Rotax 914 head bolt failure

[SDQDI]

Just thought I would share this story as I have a few minutes to type it out.

 

On the 10th of may we had a fly day at YQDI it was a beautiful day with mild temps and a high cloud base which pretty well kept the thermals away, I did the 3nm trip from home to YQDI at around 9am. Was great meeting new people and was a friendly relaxed atmosphere.

 

After lunch once things were finishing up I took a pilot neighbour for a local flight basically to have a look around and to see where his son was picking some cotton. Once we were heading home cruising at about 73 or 74 knots with everything nice and smooth I heard a slight noise which sounded like a rivet popping, passenger also heard it but thought it might have just been a patch of rough air, I immediately slowed to 60 knots as a precaution worried at the time that it could have been a wing stay bolt (it's amazing how much stuff goes through your head) all my temps were cool as usual and the engine was still purring like a kitten. We returned without further noises or incidents for a slightly bumpy landing (my specialty) at YQDI.

 

A thorough check of the airframe revealed nothing, no missing rivets, no loose bolts, no excessive movement everything looked good. It wasn't until I checked the engine that I found the problem. Firstly the nut was missing off the top front of cylinder number 1, the reason for this was the bolt at the other end of that stud had broken and the noise was that bolt head hitting my cowling with enough force to make a little impression in it.

 

So after contacting Ole and filling out all the required incident report forms (I spoke to Darren and he said they are well on the way at RAA to getting the incident and accident form set up as an online form similar I guess to the atsb one) rotax was very helpful sending all the bits to Ole and then he flew out on Friday morning and did the necessary repairs. Went for a fly Friday afternoon with no problems so am a happy boy again.

 

I will try and attach some photos, the helpful people at bert flood said it wasn't the first one that had done it so thought I would post this up here so people know what to look for.

 

Engine had 60.7 hours on the clock

 

 

 

 

 

[bexrbetter]

Glad you found it and got home safe.

 

 

[Guest Maj Millard]

First one I've ever seen fail on any 912, over torqued perhaps at some stage ?....................Maj....

 

 

[facthunter]

There is always the chance of a fault in an item such as this. The QA process reduces it but doesn't guarantee 100% unless you test each one. In a critical part I would like to do this if the nature of the structure allowed it. Rotax probably have a few suppliers... Nev

 

 

[SDQDI]

Yep could be plenty of reasons for it like I said it wasn't the first one that had done it so no doubt they are investigating as to the cause whether it was a bad batch of bolts or like maj said over torqued, they haven't told me I'm a rough operator yet lol

 

I'm pretty sure that the other problem bolt/s was also a 914, and from what I could gather it was the exact same bolt that had failed.

 

All in all it wasn't a very dramatic experience the engine still ran perfectly and the service from Ole and Rotax was well above what I expected I have no complaints at all:thumb up:

 

 

[Bob Llewellyn]

Them there bolts are actually quite sophisticated - brittleness is the enemy of high-strength alloys, and various process stages - the original alloying, forging, forming, plating, and any straightening - can predispose a bolt to failure, or simply increase its sensitivity to over-torquing. I wonder how much the bolt size increased from the 80-hp 912 to the 914?

 

 

[Guest Maj Millard]

Them there bolts are actually quite sophisticated - brittleness is the enemy of high-strength alloys, and various process stages - the original alloying, forging, forming, plating, and any straightening - can predispose a bolt to failure, or simply increase its sensitivity to over-torquing. I wonder how much the bolt size increased from the 80-hp 912 to the 914?

Probabily not at all...the 914 is the 80 HP (low compression) with a turbo attached...same case as the 80 HP......the 115 HP is achieved with 35 HP produced at the top end............Maj.....

 

 

[Bob Llewellyn]

Probabily not at all...the 914 is the 80 HP (low compression) with a turbo attached...same case as the 80 HP......the 115 HP is achieved with 35 HP produced at the top end............Maj.....

err, from the TCDSs the old 81 hp was 1211cc, the 99hp was 1352cc... same stroke... thought the 914 was a blown 1352cc?

