‘Cutting Edge Engineering’ attempts to repair a large casting

[Ches Green UKParticipant @chesgreenuk]

'Challenging CRACKED Cast Iron Housing Repair! | Flame Spray Welding' –
https://www.youtube.com/watch?v=IoL2kqel46U

Another interesting video from Kurtis and the team. The Dye Penetrant certainly highlighted the cracks.

Ches

[Ches Green UKParticipant @chesgreenuk]

[JasonBModerator @jasonb]

It was going so well until………………………..

[Ches Green UKParticipant @chesgreenuk]

Yes, Kurtis knew it probably only had a 50:50 chance of success, if that…. but when it went wrong, it sure went wrong. Quite impressive in a way.

Seems the overwhelmed local expert puts the casting in a large oven and extremely carefully controls the temperature of the casting whilst carrying out the procedure.

Some of the Commenters also noted it was possibly a poor quality casting, and that the impurities might have migrated to the top surface.

Ches

[SillyOldDufferModerator @sillyoldduffer]

Kurtis is impressive. I've watched many Cutting Edge videos in hope of catching him do something unwise; no luck so far. Really knows his materials and methods, with none of the incidental blunders and bad practice that spoil lesser Youtube efforts.

I guess Karen is the talent behind filming and editing because the videos are so well-made too. Not many can do tasty engineering and make a slick video as well.

My only complaint is the dog! Far too well-behaved, plus I'm a cat person…

Dave

[Clive FosterParticipant @clivefoster55965]

Kurtis is generally impressive but he really should have known better than to attack a casting showing that sort of crack issues with a hot welding process putting great gobs of metal into large gouges. A splendid illustration of why "metal-loc" and the various other cold, stitching, repair processes were invented.

Basic cause is that the casting was riddled with grain boundary dislocations and similar nascent cracks formed during manufacture or post manufacture heat treatment. The part has high tensile stresses across the line where the cracks have formed. Given the low tensile strength of cast iron the part is already close to failure so anything whether temperature cycling or mechanical loads will cause the material along the nascent cracks to separate reliving stresses elsewhere.

My guess is that someone cocked up a post casting heat treatment on a casting that wasn't as well bonded together internally as you'd ideally like leaving multitude of tensile stresses just waiting to turn into cracks. Heat it to nearly red and allow it to cool naturally and the inherent stresses of uneven cooling, due to wall thickness variation if nothing else, will cause more cracks to appear. Throw in significant thicknesses of even hotter, melted metal which will contract more than the coast iron as it cools and it's an open invitation for cracks to appear all over the place. Trying to figure out where the tensile forces induced by differential cooling will cause a not quite enough stress to crack to turn into an actual crack is probably impossible.

Ideally the repair needs to apply an expansive force to the cracks being fixed further relieving stresses in the main body whose combined efforts had generated the crack inn the first place. Pretty much wedging it back together.

The hot metal processes are largely fine when you are sticking a broken off bit back on given requisite care to keep the extra cooling stress somewhere safe. But dealing with main body cracks is an art form, assuming it can be done at all.

This is probably the one place where the 'field expedient" puddling technique using an arc welder with ductile rods is actually the one most likely to produce a good repair.

Start by gouging out the first crack you are going to fix for access, I'd not take out as much material as Kurtis did and I'd try to keep a sharpish bottom V rather than the near half round he used. Then using the thinnest rod you can with just enough current to bond put a thin layer of weld on one side of your gouge. Tap it with your chipping hammer as it cools peening the hot weld to largely relieve tensile stresses formed during cooling. Do the other side. Rinse and repeat a couple of times stopping if the casting shows any signs of getting significantly warm. When it does go find something else to do for an hour whilst it cools right down. When you start joining the bottom of the Vee make special effort with the peening. When the gouge is about half full you can move up to a slightly bigger rod and accept the repair area staying little warmer so things go faster but you still have to keep peening trying to drive the weld metal into compression.

If you managed a perfect job (Ha! optimist) all the cooling tensile stresses are confined to weld metal, hence the need for a ductile rod, so the main casting sits as was cold retaining the internal stress release resulting from the original crack opening up.

Rinse and repeat for the next crack. Takes forever. On a job like that fixing two cracks a day would be good going. That said most of the time you are doing other things whilst it cools down properly.

Objectively it's not a commercial technique but it does work. I've used it a time or six when needs must. Impurities are still a major bugbear. Did a bit of Victorian decorative cast iron metal work that appeared to be about 50% included carbon. I looked like coal miner by the time I'd finished and the amount of weld for simple break was stupid but got 'er done! Never, absolutely not ever, again! On the other hand a broken banjo off a SouthBend lathe went together so well that it might as well have been mild steel.

