The rise of FrankenEngine - An engine for the VP-21

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

Aviacs

Well-Known Member
Joined
Oct 21, 2019
Messages
311
That small displacement 5000 rpm screamer won't have sufficient cylinder head area for effective air cooling at high continuous HP. Heat exchanger, hoses , pumps, it all adds up. At these small power levels, the weight needed for liquid cooling ancillaries is proportionally greater than it is for higher HP installations
That could apply to many conversions of existing engines.
However, with a clean sheet, there is no imperative for that path.
Small cylinders with adequate aircooling built into the finning are certainly possible, look at all the actual MC aircooled "screamers"
Good initial design with adequate finning generally adds little extra weight. Most MC versions are not even (generall) cowled/ducted to optimize cooling efficiency.

Benchmark 1930's Pobjoy Niagaras:
I don't think a radial is going to be optimal for the OP's brief; but the 7 cylinder, 130lb, 88HP powerplant (98HP for the "High speed" 3,500 rpm version) is suggestive. 4 of those 24.7 cu in cylinders opposed on a boxer layout with"similar" gearbox should easily make 40 - 50 HP in a completely aircooled package under 100 lbs. with OHV architecture, at 26.5" width.

smt
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
7,143
Location
US
4 of those 24.7 cu in cylinders opposed on a boxer layout with"similar" gearbox should easily make 40 - 50 HP in a completely aircooled package under 100 lbs. with OHV architecture, at 26.5" width.
I agree that 1/2 hp per cu inch falls well into the normal range for air cooling, and for a direct drive prop. Present industrial engines and direct drive Lycomings, etc operate there. But if someone is considering an engine of considerably smaller displacement driven at higher RPMs (e.g your cited 5000 rpm) to get a higher specific continuous HP through a PSRU (e.g. a Rotax 912), then dedicated liquid cooling becomes necessary (Rotax 912 heads, etc).

Strictly/predominantly air cooled 4 stroke MC engines that produce phenomenal HP/cc can't do it on a continuous basis. The airplane duty cycle typically requires high continuous power levels.
 
Last edited:

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
5,248
Location
capital district NY
The big Pratt round motors do not have magnetos, they have a switch that you "mash" together with the starter switch and it creates a "shower of sparks" to aid starting. There is also another switch that you "tickle" and it squirts additional fuel into the intake system. Starting the Pratts is considered a black art by some but is relatively simple, turn engine with starter two revolutions of propeller then mash shower of sparks "tickling" the fuel is done by experience if the engines are hot and hot day they don't need anymore fuel usually and as the atmosphere and the engines get cooler they need more "tickle" till when it is really, really cold you just stand on the switch you can't make it to rich when it is that cold.

So you don't need magnetos. You can use a lighter system that is less prone to arcing out at altitude.
 

Aviacs

Well-Known Member
Joined
Oct 21, 2019
Messages
311
quick response:

(e.g your cited 5000 rpm)
My actual note was "4,000 to 5,000 range"
Plenty of aircooled MC's operate all day long at 4,000 to 4,500 rpm.
I hate to throw specific examples out there because you will fixate on the example, not the general implications.
But that was (relatively crude) push-rod cruising rpm when the national speed limit was 70.
There's been improvements since in engine technology.

Nonetheless, to pursue your notes re liquid cooling:
Though it won't help the OP cause for types of airplanes considered;
if i had another lifetime and a bit more money, a near fully-cowled metal airplane in which the skins themselves formed the radiator would be interesting.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
7,143
Location
US
Nonetheless, to pursue your notes re liquid cooling:
Though it won't help the OP cause for types of airplanes considered;
if i had another lifetime and a bit more money, a near fully-cowled metal airplane in which the skins themselves formed the radiator would be interesting.
Several teams had efforts along those lines in the Schneider Cup races. It didn't work very well for practical reasons (especially the weight and complexity of the coolant lines needed to achieve sufficient exchange area with the airflow). In practice, it is very hard to beat a well designed duct and heat exchanger. Lots of exchange area, relatively light, and very little drag if they are carefully designed.

My actual note was "4,000 to 5,000 range"
Plenty of aircooled MC's operate all day long at 4,000 to 4,500 rpm.
That doesn't tell us how much power they are making. They could do 5000 rpm standing still in the driveway.
 
Last edited:

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
5,248
Location
capital district NY
Ignition, I forget the details it was literally 50 years ago one of my high school shop teachers didn't like people leaning on his car with the nice custom paint job. He showed me how you cut a little hole in the floor and dropped a chain through to lay on the ground, a wire coming from a purpose installed coil was attached to the car body, if you leaned on the car you completed a circuit with a buzzer installed, you didn't lean on the car twice. You just can't find shop teachers like that anymore. But making a coil buzz can't be that hard I understood how to do it at the time.
 

