Van's Air Force
Don't miss anything! Register now for full access to the definitive RV support community.
A Little Eggenfellner History
- Thread starter David-aviator
- Start date
- Status
maxvogelman
I'm New Here
Quite frankly, for a community of experimental aircraft builders, I find the way these discussions turn out fascinating. Some of us find production aircraft not to fit our needs, wants and budget and have no problem building a "better" airplane like an RV which until recently could hardly be considered "proven". Yet some of those same people are adamant that the same mentality cannot be applied to the engines that go in these airplanes. For the sake of discussion I'm not saying that Subaru is better. But to say that nothing can be better than (insert choice subject) is to say you might as well give up on life and commit hara-kiri. My personal thinking is that there is always a better way. We're flying airplanes right? Of course nobody could possibly find a better way to get around than driving, right?
As far as what future builders will choose, when my father and I went looking for an engine for our airplane we knew almost nothing of the history of automotive conversions or the history of Eggenfellner. We also knew little of the technical side of the equation of which so many here get hung up on. By the time we ordered an engine we still didn't know that much. What we did know is that everything we were hearing and seeing regarding Eggenfellner made a whole lot of sense for the aircraft we had in mind. Much the same way that Vans made a whole lot of sense for the aircraft we had in mind. We now have a LONG awaited
turbo'd H6 hanging on our 7a and as we get closer to flying I become more and more convinced that we made exactly the right choice in supplying our aircraft with power. I can tell you that for anybody looking at engines today, coming from the same perspective that we were, technical data means almost nothing compared to flying examples and even more importantly, happy customers. The ONLY thing that we've been unhappy with is time delays, but all things considered it's understandable.
Also, we live at 6500' msl here in Colorado. I have full faith that this engine will pull us safely over the hundreds of cliffs, peaks and valleys that are the Rockies.
At this point I will stop wasting time here, changing nobody's opinions, and go work on getting one more auto conversion flying.
Max Vogelman
Eagle, CO
RV-7a
H6T (intercooled)
As far as what future builders will choose, when my father and I went looking for an engine for our airplane we knew almost nothing of the history of automotive conversions or the history of Eggenfellner. We also knew little of the technical side of the equation of which so many here get hung up on. By the time we ordered an engine we still didn't know that much. What we did know is that everything we were hearing and seeing regarding Eggenfellner made a whole lot of sense for the aircraft we had in mind. Much the same way that Vans made a whole lot of sense for the aircraft we had in mind. We now have a LONG awaited
turbo'd H6 hanging on our 7a and as we get closer to flying I become more and more convinced that we made exactly the right choice in supplying our aircraft with power. I can tell you that for anybody looking at engines today, coming from the same perspective that we were, technical data means almost nothing compared to flying examples and even more importantly, happy customers. The ONLY thing that we've been unhappy with is time delays, but all things considered it's understandable. Also, we live at 6500' msl here in Colorado. I have full faith that this engine will pull us safely over the hundreds of cliffs, peaks and valleys that are the Rockies.
At this point I will stop wasting time here, changing nobody's opinions, and go work on getting one more auto conversion flying.
Max Vogelman
Eagle, CO
RV-7a
H6T (intercooled)
Last edited:
L.Adamson
Well Known Member
Gee that makes me feel bad. I'm not cool like a modern fighter pilot.
My panel is kind of like a T-37's. A whole bunch of instruments, switches, and knobs, to make me look very smart.
If I was to do away with the mixture, I'd just have to add something to takes it's place.
But being from a high altitude airport area, I think mixture knobs are part of our heritage! L.Adamson --- RV6A
Kind of ironic isn't it! I have avoided adding this comment for years! Do you imagine how many people were called morons in the 70's and 80's for building an RV! Build what you want and put whatever engine in you want, it's your plane! Enjoy!
janeggenfellner
Well Known Member
It is a simple matter of MAP and air temperature to calculate the amount of air entering the engine. The ECU can then simply add the right amount of fuel for any condition. The altitude is not important, just the MAP and air temp.
Jan
I am not cool; I still think old farm tractor technology, mags and mechanical FI or Carb are still cool in their own way, even if the cool kids don't like it. ha-ha
George are you coming to Reno this year? There is some cool stuff being prepped for Sport Class. Maybe you can recover some of your mojo there.
<<I tell it like it is and don't sugar coat what I see.>>
That's why I love you brother <g>
<<I have serious TV at two rpm points not within my important operational ranges.>>
Yeah, and as you know it worries me.
<<almost every engine package has a bad TV range somewhere between zero and redline rpm- even highly developed auto drivetrains.>>
Right. The vital question is amplitude. Material stress in steel parts should be below the knee in the SN curve.
<<It may be acceptable to simply not operate there.>>
You can operate there continuously if you know stress is below the knee. If it is not below the knee, total number of stress cycles becomes the issue. Stress level determines how many cycles are in your piggy bank; few if high, more if lower. When you use them up, the part fails.
Based on your description, stress sounds high. Each time you run through that resonant RPM, you spend a few more of your available cycles.
Can't calculate or measure stress? Want to rely on test hours instead? Fine, run your 200 hours at 1350 RPM. In your case I'm pretty sure component stress is far higher at 1350 than at wide open throttle. It will empty your piggy bank at a rate of 162,000 cycles per hour (1350 x 2 x 60), or 32.4 million in 200 hours. Available cycles depends on material choice, heat treat, and surface finish....but even at the lowest above-the-knee stress levels a lot of the choices don't give you much more than 10 million.
"But wait," you say, "In real life I shove the throttle up quickly and push right through it. I don't run there for more than a few seconds." Ok, let's pick four seconds, or 180 cycles. Pushing up the manifold pressure also pushed up the vibratory amplitude; higher stress, less available cycles. The very high stress part of the SN curve often doesn't give you more than 100,000 cycles. 100,000/180 = 555, and you use a few every time you taxi out to fly.
Lies, **** lies, and statistics; the above makes obvious assumptions and SN curves are far from exact. However, none of it is real far from reality, so yeah, let's experiment with some ways to lower the amplitude.
That's why I love you brother <g>
<<I have serious TV at two rpm points not within my important operational ranges.>>
Yeah, and as you know it worries me.
<<almost every engine package has a bad TV range somewhere between zero and redline rpm- even highly developed auto drivetrains.>>
Right. The vital question is amplitude. Material stress in steel parts should be below the knee in the SN curve.
<<It may be acceptable to simply not operate there.>>
You can operate there continuously if you know stress is below the knee. If it is not below the knee, total number of stress cycles becomes the issue. Stress level determines how many cycles are in your piggy bank; few if high, more if lower. When you use them up, the part fails.
Based on your description, stress sounds high. Each time you run through that resonant RPM, you spend a few more of your available cycles.
Can't calculate or measure stress? Want to rely on test hours instead? Fine, run your 200 hours at 1350 RPM. In your case I'm pretty sure component stress is far higher at 1350 than at wide open throttle. It will empty your piggy bank at a rate of 162,000 cycles per hour (1350 x 2 x 60), or 32.4 million in 200 hours. Available cycles depends on material choice, heat treat, and surface finish....but even at the lowest above-the-knee stress levels a lot of the choices don't give you much more than 10 million.
"But wait," you say, "In real life I shove the throttle up quickly and push right through it. I don't run there for more than a few seconds." Ok, let's pick four seconds, or 180 cycles. Pushing up the manifold pressure also pushed up the vibratory amplitude; higher stress, less available cycles. The very high stress part of the SN curve often doesn't give you more than 100,000 cycles. 100,000/180 = 555, and you use a few every time you taxi out to fly.
Lies, **** lies, and statistics; the above makes obvious assumptions and SN curves are far from exact. However, none of it is real far from reality, so yeah, let's experiment with some ways to lower the amplitude.
You can search the net for any engine +"problems" and the result is the same. This doesn't indicate anything particular.
It is not only burniing 35% less fuel, it is burning 35% less of a fuel that cost 50% less.
Jconard
Well Known Member
It is a simple matter of MAP and air temperature to calculate the amount of air entering the engine. The ECU can then simply add the right amount of fuel for any condition. The altitude is not important, just the MAP and air temp.
This is an interesting piece of data...I take it that this is the implementation of the new ECU, not Subaru's, I think it is made by Ross, right?
With that simple calculation and from those data it would appear you have a static air fuel ratio which be the same for a given density altitude and temperature, regardless of the flight regime.
So full throttle at a high altitude airport...say 19 inches of pressure would yield the same fuel flow as wide open throttle in cruise (assuming CS prop) at an altitude that resulted in the same MP?
Or are there other sensor that trim the fuel back, or is there a switch of some sort to allow different fuel maps to be chosen?
This is an interesting piece of data...I take it that this is the implementation of the new ECU, not Subaru's, I think it is made by Ross, right?
With that simple calculation and from those data it would appear you have a static air fuel ratio which be the same for a given density altitude and temperature, regardless of the flight regime.
