RV-8 Turbine update

[airguy]

With no spool-up time, since it's always operating at 98-100% RPM, and a FAST actuating hydraulic prop, that sounds like one wonderful high-power engine for a small aerobatic aircraft - but in it's current state, even I'm scared of it, and I'm a very pro-alternative kind of guy. I've got some turboprop and some true turbine time, and you're right - it's got to be the easiest engine management you'll ever find. This particular setup on the Innodyne scares me.

[John Clark]

I, too would like the turbine dream to be true. With something over 11000 hours flying turbine power, I share the fantasy of of kerosene power for my RV8. Not gonna happen anytime soon.

The whole Innodyn thing has a familliar ring, I will be polite here and say that the people involved in the Innodyn project are, um, optimistic. With the time they have spent I find it hard to use the term "scam" but if they aren't kidding the public, they are kidding themselves.

The technology they are pushing is rehashed 1960s APU (auxillary power unit) science. Google "Solar T-62" and read all about it. One of the more interesting reads is at http://www.provide.net/~pratt1/ambuilt/turbine.htm

This isn't really a technology problem, it is a financial one. The market is so small and the development costs so high that it is unlikely that it will ever happen. What we need is some available "spin off" technology. Anyone know of a 200 HP UAV turbine?

Nice weather today, I'm going to go fly my Lycoming.

John Clark
RV8 N18U
KSBA

[OldAndBold]

OK I ...

1) am a low time pilot
2) know very little about turbines or flying turbine aircraft
3) know nothing about Innodyne's actual software/hardware implementation.

But I am concerned with what might happen if the computer controller goes out to lunch. I have little concern of a software error, but what would/could happen if the computer (and I am assuming there is only one processor) goes out at a critical time? What if it goes out and leaves the throttle full open? Are there any efforts made at redundant processors?

I am just writing off the top of my head here, (I am an avionics software engineer) if I were designing such a system I would be planning on like three separate redundant processors on isolated independent power systems each monitoring the other two for errors/differences. And each processor with separate sensors. And then I would still want some sort of mechanical override.

Or would all of this be overkill? If something goes wrong with a turbine, how much time does a pilot have to correct the problem before a disaster occurs?

I would sort of think that if they had a redundant controller design here they would have highlighted it as a feature. That they didn't suggests to me that there is not such a feature.

--John Babrick

[John Clark]

John,
You are right on target with your concerns. The engine in question was designed as an APU (auxillary power unit in an aircraft) or in some cases a GPU (ground power unit on a cart.) In both uses the engine would run at a constant "100%" power. That is, governed to whatever the designer deemed the best for output and service life. In this application the "computer" needs only to regulate the fuel to maintain RPM if the load varies. As a APU, the unit can be set up to shut itself down if something overheats, overloads or whatever. So the airplane gets dark sitting on the ground, big deal. Engines designed to power aircraft have redundant systems to protect the engine. Remember, a turbine engine is controlled only by fuel available. Unlike a piston engine, if the controller goes nuts and really pours the fuel in it will over rev, overheat and very likely come apart with a bang. It had been a while, but I recall that a PT6 has 3 independant systems to protect against a situation. I only see one on this engine. I think that the suggestion that the engine can be run at 100% RPM and controlled only with the prop is the result of not having the technology to properly control the engine.

Again, all this could be made to work given enough time and MONEY but I really don't see the numbers that would make it happen.

If you haven't read it yet, this is interesting:
http://www.provide.net/~pratt1/ambuilt/turbine.htm


John Clark
RV8 N18U
KSBA

[rv6ejguy]

Model airplane gas turbines now generate up to 50 lbs. of thrust, run on kerosene and have their own FADECs, all for about $7K US. This sort of digital control is easy compared to those for piston engines. Can sense N1 or EPR and TIT for limiting. Should not be too expensive to develop. I don't think this aspect is the show stopper.

Having the thing go 1000+ hours between overhauls might be more difficult.

http://www.cat-ing.de/turbines/jetca...hlturbinen.htm

[John Clark]

The model turbines are very impressive but you need to remember how the size effects the design and materials. Back in my youth I had model airplanes with piston engines that produces amazing horsepower at very high RPM and were dead simple. Plain bearings, no piston rings, and brutally simple induction systems. The reason this all worked was the size (or lack of it.) You can't just scale up an engine, piston or turbine and have it work the same way. Expansion rates and material properties don't "scale up" well.

There is no reason a GA/experimental turboprop couldn't be developed. The real issue is market and cost. Here is part of a release from P&W Canada:

Last Updated: Wednesday, December 13, 2006 | 11:57 AM ET
CBC News
Pratt & Whitney Canada Corp. said Wednesday it plans to spend $1.5 billion, including $350 million from the federal government, to develop lighter, more fuel-efficient gas-turbine engines.

Ya gotta build and sell a lot of engines to recoup that kind of investment.

