Jabiru engine failures

[Jabawocky]

Mach E, this is not really thread drift, it is part of the core of the issues.


I should not have flicked back a page, but talk about funny.

I asked this,

if you have an engine at peak EGT and you make it richer, which by your example means it will get cooler, why then does it get hotter (CHT) when the EGT is falling? And once past about 50 or so dF CHT will drop again.

And shortly there after our resident expert on all things from wind farms to law and valve failures responds with,

Why dose cht rise. Because the energy is being released into the Cylinder chamber and not out the exhaust. Hence that's why you don't get burnt exhaust valves when running rop. But then again what would I know !!!

Let me repeat this again, if at Peak EGT (where many have in the past recommended you run, and then richen it up a bit like yr right wants you to do, why does the CHT go up while the EGT gets lower?

I will offer a tip here, despite what yr right thinks, everything is actually 100% opposite to what he posted above. Including the valves burning from LOP ops. The reason the do that is either poor fit or poor valve guides. Running at high CHT's accelerates the problem. LOP ops can't do it any more than the other in fact the logic of it all would suggest less so due less pressure and temperature.

So please explain where Pratt & Whitney, Curtis Wright, (TCM and Lycoming also) and a whole bunch of others got it wrong.

[andrewr]

I’m just fascinated to know what it is you propose to actually do about your concerns, and how you will be able to work out whether your concerns have actually been addressed. Presumably as some stage you’ll be sitting in a cockpit in proximity to a reciprocating engine, with some knobs to play with and some dials to watch.

Good question. Where would you get operating procedures for one of these engines, if you didn't know it all already? Ideally, it would be from someone who had been intimately involved in the designing and building of the engines. Someone involved in matching and testing engines for particular airframes, and who receives feedback and statistics about any problems and failures. Where would you find that sort of information?

No, I am not a subscriber to the John Deakin view that Lycoming know nothing about operation of their engine.

However

how you will be able to work out whether your concerns have actually been addressed

This is an excellent question for CASA on the original topic i.e. Jabiru. The stats quoted are a failure rate of about 1 in 1500 hours. This could be a problem in evaluating any fix. If there isn't an obvious cause of the problem, at that failure rate it might take 20-30 engines and tens of thousands of hours to concude whether a fix is effective. (Need a statistician to work out the real numbers.)

That assumes that the stats don't show an obvious cause where you could go back to the old configuration e.g. hydraulic lifters, as some have suggested.

[Creampuff]

Where would you get operating procedures for one of these engines, if you didn't know it all already? Ideally, it would be from someone who had been intimately involved in the designing and building of the engines. Someone involved in matching and testing engines for particular airframes, and who receives feedback and statistics about any problems and failures. Where would you find that sort of information?

I see your point.

The millions of operating hours of data collected about the effect of mixture on EGT and CHT only apply to these piston engines:
- Briggs and Stratton
- Honda
- Jacobs
- Ford
- Chrysler
- Franklin
- Pratt and Whitney
- Wright Aircraft
- General Motors
- Harley-Davidson
- Lycoming
- Continental.

The laws of physics don’t apply to Jabiru engines. Best to get yr right to sort out the problem.

[andrewr]

For Jabiru engines typically the only control the pilot has is the throttle.

It is also possible to have an endless debate about Jabiru throttle settings - not too low etc. And for Rotax 912(S) there is the "avoid cruising below 5000 rpm" camp...

[Creampuff]

If I wanted to figure out what’s breaking a particular kind of piston aero engine (other than purely random failures), I’d set one up so that I could be sure that all cylinders are operating between about 40 and 50 degrees rich of peak on the lean curve, at high power and a relatively low cruise RPM.

That way I’d be giving the engine the toughest pounding that I could possibly give it. (Ssshhhhhh: Don’t tell yr right, but that’s ROP….)

I’d then apply time and observation.

[rutan around]

Andrewr

How does lambda at peak EGT compare to the lambda for maximum power? Maximum efficiency?

I don't know why you're pissing around with the fuel air equivalence ratio. You're dealing with the fuel pumped into your aircraft so fuel air ratio is good enough. ie the mixture

1 Maximum efficiency occurs when all the fuel is burned and some excess oxygen is in the exhaust gas. ie a lop mixture. 50* - 80* is good to ensure absolutely all the fuel is burnt.

