ROTAX Engines - Quirky or some bad design choices?

[FastEddieB]

In another thread I mentioned I might start this one, so I am.

 

As background, I’ve owned 6 aircraft with Continental and Lycoming engines, and these are not without their quirks and foibles and design compromises as well. I do not in any way mean to imply the the ROTAX engine is any worse - or better, necessarily - than these overall in terms of design or reliability.

 

When asked my take about the ROTAX, I tend to say that in my experience so far, they are neither more nor less reliable overall than the more familiar Continental and Lycoming engines many pilots are used to, but they do have certain “quirks” you have to be familiar with.

 

I thought I’d list some things that I see as design defects, or at least compromises or things that could be improved upon. As was pointed out in the other thread, I am not an engineer, but I think end users can have insight into the success of certain design choices, or lack thereof.

 

My list:

 

1) If I was designing an aircraft engine, and the engineers came back with a  design where turning the engine over backwards could cause huge problems, I think I’d send them back to the drawing board. If the problem was sucking air into the system, I’d ask them to consider check valves or accumulators or reservoirs or something, such that a prop could safely be turned backwards. To require a complicated purge with somewhat specialized equipment to put the airplane back in service after such an innocuous action just seems like poor design to me.

 

2) When my ignition modules bit the dust, to confirm no spark I used the tried and true method of grounding the plug body with the plug removed to see if there was spark. I did not see one, and Roger told me that not only was the spark very difficult to see, but warned me that if the plug ungrounded in the process the ignition module could be damaged.

 

Assuming that that is a real possibility, would not the same damage to the ignition module be possible any time a plug wire blame disconnected? Or if corrosion compromised the connection at the plug? Again, I think this goes beyond a quirk - an ignition system should not self-destruct when a plug wire disconnects.

 

3) I have no experience with bad sprag clutches, but it appears the problem crops up when a ROTAX “kicks back”, and they have a tendency to “kick back” with at least some regularity if everything is not “just so” or if the pilot holds the starter engaged for not just the right amount of time. I’d want my engineers to come up with a solution that could handle that sort of abuse and not fall apart from kickback.

 

4) ROTAX sourced their carbs from BING, who designed them originally for BMW motorcycles. In that setting, debris was already known to be an issue, sometimes holding the needle valve open and causing a too-rich mixture or an overflowing float bowl, or sometimes making it past the needle valve only to get sucked up in the main jet. The problem was known for decades, and many BMW owners resorted to inline filters right at the carb to catch any crud making its way through the fuel system.

 

I think a proper inline filter, complete with a bypass mechanism, right at the carb could avoid many of the problems owners have been faced with, some leading to accidents. Instead, they provide no filtration to the incoming fuel at all, relying on the airframe designers to take care of it. We’re repeatedly seen the repercussions of that design decision.

 

Anyway, its been stated by Roger and others that the problems lie mainly with the owners. That reminds me of Steve Jobs reaction to “antennagate”, saying iPhone users were “holding it wrong”. I think good design anticipates less-than-perfect operation by owners.

 

Open to any feedback, or other examples of things you think ROTAX might improve upon.

[Ed Cesnalis]

Carbed or injected, 912's resort to full rich above 92% throttle and force you to choose between cruising rich or cruising at a low power setting when at high altitudes.

[Anticept]

1) Irritating, but maybe that's the best they could come up with. More parts means more points of failure.
 

2) Actually, this is an issue with nearly all ignition systems, magnetos included. We're talking about tens of thousands of volts. It's unbelievably difficult to insulate against that.

 

Check out this 30kv wire! This is the size of the leads that go from power poles to industrial complexes!

 

Here's a 10kv wire too!

 

In mags, the spark can jump across the distributor cap, leaving carbon tracks and putting stress on the coils and condenser, as it will build the flux in the coils to a far higher potential than they are designed to handle. It might not self destruct the first time, or the 100th time, but it can.

