Debate about Mark Langford's 3rd crank failure

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

Ramrod25

Member
Joined
Jun 4, 2007
Messages
10
One of the issues that MAY have some relevance here is the type of flying that Mark is apparently skilled enough to do.

By that I mean aerobatic maneuvers that impose significant G loads on the internal components of the engine.

Both Dan Weseman and Mark are apparently skilled pilots and have airplanes capable of performing high

G maneuvers. Just a thought.

After reading both WW's comments and Mr. Langford's, it occurred to me that there probably haven't been many
crankshaft failures on slow flying aircraft with nitrided cranks and 5th bearings installed.

Bringing any engine designed for automobiles along the path of increased reliability for use in aircraft is necessarily

a long, involved learning experience, punctuated by engine and engine component failures. These failures point

us toward engineering solutions which necessarily put us back at the beginning point on some learning curve.

Someday, perhaps with a new crankshaft composed of modern steel and machining methods, 5th bearings and
the other refinements all ready proven, we will get past crank problems and quite possibly move on until we
identify the next weakest component in the engine to fail. Guess what - that puts us right back at the beginning again;
engineering solutions, testing and improving.

To me, the sad part of this is when conflicts arise, as I guess they inevitably must, among the people who have given
so much of their time, energy, effort and $$ to bring us an engine worthy of our trust and $$ to hang on our own airplanes.

May I suggest that who said what, when or where is not all that important. Those of us looking to build, or who are
building Corvair engines for their aircraft are quite willing to acknowledge that the combined expertise of WW, ML and DW
is the main reason that there even exists a Corvair movement today.

What we need to acknowledge is that there appears to be a continuing issue that needs to be solved. Perhaps there
will always be continuing issues to solve which is why we AD's and SB's on certified engines. That said, don't we need
to push forward and simply solve the problems? We all want the safest engine we can build to put in our airplanes.


Regards
Rodney
 

BBerson

Well-Known Member
HBA Supporter
Joined
Dec 16, 2007
Messages
12,823
Location
Port Townsend WA
Torsional resistance is mostly provided by crankshaft stiffness. Nitriding improves the surface hardness but does not change the deflection stiffness and therefor does not reduce torsional vibration. Aircraft engine crankshafts are almost always larger for stiffness (and hollow for lightness) compared with solid automotive cranks.

Again, from the book: THE AIRPLANE ENGINE 1922

"Crankshafts are dimensioned for stiffness as well as for strength. Stiffness for given dimensions depends only on the modulus of elasticity and not at all upon tensile strength. As the modulus of elasticity is practically constant for all steels, there is no advantage, so far as stiffness is concerned, in using a steel of very high tensile strength for shafts."
 

orion

Well-Known Member
Joined
Mar 2, 2003
Messages
5,800
Location
Western Washington
About four months ago I started looking into alternate engines for one of our customers. The Corvair seemed to be a pretty good candidate given the simplicity, power and weight. Since I've owned two Corvairs (many years ago) I was familiar with the engines and felt reasonably assured that this is a potential candidate. But looking closer I found that I did not fully agree with some of the assessments regarding the design and especially the reasoning behind the fifth bearing that is being installed in these conversions. The theory is that this fifth bearing will provide the prop end of the crankshaft with sufficient support to handle all flight loads, thus protecting the crank from what might be damaging loads and deflections. But in practice this is only partially accurate.

The fifth bearing does provide end support but it primarily addresses thrust loading and loads perpendicular to the crankshaft axis - it however is incapable of handling any other imposed load. When installed in a car, whether clutched or with an automatic, the crankshaft sees only pure torque. All components bolted to, or taking power from the crankshaft are rigid and piloted so the crankshaft end sees only axial moments. Yes, there is a small side load encountered during starting (due to gear mesh of starter with ring gear) but this is short period and pretty minimal and thus not significant.

