed, how much full power climbing can one do with class A starting at 5500 feet?

As for that Vne climb, I reckon I can do better than 1000 fpm ;):

The -3 needs about 160 hp for Vne, which means that I’d have 120 hp left for climb.

Switching to proper units, that’s nearly 90 kW

Suppose 80% prop efficiency, this means just over 70 kW for true climb performance.

m * g * dh/s = 70 * 10^3

So dh/s = 70,000 / (500 * 10) = 14 m/s, or 2750 fpm.

At Vne….

Yeah, this should be fun.... :cool:
 
Looks good. I might stagger the coolant inlet and outlet to get more even flow across the tubes.

Be sure not to wet the tanks or top/bottom supports with airflow, just more drag.

Dimensions are about exactly what I'd choose here for your mission and climate.
 
Looks good. I might stagger the coolant inlet and outlet to get more even flow across the tubes.

Be sure not to wet the tanks or top/bottom supports with airflow, just more drag.

Dimensions are about exactly what I'd choose here for your mission and climate.

Thanks, and good point about the stagger!
 
A few shots from mine which you've probably seen.
 

Attachments

  • 8400.jpg
    8400.jpg
    69.5 KB · Views: 76
  • 8935.jpg
    8935.jpg
    64.4 KB · Views: 72
  • 8404.jpg
    8404.jpg
    65.8 KB · Views: 67
Burying a reasonable fraction (half) of the radiators inside the existing mold lines really helps with trying to get minimum cooling system drag (or even a little bit of thrust). Not that it can't be done fully external -- Ross's is a good example, but it is much easier to get good performance if it is partially buried.

Yes, it would be a bit complicated to dig into the fuselage behind the pilot - mostly figuring out how to re-route the elevator control system. But it could be done.

Much easier to go into the wing behind the spar. The diffuser wall would sweep up inside with adequate clearance on the aileron pushrod. The radiators would be fairly wide, but not very deep. You can probably get almost 5" of it up inside the wing, and then have the enclosure with inlet and nozzle extend about 5" below the wing OML.

Do keep them as far inboard as possible - it is not great to hang a lot of weight on the aft portion of a wing for flutter considerations, but as long as it is well inboard (in the rib bay just ouboard of the wing-walk ribs) the wing is plenty stiff. Also, remember that the diffuser wall helps complete the torsion box in the area where the wing skin has been cut out, so be sure that it is attached to wing structure along all edges. If the radiator does not span across the full rib bay, then make additional ribs to form the side walls of the passage and tie the diffuser wall to those ribs.

Putting the radiators in the wings has the added benefit of keeping the heat out of the cockpit. With P-51 style radiator installation, you end up having a lot of heat come through the diffuser wall, radiator upper tank, and nozzle wall, into the cockpit, requiring a lot of insulation. Ask anyone who flies a Thunder Mustang how much heat comes into the cockpit.
 
Last edited:
Heat in the cockpit would indicate a sealing issue with the duct.

The Thunder Mustang duct wasn't laid out well internally, though it looks beautiful on the outside. Notice the lack of attention to wetting just the HX here with that flat sheet metal in the breeze. No noticeable heat comes through the floor in my RV with the belly skin forming the top of the duct aft of the HX.
 

Attachments

  • thunder.jpg
    thunder.jpg
    208.2 KB · Views: 78
Last edited:
Here's some really old pics of me building the radiator pods underneath/ inside my Jodel wing. (Wing is upside-down on the photos).

They work well. Bit hesitant to cut into the wing skin of the RV though. On my Jodel, it was just fabric. Plus with the RV3 looking like a mini mustang as it is, might as well go all the way then ;)

When doing wing radiators, they should be underneath the wing walks in order to keep them within the propwash, or there will be hardly any ground cooling. With them being here, I can idle all day long in all weather conditions without the temps ever getting 1 degree above what the thermostat wants them to be.

mga-radiatorpod2.jpg


mga-radiatorpod4.jpg
 
Burying a reasonable fraction (half) of the radiators inside the existing mold lines really helps with trying to get minimum cooling system drag (or even a little bit of thrust). Not that it can't be done fully external -- Ross's is a good example, but it is much easier to get good performance if it is partially buried.