PMD

Well-Known Member
Joined
Apr 11, 2015
Messages
720
Location
Martensville SK
interesting. Since it has 2 crank shafts that are then geared to final shaft, I assume it would not be too difficult to have different ratios for the final gearing.

high torque at low rpm, low power pulses, can gear to low propeller rpm
While the Neander (developed BEFORE involvement with Yanmar - but Yanmar dropped their own diesel outboard design and went with Neander) is extremely interesting, and two cycle diesels ARE the absolute best answer to powering boats and airplanes (what makes a good boat engine happens to make a good airplane engine) the things missing on the Shark are the simplicity of what CAN be done and what is there that adds a bunch of weight and costs a bunch of efficiency is the cylinder head. This is from a world of alternative designs that goes back over a century, but the ideal way to do it is with opposed pistons, opposed cylinders (OPOC) as did Hugo Junkers many moons ago. THAT allows for a gear train that can easily deliver whatever RPM drive you want/need and since it has no cylinder head, the parts count is considerably less, thus less weight and no thermal loss through cylinder head.

Just sayin'.
 

WonderousMountain

Well-Known Member
Joined
Apr 10, 2010
Messages
2,508
Location
Clatsop, Or
How about designing VW heads using electronically actuated valves thus eliminating the cam, pushrods, pushrod tubes and seals, rockers, cam chain/gears? The weight savings could be significant. Also allows variable valve timing and repositioning the intake and exhaust valve locations to suit the application.
You can't get them, but if you could it would be good to try out.
I wrote both companies asking about four to replace exhaust
side cams & still run intake. Nothing, but maybe they'll like you.
 

Topaz

Super Moderator
Staff member
Log Member
Joined
Jul 29, 2005
Messages
14,760
Location
Orange County, California
Shortest line from here to there, mate.

And yes, 40hp will easily fly a single-seater. I don't think a two-seater could be done on 40hp and still respect modern ideas about a safe climb rate "hot and high", but planes like the Aeronca K, C-3, and other put the lie to the idea that airplanes need a ton of power to just fly. More power is always desirable, but if a major constraint is cost, then it shows that there's room to play with.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
7,143
Location
US
How about designing VW heads using electronically actuated valves thus eliminating the cam, pushrods, pushrod tubes and seals, rockers, cam chain/gears? The weight savings could be significant. Also allows variable valve timing and repositioning the intake and exhaust valve locations to suit the application.
I think the idea may run afoul of one of the stipulations in the OP:
Simple. No dual OHC, VVT, variable compression, 12 valves per cylinder or antimatter containment fields.
Electronic-piezo-phlogostonic valves might simplify some (a lot) of mechanical things, but would sure introduce a new electronic nest-o-troubles.
 

Aviacs

Well-Known Member
Joined
Oct 21, 2019
Messages
311
That doesn't tell us how much power they are making. They could do 5000 rpm standing still in the driveway
I see what you are doing - you keep selectively editing my posts, clipping relevent information to change the focus and make your own point!

My actual quote was:
Plenty of aircooled MC's operate all day long at 4,000 to 4,500 rpm.
I hate to throw specific examples out there because you will fixate on the example, not the general implications.
But that was (relatively crude) push-rod cruising rpm when the national speed limit was 70.
There's been improvements since in engine technology.
You can probably interpolate that era to aprox 500cc to 750 twins weighing aprox 400-ish lbs. A smart guy like you can approximate the HP pushing an unfaired machine with a typical load at 75 mph (70 speed limit plus the more or less safe cheat) for endless hours.

Several teams had efforts along those lines in the Schneider Cup races. It didn't work very well for practical reasons (especially the weight and complexity of the coolant lines needed to achieve sufficient exchange area with the airflow). In practice, it is very hard to beat a well designed duct and heat exchanger. Lots of exchange area, relatively light, and very little drag if they are carefully designed.
By the time of the skin-cooling rudimentary experiments, the necessity was to eject the waste heat from 500+ HP engines.
40 HP in a similar planform/size airframe give more room for work.

It will "probably" add weight. (Maybe, maybe not necessarily)
Adding weight, there is room to investigate radiator placement inside wings, to minimize cooling drag losses to strongly benefit total aerodynamic drag reduction. E.g. the very successful Mosquito.