2 Peak EGT all the fuel and all the oxygen is used up ( theoretically )

3 Maximum power occurs about 40* rich of peak.To ensure all the oxygen is burnt a small excess of fuel is used (slightly rich mixture ) The unburnt fuel goes down the exhaust. Efficiency is fairly good but not as good as at peak and less again than lop. 40* rop is very hard on the engine.

40* rop gives the fastest mixture burn (due to a shorter ignition latency period after the spark event ) This means all the mixture is burnt sooner than than is good for the engine. Sometimes the burn is complete at or just after top dead centre and sometimes even worse before TDC. This means maximum pressure occurs when the swept volume of the cylinder is smallest and the mechanical advantage (ability to impart rotational energy) is zero or actually negative.

This is very hard on the mechanical parts of the engine and it creates a lot of heat which you read on the CHT gauge. Imagine riding a bicycle and applying maximum pressure to the pedal at the top or just before the start of the down stroke. You may push as much as you like but nothing much happens till you obtain quite a few degrees of mechanical advantage.

The ignition timing on an engine is set so that maximum pressure in the cylinder occurs 5* to 15* after top dead centre. As it is a fixed ignition system a compromise has to be made and so it is set for normal mixtures.ie 40* LOP or more or 100* + ROP

[andrewr]

If I wanted to figure out what’s breaking a particular kind of piston aero engine (other than purely random failures), I’d set one up so that I could be sure that all cylinders are operating between about 40 and 50 degrees rich of peak on the lean curve, at high power and a relatively low cruise RPM.

That way I’d be giving the engine the toughest pounding that I could possibly give it. (Ssshhhhhh: Don’t tell yr right, but that’s ROP….)

I’d then apply time and observation.

I suspect most engine manufacturers are on top of that, and already perform similar tests.

It is probably something more obscure. Maybe something maintenance related, maybe something assembly related, maybe thermal cycle related. Someone hypothesized that it could be due to the mass of the hydraulic lifters because they are from a much larger engine.

The first thing we need to know is what are the common failures? Through bolts seems to be the main one being discussed.

[Arnold E]

I find it interesting that this debate even occurs. I worked for many years in the 'experimental engineering' division of a major motor vehicle manufacturer. Now it appears to me that Lyco and Conti would have similar divisions. It also appears to me that some of you are saying that we did not have any idea what we were doing, (despite being employed in this are for many years) and a lot of you know more than the people that actually design and build the equipment. Jabba, I wonder how many people know more about your business than you do. On your standard, a fair few I would guess. (first time I have commented on this debate).

[Ultralights]

And for Rotax 912(S) there is the "avoid cruising below 5000 rpm" camp...

Theres good reason for that!

it avoids this..

This is very hard on the mechanical parts of the engine and it creates a lot of heat which you read on the CHT gauge. Imagine riding a bicycle and applying maximum pressure to the pedal at the top or just before the start of the down stroke. You may push as much as you like but nothing much happens till you obtain quite a few degrees of mechanical advantage.

very similar effect on internal pressures as running LOP..... sharp drop of in pressures...and hence, temps..


the events described above, the high pressures, running at the worst ROP mixtures, i think are part, if not most of the issues facing the Jabiru engine failures... how do you snap a bolt under tension? with a lot of pressure.. repeatedly..

[Jabawocky]

rutan

You are getting your numbers mixed up, the max power is around 75-80dF ROP, and the highest ICP is around 40dF (30-50).

Cheers mate

[Andy_RR]

Wow, there's a lot of pilots here with opinions about engines with little data to back up said opinions.

Sadly, it appears that even CASA don't have much data, or at least they're not sharing the data to support their proposed opinion.

Maybe that's the way engineering works in these days of the service economy?

[Creampuff]

Andrew: All of those suggested causes relate to design, manufacture and maintenance. Surely it can only be the way the pilots are operating them that is causing the damage? (Just kidding. I added that just to goad yr right.)

Assuming that the cause/s is/are one or more of design, manufacture (including assembly) and maintenance, I’d still run the engine at the settings at which it’s getting the hardest pounding I could give it (40 to 50 ROP, lower RPM). That would also mean the thermal cycle would have the highest peaks.

I suspect most engine manufacturers are on top of that, and already perform similar tests.