So full throttle at a high altitude airport...say 19 inches of pressure would yield the same fuel flow as wide open throttle in cruise (assuming CS prop) at an altitude that resulted in the same MP?
Or are there other sensor that trim the fuel back, or is there a switch of some sort to allow different fuel maps to be chosen?
You are welcome to believe the hype if you wish. You find out if ANY 1.7 has reached 2400 hours. I think you will find the answer is no.
Jet and avgas are essentially the same price over here. Downtime waiting for engine changes will eat up any savings in fuel. When diesel aero engines actually go 2400 hours without a ton of maintenance hours, they will have arrived. Nobody has proven that yet.
Rotary10-RV
Well Known Member
Certified or Experimental?
Yes Bill, it should if the sellers are selling the engine as CERTIFIED. I am a rotary engine (Wankel) user, and have been following Mistral Engines with interest. (www.mistralengines.com) They have been running 2 engines in a FAA cert program for a couple of months now. The thing is this is so all-consuming that they haven't had time for a lot of communication with parts buyers. Some people were thinking they were on the ropes. (NOT true they even have new investors and testing is going great.) A engine, and psru that is sold as certified SHOULD run to TBO without problems. So I would expect that if Mistral succeeds their alternative engine and psru would make TBO.Even with certified engines some do and some do not. (Note the Piper Turbo Mirage class action suit.) An experimental is not LEGALLY required to meet these standards. A experimental engine, and that includes the Lycoming clones, is not tested to those standards. If you require that level of testing you should buy a certified example. You should note though that even a Manittuck assembled Superior IO-360 clone is an EXPERIMENTAL engine and the combination of parts may never have been tested to certified standards.
If you ask for a certified engine the price instantly goes up several thousand dollars, even if the same parts were used you are now paying for the paper trail. I'd just as soon avoid that. A Timken bearing for the front wheel of a 1972 DODGE is about $9.00 at the auto parts store, The exact same bearing for Grumman Tiger with a single different character etched on the bearing race goes for $50.00+ if you can even find a place that sells them. Just bear that in mind.
Bill Jepson
Yes Bill, it should if the sellers are selling the engine as CERTIFIED. I am a rotary engine (Wankel) user, and have been following Mistral Engines with interest. (www.mistralengines.com) They have been running 2 engines in a FAA cert program for a couple of months now. The thing is this is so all-consuming that they haven't had time for a lot of communication with parts buyers. Some people were thinking they were on the ropes. (NOT true they even have new investors and testing is going great.) A engine, and psru that is sold as certified SHOULD run to TBO without problems. So I would expect that if Mistral succeeds their alternative engine and psru would make TBO.Even with certified engines some do and some do not. (Note the Piper Turbo Mirage class action suit.) An experimental is not LEGALLY required to meet these standards. A experimental engine, and that includes the Lycoming clones, is not tested to those standards. If you require that level of testing you should buy a certified example. You should note though that even a Manittuck assembled Superior IO-360 clone is an EXPERIMENTAL engine and the combination of parts may never have been tested to certified standards.
If you ask for a certified engine the price instantly goes up several thousand dollars, even if the same parts were used you are now paying for the paper trail. I'd just as soon avoid that. A Timken bearing for the front wheel of a 1972 DODGE is about $9.00 at the auto parts store, The exact same bearing for Grumman Tiger with a single different character etched on the bearing race goes for $50.00+ if you can even find a place that sells them. Just bear that in mind.
Bill Jepson
gmcjetpilot
Well Known Member
My take, no garentee in life and life is not fair?
As far as the crankshaft issue (very embarrassing with no real excuse), before Lyc had their crank issue in the late 90's, Continental went through a similar thing in the early 90's. You think Lyc would have seen that and been more careful? Continental changed their crankshaft manufacturing process. Doha! That "golden" process got corrupted or changed by ill advice. The safety net, QC like random testing (destructive, cut it up & inspect it) was either not done or done properly. Still your point is well taken, even certified is no guarantee of perfection or zero chance of failure. However when all the processes are done properly, it offers some safety, QC and statistical consistency. This does not mean other processes out side the FAA paper trail can't produce reliable parts. To be fair both Continental and Lyc made their cranks for decades before and since with out issues. It was more manufacturing than design, but it shows a small change can make a difference. So don't fix it if it ain't broke.
THIS MAY INTEREST YOU NEW TURBO GUYS WITH THE EGG 3.6 SUBIE H6 - "Watch those temps"
The PA46-310 and PA46-350 is an interesting case history. The first Malibu's came with a 310hp Continental but a string of accidents due to crank failures, 1984 - 1986, forced Piper to "re-brand" its flagship aircraft to Mirage, dumping the Continental for the Lyc TIO-540 350HP changing. Speed went up but range went down. To improve range the Lyc was allowing higher TIT (turbo inlet temps). There where no Lyc crank issues on this engine, but with higher TIT, achieving good engine longevity became a new issue for the PA46. The issue is you have to have real discipline and pay attention to operate these engines. Many private pilots where not being too careful, despite having an engine monitor. Also TIT probes where found to be bad, reading low! Lesson calibrate your instruments, often? They now call for 250 hr TIT checks. Some pilots even where intentionally going on the LEAN side of TIT. It saves a little fuel, but it eats the engine. So Lyc issued a SB (below) on the Piper Mirage engine with operational limits & techniques. It's interesting, they go into LOP operations and temp limits like CHT = 400F max for longevity. Bottom line TEMP = Engine Life, any engine.
http://www.lycoming.textron.com/support/troubleshooting/resources/SSP400.pdf
(This SB is just real world limits & operational issues, but it applies to all new engines or applications. I see where water cooling could mitigate temperature issues and simplify operations. Water cooling + turbo charging + flying at high altitude has advantages. On the other hand sport planes are not always flying in the high "teens" where a turbos and oxygen are needed, and sport planes don't always launch on +500 mile trips every flight. An argument for "auto-mixture" of FADEC may be more than convenience, it could protect the engine. I'll be following the Egg developments closely with anticipation.)
Bill I agree with your comments; well said. To be fair, even certified engines are not TESTED, except many be initial break-in. "Compliance" is making sure every part is made exactly the same way it was made on the "approved" test engine'(s) certified. Every engine after is a "CLONE". It's how the parts are made, the process, paper trail that gives the pedigree. Even the CLONE engines have certified parts. If you buy a new Lyc cylinder regardless of the vendor, it can go on a "clone" or certified engine.
As far as the crankshaft issue (very embarrassing with no real excuse), before Lyc had their crank issue in the late 90's, Continental went through a similar thing in the early 90's. You think Lyc would have seen that and been more careful? Continental changed their crankshaft manufacturing process. Doha! That "golden" process got corrupted or changed by ill advice. The safety net, QC like random testing (destructive, cut it up & inspect it) was either not done or done properly. Still your point is well taken, even certified is no guarantee of perfection or zero chance of failure. However when all the processes are done properly, it offers some safety, QC and statistical consistency. This does not mean other processes out side the FAA paper trail can't produce reliable parts. To be fair both Continental and Lyc made their cranks for decades before and since with out issues. It was more manufacturing than design, but it shows a small change can make a difference. So don't fix it if it ain't broke.
THIS MAY INTEREST YOU NEW TURBO GUYS WITH THE EGG 3.6 SUBIE H6 - "Watch those temps"
The PA46-310 and PA46-350 is an interesting case history. The first Malibu's came with a 310hp Continental but a string of accidents due to crank failures, 1984 - 1986, forced Piper to "re-brand" its flagship aircraft to Mirage, dumping the Continental for the Lyc TIO-540 350HP changing. Speed went up but range went down. To improve range the Lyc was allowing higher TIT (turbo inlet temps). There where no Lyc crank issues on this engine, but with higher TIT, achieving good engine longevity became a new issue for the PA46. The issue is you have to have real discipline and pay attention to operate these engines. Many private pilots where not being too careful, despite having an engine monitor. Also TIT probes where found to be bad, reading low! Lesson calibrate your instruments, often? They now call for 250 hr TIT checks. Some pilots even where intentionally going on the LEAN side of TIT. It saves a little fuel, but it eats the engine. So Lyc issued a SB (below) on the Piper Mirage engine with operational limits & techniques. It's interesting, they go into LOP operations and temp limits like CHT = 400F max for longevity. Bottom line TEMP = Engine Life, any engine.
http://www.lycoming.textron.com/support/troubleshooting/resources/SSP400.pdf
(This SB is just real world limits & operational issues, but it applies to all new engines or applications. I see where water cooling could mitigate temperature issues and simplify operations. Water cooling + turbo charging + flying at high altitude has advantages. On the other hand sport planes are not always flying in the high "teens" where a turbos and oxygen are needed, and sport planes don't always launch on +500 mile trips every flight. An argument for "auto-mixture" of FADEC may be more than convenience, it could protect the engine. I'll be following the Egg developments closely with anticipation.)