Clive

Edited By Clive Foster on 22/09/2023 20:38:13

[Clive FosterParticipant @clivefoster55965]

Postscript about strength of puddled repairs as described in previous post.

There is no reliable way to know how strong a puddled repair is. What you are effectively trying to do is wedging the casting back together where it has cracked. Hopefully receiving some of the stresses behind that helped open the crack.

If you have pure compressive stresses its at least as safe as the original casting.

Any tensile or bending loads and you are in the lap of the gods. It will be stronger than the cracked casting but no way of knowing how much by as the bond made on the first puddling layer must be weaker than a deeper adhesion from a brazed or more sophisticated weld. Never as strong as a good casting would have been. But most successful "ordinary" castings are low stressed anyway. Crankshafts et al are a whole different world.

Clive

[Chris MateParticipant @chrismate31303]

Primary:Factory
This was an interesting one, looking at the failures//cracks before he attempted to repair, my thoughts were poor quality casting by factory, maybe a batch like that, and the torgue spec applied to all those bolts and over time heat & cold cycles it cracked like that. Maybe get that cast iron tested and compared to another similar part not cracked.

2ndary:Repair attempt
After I saw how a cracked head was repaired in a 4th world country, where they heated up the complete head in a coal fire, wild job, I think he's heat was not enough for the complete casting doing such a repair, only then one would know if its actually repairable.

[Kiwi BlokeParticipant @kiwibloke62605]

Clive – fascinating! Please tell us more!

I assume your 'puddling' technique is stick welding – correct? What kind of rod is suitable? Is there any use for TIG welding in this application? I'm thinking that its smaller heat-affected zone could be either a blessing or a curse…

[Bo’sunParticipant @bosun58570]

I take my hat off to Kurtis for giving it his best shot. He did at least warn the customer it may not work. We could all learn a thing or two from his approach to repairing stuff.

[Ady1Participant @ady1]

So much cracking basically means end of life for the metal

At 5.46 the entire boss is trying to rip itself out of the casting so the metal in the metal casting is actually failing to deal with the loads

If he'd fixed it perfectly to 200% another bunch of cracks would appear just beyond his perfect fix

If the machine isn't too old then it's a bad casting

I've seen it with bad steel, cracks appearing magically in the middle of a plate under load

edit: If a lot of users start getting this problem then the manufacturing company has possibly taken the thin-ness of the casting to it's limits to save on costs and vehicle weight

It's also possible that modern castings have a built in design life, making it a consumable, unlike the old days when they were made to last 100 years

Edited By Ady1 on 25/09/2023 10:51:36

[Clive FosterParticipant @clivefoster55965]

Kiwi Bloke

Regrettably I know little more about the pudding process beyond discovering that it worked well after being told that was the way to repair the banjo casting on my SouthBend Ten after discovering a previous owner had over-tensioned the bolt snapping one side of the loop off the arm. I used the DIY market arc welding rods I had and made the repair around a mandrel of the same diameter as the SouthBend mount. The result was, so far as I could see, as round as the original. The banjo moved freely without rock and clamped up properly with moderate bolt torque. Which confirmed that very little, if any, contraction occurred a the repair joint. Any contraction would have opened things out and pulled the hole out of round.

Subsequent efforts have proven to my satisfaction that it's a suitable process in the right circumstances. Takes forever though.

Raising the subject with folk who do know cast iron welding, whether in person or on line gets basically this response.

Expert "Naughty boy. You really shouldn't do this. Totally wrong way to go about things. It has never been an approved technique although folk used to do it back in the dark ages."

Me "But it works just fine if you are careful."

Expert "Of course it does. Can be more than decently strong too."

Me. "??????"

Expert "But you must never, ever do it. Its wrong."

Me "OKay."

Thinks naughty words and chalks it up to yet another time when ignoring the expert is the way to go. One thing I learned in 40 odd years as a Scientist / R&D engineer is that experts are very handy people to have around but there are limits.

Clive

[SillyOldDufferModerator @sillyoldduffer]

Posted by Clive Foster on 25/09/2023 13:09:34:

…

Thinks naughty words and chalks it up to yet another time when ignoring the expert is the way to go. One thing I learned in 40 odd years as a Scientist / R&D engineer is that experts are very handy people to have around but there are limits.

Clive

Experts may not be infallible, but faulty logic to think they can be ignored because ignorant guesses are just as valid, They're not – experts are more likely to get it right than anyone else. Judgement beats luck on average, which is why I don't waste money on lottery tickets. Experts tell me the odds against winning big are about 14 million to one; the fact that one person wins does not alter the fact that there are also 13,999,999 losers!