Yes, it would be a bit complicated to dig into the fuselage behind the pilot - mostly figuring out how to re-route the elevator control system. But it could be done.

Having to modify the control system would be one thing, but I'd be more worried about cutting into that semi- monocoque fuselage structure. I don't believe that those skins are just there to keep the air out, but rather serve a structural purpose as well. Or am I seeing this all wrong?
 
Heat in the cockpit would indicate a sealing issue with the duct.

The Thunder Mustang duct wasn't laid out well internally, though it looks beautiful on the outside. Notice the lack of attention to wetting just the HX here with that flat sheet metal in the breeze. No noticeable heat comes through the floor in my RV with the belly skin forming the top of the duct aft of the HX.

Ross, the TM I flew had a well-sealed duct, and as I recall looking inside, the internal duct did only expose the radiator core. The bar that is visible in your picture is the edge of another heat exchanger in front of the coolant radiator, right? (oil cooler?). Not much you could do about that except put a leading edge lip on it. The diffuser duct curves up well up out of view in your picture and the upper edge of the radiator core is out of sight in your picture.

The cockpit heat was all from conduction through the floor. It is worse at 300 MPH of course because there is significant temperature rise in the diffuser even before you get to the radiator. But the radiator is also really hot of course and conducts heat up through the mounting system.
 
Last edited:
Can the longerons and firewall brackets handle the extra pull force from the massive increase in power? As much fun as it sounds, I'd hate to see broken airframe parts.

I don't know the answer, but it's something I'd look into.
 
Having to modify the control system would be one thing, but I'd be more worried about cutting into that semi- monocoque fuselage structure. I don't believe that those skins are just there to keep the air out, but rather serve a structural purpose as well. Or am I seeing this all wrong?

You would not cut the longerons, of course. The whole cooler duct system would be enough narrower than the fuselage to fit in between the longerons.

The principle role of the belly skin is torsional stiffness. You would make fuselage 'ribs' to form the side walls of the duct, and a new "floor" sheet to form the upper wall of the diffuser and the nozzle. Also modified bulkheads to support the duct upper wall, ahead of, at, and behind the radiator, and connect all that to the fuselage sides and longerons.

If done right, this would pretty much recover all the torsional stiffness of the belly skin. This is all fairly straightforward.

Then you have to devise a couple of bell cranks to get the elevator push rod up above the cooler duct work.
 
With enough power to exceed Vne in level flight easily, it would not seem to be worth the trouble to recess the HX into the fuselage given the number of mods required, especially to the control system which would all have to be analyzed.
Not trivial for a non-engineer.

One mod often creates multiple issues down the line.

Ram rise at 182 knots is inconsequential and even lower at the normal flights speeds of an RV-3.
 
The bar that is visible in your picture is the edge of another heat exchanger in front of the coolant radiator, right? (oil cooler?). Not much you could do about that except put a leading edge lip on it. The diffuser duct curves up well up out of view in your picture and the upper edge of the radiator core is out of sight in your picture.

The cockpit heat was all from conduction through the floor. It is worse at 300 MPH of course because there is significant temperature rise in the diffuser even before you get to the radiator. But the radiator is also really hot of course and conducts heat up through the mounting system.

Yes, the bar and plate HX you see there is the oil cooler which is a bad choice again if you want to reduce drag. My flow bench tests show about 35% more pressure drop (drag) for this type of HX over a typical tube and fin HX (round nose and tail).

You see in the photo below that the TM rad is much taller by scale than a stock P51 which would mean the diffuser angle is much steeper, probably not good given what I saw with my internal tuft testing. You see also it uses an angled linear actuator which isn't well sealed to the duct. With the exit door closed, lots of delta P between there and the inside of the cockpit. The Bowden cable Hans plans to use in his drawing is what I used. Simple, light, failsafe and easy to seal to the duct wall.