Last edited:
I know that the Thielert have had some issues, but it is nowhere near this big catastrophe as you describe it to be. The TBO is not 2400 h yet as far as I know. It is 1000 h, and then you get a replacement engine for free. At 2000 h, you get a new engine at reduced cost. The aim is more than 3000h TBO in the future.
As I said, the press hyped the Thielert to death even before it flew, passing along this 2400 hr. figure which was totally unrealistic as we now see with their much lower actual TBR. I'm trying to find a post from a guy in Florida who flies both single and twin engined Diamonds for a flight school there. Their experience has also been shockingly bad. I recall one engine having to came out with 325 hours on it after oil consumption exceeded 1 qt./hr. He listed other premature removals for similar problems way before 600 hours.
The idea of a simple, inexpensive, reliable aero diesel is fantastic in my view. The Thielert 1.7 is just not that engine. I hope the 2.0 is.
And we hear all these complaints about non-certified alternative engine guys being beta testers for under developed products? $200M worth of engineering, development, testing and certification did not guarantee results in this case. They are back to the age old standby of learning in service and applying those lessons learned for the next iteration.
Update: Here are a couple posts from a pilot flying Diamonds in Austria-
The power losses our engines had were almost alway electrically induced. wiring harnesses that weren't completely waterproof being one example.
The engines we had to exchange, weren't exchanged due to no more go, but due to "wayyy to much oil". normal oil consumption is 0.1 liters per hour and engine. in the end we wound up with over ONE LITER (about a quart) of oil consumption per hour and engine.
That's when diamond sent us a new engine ...
*********************************************************************************
How the **** should I be objective when I've flown 140 hrs in diamonds with thielerts, have had many problems with these engines (power loss down to 30% in 13.000 ft over the alps (very high mountains in europe for the geographic geniuses out there )
was grounded for weeks (added up) because engines failed ( and thielert couldn't deliver a new one), had to be replaced (see before), had electrical problems and so forth. I have posted before what problems we have had with the thielerts.
i'd love to try the 4.0 in the cirrus, just to see if its any better, but from what i've read in magazines about the engine, it doesn't really cut it any better, sadly.
********************************************************************************
I've just had my CPL IFR check ride, and I have spent the last 70 hrs flight time on the left seat of both a DA40 and DA42, as well as another 70hrs on the backseat, with my friend doing the same. so a total of 140 hrs thielert time.
So far we've had :
ECU A FAIL
DUAL ALTERNATOR FAIL
POWER LOSS TO 35% (limp home mode) emergency descent in IMC over the alps (MEA 12000, MORA 15000)
A TOTAL OF THREE ENGINE SWAPS (1 DA40, 2 DA42)
TWO ALTERNATOR SWAPS (DA42)
Yes i like the engines, no i don't like the way they don't hold up. no engine ever has met the 1500 TBR, the most we have managed so far is 960 hrs, and Thielert is now giving us two brand spanking new 2.0 in exchange for the 1.7, as they want to know why they held up so long
*******************************************************************************
I'm not making this stuff up.
Last edited:
Rotary10-RV
Well Known Member
TBR time between REPLACEMENT not TBO
B,
I hope Thilert can fix some of the issues they have had. I'm am almost a dogged champion of alternate engines, I do worry though that Thielert had to certify as TBR, Time Between Replacement, rather than the common TBO. I would like to find out why this restriction went into place. Like Dan H has mentioned is it the number of cycles? (heat, vibration, otherwise) I worry any time that the base engine core isn't suitable for re-use. I would also like to know if this was imposed by Thielert or the FAA. What part is marginal enough to need the engine replaced? Or to turn the other side of the coin are they doing this as a simple precaution to check out the early engines wear points? Any time you see a different requirement you wonder about the reason for the different requirement. Consider it like Dan's request for the vibrations studies from Eggenfellner so he could satisfy himself it was done correctly. I do believe that Diesels will become a common GA engine it just makes sense.
Bill Jepson
B,
I hope Thilert can fix some of the issues they have had. I'm am almost a dogged champion of alternate engines, I do worry though that Thielert had to certify as TBR, Time Between Replacement, rather than the common TBO. I would like to find out why this restriction went into place. Like Dan H has mentioned is it the number of cycles? (heat, vibration, otherwise) I worry any time that the base engine core isn't suitable for re-use. I would also like to know if this was imposed by Thielert or the FAA. What part is marginal enough to need the engine replaced? Or to turn the other side of the coin are they doing this as a simple precaution to check out the early engines wear points? Any time you see a different requirement you wonder about the reason for the different requirement. Consider it like Dan's request for the vibrations studies from Eggenfellner so he could satisfy himself it was done correctly. I do believe that Diesels will become a common GA engine it just makes sense.
Bill Jepson
As far as I know (could be wrong) this is more of a business strategy, you get an overhauled or rebuilt engine where only the worn parts are new, then you need 100% control of all the parts. I guess it is more practical, economical and better quality vise to ship the entire engine to one (or a handfull) service facility with all the proper tools and knowledge instead of training thousands of mechanics. It is also the only way to get proper measurements of the wear for proper feedback for eventual re-design of parts. This is a brand new engine, and the factory cannot possibly know how it will hold together under various conditions, there are just too many unknown variables. Their aim is 3000h, and this is not an easy achievement. The certified Rotax'es are still only 1500h.
Yes, I believe this is their thinking. Overhauls or replacement and upgrading are all done at the factory. A very good idea in the case of an unfamiliar design. This way everything is done right, they can tear down and see what is wearing and there is less potential for reputation problems if a field service center screws up. Rotax has gone a different way which also seems to work well- setting up a few factory trained service centers in each country.
I personally think the replacement idea is best for Thielert at this stage. I too hope they succeed in getting the bugs out and achieving regular 3000 hr TBRs.
Open loop control
There are two kinds of system control.
Without an O2 sensor (or some other type of feedback to allow closed loop control), the engine likely won't run as efficiently as it would in a Subaru (equipped with O2 sensor).
Edit:
After thinking about this some more, an open loop computer controlled system could actually be less efficient than a standard lycoming with an engine monitor running LOP. In the LOP scenario, that's essentially a closed loop system...it's just that the pilot is closing the loop.
There are a couple factors that could help Lyc's run more efficiently:
There are a couple other concerns about using a MAP sensor. I've had personal experience with at least three cars (none Subaru) that would not run with a failed MAP sensor. It makes me wonder what would happen if the MAP sensor on the Egg failed low, high, or anywhere in between. This is the kind of thing I was referring to when I asked about FMEA in the previous post.
There are two kinds of system control.
- Open loop control - The system generates an output (fuel quantity) based on one or more inputs (MAP and air temp in this case). Essentially it's a best guess methodology.
- Closed loop control - Works just like open loop control, but there's an additional input...an O2 sensor. This provides a feedback correction to compensate for other variables. So the computer says, ok for this MAP and air temp I think I need to supply X amount of fuel. Then it checks the O2 sensor and says hrm, it looks like my guess was a bit too high (or low), so I'll reduce (or increase) the fuel flow a bit. The method of correction is historically a PID loop.
Without an O2 sensor (or some other type of feedback to allow closed loop control), the engine likely won't run as efficiently as it would in a Subaru (equipped with O2 sensor).
Edit:
After thinking about this some more, an open loop computer controlled system could actually be less efficient than a standard lycoming with an engine monitor running LOP. In the LOP scenario, that's essentially a closed loop system...it's just that the pilot is closing the loop.
There are a couple factors that could help Lyc's run more efficiently:
- Engine efficiency is very directly related to compression ratio. Since the Egg runs on auto gas, it will require a lower compression ratio than an engine running 100LL. Thus, any Lyc with a higher compression ratio than the Egg will benefit from this.
- As a general statement, for a given throttle position an engine will produce more power per stroke at a lower RPM than at a higher RPM due to losses in the intake and exhaust system. Certainly there are lots of other variables at play here, but this could also give a lower RPM Lyc an advantage over a higher RPM Subie.
There are a couple other concerns about using a MAP sensor. I've had personal experience with at least three cars (none Subaru) that would not run with a failed MAP sensor. It makes me wonder what would happen if the MAP sensor on the Egg failed low, high, or anywhere in between. This is the kind of thing I was referring to when I asked about FMEA in the previous post.
Last edited:
Fatigue limit only applies to steel
Dan,
You're probably already aware, but for others, this "knee" only applies to steel. Aluminum has no fatigue limit below which can be assumed infinite life. It appears that titanium also has a fatigue limit.
Dan,
You're probably already aware, but for others, this "knee" only applies to steel. Aluminum has no fatigue limit below which can be assumed infinite life. It appears that titanium also has a fatigue limit.
Last edited:
Open loop control can be quite precise assuming engine wear does not enter into the picture over a long period of time. Using RPM, MAP and IAT we know mass flow, by programming for the desired AFR using a wideband, we can nail all the points in the map. If any of the 3 determining parameters change, the ECU will adjust to achieve the desired AFR.
