Welding on an Aluminum Hydraulic Tank

Some flash points you can get from http://www.engineeringtoolbox.com/flash-point-fuels-d_937.html. The aluminium melting point is 1220.58 °F that is high above the flash point of any oil, above the fire point to. Aluminium is transmitting heat wery well so if the tank wall is not wery tick, you will be above it.

If you don't purge with nitrogen, weld well BELOW the fluid level - note BELOW

If you are above the red hot metal has a high probability (certainty) of igniting the vapours, if you are below the fluid level the fluid itself will draw the heat away very quickly preventing ignition.

I have had to weld many hydraulic tanks and related components on things that could not be practically drained or purged all the way. What I do is make sure the vents are open all the way in case it decides to burp.

So far I have never found a hydraulic system that ever put out enough combustible vapors to ever ignite no mater how hard I tried. Even when welling below the fluid level at most all I have ever got was some smoke from localized fluid boiling but still that in itself would not ignite and hold a flame by itself.

Now fuel tanks are a whole different beast to weld on!

You should be able repair the tank without completely draining, unless it is some exotic hydrolic fluid.

Backpurging with argon (whatever shield gas the tig machine is using) is the right thing to do, though I have seen several places where backpurging is not standard practice.....and not just with hydraulic fluid, but commonly on diesel fuel tank repairs as well, with white diesel fumes pouring out of the refueling line. I wouldn't do it without backpurging, but that is a quality not safety issue.

If the repair area is thick walled or you will be repairing in close proximity to the fluid, heat is more of a concern. Consider clamping a large thermal mass of copper to the outside of the tank wall between the repair area and the approximate level of fluid in the tank, to avoid cooking the fluid to sludge.

A good pulser is helpful, but I'd say that about any tig job.

You can't use die-penetrant, so pressure testing would be the way to go.

One thing you said concerns me; you mentioned planning to 'patch over crack'. That is not a good way to repair. It is a really bad way to repair aluminum, especially aluminum with cyclic stresses.

The crack must be removed, or it will most likely continue to grow and soon ruin the patch.

100% penetration is very important for aluminum under cyclic stress.

Additional gussets or reinforcement in the form of a patch often has unintended consequences, creating new problems often just outside the new support. It is definitely not a 'more is better' scenario.

Tell us how it goes.

Use your TIG torch to fill the tank with argon gas, put a moderate seal with a weight (not threading tap) and work fast.

What a lot of presumptions. That is a BIG risk! What are the gases that will come out when the hydraulic fluid is overheated? Can they recombine and explode? What is the pressure? Why did it crack? How will you test it, with it full to the brim to prevent bursting all over the place? Hmmm.

Ormondotvos:

My comments were based on personal experience and 1st hand observations, not presumptions.

No appreciable concentration any gas from heated hydraulic oil will be allowed to build with proper purge set up and flow.

Pressure testing for leaks requires only determining if a (very) low pressure differential can be maintained over an extended period, say 24 hours. A high fluid level, minimal change in temperature and increasing time period increase sensitivity, allowing for minimal pressure differential. Pressures would be similar to what might be expected in service.

Pressure test is a go /no-go test. It does not locate the leak, it only tells you that you have a leak or that you don't. Being easy cheap and reliable it is useful.

How will you check for leaks? This has to be done whether the tank is drained and cleaned, or not. Will you send it out without testing? Pay for something like radiographic testing? Always respond to need for repaairs by ordering a new replacement.

Presumptions? 'Presumption' might more accurately describes the route taken to your stated opinions. Were your statements based on experience tig repairs on tanks, or were you 'pre-assuming'?

Your query is well stated, unfortunately CR4 does receive a lot of ill informed posts which should be be corrected.

Regards testing the weld as this is a storage tank if possible all openings could be sealed and the argon very slightly pressurised and do a bubble leak test on the weld. Even if there is a slight leak of gas from any ports it only takes a few seconds to check the repair, If there is a problem the weld can be repaired without having to re purge the tank with expensive gas. Also I do not recommend CO2 if a good weld is to be had.

Hopefully one weld is all that is required.

You will find that the crack is contaminated with oil and a good weld cannot be made unless you groove out the crack and clean throughly , also the tank should be purged with Argon if you want your root weld to be effective.

There must be a design fault with the tank to have cracks appear or are they from the origonal shaping of the tank where stress points were caused?

I have welded on many tanks with either tig or mig and I always connect a vacuum to the tank to evacuate vapors and fumes while I am welding. There is an upper and lower explosive limit to any combustible mixture and I figure why take a chance.

I'm not talking about vacuum that will collapse a tank. I'm talking about a shop vacuum that you keep the nozzle slightly away from the opening and by the way I have welded on gasoline tanks that still had a small amount of fresh gasoline still in the tank. Never had a problem with either collapsing a tank or an explosion because if the vapors are evacuated there is not enough mixture to cause a problem. I lost track of how many times I have used this method and I have been welding for a lot of years.

In addition to what Garth wrote about grooving out the crack, I would add that prior steps would be:

1. Clean all coatings off of the area of the crack

2. Do a dye penetrant check of the crack and surrounding area (the crack can be significantly bigger than what the naked eye reveals)

3. Once the extent of the crack is determined, center punch and then drill-stop the crack at it's extremes with I'd say a 3/16" or so drill (depending on the thickness of the plate). The radius of the drilled hole will relieve the immediate area of stress that caused the crack so that it isn't compounded by the stress added when it's welded up.

4. Then vee out the crack for full penetration, blanket the tank with an inert gas as mentioned before and weld'er up.

There is no guarantee that you won't get another crack, as it depends on what caused the first one.

thank you for all answers ,the weld job came out good. with good venting diamond patch worked perfect