In aircraft application the transmission is of course removed and a prop is generally bolted on the end of the crankshaft. The main load of the prop is, as with the car application, torsional. However we have two other loads to deal with now. The first is a force vector perpendicular to the axis of rotation and it represents the weight of the prop. The prop is displaced from the end bearing so its weight times that distance is a bending moment imposed on the crank's end. The prop's weight is then multiplied by a combination of turning Gs and pitch (or yaw) accelerations. These multiply the prop load, significantly increasing the overhang moment.

But the main and very significant moment that's created here is the reaction caused by the rotation of the prop and the pitch or yaw of the plane. Even relatively mild turbulence for instance can generate some pretty impressive rotation rates, resulting in sizable "gyroscopic" loads. So while the bigger end bearing can handle the axial vector forces, it is generally incapable of doing anything with the moments. Even a double row bearing has insufficient rigidity and distance between the contact points to counter what can be a sizable bending load imposed on the end of the crank, thus allowing measurable crank deflections. Given the typical geometry, it is no surprise that these conversions are seeing failures.

The fix for this would be to isolate the end of the crank completely from the loads of the prop. This would take a somewhat elongated housing with two bearings, the pair separated by some distance. How much? Hard to tell without any analysis but as a first cut rough guess, I'd suggest at least four inches. This would then create a reaction couple, which would get transmitted by the bearings to their housing and then in turn into the block, and then the engine mount. This way all vector and moment loads would be handled by the bearings, thus isolating the crank. This type of mounting would also provide the opportunity to create a flexible coupling between the crank and the output prop shaft, thus also addressing the torsional feedback issues, if any.

But as things sit right now, for our project I decided to pass on the Corvairs.
 

Autodidact

Well-Known Member
Joined
Oct 21, 2009
Messages
4,513
Location
Oklahoma
Orion, would the crankshaft, when coupled to the prop drive by a flexible coupling, also need a flywheel?
 

orion

Well-Known Member
Joined
Mar 2, 2003
Messages
5,800
Location
Western Washington
The flexible coupling is flexible only from a torsional standpoint and only as compared to a rigid steel on steel connection. As such, probably not but don't quote me on that - I haven't gone that far in my research on this. I know other engines that use any type of dampened coupling don't use a flywheel or if they do, it is substantially machined down, so the Corvair application might need a bit more investigation.
 

Jan Carlsson

Well-Known Member
Joined
Jan 11, 2009
Messages
1,860
Location
Sweden
M.T.

The small spocket must be unloaded from the belt tension and drive force, so it don't put a side/radial load on the crank, and a flex coupling, like centaflex, to take the T.V. the belt will not flex much if any. consider it can be a need for 10-30 degree of tortional movment at some rpm. Flywheel - you need a starter ring anyway, but designing a PSRU isn't for the experimenter, it need the knowlege and experiance of a pro, or, a lotto miljonär's amount of pure luck.
there are people that know this on there five fingers.
 

Dan Thomas

Well-Known Member
Joined
Sep 17, 2008
Messages
5,091
Someday, perhaps with a new crankshaft composed of modern steel and machining methods, 5th bearings and
the other refinements all ready proven, we will get past crank problems and quite possibly move on until we
identify the next weakest component in the engine to fail. Guess what - that puts us right back at the beginning again;
engineering solutions, testing and improving.
Developing a new, strong crank might cost so much (especially for so few crankshafts) that it would probably cost the end user as much as buying a nice new Lycoming.

I would wonder, after the new crank is in, how long the bearings and case will last. Those parts were designed for handling only torque, not dealing with thrust and precession.

As far, again, as broken Lyc or Continental cranks go, an insurance company will now usually demand a complete engine teardown and NDI of all components if an accident involving a propstrike occurs. They'll often pay for it. They have found that it's cheaper to do that than to continue to insure an airplane that has had a propstrike without any inspection. Even if the prop flange dials out OK, cracks have often started far back along the crank and it eventually fails and drops the airplane and occupants into a forced landing that sometimes gets really expensive. Small Continentals are known to crack between the #1 and #2 journals, which in that engine are at the back, far from the prop. My inflight break was right there, too, and of the O-200s we had propstrikes with (flight training accidents) we found cracks starting in that same spot. No amount of checking the prop flange runout will find that.