Traditional narrow band sensors driving closed loop control target 14.7 AFR which is generally too lean for sustained high rpm, high output conditions seen in aircraft. It also results in peak EGT and nearly a 15% loss in power. The OEMs jump out of closed loop under these conditions.
With the advent of wideband sensors today, many OE ECUs use a targeted closed loop strategy right to WOT/ max rpm. They may target 12.8 for max power and 17 to 1 for light throttle cruising (LOP).This is the ultimate strategy if the sensor does not fail. 100LL is generally unkind to sensors but by adding Decalin lead scavenger or similar products, life has been shown to be acceptable in some cases.
Automotive engines have a much lower octane requirement today than aircraft engines. Witness many engines today with CR around 10 to 1 which run on 87 octane. Modern combustion chambers and liquid cooling make this part much better optimized than their older air cooled cousins. We have many designs today topping 11 to 1 which run on 91 octane.
The volumetric efficiency of the modern Subaru surpasses the Lycoming by a substantial margin due to AVCS, induction pulse tuning and 4 valves per cylinder. Where it loses out is in frictional losses at high rpm. This negates any advantages it might have due to modern features and technology so we don't see auto engines offering any improvements in SFC at this time.
Aviation spec ECUs are programmed with default values in the event of IAT, CLT and MAP failures (open or shorted). MAP is pulled high (EM46) and defaults to a fueling value of around 90% power. IAT is defaulted to 70F and CLT to warm engine.
Last edited:
Mike S
Senior Curmudgeon
Well said Ross.
One thing that always seems to get lost in the traditional vs. auto engine discussions, is that most if not all of each designs "pluses" result from the engine fitting into the operational environment it was designed to operate in.
Just for a little mental distraction,
try to figure out how effectively to use a Lyclone in a modern auto. Dont forget smog/safety/CAFE/etc requirements. And, oh yes, you need to obtain equal or better performance, longevity, and fuel economy.janeggenfellner
Well Known Member
Someone pointed out that they had a tough time installing the specialty STI engine package we made a few years back. Here is an example of an E-6 engine being installed last week. This is the secont KIS Cruiser we have provided power for. The first one won awards at Arlington.
http://www.kisbuild.r-a-reed-assoc.com/TimSahagun.html
Jan
http://www.kisbuild.r-a-reed-assoc.com/TimSahagun.html
Jan
Last edited:
gmcjetpilot
Well Known Member
Lycs in cars
As I understand it, the Germans pre WWII, envisioned the VW car engine as a work horse. In 1933, Ferdinand Porsche proposed the "Volks-Wagen" ("people's car" in German). Rumor has it, the VW engine was made to come out easily so it could drive other applications like a small airplane, generator or equipment. Of course the VW has had success as an airplane engine, and apparently in parts of the world its common to see VW engines driving air compressors at remote mining sites.
Mike Starkey, that is a distraction but you are right. Noise - Lyc would be too loud in a car by today's standards, even if you ran full mufflers and tail pipes; you would hear the clacking with out water jackets. My Acura can't be heard running. Lyc could not meet noise regulations. Pollution wise, yep the Lyc could not pass the state emission test, but than a plane doesn't need to meet this test. Acceleration and deceleration over and over would not be desirable with the Lyc-O-car. A car engine with it's smaller inertial mass, pistons, valves, displacement is a higher revving design, better suited for drag racing off a stop light, with fast acceleration, over and over all day. Not something we do in a plane, which is made to run at constant RPM's. A Lyc would be a dog on the drag strip. That just shows you the different mission and design. It's not a put down of the Lyc, but shows its made specifically for this one application. In all designs there are compromises. A Lyc would make a terrible car engine. To the Subaru's credit, it's a great car engine, and works well in a plane. The Subaru is very versatile engine. Now if you modified and adapted a Lyc in a car, you could get it to "work", go down the road. It would be unique. I guess the "air-boat" is the best adaptation of a small aircraft engine. A 18 wheeler truck engine, Formula Race car, Drag racer, Large Farm equipment, boat, plane, motorcycle all have different missions and their engines are compromises or optimizations towards that mission.Just for a little mental distraction,
As I understand it, the Germans pre WWII, envisioned the VW car engine as a work horse. In 1933, Ferdinand Porsche proposed the "Volks-Wagen" ("people's car" in German). Rumor has it, the VW engine was made to come out easily so it could drive other applications like a small airplane, generator or equipment. Of course the VW has had success as an airplane engine, and apparently in parts of the world its common to see VW engines driving air compressors at remote mining sites.
Last edited:
Jconard
Well Known Member
One thing that always seems to get lost in the traditional vs. auto engine discussions, is that most if not all of each designs "pluses" result from the engine fitting into the operational environment it was designed to operate in.
Just for a little mental distraction, try to figure out how effectively to use a Lyclone in a modern auto. Dont forget smog/safety/CAFE/etc requirements. And, oh yes, you need to obtain equal or better performance, longevity, and fuel economy.
Couldn't agree more, actually when I started thinking about adapting an aircraft engine to automotive use, I quickly got to the point of "why bother?"
An excellent metal exercise. But I think the analysis goes both directions...
Back to the fuel maps...I have always thought that there were engine operating regimes where you would want to run more rich than normal, and regimes when you would run more lean than normal. With just the inputs metioned it seems that you would have the same ratio all the time. WOuldn't you need some other sensors or a fuel trim knob to handle those other special situations?
If you put on a fuel trim knob, could you paint it red?
It sort of becomes an electronic carb...put in a full throttle microswitch to give a little extra fuel during takeoff for example, then add a mixture trim knob for when you can run super lean. Kinda like an economizer valve and a mixture cable.
Just for a little mental distraction, try to figure out how effectively to use a Lyclone in a modern auto. Dont forget smog/safety/CAFE/etc requirements. And, oh yes, you need to obtain equal or better performance, longevity, and fuel economy.
Couldn't agree more, actually when I started thinking about adapting an aircraft engine to automotive use, I quickly got to the point of "why bother?"
An excellent metal exercise. But I think the analysis goes both directions...
Back to the fuel maps...I have always thought that there were engine operating regimes where you would want to run more rich than normal, and regimes when you would run more lean than normal. With just the inputs metioned it seems that you would have the same ratio all the time. WOuldn't you need some other sensors or a fuel trim knob to handle those other special situations?
If you put on a fuel trim knob, could you paint it red?
It sort of becomes an electronic carb...put in a full throttle microswitch to give a little extra fuel during takeoff for example, then add a mixture trim knob for when you can run super lean. Kinda like an economizer valve and a mixture cable.
The beauty of the ECU is it looks at this stuff all the time and adjusts for best AFR and timing based on extensive testing and tuning.
Since long term LOP operation is not proven on these engines at high power settings, the current preference is do what is safe and proven to date and that is ROP for these engines.
I do have a mixture knob in my 6A but have the knowledge and instrumentation to use this properly. Some vendors do not want the liability of having the pilot able to diddle with mixture. If the engine fails due to excessive leaning, the finger pointing ends up towards the vendor invariably. Dan Hawken who built the Suzuki V6 conversions for the Titan T51s did it this way as well.
Since long term LOP operation is not proven on these engines at high power settings, the current preference is do what is safe and proven to date and that is ROP for these engines.
I do have a mixture knob in my 6A but have the knowledge and instrumentation to use this properly. Some vendors do not want the liability of having the pilot able to diddle with mixture. If the engine fails due to excessive leaning, the finger pointing ends up towards the vendor invariably. Dan Hawken who built the Suzuki V6 conversions for the Titan T51s did it this way as well.
Mike S
Senior Curmudgeon
Of course it does.
That is the entire reason I posed the question.
Just trying to shake folks out of the tunnel vision thing.
Now, consider we are looking at a product----Egg subie in this case----that has traveled down a long hard road of development, how would you folks like to be the one starting on that path????
Or, to put it another way, mental distraction, take 2-----contemplate being the first one to try using a modern car engine in an airplane.
Octane and ignition advance
You beat me to it , Ross. From the reading I've done, I understand that combustion chamber shape, surface temperature, spark plugs (one vs two)and even material (CI vs AL) make a difference in rate of flame propagation and detonation sensitivity.
Which leads me to a question. I've think I've read data that suggests the popular EIs for aviation have ignition advance curves like cars. Now, what really matters is where peak combustion pressure occurs relative to crank angle. Does an automotive advance curve work well (safely) with a hot aluminum dual plug head?
Sorry this isn't exactly on thread topic.
You beat me to it , Ross. From the reading I've done, I understand that combustion chamber shape, surface temperature, spark plugs (one vs two)and even material (CI vs AL) make a difference in rate of flame propagation and detonation sensitivity.
Which leads me to a question. I've think I've read data that suggests the popular EIs for aviation have ignition advance curves like cars. Now, what really matters is where peak combustion pressure occurs relative to crank angle. Does an automotive advance curve work well (safely) with a hot aluminum dual plug head?
Sorry this isn't exactly on thread topic.
Glossary
Well I had a helluva time reading that post with all the TLAs (three letter acronyms)
Here's a glossary of terms:
All that sounds really freakin' complex. I'm sure glad we don't have to deal with that level of complexity on traditional aircraft engines.