My RV-6 setup has much more in common with the RV-3 (aluminum vs. composite, similar speed and rad layout) compared to a Thunder Mustang.
 

Attachments

  • reno050092.JPG
    reno050092.JPG
    586.2 KB · Views: 44
Last edited:
Steve, with your remarks on heat you bring up an important point that I had not given much thought yet. My Jodel is made of wood, which doesn't have this issue, but on the aluminum RV this will certainly be different. So I'll have to take consideration to heat transfer when mounting the radiator or tubes. Thank you!

Ross has made a good point in remarking that in my particular case (moderate climate, Vne making it impossible to use more than 60% power in level flight, class A airspace starting at 5500 ft overhead (and not higher than 9500 ft when wandering farther away from home), there is only so much sustained climbing that I can do. In other words - this will be a 160 hp aircraft with short bursts of 280 hp for fun, rather than a 280 hp continuous one. For this, I can get away with cooling that is dimensioned more towards 160 hp than to 280...

This can be done with a duct that sits completely outside of the fuselage, saving me a ton of modifications and potential screw-ups. So underslung it will be. Better is the biggest enemy of good enough....
 
Can the longerons and firewall brackets handle the extra pull force from the massive increase in power? As much fun as it sounds, I'd hate to see broken airframe parts.

I don't know the answer, but it's something I'd look into.

I've done this study when putting 170 hp on my 85 hp Jodel. When calculating the stresses on the structure from the engine hanging off its mount at 6G and comparing it to the extra pull from the increase in horsepower, the latter all but disappears from the equations. Discussed this with the (then 80+ year old) designer of the Jodel and he fully concurred.

The structure and dimensions of the Jodel are VERY similar to that of the RV. It's wood rather than aluminum, but that's about the only difference. And here too the extra pull is of no consequence.... Thanks for raising the point though!
 
Steve, you mentioned that you've flown the Thunder Mustang.

How is that plane? How does its handling compare to that of our beloved RVs?

Most exotic aircraft on my resume is the T6, which flies nothing like the RV. It's a bit like an old farm dog that has one or two wolves among its ancestors. On the ground it is mostly big and cumbersome. I've flown the 'A model, so with the least refinements. Need to manually power the hydraulic system prior to being able to set the flaps lower the gear (and not at the same time). Need to arrest the flaps, as they will go to full down if you don't. Can't shove the throttle in, as it does not have a waste gate and you WILL blow up the engine. Don't get behind the power curve as it is one big, bulky beast. And the first time I did an accelerated stall in one, it snapped into a spin with such ferocity that my knees literally swiped my headset off my ears.

So yeah, good fun ;)

Really wondering how you'd rate the Thunder Mustang?
 
On the RV, you have the floor ribs separating the belly skin from the seat pan skins with a couple inches of air space so no way much heat is reaching the interior.
 
On the RV, you have the floor ribs separating the belly skin from the seat pan skins with a couple inches of air space so no way much heat is reaching the interior.

Unless I'd bolt the hot radiator flat against the fuselage skin (as I had planned), allowing for a lot of heat transfer into the aluminum. Or at least, it might.

How did you mount yours Ross?
 
On the RV, you have the floor ribs separating the belly skin from the seat pan skins with a couple inches of air space so no way much heat is reaching the interior.

Yes, that is certainly true for a fully underslung system. I was more referring to problems with the partially internal system. And you are right, on a typical TM, that actuator does provide a big hot-air leak. It needs to have a boot on it.
 
A bit of thread drift, but yes, the Thunder Mustang is very delightful to fly. Nothing super-light and quick like an RV, but medium-light, well-harmonized and responsive. (nothing at all like the airplane it was inspired by). I didn't really push it, but I don't think it has any violent tendencies like the T-6 is famous for. It rolls very nicely.