To make a crackproof crank might be possible but it would be massively heavy. Using a wooden propeller is a better way to avoid it. It absorbs the shock and splinters when it hit something and saves the crank, and imparts much less precession loading.

The FAA has an AD against Lycomings that have suffered any propsrike, including encounters with water or tall grass. Such decelerations cause the bolt that retains the accessory drive gear on the crank to come loose, and can partially shear the drive dowel. That's because of the inertia of the mags and other stuff back there, but the same AD also warns of crank cracking and the AD applies even if the prop wasn't turning and ran into a hangar door, for instance. Things bend a bit and start cracks. And that's on a really strong aircraft crank, not a much weaker auto crank. A well-designed PSRU sounds like a much better bet to me.

Dan
 

Doug2233

Well-Known Member
Joined
Jan 10, 2011
Messages
48
Location
Sydney, Australia
There are a few reasons cranks can break (at least so far as I am aware) beyond just a weak or damaged crank itself. Cases can become warped, can flex, can suffer non linear distortion as they heat. Increasing the cylinder forces with increased power will also result in case flexing.

I suffered a crank failure in a vw engine, thankfully in a car not an aeroplane. This was just a 40 hp engine too! I put that down to not having the case main bearings line bored prior to reassembly.

Lycoming and continental use enormous steel through studs to stabalise the crankcase halves, even then fretting is sometimes discovered on the mating faces due to movement. Does the Corvair have through studs?

Personnally I doubt just a forged crank in a more appropriate steel will do the job. The case needs to be reengineered as well. Nothing less than a full FEA model and analysis of the bottom end of the engine.

I believe the Corvair is the genesis of a great aero engine an fits a niche, but it is just that.
 

autoreply

Moderator
Joined
Jul 7, 2009
Messages
10,733
Location
Rotterdam, Netherlands
Small Continentals are known to crack between the #1 and #2 journals, which in that engine are at the back, far from the prop.
Any idea why they would crack there first? My "seat of the pants" feeling says they would be loaded most of the forward journals?
A well-designed PSRU sounds like a much better bet to me.
The clutch (as used on many of the Rotax engines) apparently works great too.
 

psween

Well-Known Member
Joined
Jan 20, 2011
Messages
100
Location
7MN3
Might be an interesting follow-on post from Mark Langford Crank Break #3
Way down at the bottom is a follow up detailing inspection of the crank radii on his most recent break. It shows that none of the radii were up to the original spec, and showed regrind errors and grooves that would seem to create multiple stress risers. His recommendation (which I agree with) is to do a detailed inspection of any crank prior to assembly, especially those reground, and consider waiting for an aftermarket crank. Apparently the first batch of 6 aftermarket cranks are in works now, from Dan Weseman, so they might become reality. I don't have a dog in this show, but am leaning strongly toward a Corvair to power my KR derivative. Considering that the alternatives for me are VW (not quite the power), Rotax 912 ($$$), or o-200/o-235 (weight and $$$), I'm very interested to see what comes of these new cranks. Even if they add $3000 to a build, they would still result in a cheaper package than all but the VW, and might rival the reliability of a certified motor. Only getting them into airplanes in significant numbers and flying a lot of hours will prove them out, but I've got time.

Patrick
 

Marc Bourget

Well-Known Member
Joined
Feb 28, 2011
Messages
354
Location
Stockton, California
I have an engineer friend who was one of the memebers of the core design group for an engine. Another friend who had switched to a larger engine used the first engine's Harmonic Balancer on the second - and the Balancer "grenaded" in Rush Hour Traffic, making the evening news as he was stalled in the middle lane of a 6 lane freeway in LA on Friday. Well, when we started discussing the situation I now discovered from the first friend that there were things like "Sixth Order Vibrations" and the like. He made my head hurt.