Unless I missed something, it still sounds like the Egg is running open loop. Not necessarily a bad thing, but be aware that closed loop control is one of the major achievements of automotive design that makes an engine "modern".
But then again...I'm not an engine expert...just a guy that knows the basics.
Thanks for chiming in Ross. That was a very educational post.
Well I had a helluva time reading that post with all the TLAs (three letter acronyms)
Here's a glossary of terms:
- RPM - revolutions per minute (duh)
- AVCS - active valve control system
- MAP - manifold air temperature
- AFR - air fuel ratio
- WOT - wide open throttle
- LOP - lean of peak
- IAT - intake air temperature
- CLT - coolant temperature
- ECU - electronic control unit
- SFC - specific fuel consumption
All that sounds really freakin' complex. I'm sure glad we don't have to deal with that level of complexity on traditional aircraft engines.
Unless I missed something, it still sounds like the Egg is running open loop. Not necessarily a bad thing, but be aware that closed loop control is one of the major achievements of automotive design that makes an engine "modern".
But then again...I'm not an engine expert...just a guy that knows the basics.
Thanks for chiming in Ross. That was a very educational post.
Jconard
Well Known Member
The beauty of the ECU is it looks at this stuff all the time and adjusts for best AFR and timing based on extensive testing and tuning.
It does not seem like it would adjust much of anything in cruise for example where rpm, throttle and DA are constant, and in that regime would have the same A/F ratio as in takeoff or desent...
But am I wrong that it looks at MP, AF, and RPM only? If so, then a full throttle mixture at a certain airflow and MAP would either be too rich under most conditions, or too lean at some. Again, 22", full throttle with some airflow in cruise against a cs prop qwould have different requirements than the same parameters on a takeoff roll...wouldn't it?
I mean getting rid of the O2, tps, etc.. makes this a very simple map that would actually be less flexible than what a pilot can do with throttle, mixture and prop.
what am I missing
It does not seem like it would adjust much of anything in cruise for example where rpm, throttle and DA are constant, and in that regime would have the same A/F ratio as in takeoff or desent...
But am I wrong that it looks at MP, AF, and RPM only? If so, then a full throttle mixture at a certain airflow and MAP would either be too rich under most conditions, or too lean at some. Again, 22", full throttle with some airflow in cruise against a cs prop qwould have different requirements than the same parameters on a takeoff roll...wouldn't it?
I mean getting rid of the O2, tps, etc.. makes this a very simple map that would actually be less flexible than what a pilot can do with throttle, mixture and prop.
what am I missing
Lots of questions.
In a speed/ density system which uses a MAP sensor for load, we don't care about throttle angle. We have an rpm fuel value to correct for VE differences with rpm, we sense MAP (manifold absolute pressure) which takes care of altitude and throttling effects from the plate simultaneously and we have IAT to correct for air density changes related to temperature. If these parameters are set correctly, any combination of the 3 will result in the targeted AFR.
Cut back rpm and ECU delivers less injection pulses, increase MAP and ECU lengthens injector pulse, increase IAT and ECU reduces injector pulse. In this, we have to assume the user has run the actual engine under a loaded condition in flight or on a dyno to get the values correct using EGT and/or a wideband meter.
True, without a mixture control, we have less flexibility to over lean or over richen compared to conventional controls but we also can't hurt the engine. We can target any desired AFR anywhere in the operating range using a wideband sensor in closed loop but it all rides on the sensor living in a 100LL environment- not well proven at this time long term.
As far as timing goes, we see OEM automotive ECUs generally rely heavily on a knock sensor to achieve peak cylinder pressures. We think a realistic timing curve with the knock sensor acting as a safety backup is more practical for aviation use. Much of the highly advanced timing in automotive use is to keep emissions low.
Typical Subaru engines require relatively little advance for max power due to the smaller chamber, high CR and centrally located plug. We have found minimal gains in advancing the spark at lower MAP. The optimal ignition curve seems to vary less with load than what we are led to believe is used on traditional air cooled engines with EI or FADECs. We do need to make sure that detonation or pre-ignition never takes place as in a constant, high power application like this, even liquid cooled engines can have piston damage happen pretty quickly. We do have to be concerned with balancing timing with resultant EGT as well so much testing has to be done to make this all automatic.
When done properly, it can be demonstrated that a FADEC/ ECU can do a better job with regards to fuel used per mission than a pilot can with conventional controls with far less monitoring. More time to fly the plane and look for traffic. We see the biggest savings in startup, taxi and climb regimes. In a steady state, cruise condition, a pilot can do a good job at setting up for max fuel economy.
Closed loop control was originally designed for emissions to satisfy the requirements for catalytic converters and secondarily for improving fuel economy and providing a means to correct AFRs long term with engine wear. It never had a performance aspect until wideband sensors arrived and allowed engineers to target specific AFRs vs. hp/rpm/load conditions permitting closed loop operation through the entire operating range- something narrow band sensors do not allow.
The resistance to new technology is interesting in aviation. Certain things are immediately embraced while other things are not. Maybe this has something to do with keeping the fan turning. In Reno Sport Class, convincing pilots to switch over to FADECs has been relatively slow until faced with the realities of the advantages. Once the mechanical systems are beyond their design limits to meter fuel accurately or pilot workload starts to compromise safety, things become more clear. Every year now sees another entry dump the old guard for EFI. While this technology is not for everyone or every engine, it is the logical choice for modern automotive based powerplants. I shake my head at the Unlimited Class Merlin guys who keep burning pistons using carbs and a log manifold. They just do it over and over...
With regards to FMEA, most of us in this world do plenty of thinking about what ifs with regards to system designs and if anything, learning failure modes based on experience is the primary tool in improving the design. So while we might not call it FMEA, we are doing the same sort of things. Conservative, evolutionary approaches result in the best probability of success. There is no substitute IMO for weeding out the bad points of a design or system based on lots of flight time. If we can combine this with better, cheaper analytical tools and models in the design and testing stage, things should improve at a faster rate with less waste and risk. Designing without a solid basis in theory/ engineering often has its negatives.
It is interesting what a professional design team comes up with sometimes though. This afternoon I was looking at a dismantled Thielert clutch coupling which had started to shed parts at 177 hours. One look was enough to make me wonder what these guys were thinking when they chose this design. The same thing has been shown to be unreliable in other aircraft PSRUs a decade ago but these guys did not know that I guess. There have been multiple failures of this part and two power loss accidents that I know of. I hope they get a better design on the new 2.0 engines.
In a speed/ density system which uses a MAP sensor for load, we don't care about throttle angle. We have an rpm fuel value to correct for VE differences with rpm, we sense MAP (manifold absolute pressure) which takes care of altitude and throttling effects from the plate simultaneously and we have IAT to correct for air density changes related to temperature. If these parameters are set correctly, any combination of the 3 will result in the targeted AFR.
Cut back rpm and ECU delivers less injection pulses, increase MAP and ECU lengthens injector pulse, increase IAT and ECU reduces injector pulse. In this, we have to assume the user has run the actual engine under a loaded condition in flight or on a dyno to get the values correct using EGT and/or a wideband meter.
True, without a mixture control, we have less flexibility to over lean or over richen compared to conventional controls but we also can't hurt the engine. We can target any desired AFR anywhere in the operating range using a wideband sensor in closed loop but it all rides on the sensor living in a 100LL environment- not well proven at this time long term.
As far as timing goes, we see OEM automotive ECUs generally rely heavily on a knock sensor to achieve peak cylinder pressures. We think a realistic timing curve with the knock sensor acting as a safety backup is more practical for aviation use. Much of the highly advanced timing in automotive use is to keep emissions low.
Typical Subaru engines require relatively little advance for max power due to the smaller chamber, high CR and centrally located plug. We have found minimal gains in advancing the spark at lower MAP. The optimal ignition curve seems to vary less with load than what we are led to believe is used on traditional air cooled engines with EI or FADECs. We do need to make sure that detonation or pre-ignition never takes place as in a constant, high power application like this, even liquid cooled engines can have piston damage happen pretty quickly. We do have to be concerned with balancing timing with resultant EGT as well so much testing has to be done to make this all automatic.
When done properly, it can be demonstrated that a FADEC/ ECU can do a better job with regards to fuel used per mission than a pilot can with conventional controls with far less monitoring. More time to fly the plane and look for traffic. We see the biggest savings in startup, taxi and climb regimes. In a steady state, cruise condition, a pilot can do a good job at setting up for max fuel economy.
Closed loop control was originally designed for emissions to satisfy the requirements for catalytic converters and secondarily for improving fuel economy and providing a means to correct AFRs long term with engine wear. It never had a performance aspect until wideband sensors arrived and allowed engineers to target specific AFRs vs. hp/rpm/load conditions permitting closed loop operation through the entire operating range- something narrow band sensors do not allow.
