Advice sought - exhaust cracking

[Nev]

Hi guys,

I need some advice. My left exhaust pipe has cracked in the exact same place 3x now which has got me scratching my head. The first time I reasoned that it was the result of the cylinder issue that occurred 14 months ago. However we did another repair on it a few months ago with reinforcing, and it has now cracked again after only 30 hours. The welding was done by a professional aircraft welder and each time it has cracked beside the weld. It's occurring on the aft plate of the heat muff where the pipe attaches. I can think of a few reasons why it might happen, but I'm very interested in "crowd sourcing" potential causes and fixes from the group. One possibility is that I have added a short section of turned down exhaust tips on recently, although the first cracking occurred before this change. I'm now starting to wonder if it's more to do with the engine movement on the mounts during starting and power applications. It's currently supported from the engine mount so I might see if it can be supported by the engine itself instead. Any ideas appreciated.

IMG_1466.jpg

IMG_3026.jpg
The photo below is of the righthand exhaust, but the mounting and attachment is in the same place.

IMG_3027.jpg

The photo below was taken during the build and shows the universal joints.

IMG_3025.jpg

[huntaero]

Hi Nev
Do you have the hot air shutoff valves set to vent overboard to assist in cooling the exhaust? I have my aft exhaust mounted to the eng so there is no relative motion as the complete exhaust moves with the engine, i have a 100 hours on it now but time will tell.

[Nev]

On the side with the issue I have removed the outer casing and no longer use it to provide heat. (I discussed it with Clint at the time). So the inner structural section is open to the airflow. But I'm thinking the same thing of mounting it to the engine, that might be the key.

[huntaero]

Hi Nev
This is how I supported my aft end of the exhaust to the engine so everthing moves with the eng no relative movement. 100 hours on so far so good.



IMG_4005.jpg



67341234829__5BBC0664-B7C8-435E-B56A-6B1E092757F6.jpg 67348126459__DC561E44-8FCB-4595-8322-8A2894553641.jpg
IMG_3978.jpg IMG_3977.jpg

[Nev]

That looks like a very good way to support it.

[jaredyates]

More support is definitely worth considering. Also, maybe see what your welder thinks about adding a few beads perpendicular to the one that keeps failing, especially since the crack is so close to the bead? I drew an example in red.

[Nev]

Good ideas, I'll run them past him. Thanks Jared.

[huntaero]

It might have to be backup on the inside with ARGON also, as SS needs to have an inert fgas on both sides when welding. Just a thought.

[Battson]

I would guess the original crack is related to the original engine failure.

Welding over a crack is a notoriously unsuccessful repair method - no matter how good the welder.

For welding over a crack to work, the entire crack has to be completely removed first - even at a microscopic level - no crack can remain, or the remains of the crack will grow along the weld (usually at the edge of the HAZ, which is what you're seeing). Just welding over the crack has poor odds of success, because often some small part of the crack remains and grows.

Usually a successful repair involves extensive and invasive grinding of the crack, until you're sure you are into good metal - then grind a little further, just to be sure. At risk of making a presumption about the repair, the part looks very challenging to grind out properly. Unless someone spent a long time being very careful with a die grinder, I am assuming the crack was simply "welded over", which almost always leads to re-growth of the crack.

[Nev]

Thanks Jono – all very interesting.

Last time it cracked it actually separated into two pieces, which made it easy to clean and prep properly with a die grinder before rewelding. I then made a small conical sleeve (visible in the top photo) that slides over the joint and is welded top and bottom to act as reinforcement and spread the load path. This time, though, the crack has formed around the lower edge of that reinforcement.

After talking it through with Graham Johnson, I’ve now made another reinforcement that should add a fair bit more strength. I’m also planning to move the support attachments off the engine mounts and onto the engine itself, in the hope that it reduces relative movement and vibration.

We’ll see how it behaves!

[Battson]

That is critical information which completely changes the assumptions. It sounds like overloading is more likely.

Its still cracking after the doubler, in that case, the doubler may have worsened things by stiffening one part and not the other.