But other discussions focused on the Journal Radius. If you're serious about longevity, a diligent Crankshaft manufacturer will "roll" the radii with Hardened Rollers, under great force, similar to shot peening, to eliminate the stress risers. I think this will be hard to find from suppliers of after-market and nitrided crankshafts, but, having suffered a catastrophic driveline failure (transaxle) for this "inside radius" problem, I wouldn't run a Corvair crankshaft without the technique. Take away the "opportunity" (and maybe run a wooden propeller) and the stock crankshaft may be just fine.

Onward and upward

Marc Bourget
 
Last edited:

john affleck

Member
Joined
Jan 10, 2013
Messages
5
Location
ocean view Hawaii usa
The people getting interested in the new aftermarket crankshaft for the Corvair need to read this; as it correctly states the facts. Mark Langford stated that 4340 would be twice as strong as 5340 in a crankshaft; this is ridiculous. He has a talent for getting the wrong answer; I've communicated extensively with him by e-mail; I'm the only person qualified to speak on this subject who has communicated with him; but he decided not to listen. at that time he was in love with his guru WW; who is a MECHANIC; and not an engineer. He's a good mechanic and a disaster as a engineer; but that's what you would expect. I'll explain exactly why his crankshafts broke on another post here which you can look up; I'm not going to write the whole thing twice. DO NOT BUY THE AFTERMARKET CRANKSHAFT; this is just more amateur screwing around. Okay; this was intended to follow on the man's post to the effect that only the modulus of elasticity is relevant; but I don't understand how forums work; so this is in the wrong place.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
4,561
Location
US
I'm the only person qualified to speak on this subject who has communicated with him
WW; who is a MECHANIC; and not an engineer. He's a good mechanic and a disaster as a engineer; but that's what you would expect.
Umm, well where to begin? Welcome to the forum. Please don't be a shrinking violet, let us know your opinion on things. Mark Langford and William Wynne are names we know. They've produced much, written much, tried many things, and have been very clear about their qualifications. That helps everyone know how much weight to put on the things they say.
 

stol

Well-Known Member
Joined
Feb 15, 2008
Messages
497
Location
Jackson Hole Wyoming
The people getting interested in the new aftermarket crankshaft for the Corvair need to read this; as it correctly states the facts. Mark Langford stated that 4340 would be twice as strong as 5340 in a crankshaft; this is ridiculous. He has a talent for getting the wrong answer; I've communicated extensively with him by e-mail; I'm the only person qualified to speak on this subject who has communicated with him; but he decided not to listen. at that time he was in love with his guru WW; who is a MECHANIC; and not an engineer. He's a good mechanic and a disaster as a engineer; but that's what you would expect. I'll explain exactly why his crankshafts broke on another post here which you can look up; I'm not going to write the whole thing twice. DO NOT BUY THE AFTERMARKET CRANKSHAFT; this is just more amateur screwing around. Okay; this was intended to follow on the man's post to the effect that only the modulus of elasticity is relevant; but I don't understand how forums work; so this is in the wrong place.
Let me get this straight..................

according to the sig line this is your first post on HBA and yet you say

"disaster as a engineer; but that's what you would expect. I'll explain exactly why his crankshafts broke on another post here which you can look up; I'm not going to write the whole thing twice"..

care to share the link to that post ?
 