The resistance to new technology is interesting in aviation. Certain things are immediately embraced while other things are not. Maybe this has something to do with keeping the fan turning. In Reno Sport Class, convincing pilots to switch over to FADECs has been relatively slow until faced with the realities of the advantages. Once the mechanical systems are beyond their design limits to meter fuel accurately or pilot workload starts to compromise safety, things become more clear. Every year now sees another entry dump the old guard for EFI. While this technology is not for everyone or every engine, it is the logical choice for modern automotive based powerplants. I shake my head at the Unlimited Class Merlin guys who keep burning pistons using carbs and a log manifold. They just do it over and over...
With regards to FMEA, most of us in this world do plenty of thinking about what ifs with regards to system designs and if anything, learning failure modes based on experience is the primary tool in improving the design. So while we might not call it FMEA, we are doing the same sort of things. Conservative, evolutionary approaches result in the best probability of success. There is no substitute IMO for weeding out the bad points of a design or system based on lots of flight time. If we can combine this with better, cheaper analytical tools and models in the design and testing stage, things should improve at a faster rate with less waste and risk. Designing without a solid basis in theory/ engineering often has its negatives.
It is interesting what a professional design team comes up with sometimes though. This afternoon I was looking at a dismantled Thielert clutch coupling which had started to shed parts at 177 hours. One look was enough to make me wonder what these guys were thinking when they chose this design. The same thing has been shown to be unreliable in other aircraft PSRUs a decade ago but these guys did not know that I guess. There have been multiple failures of this part and two power loss accidents that I know of. I hope they get a better design on the new 2.0 engines.
Last edited:
Other fuel control units...
The EC2 controller (by Tracy Crook) that I am using with my Mazda 13B comes with a mixture control knob so the pilot can leave it at the 12 oclock position and let the computer set it where it should, or manually adjust it to a leaner or richer point. The rotary of course doesn't have valves to burn out if running too lean of peak and Tracy reports that increased fuel economy can be had by cruising the rotary in that regime on cross country flights. I'm glad Tracy's unit gives me the option since it is just another thing that makes this "experimental" thing fun and educational.
Doug Lomheim
RV-9A, 13B
OK City, OK
The EC2 controller (by Tracy Crook) that I am using with my Mazda 13B comes with a mixture control knob so the pilot can leave it at the 12 oclock position and let the computer set it where it should, or manually adjust it to a leaner or richer point. The rotary of course doesn't have valves to burn out if running too lean of peak and Tracy reports that increased fuel economy can be had by cruising the rotary in that regime on cross country flights. I'm glad Tracy's unit gives me the option since it is just another thing that makes this "experimental" thing fun and educational.
Doug Lomheim
RV-9A, 13B
OK City, OK
The EC2 controller (by Tracy Crook) that I am using with my Mazda 13B comes with a mixture control knob so the pilot can leave it at the 12 oclock position and let the computer set it where it should, or manually adjust it to a leaner or richer point. The rotary of course doesn't have valves to burn out if running too lean of peak and Tracy reports that increased fuel economy can be had by cruising the rotary in that regime on cross country flights. I'm glad Tracy's unit gives me the option since it is just another thing that makes this "experimental" thing fun and educational.
Doug Lomheim
RV-9A, 13B
OK City, OK
I am also using the EC2 controller and like the mixture knob. I also went to the automotive store and bought a O2 gauge so I can monitor how the engine is running.
gmcjetpilot
Well Known Member
Give me Oh two!
Score! Here is a little article. $30 gauge and $30 sensor and your in business. I guess you have to weld a bung on to screw the sensor into the exhaust. LINK (they have a $270 version) How useful would this be on a Lyc? Well probably not much and what little I know about "narrow band" and wide band O2 sensors this product "wide band" kit would be better but more money. LINK I guess the down side is they last about 150 hours, may be less in 100LL? So I guess that answers my own question. It might be fun to play with the cheap sensors and play around, but welding bungs in your exhaust is not something you want to do casually.
You have to find a place where the probe is at least at 580-680F. If the O2 sensor is too far down stream or its not going to be accurate or sensitive, kind of like a EGT probe. If its hotter it's better for real time close loop input, but if we are just using it to monitor fuel/air ratio 600F is about the min temp to work. Not sure where that would be along the exhaust pipe. If you went with a 4 wire heated O2 sensor it would give more flexibility in locating it and should be more consistent, but those probes cost more, like $80 each and up. However if you got it all set-up, calibrated, it could be a great aid in setting mixture. Would you ignore EGT?
Bottom line it would be more for fun and experimentation than practical. With the short O2 life with leaded fuel and the aforementioned installation issues it's probably more trouble than its worth. Move that red knob and watch the EGT's seems to work pretty well.
You could actually put a O2 sensor on a Lyc, but how long do they last with 100LL. I see they are as cheap as $30. You could make a little meter to read the volts and calibrate it to be meaning full? hummm any lyco drivers do that?
Score! Here is a little article. $30 gauge and $30 sensor and your in business. I guess you have to weld a bung on to screw the sensor into the exhaust. LINK (they have a $270 version) How useful would this be on a Lyc? Well probably not much and what little I know about "narrow band" and wide band O2 sensors this product "wide band" kit would be better but more money. LINK I guess the down side is they last about 150 hours, may be less in 100LL? So I guess that answers my own question. It might be fun to play with the cheap sensors and play around, but welding bungs in your exhaust is not something you want to do casually.
You have to find a place where the probe is at least at 580-680F. If the O2 sensor is too far down stream or its not going to be accurate or sensitive, kind of like a EGT probe. If its hotter it's better for real time close loop input, but if we are just using it to monitor fuel/air ratio 600F is about the min temp to work. Not sure where that would be along the exhaust pipe. If you went with a 4 wire heated O2 sensor it would give more flexibility in locating it and should be more consistent, but those probes cost more, like $80 each and up. However if you got it all set-up, calibrated, it could be a great aid in setting mixture. Would you ignore EGT?
Bottom line it would be more for fun and experimentation than practical. With the short O2 life with leaded fuel and the aforementioned installation issues it's probably more trouble than its worth. Move that red knob and watch the EGT's seems to work pretty well.
Last edited:
<<a dismantled Thielert clutch coupling which had started to shed parts at 177 hours. One look was enough to make me wonder what these guys were thinking when they chose this design. The same thing has been shown to be unreliable in other aircraft PSRUs a decade ago but these guys did not know that I guess. >>
I'll bite! What sort of clutch was it?
I'll bite! What sort of clutch was it?
pierre smith
Well Known Member
Allow me....
.......to simplify things, George. Recently I was flying my buddy's -4 with a Dynon 180. Cruising along, I did what I've been doing for 41 years, leaned his Lyc 160 'til a slight RPM loss and enriched it a tad...1/4" maybe. He was in the back seat and pointed out that in the upper right corner, the Dynon said ROP!!
Some things never change...
Ross...thanks a lot for your very well presented schooling on ECU and related setups without O2 sensors...very, very good reading,
.......to simplify things, George. Recently I was flying my buddy's -4 with a Dynon 180. Cruising along, I did what I've been doing for 41 years, leaned his Lyc 160 'til a slight RPM loss and enriched it a tad...1/4" maybe. He was in the back seat and pointed out that in the upper right corner, the Dynon said ROP!!
Some things never change...
Ross...thanks a lot for your very well presented schooling on ECU and related setups without O2 sensors...very, very good reading,
We like the PLX wideband kits which come with box, harness, sensor and display for under $300. Great company to deal with too. http://www.plxdevices.com/
For non-vendor and automotive use, SDS units come a with mixture knob and a backlit 3 1/8 panel mount programmer/ gauge head. I think it is a good idea to have the knob for the educated pilot but totally understand the vendor's position as well where they get blamed for any problems a pilot might cause.
The Thielert uses a sprung center, clutch disc with sintered metallic lining. Looks very race car like but the springs are nowhere near large/ stiff enough IMO to do anything but bottom out. They load these with some Belleville washers which disintegrate under the pounding of the diesel driven by a very lightweight flywheel. The PSRU was not apart, just off the engine but it looks pretty puny externally for this application. I don't know what, if any damper/ absorber might be inside there- perhaps a quill shaft as there does not look like room for anything else and it is bathed in oil. Cost of the disc is 2000 Euros! My friend had no compliments for these engines nor the support of the factory nor the flat rate times given by the factory to change things like crank sensors (15 minutes including removing the cowling, replacing it, making the log entries and running it up!)
I use this type of thing to illustrate my point that just because something is professionally engineered and certified does not mean it is reliable. The proof is in the long term pudding. Maybe some should not be so hard on Jan's designs. I think the E6/ Gen 3 combos are going to last longer than 177 hours.
For non-vendor and automotive use, SDS units come a with mixture knob and a backlit 3 1/8 panel mount programmer/ gauge head. I think it is a good idea to have the knob for the educated pilot but totally understand the vendor's position as well where they get blamed for any problems a pilot might cause.