 

3) I have no experience with this. In fact, I've never seen a kickback. Maybe it's possible in the older engines, but not nearly as easy in the ones with the new flywheel due to the ATDC timing.

 

In addition, rotax is not unique to this. A kickback in a lycoming or continental can damage the vacuum pump. It can shear the starter's safety shaft too.

 

4) Not entirely correct. The fuel pumps have a very fine screen, which is your "last chance" filter. And yes, it is the engine AND airframe manufacturer's responsibility. And no, it's not unique to bing. Debris can clog any other carb too. We just don't see it happen with GA aircraft because fuel gets filtered so many times before it goes in the aircraft.

 

Fuel that comes from gas stations is NASTY. We filter our fuel. There's always debris. If anything, there should be a sump in flight design's wing tanks. There's crud that sits in the corner sometimes. Cars have the fuel filter that they do because of how nasty that gas station fuel is.

 

 

 

CharlieTango: I don't know why you're stuck on this, but it's not fixed at 92% throttle. There is absolutely NO linkage whatsoever between throttle position and the jet needle. It's based solely on mass air flow whoops, force of habit, I meant volumetric air flow. Maybe at sea level, it goes full rich at around 92% throttle. I don't know exactly when, or if, it will even go full rich for you at high altitude.
 

[FlyingMonkey]

Rotax does everything the right way, every other manufacturer that differs in any way has gotten it wrong.  Seems obvious.

 

 

 

 

[FlyingMonkey]

Since you might not believe my previous post, I'll address your concerns:

 

1)  Why do you *need* to be able to turn an engine backwards?  I can't think of any reason to do so, other than "I might want to".  It never occurred to me to do so until somebody mentioned to me that you can do it on a TCM or Lycoming engine without harm.  Well, cool.  Unnecessary, but cool.  Seems like more a training and familiarity issue than a design defect.

 

2)  Don't know enough about this to comment, but I don't think the module will fry if a plug wire is off.  If it does, that is a problem.

 

3)  To my understanding, a kickback is almost always a result of insufficient voltage.  This can be checked on engine instrumentation before attempting a start.  I always verify that the voltage is good before turning the key.  I think I have had one kickback since I've owned my CT, when I stupidly tried to start an engine at around 40°F without pre-heating first.  A lot of kickbacks could potentially be solved with a circuit that disallows starting with insufficient voltage.  But that could cause an issue in the air with restart if the engine dies and your battery is marginal...

 

4)  You have more experience with Bing carbs than I do, but just wondering...how much fuel filtration do you need?  I have a fuel filter inline that filters fuel from both tanks.  Would additional filters before each carb introduce better reliability, or just more points of failure?  I have never had an issue with junk clogging my carbs, and every time the bowls have been dropped they have had no debris in them, but I don't know how many others might have had issues.  

[Ed Cesnalis]

CharlieTango: I don't know why you're stuck on this, but it's not fixed at 92% throttle. There is absolutely NO linkage whatsoever between throttle position and the jet needle. It's based solely on mass air flow. Maybe at sea level, it goes full rich at around 92% throttle. I don't know exactly when, or if, it will even go full rich for you at high altitude.

 

 

If you don't know why are you correcting me?  Are you saying that the needle jet circuit cannot relinquish mixture control to both the idle circuit and the main jet circuit because of lack of linkage?

[FlyingMonkey]

Carbed or injected, 912's resort to full rich above 92% throttle and force you to choose between cruising rich or cruising at a low power setting when at high altitudes.

 

That does seem to be an issue, or at least an annoyance, especially for you high-altitude guys.  It seems if the mixture is set right it should prevent detonation at all RPMs.  Maybe the circuit could go progressively *slightly* more rich at high throttle, to balance engine safety with performance a bit better than the current setup.  Or have an alternate "mixture bypass" something like a HACMan that a pilot could use to take over manual mixture control.  Of course added complexity means more potential failures.