John Clark
RV8 N18U
KSBA

[David Johnson]

China could make cheap turbine power happen. Labor is cheap, lots of engineering talent. If I were interested in making an experimental turbine, I'd hire a Chinese engineering firm with the necessary experience and have it manufactured there. I'd call my company Harbor Freight Turbines

[rv6ejguy]

Originally Posted by John Clark

The model turbines are very impressive but you need to remember how the size effects the design and materials. Back in my youth I had model airplanes with piston engines that produces amazing horsepower at very high RPM and were dead simple. Plain bearings, no piston rings, and brutally simple induction systems. The reason this all worked was the size (or lack of it.) You can't just scale up an engine, piston or turbine and have it work the same way. Expansion rates and material properties don't "scale up" well.

My comment was directed towards the FADEC side, not the mechanical design.

I believe that it is possible to use off the shelf industrial turbocharger parts to construct a reliable experimental gas turbine. This was done several years ago as documented in Aviation Week for use on UAVs using Garrett TO4 parts and many internet plans to build gas turbines have resulted in backyard successes.

The compressor and turbomachinery is is well proven and cheap, development would center on burners and reduction gearing mainly which are no small tasks themselves but with some actual engineering applied by a talented small team, I don't think hundreds of millions need to be spent. P&W is working on certified engines, not ones for experimentals.

[OldAndBold]

As to a FADEC, here is my feeling about the software/hardware.

What the FADEC needs to do is probably not very complicated. I would expect to be able to program it to do anything it needed to do with some high speed modern version of the 8051 family - an 8/16 bit processor. At 50 mhz speeds it would be plenty fast to control the engine. One would want three of them, in separate armored modules with separate sets of sensors (temp, rpm, oil pressure, oil temp,thrust lever, etc) and power supplies. Probably some sort of small battery backup for each in case the main airplane power goes out. The processors are small - on boards about five or six square inches. The ones I am thinking of already have built in ethernet and webservers, so some simple ethernet wiring and literally some simple webpage programming would provide you with an instant detailed realtime webpage diagnostic system. In any event, the ethernet would allow the processors to interact and compare results to determine if any of the three have failed and to negotiate what to do if one or two do fail.

An additional processor could probably be assigned the task of monitoring vibration and fire protection - essentially a digital signal processor that looks for vibration signatures and temperature out of normal. Could raise an alarm to the other processors to warn the pilot of impending engine failure.

I would envision a separate processor mounted on the control panel which would have some small color LCD video screen which would be connected to the FADEC processors via the ethernet and which would display the engine parameters in realtime.

The software would operate in one of several modes - startup, taxi, takeoff, climb, cruise, descent, approach and shutdown. You probably wouldn't need to use the thrust lever much - instead, you would have several buttons on the control panel labeled with these modes that you would push to tell the engine what you wanted it to do. The controllers would also need to control the propeller pitch. Couple all of this with some sort of GPS coupled FMS and autopilot and you have an extremely automated turbine powered RV.

Really, I can envision implementing the hardware for this kind of FADEC for under $1000. The boards I am thinking of ( www.rabbitsemiconductor.com ) are ready made and cost like $200 - $300 each. Perhaps the sensors would run the bill up a bit, but the computers would not. Very, very doable. I wish I had a turbine engine to try this out on.

I guess my question is, (and I have no idea of the answer) how difficult is it to actually build the turbine? One shaft with a compressor on one side and a turbine section on the other side, able to stand the heat, pressure, torque and vibration? A lubrication system? Fuel injectors? Gear reduction? Alternator drive? Propeller pitch control?

??? I don't know. But I do know that the computer control for such a thing would be straightforward.

--John Babrick

[Gsuit]

Last year when I was making my engine decision, I was lured by the Innodyn turbine. After a lot of conversation with Innodyn, the guys at Firewall Forward, and even the guy making the throttle quadrant and guage package for Innodyn - I went so far as to plop down the $100 for the engine "manual" (and I'll use that term loosely). I seriously could write a book here telling you of the ups and downs of this engine and the company, but this is what it all boiled down to for me.

It was mentioned in an earlier post how the pilot must control the throttle and prop simultaneously... but the deeper and more dangerous issue there is how the pilot has to control the prop. It's a freakin toggle switch on the instrument panel!!!

Now you can belabor the pros and cons of the efficiency or operating limits of the turbine... but when you start trying to fly an airplane and control your speed with a toggle on the panel, you're asking for trouble. And that's exactly what Innodyn found out early in their testing.

BUT - when I discussed this with them, and the need to develop a way to control rpm and pitch with a quadrant, they agreed but didn't see it happening. Their answer to me was to wire the prop control into one of the flight stick grip switches?!?! They did this - and again had some bad results - but continued with this approach and I believe their flying test aircraft have this configuration to this day.

In the end, I think Innodyn chose the route they thought best. Limited finances allowed for one direction. They put all of their dollars into this "super-atomizing" fuel controller, and no one stopped to ask a pilot what would be the safest way to actually fly the thing (prop control). This coupled with their idea that "they'll build the engine, someone else will build the motor mount and the firewall forward" will result in a failed company; agian, in my opinion.

So, to the group who swoops in and picks up Innodyn for pennies on the dollar, I say this. Develop a working engine mount and firewall forward kit. Engineer an electric or hydraulic option to control the prop like normal turbine aircraft do (think T-34C) and you might sell some of these. RV builders already like this idea - but you've GOT to deliver a useable product. Innovaton from your initial offering will come - later.