I prefer Clive's earlier comment: 'If you have pure compressive stresses its at least as safe as the original casting. Any tensile or bending loads and you are in the lap of the gods. It will be stronger than the cracked casting but no way of knowing how much by as the bond made on the first puddling layer must be weaker than a deeper adhesion from a brazed or more sophisticated weld. Never as strong as a good casting would have been.'

Experts can't recommend the technique because it doesn't handle common stresses, and is unreliable to boot. If cast-iron has to be repaired, they would argue, why not do it properly? The answer is that gambles are sometimes worth trying, especially at home. If a chancy repair works, hurrah. Otherwise no worse off than before.

But thanks to Clive's explanation, I'm now more expert than I was on puddling! The education makes it possible for me to better judge if the method is worth trying should I need to repair cast-iron.

I guess faulty heat-treatment was the problem with the Kurtis' casting. Large, complex, thin castings are likely to develop internal stresses if they are cooled unevenly. Who knows, perhaps in the factory it was last out of the oven at close of play, and was brought out 20 minutes early so the shift could go home. In the rush to leave they may have dropped the temperature extra quickly by opening all the doors on a cold-day.

We live in an imperfect world. Remember the British workers who faked Nuclear Fuel certificates so they could watch football on TV?

Dave

[Clive FosterParticipant @clivefoster55965]

Ady1

I'm not convinced that such cracking is necessarily end of life for an iron casting on a hefty piece of equipment.

Way I see it is that theoretically a casting is a nice essentially homogeneous lump of iron which has cooled evenly so all the little grains are stuck nicely together and just sit there without mutual stress.

If anyone thinks that ever happens in practice I have a wonderful deal on a bridge. Used notes only.

In reality there is a tensile stress across every grain boundary, a necessary condition of the formation of the surfaces that define a grain, which could expand to a crack given sufficient force in the right place. In the body of a decently solid casting everything is decently well balanced. No sufficiently large forces to start a crack and drive it past grain boundaries can exist. So all is fine. Important to remember that although the stresses across a grain boundary could be crack starters the boundaries are also crack stoppers. I'm told the maths and science is really, really complicated.

Cut a hole in the casting (or drastically change its section) and a new boundary is created taking things out of balance. As material is removed the balance is most likely changed to increase the tensile stress at grain boundaries around the hole. In a good, well designed casting all is fine.

However if a crack does start inside and run across into a hole or surface it's proof that somewhere there is enough tensile force to break the casting along the line of the crack. The crack opening up has relieved some of the stress further back. So, in some ways, the casting may overall be stronger as the points at which opening crack has received the stress can take more load before breaking.

Its no different to the bottom rivets in Kiplings "Ship That Found Herself" giving a little so the garboard strake gained a fraction of an inch of play letting the ship go along much more easily.

Significant difference between giving a little and failing completely of course.

An effective way visualise whats going on with such cracks is to consider a seesaw held level by ratchet straps at each end. Tightening the straps just bends the seesaw a bit but things stay stable. But clearly it will now take less load to break the seesaw. The tighter the straps the effectively weaker the seesaw becomes as more and more of its strength is absorbed in resisting the straps.

By analogy this is the same as a piece of casting strained between two potential cracks.

If one strap is sufficiently weaker than the other to fail before the seesaw breaks it will snap. That end of the seesaw goes up. The other goes down to the ground. The bending stress in the blade disappears, so its full strength is available to resist loads. Stick a scaffold pole or Acrow prop under the upward end and everything is stable with the down and fully supported by the ground. Overall it's actually stronger than before. But a, hopefully slightly, different shape.

Conversely put a new stronger strap on and winch things up good'n tight. Then add a bit more for good measure. Odds are the other strap will break and you are back where you started except t'other way about.

Going back to our casting the crack opening up to the hole has relieved some internal stresses. So if you repair it by a cold lock or puddling process putting minimal closing stress on the crack balance will be restored and overall things may well be a little stronger.

A hot process putting significant quantities of molten metal into the hole that contracts when it cools is effectively the same as the stronger strap in the seesaw analogy. Fixing the crack has put more stress on the nascent one at the other end due to contraction of the filler. Maybe enough to turn that into a real crack.

Real castings are way more complex than the seesaw but, as Kurtis found, adding significant contraction forces when repairing cracks upsets the balance elsewhere in the casting forming lots of new cracks.

If you heat the casting before hot repair contraction stresses due to cooling can be much less so the repair is satisfactory. But there is always some extra stress involved. Knowing what to do to minimise this added stress is an art form.

Clive

[Peter Jones 20Participant @peterjones20]

Guess I’m not the only one who spends too much time on YouTube?

I posted he should have tried the PK Truck method of pre-heating, fill it with charcoal so the entire thing gets hot.

Seeing some of the repairs they (PK Truck) do that I would class impossible if I hadn’t seen it (although I’m also sure they have a high failure rate)

[Author]

Posts