[Anticept]

If you don't know why are you correcting me? Are you saying that the needle jet circuit cannot relinquish mixture control to both the idle circuit and the main jet circuit because of lack of linkage?

I'm saying it's inaccurate. It does have a full rich setting that is limited by the main jet, however it's based on mass air flow whoops, force of habit, it's volumetric air flow, indirectly related to throttle setting. I may not know exactly when it goes full rich, but I do know it's not based on throttle position alone. The carb uses a calibrated spring, which is counterbalanced by a vacuum driven by MAF and a chamber which uses ambient air pressure.

[FlyingMonkey]

I'm saying it's inaccurate. It does have a full rich setting that is limited by the main jet, however it's based on mass air flow, indirectly related to throttle setting. I may not know exactly when it goes full rich, but I do know it's not based on throttle position alone. The carb uses a calibrated spring, which is counterbalanced by a vacuum driven by MAF and a chamber which uses ambient air pressure.

 

Sounds like at some critical altitude there would be insufficient air mass to force it full rich?

[Tom Baker]

I too have owned 10 different aircraft, and operated 8 of them. I have been a aircraft mechanic for about 35 years, and have worked on 100's of different airplanes. With the exception of the carburetor set up Rotax issues are not that much different than others.

 

1. While I don't know of any others that have oil issues with turning the engine backwards, there are other issues. You can damage the starter drive on certain Continental engines. Any engine with a dry air pump (Vac Pump). The pump can be damaged by turning the engine backwards. It will shatter or crack the carbon vanes in the pump causing a failure. Think JFK Jr. type accident.

 

2. When a ignition system tries to fire a spark plug and the lead is disconnected all that energy has to go somewhere. While on the Rotax ignition it damages the internal electronics of the little black box it also will damage magnetos. With a magneto the spark will likely arc across to a different spark plug. There have been cylinders depart engines, because of this arcing from have a wire loose. A more recent problem has been a large number of Champion aviation sparkplugs that have had resistors fail. The engine will start and run OK on the ground, but when at altitude it will misfire causing the same type of damage.

 

3. On this past New Years Day I received a call from a friend who happened to be at a different airport than mine. There was a fellow stranded there with a engine that wouldn't start. It was a Cessna 182 with a Continental IO-550 engine conversion. It had kicked back breaking the starter drive. This is not an uncommon problem with the big Continental engines. It is not a easy problem to fix, because to replace the adapter the engine must be disconnected and the engine removed from its mounts.

[Anticept]

Sounds like at some critical altitude there would be insufficient air mass to force it full rich?

I believe so, yes. To create such conditions on the bench will take sone macgyvering, but I intend to do it someday.

 

EDIT: force of habit. I said mass air flow when I'm supposed to be talking about volumetric air flow. Still, the carb design takes altitude into account, and there is an altitude where you likely can't get full rich.

[gbigs]

1) turning the engine over backwards could cause huge problems,

 

2) When my ignition modules bit the dust, to confirm no spark

 

3) I have no experience with bad sprag clutches,

 

4) ROTAX sourced their carbs from BING,

 

 

Open to any feedback, or other examples of things you think ROTAX might improve upon.

 

You are forgetting how easy it is to shock cool the other engines.  How hard starting they are when hot.  How complicated they are in ascent and descent (ROP vs LOP).  How much more fuel they burn (and have to burn the more expensive 100LL).  How much more they are to maintain.  And how heavy they are....  versus the high-tech fuel-injected 912iS.

 

My last trip to Woodland Aviation (Cirrus Repair) I saw table with a new Continental TSIO550k spread all over it with 30 hours on the engine.  Two cracked cylinders and a cracked case - why?  Shock cooling mishap.

[FastEddieB]

Many esteemed experts hold that shock cooling is a myth, or at least way overstated.

 

Two such:

 

http://www.avweb.com/news/pelican/182107-1.html

 

http://www.avweb.com/news/maint/182883-1.html

 

The point of the thread was not to compare Lycoming, Continental and ROTAX engines, as I thought I made clear early on.

