What's new
What's new

Patch material for aluminium casting repair suggestions


Hot Rolled
May 9, 2004
Plymouth, Devon, England
Part to be repaired is a Japanese motorcycle crank case manufactured around year 2000, this is sand cast aluminium of very good quality but unknown composition though I assume fairly high silicon content.
Unfortunately some Herbert went at it like the proverbial bull in the china shop & there is a large crack alongside the full length of the inside angle & across to the culprit weld.

The casing is water cooled beneath the area in question so I have no access from behind, my intention is to machine out the damaged area to the extent of the cracks & machine up a semi circular patch plate from suitable bar stock & weld in. I do have experience in this type of repair so am familiar with techniques to minimize distortion etc all of which will be employed in my attempt, nonetheless experienced comment would be welcomed.

My question is ..... what grade of aluminium bar stock might be compatible with the casting in terms of composition & thermal properties, weldability will be important as I certainly need to minimize the weld lapping up the heavier section vertical side of the inner corner.




  • DSCF0001s.jpg
    59.5 KB · Views: 1,277
  • DSCF0007s.jpg
    52 KB · Views: 1,402
You sure its sand cast aluminium? A lot of modern bikes are a cast higher magnesium content alloy which is why so many people find them harder to weld.
Thanks SWB, the chart is certainly worth keeping.

Adama, I have wondered lately where the magnesium / aluminium dividing line has moved, I occasionaly weld magnesium but prefer not to as the newer stuff seems to be somewhow neither one nor the other.
Fortunately I will have the scrap section that I machine out of the casing upon which to practise a little.


Aluminum and mag are soluble in each other up to ten percent, but that doesn't affect the weldability. If the crankcase is not oil-soaked it's probably metal impurity from other elements making problems.
Crack happened when the rod hit it the first time

Prolly can repair with mechanical means loc stitch etc.

No chance for a succesful weld repair.

May also do a proper wicking grade locktite repair.
Tim....looks to me like someone went in with bugger all pre-heat, prep or aforethought. Looks like mag filler rod blasted on with way too many amps.
All my repair work is done with TIG.

Heavy....just got to hope I can prove you wrong regarding "no chance for a succesful weld repair" ;)


Last edited:
This should be repairable.

Firstly, I would highly recommend sending the casting out for thermal cleaning. This will basically bake out most of the crud soaked into the casting.

Secondly, you MUST PREHEAT AND POST HEAT. You will want to keep the casting in an oven at several hundred degrees for long enough to have the whole casting completely evenly heated, inside and out. The temperature isn't too critical, though the hotter the better as long as it doesn't interfere with handling the casting and doesn't exceed whatever the critical point is for that particular alloy. With a large section like that, you will likely need to weld a section and then stick it in the oven to keep the heat in. Firebrick will also help to keep the heat in, instead of setting the casting on a cold welding table.

You MUST find the ends of the crack (possibly farther than what you can see at room temperature) and then drill out the ends of the crack to prevent progression of the crack after a topical repair.

Grind out the crack, but I think you already know that.

Hopefully your TIG machine has some wave shaping functionality. You'll want to bias the output for more cleaning, as even after thermal cleaning there will still be a lot of crud left and the added cleaning action will help deal with that. You'll also want to use the pulse function, but not in the "cosmetic" way. Running at about 33PPS (pulses per second) with about 33% background current will allow you to up the current to help blast the crud off without overheating the weld local.

Post heat. Stick it back in the oven and ramp down the temperatures over a few hours. It makes a huge difference to keep it warm longer and cool it down much more slowly than you might think is necessary.

Filler metal would be best to experiment with, if you can find junk casting to practice on.
What he said, if it's a modern casting.

If oit's an antique, something I haven't seen mentioned here is "buttering." A 1930s Harley case is an example of cast aluminum so porous that after you carefully grind out the weld, the first application of heat brings oil bubbling out of the material. The cleaning action of the arc won't do much to address this problem. You just have to work as fluid as possible so it will float out into the puddle. Then let the first pass freeze with all the crud trapped in it. Then you grind the first pass out, and the entrapped crud with it. Multiple repetitions may be required before you are dealing with clean metal. On a Harley of that vintage your weld will be by far the best part of the case.
Having welded many,many to many ,2 stroke cases , cylinderheads and various high and low quality alum castings. Bolt the crankcases together and put in an oven at about 500*F for about 4 Hours.
I say to bolt the cases together because the damaged case tends to distort if heated by its self. Let most of the entrapped oil burn off. Let the cases cool slowly. Grind clean the bake again till all of the residues are gone. Then take you a sheet of copper , and make a backup bar to hold your weld filler in the damaged area, it also helps hold you Argon-Helium Mix shielding gas . Devise a way to hold it in place with the cases bolted together, and put back in the oven and preheat to about 350*F , 2 hours should do it. Trace around on the repair area with your arc to help clean and burn off some of the remaining contaminants. I usually try and boil the crap to the top and brush it clean with a virgin SS wire brush. I use 4145 or 4047 filler but they are pricey, 4043 will work also , but it is less tolerant of remaining contamination. Like Oldwrench said , weld it ,freeze the crap in the puddle and grind or brush it out , till it fills in good . You will probably have more problems with outgassing than any thing else. Good Luck
This should be repairable. Filler metal would be best to experiment with, if you can find junk casting to practice on.

Thanks for your detailed reply, most of which are similar to techniques I use though I will say that I have never thought to use pulse for the purpose you describe :scratchchin: makes very good sense & I will do some experimenting.

TIG set is a Kemppi Mastertig 2500 with all bells & whistles so plenty of adjustment available. Sadly the damaged casing is about as rare as can be so no chance of a junk one, I will have only the small section that will be machined out.

Oldwrench......I do work on some older stuff occasionaly & buttering is invariably the only way to get the metal clean, I have spent many hours with the carbide burr ;)


Last edited:
I say to bolt the cases together because the damaged case tends to distort if heated by its self. Good Luck

Again, thank you for your detailed reply & sound suggestions. The case can not be worked on whilst bolted to the upper but am machining up a 1" thick support plate to which the casing bolts will be torqued, I find this method very good for minimizing distortion in large casting repairs & the extra mass helps keep the heat in the job.


Dunitagain has given sound advice, even if all goes well I would still be leery of the crack in the web of the case. all of the torque is transmitted to that exact spot from the crankshaft. Well at least half of it.

His methods are spot on, and will work. You need to grind out the original weld.... That definitely looks like stick rod repair.

Best of Luck