I would suggest welding in ribs - perhaps 6 or 8 - around the circumference to strengthen the whole assembly. For example, currently it appears like the left image, install ribs as shown in the righthand image.

image.png

[Nev]

This is where we're currently at, having welded over the crack, and then added a doubler several mm up the pipe.
I think it's probably crucial to now remove the source of the issue so I plan to move the support struts away from the engine mount and onto the engine.

I've also shortened them slightly in the hope of changing any frequency issues.

Let's see how it works out.

IMG_1469.jpg

CBC1C724-EA78-4C03-9877-A6A66F808147.jpg

[svyolo]

I have done a reasonable amount (many dozens, <100) of custom engine installations. Only one airplane so far. I will just say I have the same exhaust as you, and never considered mounting those struts off anything but the engine, in my case I used the sump bolts.

It is common for engine headers to be supported just like ours, off the engine. If it is a front engined car, after that it may have some type of flex joint, and then supported aft of that with VERY flexible supports, at the mid point, and/or the rear of the vehicle. But the aft end of the exhaust header/manifold, if tubular, is almost always supported at the engine.

I remember reading people mounting those flex struts off of the mounts. I can't remember if I commented on it. Maybe I did.

But just like JB said, if you just strengthen/stiffen the part that failed, and do nothing else, it will just fail either upstream or downstream. I have also done that too many times.

[dramsey]

Interesting. I assume that those mounting the exhaust off the engine still incorporate the hose or other flex device. The exhaust changes from ambient to 1400 f and the support strap changes from ambient to ABOUT 150-180F. That difference in thermal expansion can introduce significant stress. I had a builder in my neighborhood mount off the engine with no flex incorporated and I told him it was going to break. It did break on first flight setting up a vibration and scaring him. Obviously some of you have found a way to make it work.

[Nev]

Thanks for the input. Yes I've incorporated the flexible hose mounts - actually replaced them too. Will see how it goes now.

[fairchild1934]

Two things I can think of ---
one is wondering what SS alloy rod was used for the weld bead. There is one specific alloy that is preferred for this job. But it depends on what the alloy of the base metal is. It ---- LOOKS---- like the transition area between the bead and the base material has become hardened--- and brittle. It could be that this area needs to be annealed prior to use. Doing an anneal on SS would be the opposite of annealing a carbon steel
material. I have not heard of doing an anneal on something like this---- but I see that MANY SS exhaust parts get cracks like this pretty fast in use. But the fact that it has cracked like this would seem to indicate that it has become brittle there. if the rod and base material are dis-similar---
then the chemistry at the junction becomes a wild card. Thats why you try to match the alloys. Going from memory- I think I remember that many factory parts are alloy 347. As was said above--- if the back side was not sealed up and purged with argon-- an nasty Very rough and porous oxide forms on the back of the weld and likely gives the bead LOTS of opportunity to start cracks. The thin metal in this very harsh vibration environment will be crack prone under the best of conditions. (like a loudspeaker at a rock concert--) But if the back side was not purged with argon---- that would likely be your problem of all most immediate crack propagation.

I think the idea of short struts (like spokes in a wheel) to support the inner and outer piece against each other might help greatly in case the end cap that was welded - is flexing like a diaphragm and that flex has a fulcrum right where the crack is happening near the base of the inner pipe. Also possibly a strut out of tubing to the engine mount with a rubber isolator on the cold end might dampen movement of the whole thing.

Just throwing out ideas here----- might be cut and try--- somewhat----

You could call Acorn welding in Canada and ask about the idea of annealing after welding. How to do it as far as a temperature schedule- and
if it is a thing they do to prevent cracking. Allowing it to cool slowly may be partly a problem as that would tend to harden the material. But a lot depends on the exact alloy I would bet.

[Nev]

Two things I can think of ---

Clint from Vetterman exhausts recommended using either 347 or 308 rods for the welding, so you're spot on there. I didn't actually watch the welder do this job, however I know he is very particular about using argon to seal and purge, and I've watched him do it on previous jobs. He has been a full time welder for his entire career, RNZAF trained, and worked on aircraft most of his career. The past few years he worked full time welding medical equipment. The second time it cracked he suggested that there might be a torsional force at play. The idea of annealing is an interesting one. I'll ask around. Thanks !