The Thielert uses a sprung center, clutch disc with sintered metallic lining. Looks very race car like but the springs are nowhere near large/ stiff enough IMO to do anything but bottom out. They load these with some Belleville washers which disintegrate under the pounding of the diesel driven by a very lightweight flywheel. The PSRU was not apart, just off the engine but it looks pretty puny externally for this application. I don't know what, if any damper/ absorber might be inside there- perhaps a quill shaft as there does not look like room for anything else and it is bathed in oil. Cost of the disc is 2000 Euros! My friend had no compliments for these engines nor the support of the factory nor the flat rate times given by the factory to change things like crank sensors (15 minutes including removing the cowling, replacing it, making the log entries and running it up!)
I use this type of thing to illustrate my point that just because something is professionally engineered and certified does not mean it is reliable. The proof is in the long term pudding. Maybe some should not be so hard on Jan's designs. I think the E6/ Gen 3 combos are going to last longer than 177 hours.
"I use this type of thing to illustrate my point that just because something is professionally engineered and certified does not mean it is reliable. The proof is in the long term pudding. Maybe some should not be so hard on Jan's designs. I think the E6/ Gen 3 combos are going to last longer than 177 hours."
I would agree with this statement except that I would prefer a professionally engineered and certified product that stands the test of time over a product from an undercapitalized company that changes PSRUs and basic engines every few years. Currently I would buy neither the Thielert or the Eggenfeller product although I have been intrigued by the idea of putting a water cooled production engine in an aircraft for more than 30 years. That being said I am not into experimenting with an engine package although I applaud those who like to and wish them well. Thinking something will last more than 177 hours is not a basis on which I would buy an engine. I hope both Theilert and Eggenfeller come out with proven designs that stand the test of time and become the standards. I know it's possible to do but to be done right will take plenty of money. And that is the issue, the aircraft market is probably too small to entice the big players like Honda.
I was hoping Honda would jump into the engine building arena as they are an engine company. Their design made sense and I would have been happy to buy an engine from a company with credentials such as theirs. Alas, it is not to be. Maybe we can convince Subaru to build a dedicated aircraft engine. That too is something I would buy. Otherwise, I guess I am stuck with my Lycoming. In any case, I will read these threads with interest and hope someone will come out with a proven watercooled design that will stand the test of time.
I would agree with this statement except that I would prefer a professionally engineered and certified product that stands the test of time over a product from an undercapitalized company that changes PSRUs and basic engines every few years. Currently I would buy neither the Thielert or the Eggenfeller product although I have been intrigued by the idea of putting a water cooled production engine in an aircraft for more than 30 years. That being said I am not into experimenting with an engine package although I applaud those who like to and wish them well. Thinking something will last more than 177 hours is not a basis on which I would buy an engine. I hope both Theilert and Eggenfeller come out with proven designs that stand the test of time and become the standards. I know it's possible to do but to be done right will take plenty of money. And that is the issue, the aircraft market is probably too small to entice the big players like Honda.
I was hoping Honda would jump into the engine building arena as they are an engine company. Their design made sense and I would have been happy to buy an engine from a company with credentials such as theirs. Alas, it is not to be. Maybe we can convince Subaru to build a dedicated aircraft engine. That too is something I would buy. Otherwise, I guess I am stuck with my Lycoming. In any case, I will read these threads with interest and hope someone will come out with a proven watercooled design that will stand the test of time.
Very interesting reading:
http://www.haverikommissionen.se/virtupload/news/rl2006_08e.pdf
A few personal opinion comments.
The interesting thing about this system is that the clutch is intended to slip at every start and shutdown:
The engine delivers a torque of 245Nm at max take-off power. The clutch is designed to damp overtorque of 35-40Nm, which occur during engine starting and turning-off the engine
Such a design guarantees wear. I haven't done enough other reading to be sure, but apparently operators are required to change the clutch assembly every 300 hours. Note that "300 hours" doesn't take the actual number of starts and stops into consideration. For example, the operator may have conducted 600 start/stop cycles to make flights of 1/2-hour duration, or 75 start/stop cycles to make flights of 4-hour duration. At 300 hours, the first example would evidence 8 times as much clutch wear.....and there is no way to check it without disassembly. The replacement interval should probably be based on number of start/stop cycles, not hours.
The wear rate at each cycle would be related to many factors, not the least of which is the actual slipping torque. The evidence suggests slipping torque is not stable with this design; the clutch is set and checked with a torque wrench at 290 Nm at assembly, but apparently, being a dry clutch, slip torque becomes higher after the first few slip events. That means the hub springs, as noted by Ross, probably get fully compressed (with possible coil bind) during all start/stop events, at least until the clutch wears enough to drop back to a lower slip value. The gearbox also see 70% more torque:
Following this and the event in question, the properties of the clutch
plate have been studied by the engine manufacturer. These studies, which
were carried out together with the manufacturer of the clutch plate, have
shown that the slipping torque of the clutch, which is set at approximately
290 Nm at assembly, increases to approximately 410 Nm during the first 50
or so operating cycles. This is explained by the formation of a surface layer
on the friction coating of the plate that acts to increase the coefficient of
friction. The recycled clutch plates that slipped had all been cleaned before
delivery by the engine manufacturer and this surface layer had been washed
off. Tests made with cleaned plates show that after cleaning, the slipping
torque increased to as much as approximately 390 Nm and that after two
cleanings to as much as 340 Nm after 120 cycles but with a fairly stable
value of approximately 300 Nm between 25 and 110 cycles.
Eventually slip torque drops back to a lower value; the clamp force provided by the Belleville washers reduces with clutch plate wear...and the plate does wear:
SHK was present at a planned clutch change on another aircraft and
there noted that the thickness of the new clutch plate was approximately
0.18 mm greater than the thickness of the plate being replaced, which had
been in service for about 300 hours.
I'm not impressed with the design overall because of the above, but to ice the cake, this is a system with no limp-home mode at all. When it starts to slip (as in this accident) it self-destructs; the Bellevilles can't possible provide enough pressure plate travel to keep up with runaway clutch wear.
Quite a few "homebuilders" have designed better systems than this one, including Jan.
http://www.haverikommissionen.se/virtupload/news/rl2006_08e.pdf
A few personal opinion comments.
The interesting thing about this system is that the clutch is intended to slip at every start and shutdown:
The engine delivers a torque of 245Nm at max take-off power. The clutch is designed to damp overtorque of 35-40Nm, which occur during engine starting and turning-off the engine
Such a design guarantees wear. I haven't done enough other reading to be sure, but apparently operators are required to change the clutch assembly every 300 hours. Note that "300 hours" doesn't take the actual number of starts and stops into consideration. For example, the operator may have conducted 600 start/stop cycles to make flights of 1/2-hour duration, or 75 start/stop cycles to make flights of 4-hour duration. At 300 hours, the first example would evidence 8 times as much clutch wear.....and there is no way to check it without disassembly. The replacement interval should probably be based on number of start/stop cycles, not hours.
The wear rate at each cycle would be related to many factors, not the least of which is the actual slipping torque. The evidence suggests slipping torque is not stable with this design; the clutch is set and checked with a torque wrench at 290 Nm at assembly, but apparently, being a dry clutch, slip torque becomes higher after the first few slip events. That means the hub springs, as noted by Ross, probably get fully compressed (with possible coil bind) during all start/stop events, at least until the clutch wears enough to drop back to a lower slip value. The gearbox also see 70% more torque:
Following this and the event in question, the properties of the clutch
plate have been studied by the engine manufacturer. These studies, which
were carried out together with the manufacturer of the clutch plate, have
shown that the slipping torque of the clutch, which is set at approximately
290 Nm at assembly, increases to approximately 410 Nm during the first 50
or so operating cycles. This is explained by the formation of a surface layer
on the friction coating of the plate that acts to increase the coefficient of
friction. The recycled clutch plates that slipped had all been cleaned before
delivery by the engine manufacturer and this surface layer had been washed
off. Tests made with cleaned plates show that after cleaning, the slipping
torque increased to as much as approximately 390 Nm and that after two
cleanings to as much as 340 Nm after 120 cycles but with a fairly stable
value of approximately 300 Nm between 25 and 110 cycles.
Eventually slip torque drops back to a lower value; the clamp force provided by the Belleville washers reduces with clutch plate wear...and the plate does wear:
SHK was present at a planned clutch change on another aircraft and
there noted that the thickness of the new clutch plate was approximately
0.18 mm greater than the thickness of the plate being replaced, which had
been in service for about 300 hours.
I'm not impressed with the design overall because of the above, but to ice the cake, this is a system with no limp-home mode at all. When it starts to slip (as in this accident) it self-destructs; the Bellevilles can't possible provide enough pressure plate travel to keep up with runaway clutch wear.
Quite a few "homebuilders" have designed better systems than this one, including Jan.
gmcjetpilot
Well Known Member
Flying the dinosaures
, you're right, thanks. O2 silly idea, K.I.S.S. Heck young whipper snappers are shocked that we once flew with out even one CHT / EGT gauge, in the "old days". If you don't have a 20 channel engine monitor today
you are a flying ........
