Rotary Engine (Mazda/Wankel) conversions Hummmmmm!

[rv6ejguy]

Divergent ducting with a smooth transition not exceeding 6 degrees has shown to be the best. Converging the flow after the rad may help to recover the velocity before dumping it back out into the freestream. The front mounted rads used by Tracy Crook and Eggenfellner are there for packaging and ease of installation, not efficiency. The short duct length MUST cause poor pressure recovery and dumping the exit flow into the turbulent cowling flow causes higher drag. These installations cool because of high rad surface area and volume, not because they are great designs. The tight fin and tube spacing of GM evaporator cores has shown to offer high drag per unit volume and poor heat rejection at typical RV climb speeds. They work ok at higher speeds as far as heat rejection goes.

They do work and the front mounted rads keep plumbing short and solve many packaging problems on the RVs with deep spars. The cowlings can be kept near stock also. These are real considerations as well for any installation.

For the lowest drag, a proper divergent/ convergent duct would be used with an inlet area 25% of the rad face area, as the flow diverges, it slows down and gains pressure. Pressure differential is one of the most important factors in heat exchanger performance. The rad core needs to be relatively open with widely spaced oval (not flat) tubes for lowest core drag. Flow would then converge through a sealed duct, exiting through a controllable exit door with areas adjustable between 50-150% of the inlet area. This allows matching of flow velocities with the freestream at different airspeeds, increases airflow through the core during ground and climb conditions where mass flow is low and a large reduction in drag in cruise flight where mass flow is high.

On an RV, extensive cowling mods will be required to fit something like what is described above, ahead of the firewall. Other installations have fitted massive rads a few inches from the firewall and relied again on sheer size to get proper cooling. Very few recent installations have employed the lessons learned in WWII. I can think of Reg Clarke's Sube Dragonfly which eventually ended up with a really small belly scoop and excellent cooling and one Ford V6 powered Velocity with a P51 style belly scoop with an inlet area of only 22 square inches which also cooled very well.

A belly scoop is possible on the RV10 due to the central tunnel for internal routing of coolant lines. This allows a proper duct length and exit door configuration. It also allows cowling space for intercoolers. A word of note on intercoolers: these require 50-75% of the inlet area of the main cooling system to be 65-85% effective at high altitudes and are especially challenging in climb at high density altitudes.

[copterdoc]

Count me in on a rotary too! I'm planning a 13B or Renesis, fixed pitch prop, using Tracy Crooks PSRU, EC2, and EM2 at this time. I'm really hoping to see a continuous discussion here. I agree with cjenson that things will only get better within the next few years as more information and products are made available. I'm just building the wings myself so I'm a couple of years away from needing an engine for my RV-8. Which brings me to my favorite dilemma. You try to plan ahead but you know new products, ideas, or techniques will surface and some or all of your plans fly out the hangar door! Obviously the intake, cooling, and exhaust are the primary problems to over come. Those of you allready discussing the cooling and ductwork have an avid fan here! I'll be watching very closely to see how others set up their aircraft. Personally I like the KISS approach as much as possible. Massive redesigns with coolant hoses running long distances through the fuselage to a belly mounted radiator creates failure points and maint. issues that I'm not ready to accept at this time. But I'm keeping an open mind and the mental pic. of an RV-8 with a P-51 belly scoop incites the RV grin! Who is working on the aluminum end housings? This is the first mention I've heard of them but a 27 lbs savings would be substantial!

[rv6ejguy]

Hey Bill, what stage are you at on your RV10? I've got my tail surfaces done and am working on the rear fuselage now. I've got a QB so it's coming together quickly.

[Rotary10-RV]

Originally Posted by rv6ejguy

Hey Bill, what stage are you at on your RV10? I've got my tail surfaces done and am working on the rear fuselage now. I've got a QB so it's coming together quickly.

I'm just finishing up the tail feathers. I have a slow build kit. I have the wings (still in the box) No fuse parts yet. I'm a "low buck operation".
Bill

[Rotary10-RV]

[quote=rv6ejguy]Divergent ducting with a smooth transition not exceeding 6 degrees has shown to be the best. Converging the flow after the rad may help to recover the velocity before dumping it back out into the freestream. The front mounted rads used by Tracy Crook and Eggenfellner are there for packaging and ease of installation, not efficiency. The short duct length MUST cause poor pressure recovery and dumping the exit flow into the turbulent cowling flow causes higher drag. These installations cool because of high rad surface area and volume, not because they are great designs. The tight fin and tube spacing of GM evaporator cores has shown to offer high drag per unit volume and poor heat rejection at typical RV climb speeds. They work ok at higher speeds as far as heat rejection goes.

They do work and the front mounted rads keep plumbing short and solve many packaging problems on the RVs with deep spars. The cowlings can be kept near stock also. These are real considerations as well for any installation.

I'll add additional comments on Tracy and Eggenfellner set-ups. Tracy gets just adaquate cooling with his evap cores. He has the open area to allow the exit area needed.(at least better than the Subaru) I have seen the H6 Eggenfellner conversion and the right side radiator is almost touching the front of the right bank of cylinders. Not good. The exit area can be as important as inlet on a good cooling system. Tracy is limited by the small in-cowl space allowed by the RV-4. Plan for plenty of exit space for the post-radiator airstream!