[Ed Cesnalis]

I'm saying it's inaccurate. It does have a full rich setting that is limited by the main jet, however it's based on mass air flow, indirectly related to throttle setting. I may not know exactly when it goes full rich, but I do know it's not based on throttle position alone. The carb uses a calibrated spring, which is counterbalanced by a vacuum driven by MAF and a chamber which uses ambient air pressure.

 

The full rich setting is not limited by the main jet it is determined by the main jet.

 

What do you mean its based on mass air flow?  We have no sensor so how would the system base full rich on something it doesn't know?  If it did know the MAF how would it act on it to achieve full rich?

 

Answer is full rich is achieved by taking the needle circuit out of the equation and then the limiting orifice is the main jet.  To adjust the full rich setting you would need to replace the main jet.

 

You continue to say you don't know but I am wrong, can you come up with something more concrete?

 

I say you are correct in that there are 2 systems in play, the spring vs the diaphragm, the spring leans as throttle advances and the diaphragm leans as density altitude increases but this stops being true at 92% throttle  (both of these leaning systems are dependant on the needle circuit and only work at partial throttle). At this point, by design the diaphragm is out of the picture and so is the needle circuit, all you have is the main jet providing full rich condition.

 

If you are going to rest on your position that leaning is provided by MAF you are going to have to explain how it works.

[Ed Cesnalis]

Sounds like at some critical altitude there would be insufficient air mass to force it full rich?

 

 

I believe so, yes. To create such conditions on the bench will take sone macgyvering, but I intend to do it someday.

 

You guys are out in the weeds. Full rich means you are not limiting the mixture beyond what the main jet provides.

 

Rotax lacks high altitude leaning, other aircraft engines provide it, the only critical altitude is your absolute ceiling.

 

Mixture can be consistent as mass varies.

[FastEddieB]

This gets complicated, but...

 

When I climbed my Sky Arrow to 13,500', it gave no indication of running too rich.

 

I think this may have something to do with less atmospheric pressure "pressing down" on the fuel in the float bowl, automatically causing less fuel flow to the main jet and hence a leaner mixture at altitude.

 

Or not - the inner workings of these carbs do seem pretty mysterious at times.

[FlyingMonkey]

You guys are out in the weeds.

 

Because I asked a question?  I am in no way a carb expert, and I'm trying to understand how it all works.  You and Corey are saying opposite things, and if I don't ask any questions I can't separate fact from BS.

[FlyingMonkey]

 

I think this may have something to do with less atmospheric pressure "pressing down" on the fuel in the float bowl, automatically causing less fuel flow to the main jet and hence a leaner mixture at altitude.

 

Or not - the inner workings of these carbs do seem pretty mysterious at times.

 

That would be the only way you could meter anything without a MAF sensor...by air pressure (or vacuum).

[Ed Cesnalis]

This gets complicated, but...

 

When I climbed my Sky Arrow to 13,500', it gave no indication of running too rich.

 

I think this may have something to do with less atmospheric pressure "pressing down" on the fuel in the float bowl, automatically causing less fuel flow to the main jet and hence a leaner mixture at altitude.

 

Or not - the inner workings of these carbs do seem pretty mysterious at times.

 

I don't think atmospheric pressure presses 'down' but applies pressure in all directions.  If this theory worked it would work in Lycomings and such yet they all have high altitude mixture control that are heavily used when cruising above 7.5K.

[Ed Cesnalis]

Because I asked a question?  I am in no way a carb expert, and I'm trying to understand how it all works.  You and Corey are saying opposite things, and if I don't ask any questions I can't separate fact from BS.

 

 

Because you presented a potential conclusion (sounds like ...) that confuses the relationship between mass and mixture.

 

If you ran out of critical mass of air there would be insufficient O2 for combustion.  As long as you have mass you can maintain a mixture that is in range.