john affleck

Member
Joined
Jan 10, 2013
Messages
5
Location
ocean view Hawaii usa
Okay; trying to reply to jsween. Do not buy an aftermarket crankshaft. I have practically written a book about this **** crankshaft already; the intention being to save people from significant injuries and damage. I corresponded with Mark Langford via e-mail and with WW; to no avail. First let's start with the picture; it's a duplicate of the picture of Mark's first crankshaft failure; which inspired me to write to him in the first place, (years ago). The picture is completely diagnostic. It leaves no room for doubt. It leaves no room for speculation; I don't speculate anyway, I'm an engineer, I deal in facts not opinions. Let's dispose of the "torsional issue" first. It is not torsional failure. Period. There's nothing more to be said about it; the fact that Mark is referring to "beach marks" on the break face and trying to sound like he knows what he's talking about; is due to my corresponding with him in the first place. Without me, he wouldn't know a beach mark from a beach ball. I pointed out to him that engine design is a subset of mechanical engineering and that people have been studying this subject for the entire twentieth century; and I gave him the title of a two volume textbook from MIT on Engine Design; to his credit, he bought the book. He also makes the statement that all the breaks originated in the shaft radius with the cheek; this is simply false; and obviously so. THE CRANKSHAFT RADII ARE NOT INVOLVED IN THIS ISSUE. "Is the Corvair crankshaft strong enough for aircraft use"--In many ways this is a typical civilian question; it's so broad and so ill-defined it doesn't have any "answer". However, considering my audience, here; civilians; I will simply say, YES IT IS. There is a way of estimating the "strongness" of a crankshaft, obviously a very imprecise concept; and I explained that to Mark and he looked it up in the reference book and seemed to get something out of this. According to the standard engineering criteria for crankshafts; the corvair crankshaft is "stronger" than the Lycoming or Continental crankshafts. This is not an opinion; I don't do opinions; I hate opinions. As I say, according to the standard measurable parameters of crankshaft design it is stronger; period. end of discussion. Okay; so now you want to know; why do the corvair conversions break crankshafts; and for those of you who don't know; there are more of them then Mark's. Ok, by the way, when WW, "our guru" told everyone they would have to nitride their crankshaft; I immediately wrote to both him and Mark Langford and told them NO; this is not the answer. Now they will fail at 700 hours instead of 70 hours. Look up the actual numbers for yourself. this is not an accident; this is not "luck" on my part; this is call shot pool; I am not an amateur. Mark refused to listen; due to being hypnotized by his Guru. so he broke his second crankshaft. I've been waiting for the third one, with the stupid front bearing to break; but I see I missed the actual event. This is understandable because the only reward I got for trying to save these amateurs from themselves was to be told "you're not a Corvair Expert" !! This reveals a sort of divine idiocy. If you specialize in engine design as a subject in mechanical engineering and have a SAT score of 800 out of 800; then you are a Corvair expert; because the **** thing is an engine. Okay, back to why they crankshafts broke; the picture reveals beyond a reasonable doubt as they say in court; that it is a bending failure. What does this mean? Imagine a bobby pin; well maybe you're not old enough t know what bobby pin is; imagine a tuning fork; it has two legs and they vibrate relative to each other. Now look at a picture of a crankshaft; could be a corvair crankshaft, if you like; see that the structure comprising the rod jounal, the two cheeks, and the main journals, resembles a tuning fork; as we say, it is topologically identical. So, what the picture of the break is telling us is that some force that was applied cyclically, to the ends of the U shapped structure caused the WEB to break; exactly where we would predict it would and in exactly the correct orientation. It is a tuning fork fracture. it's so common, it even has a name. The provision of the front bearing moved the critical bending, (NOT TORSIONAL BENDING), back to the second U shaped structure; not too surprising. Not really, predictable; I mean I didn't predict this; as I did BOTH OF THE SECOND AND THIRD failures. Okay, now are we all clear that it has nothing to do with con. rod loads. rpm. loads, hmm? Is it straight in all our heads now? Because that's what NOT TORSIONAL means. Connecting rod load shaft failures don't look anything like this; nothing; nothing at all. Nothing reveals the commenters complete and total innocence of knowledge more perfectly than looking at this break and than wanting to talk about con. rod load, and cylinder size, and whatever was hanging on the back end of the engine. You have only to ask yourself what is attached to the crankshaft that bends it back and forth like a little tuning fork 3,300 times a minute? the prop. The Prop. is attached to this end of the crankshaft; this should be a big clue, folks! And then, if we were actually educated in this subject instead of just babbling at random; we would immediately think of a little item called "Prop. Indexing". Prop. Indexing is a known and understood subject in piston engine engine aircraft. Right now, Mark is flying a Continental 0-300 engine in his new airplane; if he takes the trouble to buy a factory overhaul manual for this engine he will find out that Continental requires that the installer of the engine INDEX the prop. What does this mean? It means the two bladed prop. only goes on the mounting flange in a certain position; it has to be CLOCKED. In the case of the corvair engine; which doesn't have a aircraft application; you have to figure this out for yourself; which I have done for you. Looking at the front crankpin of the crankshaft, install the prop. so that the blade is at ninety degrees to the tuning fork. if the front cylinder is at top dead center, or bottom dead center, doesn't matter, then the prop. MUST BE STRAIGHT UP AND DOWN. And that's it. that's the cure. Now there are no more broken crankshafts. No Nitriding, No new front bearings designed by mechancs; no; just this. This explains why some break and some don't; because the great GURU, ww, being unaware of this, doesn't specify a prop index. therefore, they are clocked at random; the builders are innocent; nobody told them any different; and from there it is strictly a matter of luck; is the propellor at ninety degrees to the little U shaped crank? or 60 degrees, or 15 degrees. Just a matter of luck. DO NOT BUY THE AFTERMARKET crankshaft; Mark's statement that the steel they have chosen is "twice as strong as 5340" is complete ignorance. the REAL PROBLEM, is that these will be manufactured by turning out of round stock, (ignorantly called "billet"by the amateur engine designers who do this), on a lathe. The original factory product was a HOT FORGING. the new replacement crankshaft will be TWICE AS LIKELY TO BREAK. I tell you t his in all seriousness, people; this is a serious subject. Spread the word around. repost this. These people do not know what they are doing. Now you know the answer; apply it and fly in confidence; ALL OTHER OPERATING PARAMETERS of the Corvair crankshaft are superior in the as installed flight application to either Continental or Lycoming cranks. Period. this is not an opinion. I wlll not be responding to any comments; don't waste your time. you have the knowledge, now use it. I've done my public duty. I'm not even faintly interested in your opinions, and if you are; you are very foolish.
 