Bill Jepson
Rotary10-RV

[cobra]

The cooling issue Im curious about concerns the density of the radiator fins used. TCrook uses a high fin count core that blocks airflow- suggests a pretty high aerodynamic drag penalty. The use of regualar aluminum auto radiator design, with a less dense core, combined with higher speed airflow, makes more sense to me, particularly if exhaust air ducting allows low-restriction flow to the exit port(s). I have no idea if/ how much drag is reduced or how much the heat rejection rate differs.

Tracy reports problems with taxi-speed overheating, less so with high speed operation. Im a little surprised at that comment given the size/effectiveness of the "fan" up front. Might the high-density fin count in the A/C cores block airflow at both speeds to some extent, overcome by the ram-air effect of 100mph speeds?

I really like the idea he tried concerning routing a water cooling line thru the oil pan to assist the oil cooler, but I wonder if an outside-the-oilpan cooling line might be significantly safer (no chance for leaking water contaminating the lubricating oil) and almost as effective.

The other issue that really concerns me is the apex seal oilers that supply engine oil (the additives burn dirty, reduce engine life). Tracy suggests mixing 2-cycle oil with the fuel. I dont like that suggestion because a little oil in the fuel greatly increases the detonation, possibly a problem with high compression or turbocharged motors. Has anyone figured out a way to use the stock oil system for lubrication the apex seals, but supply them with clean-burning 2-cycle oil, not engine oil (Id prefer to use a good synthetic, like Mobil One in the engine/redrive).

[rv6ejguy]

Few people have addressed the liquid cooling drag issues. They simply put something on to keep the temps in check and fly it. I have flow benched various rad cores to see what matrix drag is like. The GM evaporator core has high drag but is quite efficient at heat rejection. The oval shaped tubes on some aircraft oil coolers compared to the concave GM tubes, , combined with less dense tube spacing and more open fin spacing offers far less drag. This is the route that Eggenfellner has gone with his latest cores as suggested by Bill Kay.

Certainly on the ground, the GM evap cores are essentially useless with no airflow penetrating the core from the propwash. We typically see less than .25 inches of water pressure at the face at idle. These cores were designed for centrifugal type fans to force air through them with 2-4 inches of water.

In cruise, we see 4-8 inches depending on speed and altitude and the cores function fine here. Drag is likely considerable. Our experiments have indicated that an open, relatively deep core is best suited for aircraft. An efficient duct located in the outer half of the prop arc may be able to reduce core area and volume by 50% over cheek mounted rads. Ground cooling is likely to be far superior as well. The air near the prop root is pretty dead and turbulent.

Running water through the oil works but a just adds more heat to be rejected out the rads. A proper air to oil cooler would do the same job. Typically Wankel and turbo engines need a substantial oil cooler. Again, it's hard to beat the true aircraft ones here.

Few people seem impressed with Mazdas original oil metering design and most racers remove it and simply add a very small amount of synthetic 2 stroke oil to the fuel. There does not appear to be any serious effects on octane rating. John Slade has run up to 46 inches I think on 87 octane with high compression rotors!

[thallock]

Cobra,

Regarding the stock oil injection system. Tracy sells a pump from PCV Technologies that replaces the stock oil injectors. It draws oil from a separate oil supply. There are two versions, a mechanical pump that is driven by the stock oil pump interface, and an electrical version. This solves the problem of using engine oil for apex seal lubrication, and the alleged "deposits" problem.

Some guys claim to be running the stock lubrication system without any problems, but most guys simply remove the whole oil injection system for simplicity and weight (it's not that heavy).

Tracy.

[Rotary10-RV]

Originally Posted by thallock

Cobra,

Regarding the stock oil injection system. Tracy sells a pump from PCV Technologies that replaces the stock oil injectors. It draws oil from a separate oil supply. There are two versions, a mechanical pump that is driven by the stock oil pump interface, and an electrical version. This solves the problem of using engine oil for apex seal lubrication, and the alleged "deposits" problem.

Some guys claim to be running the stock lubrication system without any problems, but most guys simply remove the whole oil injection system for simplicity and weight (it's not that heavy).

Tracy.

Cobra, Tracy's info is correct, additionally 2 stroke oil has a minimal effect on octane AT NORMAL PROPER MIX LEVELS. The rotary needs even less oil for the apex seals than a 2 cycle. Guys that normally mix 2 cycle in the fuel have forgotten and flown 100's of miles with nil ill effect. Mistral who are certifing a rotary are using the stock system burning sump oil. The RX-8 passes all modern smog laws while burning sump oil.(They use a electrically controlled pump with 2 inlets per rotor.) This is really not a worry (or a big worry anyway) that should cause anyone trouble. If you want the least concern or hassle run the stock system. You can get MINIMAL improvements in performance using a very lean mixture of 2 stroke oil in the fuel. (> 100:1) All the racers use pre mix because fewer parts can break. Not a weight issue the pump couldn't weigh more than a few ounces total!
Bill Jepson
Rotary10-RV

[copterdoc]

I just finished reading (for the first of what may be MANY times) Paul Lamars' article on engine cooling with respect to the needs of the rotary engine. If you haven't read it I've attached the link. Interesting information on duct shapes and heat exchangers. Lots of info on the site. http://www.rotaryeng.net/cooling.html