Dang Pierre, when ever I have my flights of fancy, you bring me back to earth. Good thing,
, you're right, thanks. O2 silly idea, K.I.S.S. Heck young whipper snappers are shocked that we once flew with out even one CHT / EGT gauge, in the "old days". If you don't have a 20 channel engine monitor today you are a flying ........
Last edited:
Yes, there have been at least 3 accidents with the Thielert involving clutch failure although one was claimed to be incorrectly assembled by a field repair station.
I don't consider a 177 or 300 hour life acceptable on a major drive component like this in a certified system. What is the point of the TBR when the PSRU has to come off 5-8 times in that period?
Yes, it is a weak design certainly. It comes down to not enough real world testing IMO. Flog the thing at full power for 2000 hours in flight and do 3000 start cycles on several examples. When they all pass, with no failures-then release it for sale maybe.
I'm no fan of clutches on PSRUs- they make little sense to me unless you are trying to protect the crankshaft against prop strike damage or take the system past some low rpm TV range. There are better, lighter and more reliable ways to do that without the liability of all these other parts in there.
The engineered/ tested solution is the best- if you can afford the end product. I'll take the tested part if I can only choose one.
I don't consider a 177 or 300 hour life acceptable on a major drive component like this in a certified system. What is the point of the TBR when the PSRU has to come off 5-8 times in that period?
Yes, it is a weak design certainly. It comes down to not enough real world testing IMO. Flog the thing at full power for 2000 hours in flight and do 3000 start cycles on several examples. When they all pass, with no failures-then release it for sale maybe.
I'm no fan of clutches on PSRUs- they make little sense to me unless you are trying to protect the crankshaft against prop strike damage or take the system past some low rpm TV range. There are better, lighter and more reliable ways to do that without the liability of all these other parts in there.
The engineered/ tested solution is the best- if you can afford the end product. I'll take the tested part if I can only choose one.
<<I'm no fan of clutches on PSRUs- they make little sense to me unless you are trying to protect the crankshaft against prop strike damage or take the system past some low rpm TV range.>>
I'd bet money a 912-style multi-plate wet clutch has a far more consistant friction setting, and would stay in the design range far longer. And there is a pretty good clutch system in the Raven drive for the little Suzukis.
But, assume Thielert didn't want to incorporate a wet clutch into their gearbox. Then why cobble up a coupling? The really crazy thing about this deal is ignoring the endless variety of commercial soft couplers made specifically for diesel applications....most of which are fail safe.
I'd bet money a 912-style multi-plate wet clutch has a far more consistant friction setting, and would stay in the design range far longer. And there is a pretty good clutch system in the Raven drive for the little Suzukis.
But, assume Thielert didn't want to incorporate a wet clutch into their gearbox. Then why cobble up a coupling? The really crazy thing about this deal is ignoring the endless variety of commercial soft couplers made specifically for diesel applications....most of which are fail safe.
Yup, I was thinking the same thing. Maybe the "not invented here" thing is too strong to fight but using low cost, proven, off the shelf solutions just seems like the first avenue to investigate before you try to redesign the wheel- badly.
Just because one of 1500 units only lasted 177 h hardly makes it a statistically significant event. Take 1500 brand new Lycomings and see how many of them have some failure (overheating, piston, ignition and so on) before 200 h. With a TBO of only 1800 h, the probability of at least one failure before 1800 h of at least one unit is as good as 100%. Also remember that the 1.7 has accumulated 350,000 h of operation, 230 h on average for each unit, and most of that from 2006. The certification process and constant feedback from the field means it will only get better as time passes.
The main thing is what the Thielert offer in reduced cost. It is an alternative that cuts the fuel cost by more than 50% (more in the order of 70%). A handfull of broken gear boxes out of 1500 units is more of academic interest compared with that. There is no other engine that can do this in the 150 hp range. The Thielert (certified Jet fueled piston engine) is one of the most important achievements that has happened to GA in the last 50 years, and will set the direction for all engine development in the next 30-40 years.
The only thing the Egg has to offer is in the subjective category; a "smooth ride" and to be different. Other than that The Egg offers nothing that cannot already be obtained better by the 50 year old tested and proven design of the Lycoming, not even in fuel consumption, efficiency or overall economic benefits. It is not without reason the Lycomings have ruled the last 50 years, it was, and still is, an excellent design. It has proven to be so good, that the only thing that will eventually make it obsolete is a transition to turbocharged diesels. This is happening now because it is only within the last 10 years that the diesel engine technology for small engines has been developed sufficiently.
This is the reality (as I see it at least
) The Egg is probably a good solution for those that want something different in their experimental aircraft. But to compare it with the Thielert is wrong, it will be an apple vs orange thing because it has none of the economical benefits the Thielert has, and the Thielert is just as smooth as the Egg.Personally I would like to have a 2 stroke direct drive diesel, maybe Wilksch at Continental will eventually make something?
<<The main thing is what the Thielert offer in reduced cost. It is an alternative that cuts the fuel cost by more than 50% (more in the order of 70%). >>
Absolutely, and we all want lower fuel costs.
<<The only thing the Egg has to offer is in the subjective category; a "smooth ride" and to be different.>>
Well, to be fair, given an argument based on economics we must consider purchase and maintenance cost.
The base 2.0 Thielert is currently priced at 30,085 Euros ($45,690 US) without accessories, or 40,470 Euros ($61,461 US) as a firewall forward for the C-172 and other certified installations. Obviously an Egg package is a lot less money. With the diesel you pay now (purchase cost). With the Egg you pay later (fuel cost) and get 40 to 60 more HP. In addition the certified installation seems to require pulling the propeller and gearbox every 300 hours for a clutch change, at official service center rates.
Europe's avgas prices help the economic argument, but at this time the advantage seems slim in the USA.
Let's set economics aside and return to mechanical.
I don't think anyone is debating the reliability of the base engines, Subaru or Thielert. The issues are the drive systems and detail design choices. Dr. Svingen, what do you think about the clutch and coupler assembly on the Thielert?
Absolutely, and we all want lower fuel costs.
<<The only thing the Egg has to offer is in the subjective category; a "smooth ride" and to be different.>>
Well, to be fair, given an argument based on economics we must consider purchase and maintenance cost.
The base 2.0 Thielert is currently priced at 30,085 Euros ($45,690 US) without accessories, or 40,470 Euros ($61,461 US) as a firewall forward for the C-172 and other certified installations. Obviously an Egg package is a lot less money. With the diesel you pay now (purchase cost). With the Egg you pay later (fuel cost) and get 40 to 60 more HP. In addition the certified installation seems to require pulling the propeller and gearbox every 300 hours for a clutch change, at official service center rates.
Europe's avgas prices help the economic argument, but at this time the advantage seems slim in the USA.
Let's set economics aside and return to mechanical.
I don't think anyone is debating the reliability of the base engines, Subaru or Thielert. The issues are the drive systems and detail design choices. Dr. Svingen, what do you think about the clutch and coupler assembly on the Thielert?
Yes, I do question the core Thielert reliability. If you go back in this thread, I have updated some of my posts with real world reports for people flying these engines, working on them and the Aviation Consumer article in Dec. 2007. My friend works on these things and listed the many problems they have seen to date on this one Twinstar in a short period of time.
Piston/ring problems lead the list causing high oil consumption and premature removal WAY before 1000 hours in almost every case.
Supporting system problems with injectors, ECUs, alternators and clutches are common.
Reduced overall cost and high utilization is the last thing these engines offer at this stage of development according to the people flying and maintaining them.
I think you will find that the majority of Lycoming O-320/360 engines get to 200 hours with zero to few major problems and probably much longer than that. It is a very mature system and still the benchmark other engine packages are judged by. I think you can agree that clutch replacement at 300 hours is not acceptable for aircraft flying frequently.
I didn't want to stray this far from topic but if we find the certified Thielert (EFI, PSRU, automotive based) offering acceptable performance and reliability then Jan must be doing a really good job with his uncertified packages.
Piston/ring problems lead the list causing high oil consumption and premature removal WAY before 1000 hours in almost every case.
Supporting system problems with injectors, ECUs, alternators and clutches are common.
Reduced overall cost and high utilization is the last thing these engines offer at this stage of development according to the people flying and maintaining them.
I think you will find that the majority of Lycoming O-320/360 engines get to 200 hours with zero to few major problems and probably much longer than that. It is a very mature system and still the benchmark other engine packages are judged by. I think you can agree that clutch replacement at 300 hours is not acceptable for aircraft flying frequently.
I didn't want to stray this far from topic but if we find the certified Thielert (EFI, PSRU, automotive based) offering acceptable performance and reliability then Jan must be doing a really good job with his uncertified packages.
Last edited:
janeggenfellner
Well Known Member
We have hundreds of engines flying and have changed the production gearbox design one time.
Jan Eggenfellner
janeggenfellner
Well Known Member
Thank you
Jan
- Status