You may well be right, but I wouldn’t be surprised if you aren’t. Testing at those settings assumes that manufacturers know that they impose the greatest stresses on the engine. If they knew that, it’s hard to explain why some POHs (albeit old ones) would still recommend operations at those settings. Sure, it’s the setting that will make the aircraft cruise very fast, but it’s not fun for the engine.

[Creampuff]

I find it interesting that this debate even occurs. I worked for many years in the 'experimental engineering' division of a major motor vehicle manufacturer. Now it appears to me that Lyco and Conti would have similar divisions. It also appears to me that some of you are saying that we did not have any idea what we were doing, (despite being employed in this are for many years) and a lot of you know more than the people that actually design and build the equipment. Jabba, I wonder how many people know more about your business than you do. On your standard, a fair few I would guess. (first time I have commented on this debate).

It’s not about what people know.

It’s about what the data prove.

The data prove what settings result in the imposition of the greatest stresses on the engine. The data prove that 40 to 50 F ROP mixture is that setting, exacerbated by reducing RPM with an engine with fixed timing.

It may well be that you and all these divisions in Lyco and Conti know this. But the problem is: They don’t say this.

It wouldn’t be so bad if they said: “You’re giving your engine the hardest pounding you can give it, at 40 to 50 F ROP and low cruise RPM, but guess what: Our engines are built to take it and we give you a money back guarantee that you’ll make it to TBO!”

It would be even better if they went on to say: “Guess what else? You’ll give your engine less of a pounding if you operate much further ROP, or LOP, and our engine can take even more of that!”

But instead, the persistent folklore results in many engines being operated in the range where they are getting the hardest pounding they can be given.

And when something breaks it is, of course, the pilot’s fault!

And even more appallingly, if you produce engine monitor data to show first hand measurements of the temps running an engine ROP and LOP, and to show that the engine has been run for hundreds of hours at cooler CHTs than they would have been if the engine had been operated ROP, the cylinder failure is still the pilot’s fault! That cylinder failed because you were running the engine LOP! It wouldn’t have failed if you’d run it (hotter) ROP! It couldn’t possibly a manufacturing or maintenance problem!

This is one of the reasons CMI’s going broke.

[andrewr]

Testing at those settings assumes that manufacturers know that they impose the greatest stresses on the engine. If they knew that, it’s hard to explain why some POHs (albeit old ones) would still recommend operations at those settings. Sure, it’s the setting that will make the aircraft cruise very fast, but it’s not fun for the engine.

Possibly because they are designed and engineered and tested to be able to run at those settings for the TBO of the engine?

There are also some assumptions about what is hard on the engine - specifically heat and high cylinder pressures. But if they were planned for in the design, are they really hard on the engine? If you are within limitations, not necessarily. Operating in ways that were not anticipated when the engine was designed may be harder.

Other things that could be hard on the engine:
- Long periods at low rpm. You may have less than optimum oil circulation etc.
- Low temperatures. Lead scavenging requires high temperatures to work properly. Lead deposit problems have been described in low compression engines operating on higher lead fuels than they were designed for, and Rotax etc. with liquid cooled heads.
- Low manifold pressures reportedly may cause problems with ring seating.

My personal opinion is that you are least likely to have problems if you operate in the way the designer expected.

[Aussie Bob]

I am going out on a limb here but is it possible that some Jabby failures are due to a machined crankcase? I am sure there are other engines using this method but I have never seen one. Even my lawnmower has a cast case.

[Arnold E]

you are least likely to have problems if you operate in the way the designer expected.

Since he is the one that knows the design intent.

[Andy_RR]

Originally Posted by Aussie Bob

I am going out on a limb here but is it possible that some Jabby failures are due to a machined crankcase? I am sure there are other engines using this method but I have never seen one. Even my lawnmower has a cast case.

Most crankcases are machined at some point in their manufacture. There's no real data to support this hypothesis, is there AB?

Originally Posted by Arnold E

Since he is the one that knows the design intent.

The design intent and the design results are often worlds apart. There's normally a development process in between which is used to minimize the characteristic flaws in a design. All engines have their weak points, some more so than others. Many of these points are attempted to be fixed by ADs and recalls, if not by "maintenance" alone. Others become "character"

[Creampuff]

Possibly because they are designed and engineered and tested to be able to run at those settings for the TBO of the engine?