[Flying Bozo]

Regarding the lack of a filter going into the Carb.. my 1959 Chevy pickup truck has a filter on the inlet to the carb..that was 57 years ago. If my Rotax had a filter in that place on each carb then I would not have had 4 emergencies where half the engine quit on climb out and I had to make an emergency return to the airport. Luckily when I reduced from full throttle the piece of crap in the float bowl was able to drop down to the bottom of the bowl and the engine ran OK at about 4000 RPM.

The piece of rubber hose coming out of the fuel pump going to the spreader tube is as important as the holy grail. If that is flaking any rubber off of the inside, sadly it ends up in the float bowl.

Larry

[Ed Cesnalis]

Injected engines use MAF sensors to caclulate how much fuel to inject to obtain the desired mixture.  It might us a MAP and rpm to calculate MAF but MAF is a metric used for injection.

 

Carbs use jets, the main jet provides constant mixture while the needle jet provides variable mixture.  The idle circuit provides another mixture and the enricher adjust mixtures.

 

Arguing that the MAF controls mixture in the carbs seems to be confusing carbed and injected engines.

 

The 92% point makes some sense in the carbed version even if it leaves no high altitude leaning but using the same point to go full lean when there is an injection system seems like Rotax is married to their full rich limitation.  Why not lean for high altitude in at least the injected system?

[Roger Lee]

You guys are going to fry some brain cells.

 

 

How did you guys get from quirky bad design Rotax parts to fuel injection versus carb fuel flows?

[Anticept]

The full rich setting is not limited by the main jet it is determined by the main jet.

 

What do you mean its based on mass air flow? We have no sensor so how would the system base full rich on something it doesn't know? If it did know the MAF how would it act on it to achieve full rich?

 

Answer is full rich is achieved by taking the needle circuit out of the equation and then the limiting orifice is the main jet. To adjust the full rich setting you would need to replace the main jet.

 

You continue to say you don't know but I am wrong, can you come up with something more concrete?

 

I say you are correct in that there are 2 systems in play, the spring vs the diaphragm, the spring leans as throttle advances and the diaphragm leans as density altitude increases but this stops being true at 92% throttle (both of these leaning systems are dependant on the needle circuit and only work at partial throttle). At this point, by design the diaphragm is out of the picture and so is the needle circuit, all you have is the main jet providing full rich condition.

 

If you are going to rest on your position that leaning is provided by MAF you are going to have to explain how it works.

Not sure what "limited" vs "determined" is supposed to mean, but we're saying the same thing. Changing the main jet changes how much fuel is metered in the full rich position.

 

As for mass air flow: I made the mistake of calling it mass air flow, it should be volumetric air flow. It's so easy to get in the habit of calling air flow "mass air flow" since it's so common in cars now.

 

However, even volumetric air flow is a little skewed in this case. These carbs are a bit more advanced than that. There is an upper and lower chamber, with the diaphragm in the middle. The lower chamber uses static air pressure, and the upper chamber uses vacuum generated by the flow of air through the venturi. I say it's a little skewed because we're using a calibrated spring to push the piston and needle down, but also that spring pressure has to fight with static air pressure in the lower chamber too, which is quite a step up from simple volumetric air flow sensors used in cars; it's self adjusting for altitude!

 

At higher altitude, there is a lower static pressure in the lower chamber. It takes more vacuum to lift the piston against the spring now. In addition, there is a lower pressure flowing into the system, so when the air goes through the venturi, it's going to have a lower velocity than you would have at sea level. Lower velocity means lower vacuum generated, and thus more difficult to raise the piston to full rich.

 

Anyways, you're asking me for more concrete evidence. Technically, the burden of proof is on you to justify that 92% claim. I've never seen this "92%" figure for the 912 carbeurated engines in any official references. I think you are mixing that up with the full rich setting for the 912iS, which IS fixed at 92%.

[FlyingMonkey]

Maybe the quirks are what makes parts so danged expensive!