Pops

Well-Known Member
Log Member
Joined
Jan 1, 2013
Messages
7,688
Location
USA.
Add me up on a Lyc crank break. Lyc-290 with about 350 hrs TT. Found out latter that it was in a batch of bad cranks. Pops


Broken Lyc or Continental cranks are most often attributable to propstrikes in the engine's history. When that prop strikes anything it sends a shock down the crank and twists it enough to start a crack. I had an engine fail this way, and have seen cracked cracks come out of engines torn down for propstrike inspections. Other than that, broken cranks are extremely rare, even when Lycoming had some problems with cranks they contracted a foundry to build for them. Those were all recalled even though only one actually failed.

Dan
 

Autodidact

Well-Known Member
Joined
Oct 21, 2009
Messages
4,513
Location
Oklahoma
john, could you at least clarify something? Do you - by "the second U shaped structure" - mean the second one forward from the oil pump drive end of the crankshaft or from the propeller end of the crankshaft?
The provision of the front bearing moved the critical bending, (NOT TORSIONAL BENDING), back to the second U shaped structure; not too surprising.
 

TinBender

Well-Known Member
Joined
Mar 4, 2010
Messages
218
Location
Charlotte, NC, USA
if the front cylinder is at top dead center, or bottom dead center, doesn't matter, then the prop. MUST BE STRAIGHT UP AND DOWN. And that's it. that's the cure. Now there are no more broken crankshafts.
From what I read here, Corvair crank failure (5th paragraph), ML's been indexing his prop 90o out since 2005, perhaps at John's suggestion. Ahh, nothing in life can be simple, can it? I agree with John's general arguments, but find his tone disrespectful.

As a student of mechanical engineering I now understand why the university is making the current crop of engineers write so many formal technical reports. It is just as important to be able to share your ideas clearly as it is to have them. A few carriage returns would do wonders to the readability of the above. "Clear and Concise," is my mantra.

Regards,
Jamie
 
2
Group Builder
Top