 

 

[Sierra]

Hello, the exact same thing happened to me 10 days ago! Same bolts missing! Same engine 2015 Rotax 914UL! How did you fix the issue???

[Area-51]

Can somebody clarify; the pictures show a "crankcase bolt" missing, and a "cylinder head nut" missing. The two items are mutually exclusive to one another.

 

Loose head nuts have been noticed on the 914 during annuals

Edited May 8 by Area-51

[Area-51]

A set of new cylinder studs for early 914 has just been measured against a set with 127hr service time; almost all removed studs showed 010" stretch; one stud showed 040" stretch. So it is possible to elongate the stud further beyond its designed yield by over or further tightening.

 

As the factory state; the cylinder studs on the early 9XX are single use only items.

[Area-51]

Sorry correction on above; the  check on the 127hr 914 provided 010" stretch on all cylinder studs. A bulk strip on a timed out 912UL provided 010" stretch on most cylinder studs, a few had 020", and one 040" stretch.

[facthunter]

Bit of a thread dredge but not good. Having all of the cylinder made of aluminium means heat will impose a greater load on the stud. This is not an unusual problem with  such a layout. Operating with a failed stud is not helping the rest of the engine but how to you know?  IF it's stretching it's exceeded it's yield point. Nev

[Area-51]

Many modern studs are now designed to be torqued up to yield; hence tightening method of initial Nm tension then series of further 90deg turns to the yield; at which point further tensioning will not provide any further clamping force, just stretch the bolt/stud further. i.e the bolt is designed to hold a specific tension at the yield.

 

Personally i do not think the solution of the 9XX cylinder design is optimum, however it does work sufficiently and is fit for its service application.

 

Regarding bolt/stud elongation there will be a permissible tolerance provided within the Rotax overhauling manual, however the public only has access to the heavy maintenance data which does not cover cylinder stud tolerance data. Most all modern high stress engines with removable barrels will have single use retaining studs where the block barrel and head are retained by a set of through fixings. As an example the early Lotus Elise engines employ approx 400mm through bolt a maximum permissible stretch and also a single use throw away specification. (we had to replace every single cylinder head gasket on every engine due to a factory stuff up on the bolt yield specification; it was too low) Thinking a stronger bolt will provide more tension can then end with component distortion due the component material itself yielding. So its a finely tuned dance of material reduction and fit for service outcome. All Cost/Weight driven

[turboplanner]

3 hours ago, Area-51 said:

Regarding bolt/stud elongation there will be a permissible tolerance provided within the Rotax overhauling manual, however the public only has access to the heavy maintenance data which does not cover cylinder stud tolerance data. 

In the days of collapsible head gaskets you could tighten the head bolts to the specified nut torque (after a series of even tightening with a socket) and that was it, you had a seal for life.

 

As aluminium became more widespread and non-collapsible gaskets were used, the sequence tightening became more critical and instead of the final tension being measured by rotational friction to trip the torque wrench, a torque wrench final socket figure is quoted then a rotation amount is specified using the thread depth to obtain the tension required.

 

For this reason you ALWAYS fit new head bolts after a head has been removed.

 

Leaks and loose bolts are the byproducts of this system for even minor mistakes. I agree the manufacturers have fallen behind in providing exact details of the process required and you have to hunt for it because not all manufacturers work on the same figures.

[facthunter]

The CAUSE of the problem here is the uneven  thermal expansion of alloy  and high tensile steel over a fairly large distance and the stress it places on the studs. Nev

[spacesailor]

RR. Had no gaskets. 

So I have been told .

Others have had a copper ring at the ' head ' .

spacesailor

[Area-51]

2 hours ago, spacesailor said:

RR. Had no gaskets. 

So I have been told .

Others have had a copper ring at the ' head ' .

spacesailor

There are no gaskets, care is required not to damage mating surfaces between head and cylinder. There is also a minimum gap specification measured between head and cylinder components post tensioning. Have not observed any copper ring and fail to see the advantage; just one more point of failure gained.