Lets go back to my simple proposition:
Fresh paper design allows cylinder head finning that will adequately reject/accommodate heat with minimal increase in wt, for the size engines under consideration by the OP. Cylinders of that concept were used in the 1930's for a successful family of engines that used a trouble free PSRU at rpms of up to 3,625 for climb & steady state aprox 3,500 @ .75HP/lb. I imagine that today the art could be extended to include higher rpm engines. Higher rpms (than 4,000) might or might not be necessary to attain the somewhat modest OP goal. Nonetheless, the advantage of a PSRU has throughout the history of aviation engines, permitted lighter & physically smaller engine systems at a given HP & displacement than the direct drive competitors. Many of them were extremely reliable.

smt
 

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
5,248
Location
capital district NY
While the Neander (developed BEFORE involvement with Yanmar - but Yanmar dropped their own diesel outboard design and went with Neander) is extremely interesting, and two cycle diesels ARE the absolute best answer to powering boats and airplanes (what makes a good boat engine happens to make a good airplane engine) the things missing on the Shark are the simplicity of what CAN be done and what is there that adds a bunch of weight and costs a bunch of efficiency is the cylinder head. This is from a world of alternative designs that goes back over a century, but the ideal way to do it is with opposed pistons, opposed cylinders (OPOC) as did Hugo Junkers many moons ago. THAT allows for a gear train that can easily deliver whatever RPM drive you want/need and since it has no cylinder head, the parts count is considerably less, thus less weight and no thermal loss through cylinder head.

Just sayin'.
If you are willing to absorb the weight in an airplane, Sure transports can absorb it and you get 6.7 lbs of fuel in a gallon container too but it would be a specialty aircraft if it is small with a diesel.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
7,143
Location
US
Lets go back to my simple proposition:
Fresh paper design allows cylinder head finning that will adequately reject/accommodate heat with minimal increase in wt, for the size engines under consideration by the OP.
Yes, we can add fin area and more effective baffling, etc. But there are diminishing returns. The issue is the finite heat flux ability of aluminum (or any solid). In our engine, heat is being added to the inner surface of the combustion chamber and piston. That heat needs to flow out to the fins where it can be exchanged. This flow occurs at a rate that depends on the cross sectional area of the aluminum and the temperature gradient. Extra fin area helps, but the physical limits of the cross sectional area of the aluminum eventually imposes a limit. When the temps inside reach about 500F, your head will live a very short life.
A thought experiment: We buy a roll of aluminum foil and unroll the whole thing. We set up a fan to blow air over this tremendous expanse of foil. If that foil was at 300F, we could easily shed thousands of watts/hr to the passing air and the foil would never get hotter. But what will happen if we put a little flame (500 watts?) on a corner of that foil? It will rapidly exceed the melting point of aluminum. Sure, the expanse of foil was capable of shedding thousands of watts of heat, but the aluminum in the immediate vicinity of the local heat application couldn't move it away as fast as it was being added, so it got hotter until it melted. We count on this same thing if we are sweating a connection on a long pipe: we add heat with the torch faster than the copper pipe can carry it away, and eventually it gets hot enough to melt solder. The long pipe easily has enough surface area to dissipate the heat from our torch to the surrounding air, but because the copper can't transport it away fast enough, our connection gets very hot. And copper moves heat much better than our aluminum head does.

Airplane engine designers are smart. If they could add more fins to their air cooled creations and thus reliably make .75hp/cu inch, they would do it. Gearing/PSRUs can let us turn a more efficient prop for a given HP output, but it does nothing to improve the thermal transfer issues that fundamentally constrain air cooled heads and cylinders. Where are all these geared air cooled airplane engines?
 
Last edited:

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
5,248
Location
capital district NY
I agree that 1/2 hp per cu inch falls well into the normal range for air cooling, and for a direct drive prop. Present industrial engines and direct drive Lycomings, etc operate there. But if someone is considering an engine of considerably smaller displacement driven at higher RPMs (e.g your cited 5000 rpm) to get a higher specific continuous HP through a PSRU (e.g. a Rotax 912), then dedicated liquid cooling becomes necessary (Rotax 912 heads, etc).

Strictly/predominantly air cooled 4 stroke MC engines that produce phenomenal HP/cc can't do it on a continuous basis. The airplane duty cycle typically requires high continuous power levels.
I was thinking more like 1/2 hp per lb than per ci. If you get 40hp from 80lbs are you happy? If so the ci doesn't matter.

I like the idea of liquid cooled better but the trade off for cowling and baffling may or may not be approximately equal all things considered like oil coolers for example.
 
Last edited:

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
7,143
Location
US
I was thinking more like 1/2 hp per lb than per ci. If you get 40hp from 80lbs are you happy? If so the ci doesn't matter.
The HP/displacement matters because, indirectly and imperfectly, it is a fair proxy for the amount of heat per square inch of combustion chamber surface area. Higher HP/cu inch correlates with higher heat intensity.
 
Last edited:

blane.c

Well-Known Member
HBA Supporter
Joined
Jun 27, 2015
Messages
5,248
Location
capital district NY
But flat head gives up efficiency for more ci at lower rpm because more torque with the ci, you trade the weight of the valve train for more surface area to get rid of heat and less hp (aka heat) per ci but more torque for a given rpm.
 
Top