Flat 4 or 6 cyl Aircraft Engine

[TSwezey]

Originally Posted by Garage Guy

Well, but... horsepower is just torque times RPM (times a constant). If you want to deliver 200 hp to a propeller spinning at 2600 RPM, no way to do it without applying 404 ft-lb of torque to the propeller shaft.

So if a diesel could get you 400 ft-lb of torque at 2626 RPM crank speed, you would have a 200hp engine that wouldn't need a prop speed reduction unit, and that would be nice. But making it light and reliable seems to be easier said than done.

Ain't that the truth!

[rv6ejguy]

Diesels usually have heavier flywheels and way more crank and rod inertia than gasoline engines so letting the clutch out at idle is not a valid method to compare extreme low rpm torque. If you look at the torque peak rpm of most auto turbo engines these days (diesel and Otto) they are often below 2000 rpm. Breathing is good at this low rpm and the turbo is already shoving lots of air in thanks to variable geometry turbine devices used today.

Turbos are where it's at for torque and hp with small displacement engines- which is why most things I own are turbocharged. Just wouldn't be fun otherwise.

I don't think anyone would even think of an atmo diesel these days because it would be a total dog. I was looking at a small, mobile diesel cement mixer a few months ago and even that had a turbo on it!

Diesels and turbos go together like bread and jam- ya gotta have 'em, especially on an aviation application where power to weight ratio and altitude performance is important.

I say Rodney, I'll have another helping of boost please.

The torsionals from an aero diesel might be hard to tame as far as the prop went without a big weight penalty. A good damper/ absorber would likely be required or a big ol' flywheel. Diesels are pretty nasty in this regard.

The common rail injection systems used today are extremely complicated from an electronics/ programming standpoint. I'm not sure these would be a good thing to have if costs were to be kept down. This is the stumbling block for the Subaru diesel in aviation use as they are tied into the chassis electronics these days to a huge degree.

[Garage Guy]

Originally Posted by rv6ejguy

The common rail injection systems used today are extremely complicated from an electronics/ programming standpoint. I'm not sure these would be a good thing to have if costs were to be kept down. This is the stumbling block for the Subaru diesel in aviation use as they are tied into the chassis electronics these days to a huge degree.

The ECU algorithms are complicated, but the common rail hardware (pumps and piezo valves) is pretty robust. And you wouldn't have to duplicate all the control and sensor complexity of the auto application to get a pretty nice FADEC system for a light plane. Even a simple software controller could give better performance than old-style mechanical injection.

[Andy_RR]

Electronic engine control (aka FADEC) is a pretty simple business at the basic level. Unfortunately, in electronics nothing ever half-fails, so most of the smarts in the controller revolve around what to do when you begin to lose sensors, or identifying duff signals and ignoring them or at least choosing to do something different.

Also, once you have the capability that a microcontroller gives you, you don't want to just half use it. Most people would be surprised at the complexity in modern engine controllers, but most of it's driven by things other than engine control itself.

I've recently finished working on a clean-sheet engine controller for an Indian customer. We had 128kB storage to play with, of which more than 30kB was taken up by the communications library alone. The (fairly sophisticated) engine control stuff was relegated to about 25kB of algorithm and 15kB of calibration data. A fair chunk of what was left was consumed by the operating system, OBD functions and other non-control-related code, all of which was necessary when you have to support tens of thousands out in the field.

On other projects I've worked on, it was commonplace to download 1MB of code and data!

I'm with Ross on this one. Avoid electronics running your engines if you can at all do so!

[leeschaumberg]

In addition to my specs the engine will be a 2 stroke diesel. Because 2 stokes fire every 360 deg of rotation the loads would be half of a 4 stroke. This requires compressed starting air. Some people are using electric pumps but a supercharger instead of a turbocharger would work also. To be like WW2s best I'd have to make it a two speed supercharger. EH

[rv6ejguy]

Originally Posted by Garage Guy

The ECU algorithms are complicated, but the common rail hardware (pumps and piezo valves) is pretty robust. And you wouldn't have to duplicate all the control and sensor complexity of the auto application to get a pretty nice FADEC system for a light plane. Even a simple software controller could give better performance than old-style mechanical injection.

The multi stage injection strategies used on most modern diesels are far more complicated than on port injected gasoline ECUs. It would take much testing and experimentation to find out what works best and implement it in software IMO. Thielert had multiple problems with software, ECU hardware, wiring and even injectors. Not saying it couldn't work and be better but it will be much more expensive than mechanical injection. I also think that pilots and A&Ps will more readily embrace a mechanical system- ask me how I know this. I think gains on an aviation diesel operating at primarily constant high load would be minimal using a FADEC over mechanical injection.

I am fan of electronic engine controls on gasoline, port injected engines because it is relatively simple and reliable and proven, low cost hardware already exists plus, that is my business so I'm biased perhaps.

Perhaps the engine could be initially fitted with mechanical injection and after some money rolls in from sales, development could be started on a FADEC system to be retrofitted later if desired. Many engine projects are doomed when introducing too many new things at once as the drawn out development burns up all the time and funding long before completion. I'd prefer to see the basic design up and running on the dyno for many hours ASAP rather than having all the bells and whistles delaying or shortening the testing and validation period- then trying to deliver a mediocre, complicated and expensive engine. Test, test, test before production is key.

[SvingenB]

Nothing is impossible. A common rail controller for an auto conversion already exist http://www.eco-motors.com/Products.htm.

[breister]

Originally Posted by Mike S

Pretty sure that is German.

Hehehe! Zoches' website. One year until certification - for the past 25 years!

[rv6ejguy]

Originally Posted by SvingenB

Nothing is impossible. A common rail controller for an auto conversion already exist http://www.eco-motors.com/Products.htm.

Very cool. Thanks for the link.

Wonder what pricing and total flight time and high time engine time is like?

Now, we just need a 175hp version for RVs.

[gtmule]

As far as doing an autoconversion, I bet that the BMW straight 6 disels would be a better fit than the subie diesel which will come with its fair share of vibration problems. The BMW should have More power (or headspace), and better torsional vibration charteristics. Possibly not ideal in terms of packaging, might even require its own airframe, I bet something like David Algie's Lp1 would be a good airframe fit as it's already setup to (hopefully) have good cooling drag, intercooling, and a long cowl.