[spacesailor]

" copper ring " .

On a very Old cast-iron motor . ( built in water cooler on the head )

Would the copper expand more than the cast-iron .

spacesailor

Ps: all had to be stripped for smelting. 

Edited May 9 by spacesailor
ps added

[Area-51]

Sorry just realised you were talking about Royce; thought you meant Remove & Replace... Yes that sounds feasible; some engines also use Wills Rings (gas filled crush rings) instead of gaskets and are located into a purpose groove.

Edited May 9 by Area-51

[facthunter]

We digress. Head sealing is not the issue being discussed but is a problem generally with many engines. That's why many are shrunk and threaded on semi permanently or have a lot of bolts. .  Nev

[SDQDI]

On 9/5/2024 at 6:46 AM, Sierra said:

Hello, the exact same thing happened to me 10 days ago! Same bolts missing! Same engine 2015 Rotax 914UL! How did you fix the issue???

G’day Sierra 

my memory is not as good as it used to be and this was 10 years ago now (where on earth did that time go!) so my recollection of the repair is basically gone. The repair was done under warranty and I did watch but I can’t remember how it was done.

 

On 9/5/2024 at 9:14 AM, Area-51 said:

Can somebody clarify; the pictures show a "crankcase bolt" missing, and a "cylinder head nut" missing. The two items are mutually exclusive to one another.

 

Loose head nuts have been noticed on the 914 during annuals

again my memory isn’t the best but I thought they were connected? Basically a through bolt sort of set up.

[RFguy]

Interesting.  I've done the calcs on thru bolts etc for jabs, and I assembled my engine(s) on thru bolt stretch, not torque.

However, that's more difficult to do with a stud ! 

 

what's the diameter of the studs ?

 

FWIW 100hp Rotax hp /litre/rpm  is about the same as the Jab. Jab has 6 head screws (5/16" from memory) .

 

FWIW the jab 7/16" thru bolts dont have much leeway between required preload tension at cold,, the tension at hot (with crankcase expansion) and yield. However, all my numbers were done on min yield specs, so the yield  point was likely higher.

The Jab head bolts , six of them, there's plenty of tension in hand there.  

Edited May 11 by RFguy

[turboplanner]

10 minutes ago, RFguy said:

Interesting.  I've done the calcs on thru bolts etc for jabs, and I assembled my engine(s) on thru bolt stretch, not torque.

However, that's more difficult to do with a stud ! 

 

what's the diameter of the studs ?

 

FWIW 100hp Rotax hp /litre/rpm  is about the same as the Jab. Jab has 6 head screws (5/16" from memory) .

 

FWIW the jab 7/16" thru bolts dont have much leeway between required preload tension at cold,, the tension at hot (with crankcase expansion) and yield. However, all my numbers were done on min yield specs, so the yield  point was likely higher.

The Jab head bolts , six of them, there's plenty of tension in hand there.  

I assume that when you set the bolt stretch cold, the manufacturer is working on that being the optimum setting at working temp.

[RFguy]

the manufacturer's advice. LOL. leads to over and under tensioning. LOL. My measurements and calculations showed that the mfr method leads ot tension errors of up to 50 %. likely +/-30%.... anyway.....  yes, initial stretch temp , temperature dependent.

 

There's a bit of a compromise between having enough tension with a crankcase temp of say 30 deg C, and having too much at crankcase temp of say , 90deg C. So,  still a good idea to take it easy until oil temp up to operating temp (which means crankcase temp , also, pretty much) . If the cold tension is too high, safer to TO with low crankcase temp. If cold tension too low, the opposite...

 

I don't want to pollute the forum, this is about 914 heads ,just wondered what the stud diameters were. I'll look it up. Looks like 8mm. 

 

I would be confident the Rotax engineers got their stud sizes correct. the rest is left to tolerances and install .....