It's too bad they seem to have such bad luck with mishaps initiated by firewall-forward issues. Each incident has been different, no particular problem area that can be easily be mitigated - although a general concern is that all the vital accessories (water pump, alternator) are driven off of a V-belt rather than a gear case, but is not unusual for auto-conversions either.

Steve, you mentioned that you've flown the Thunder Mustang.

How is that plane? How does its handling compare to that of our beloved RVs?

Most exotic aircraft on my resume is the T6, which flies nothing like the RV. It's a bit like an old farm dog that has one or two wolves among its ancestors. On the ground it is mostly big and cumbersome. I've flown the 'A model, so with the least refinements. Need to manually power the hydraulic system prior to being able to set the flaps lower the gear (and not at the same time). Need to arrest the flaps, as they will go to full down if you don't. Can't shove the throttle in, as it does not have a waste gate and you WILL blow up the engine. Don't get behind the power curve as it is one big, bulky beast. And the first time I did an accelerated stall in one, it snapped into a spin with such ferocity that my knees literally swiped my headset off my ears.

So yeah, good fun ;)

Really wondering how you'd rate the Thunder Mustang?
 
Unless I'd bolt the hot radiator flat against the fuselage skin (as I had planned), allowing for a lot of heat transfer into the aluminum. Or at least, it might.

How did you mount yours Ross?

The belly of the -6 isn't flat, slightly V shaped (guess Van's wasn't thinking about someone mounting something there...) so I had to seal the gap on both sides with fairly high density open cell foam. The rad is hung by small brackets welded to the top edge of the tanks. The duct holds it all tightly in place and the inside of the duct lines up tightly with the HX face so no tank or support edges are exposed to the air stream.

Lots of hours put into this project from flow benching multiple HXs for pressure drop and temperature drop, building the test rig, finding the right HX, making the scoop/duct, solving many issues along the way and validating performance in flight. Was happy with the results in the end though.
 
One thing to remember when using an alternative engine is that resale value will be lower. Sometimes a lot lower. I have my RV10 based at HND and there is a RV7A there that has been for sale for 10 years. No one will touch it because it has a Subaru engine in it.
 
One thing to remember when using an alternative engine is that resale value will be lower. Sometimes a lot lower. I have my RV10 based at HND and there is a RV7A there that has been for sale for 10 years. No one will touch it because it has a Subaru engine in it.

Yeah, I know. And I don't care, really.

I might at one point sell my Fly Baby, but for the kind of flying that I do, the 3-seater Jodel and the single seater RV3 make perfect sense. So they don't go, ever.

Might as well make them the way I really want them to be then :)
 
Yeah, I know. And I don't care, really.

I might at one point sell my Fly Baby, but for the kind of flying that I do, the 3-seater Jodel and the single seater RV3 make perfect sense. So they don't go, ever.

Might as well make them the way I really want them to be then :)

Agree. I built my airplane for me, not the next guy.
 
Can’t empathize how important Ross at SDS (RV6ejguy) is in knowledge of cooling and ductwork! ALSO THE SDS SYSTEM IS THE GREATEST! Do not try and
mickey mouse a car computer as it is a DANGEROUS waste of time.
 
I know, only a rough sketch of what it "could" look like, but it made me wonder... With the Radiator that far back, will the weight of it and the coolant inside be something the W&B can tolerate? I don't know what the typical baggage capacity of a -3 is (I assume it's behind the seat though?).
 
I know, only a rough sketch of what it "could" look like, but it made me wonder... With the Radiator that far back, will the weight of it and the coolant inside be something the W&B can tolerate? I don't know what the typical baggage capacity of a -3 is (I assume it's behind the seat though?).

The cooling will obviously be part of a complete redesign of the propulsion system, including a very different engine, a gearbox, a CS prop, new engine mount, new cowling, etc. And with that, the entire CG calculation is out of the window and requires me to start that from scratch.

The engine will need to be moved forward, no doubt. But that will all result from a huge excel spreadsheet where I can virtually move things around.

So yes, it will work. But requires a bit of engineering. Which is all part of the fun of a conversion like this one…
 
Back
Top