Yet so many of them don't make it to TBO.

Again, the data shows what one problem is, because there is a real life experiment going on in aviation land. A fairly homogenous group of pilots - fed about the same diet of folklore and facts during training - some flying brand X engines and some flying brand Y engines, and brand Y is suffering far more premature cylinder failures/valve problems. But what's the explanation? It couldn't possibly be the manufacture or maintenance of the cylinders and valves on brand Y engines. Gotta be a pilot problem! Get aircraft owners to pay for a fleet wide change! Further, a bunch of pilots flying brand Y LOP gets fewer pre-TBO failures that pilots flying brand Y ROP. LOP must be bad!

It all makes perfect sense.

There are also some assumptions about what is hard on the engine - specifically heat and high cylinder pressures.

"Assumptions" is an odd way to describe the realities of the metalurgy involved in what is, after all, nearly century old technology. Why did the designers put a red line on CHT?

I'm confident that there is data to show how strong the materials comprising cylinders are at various heats, and what happens when various pressures are applied to them.

Other things that could be hard on the engine:

Could be. Might not be. What do the data show?

Certainly the data show that running the engine longer and more often is better than running it 1 hour every couple of months. Not sure what that has to do with ROP v LOP though.

[Y]ou are least likely to have problems if you operate in the way the designer expected ... [s]ince he is the one that knows the design intent.

So are you saying that the designers of piston aero engines never expected them to be operated LOP?

If so, it's astonishing that those designers were blissfully ignorant of the tens of thousands of piston aero engines that have been operated LOP for millions of hours over many decades as standard operating procedure.

[Oracle1]

I Just cant Resist

Having fixed a few and flown a few here is my two bobs worth.

Jabiru engine failure modes,


1. Through bolts.

In the older solid lifter engines the tolerances would change often (for reasons I will expand on further in the next point) and unless you got the engines temperatures stable the user would have to constantly adjust the valve clearances so Jabiru decided to dumb the engine down by fitting hydraulic lifters. The mass of the solid lifter is tiny in comparison. Ergo large mass smashing back and forth pounding the thru bolts and they break.

2. Various Top End Failures.

The material chosen to manufacture the heads is ductile so that it can be CNC machined. The material has a higher plasticity when heated in comparison to the likes of a vacuum cast head such as Rotax uses. Jabiru finally recognized this when they started experimenting with vacuum cast top ends some three years ago but looked to me have made the same mistake as the coarse finned heads and didn't have enough surface area to transfer heat. Valves moving around in the guides and the seat as well as ladies waisting the valves from exhaust gases (stretching the valve in the exhaust flow, probably from insufficient diameter and surface area to transfer heat to an already hot guide) are all a result of heat retention and the ductility of the parent material.


3. Cooling problems Various

I have mates of mine who have persevered with Jabiru engines and now have many thousands of hours of reliable service from them. Anyone who knows will tell you to keep the CHT's below 110 C and the heads are a heap more stable. Then the solid lifter engine stops drifting in the valve clearance and presto no more tinkering there. Suddenly the through bolt tensions stop changing as well. These temps are easily achieved by opening the nostrils and putting extensions on the bottom of the cowls, careful checks for air leaks etc, no zoom climbs, all standard air cooling stuff.

4. Fuel Distribution

I am not going to expand on this because it is common to all aircraft engines and is the same old story, fit the engine analyzer now. In Jabiru's case they should just inject the engine, either electronically or manually.


In summary people should get of Jabiru's case. They have made some remarkable achievements on a shoestring budget using ingenuity and hard work. The air frames are a great product! Aircraft engine development is the holy grail and the big boys engines still f**k up all the time and at a much higher cost. However Rod Stiff needs to abandon the shoestring budget mentality (however altruistic) and start looking at more sophisticated production techniques and more custom parts rather than robbing cheap parts in mass production and working around them.

PS Don't bother arguing with Jaba on engines you will loose

[Aussie Bob]

Most crankcases are machined at some point in their manufacture. There's no real data to support this hypothesis, is there AB?

No data whatsoever Andy, as I said, I am going out on a limb, but who else anywhere uses machined from billet crankcases? Surley it is easier and cheaper, so why isn't everyone doing it? Just asking, I don't know myself.