|
|
The previous owner of my aluminum intercooler reconfigured the inlet/outlet at some point and welded plugs in the old holes, which were then ground flush. I was doing a boost leak test when I discovered a number of pinhole leaks. Now I'm stuck with the job of fixing numerous pinhole leaks along the welds which were ground flush. All the holes are extremely small. In fact, most of them are invisible to the naked eye; it isn't until you apply pressure and spray some soapy water that bubbles begin to magically grow from seemingly nothing.My question is, what is the standard procedure for correcting a pinhole leak? Is it necessary to first drill out every single pinhole to remove the oxides formed within the pinhole? Do I approach it as if I'm laying just another ordinary weld bead (ie - filler metal) or do I just heat the area of the pinhole until it forms a weld puddle and then allow it to cool (ie - no filler metal... autogenous or whatever its called)?Is there anything in particular that makes repairing pinhole leaks inherently problematic?
Reply:Pics would make all the difference....and the test pressure is???AL is really good at making scattered porosity and microscopic leak paths, mostlyfrom inside contamination, poor joint design, lack of prep, poor techniqueMust add filler to make new puddle that just barely fuses to the old--lots of fusion,big wet-out, allow porosity gas bubbles to float up thru the puddle (leaving a microscopic leakpath behind as they move). This exercise can be a real hair-tearer--normally isan exercise in futility. Drilling out the pores will do nothing to help, with apparently, big time, full bore, massive porosity....and most of that--you haven't even seen, since it's subsurface.Best practice is to cutout and re-do the joints properly.Next best practice is to close fit formed, prepped patches over the areas, then weld out.More likely than not you'll be doing one of the two practices mentioned--but onlyafter pissing away time trying to goober some welds-repeatedly, over the bad areas,only to find new areas leaking. Fading out very slowly with the TIG torch at each stop, to avoid craters is required.Every crater has cracks and a microscopic leak path back to the inside.Dabbing a little bit of rod to add a touch of new, wet melt onto the top of a puddle showingsome 'suck back' while fading out--helps.Do not sand/grind down the new welded areas. Sanding/grinding/hard buffing can also heat up thin AL walls with suspect welds, enoughto initiate microscopic cracking.Blackbird
Reply:I would just clean up the area and run a bead over the old one adding a bit of filler to fix the holes as I go. While welding the holes should open up and be apparent as your welding.But also if they are so small that they only make tiny bubbles under pressure it may not be a big enough leak to even worry about on a turbo car/truck. Throttle bodies leak the most on pretty much any car to the point you can hear the air escaping. Very very tiny holes don't mean anything most of the time.
Reply:The air leaking out is insignificant. The problem is that the joint is faulty and will develop into a crack after being repeatedly pressurized and it will eventually fail completely.Welding/Fab Pics: www.UtahWeld.com
Reply:I have done a lot of similar projects and would do it exactly like VPT said. Clean and prep the surface just as you would for a fresh weld, turn your AC a little more to the cleaning side than you normally would and run fresh beads over the area that is leaking. If at all possible I would leave the weld beads untouched afterwards. If you absolutely must go for the ground smooth look then make sure you run it hot enough to get good penetration and fill in the al deeper than you will be grinding back down to. If there are a lot of leaks I would use a sharpie to mark the border of the leaks and then weld out to that border. After that pressurize it again and see if there are any more leaks. If you find any then mark it with a sharpie and just hit that spot. dave powelson brings up some good points however I have never had much trouble with ICs. They are usually made out of half decent aluminum and weld up pretty nice. Even porosity areas patch up with little effort. Now if it's dirty cast or heavily contaminated on the other hand your life can turn into a nightmare.Lincoln precision TIG 275Millermatic 140 MIG
Reply:If you do attempt a weld repair of a pinhole leak in the intercooler, do your best to remove oil residue in the vicinity of the back side of the weld. Perhaps pour in solvent and swish it around, dump out, repeat as needed. You also need to probably clean out the solvent as much as you can. (Something like a heat gun or hair dryer blown through the intercooler should help evaporate the solvent)This may not be easy or convenient, but if you don't do it (and the backside is oily or with solvent residue when you begin welding), you could easily make things worse attempting a weld repair.Also, special process considerations and cleanliness may be warranted to prevent dirt, dust, or especially metal shavings from ingressing into the induction tract while you're working on the intercooler and boost tubes. Such particles ingested by a motor are not be conducive to good performance or longevity. An intercooler core (especially after end tanks are installed) seems particularly difficult to reliably remove metal shavings out of. So, the best strategy is probably to never allow the metal shavings to get inside of the intercooler in the first place.I like the other tips provided by the previous responders, but hope this helps give some extra food for thought.Last edited by jakeru; 06-23-2013 at 05:30 PM.
Reply:Originally Posted by jakeruIf you do attempt a weld repair of a pinhole leak in the intercooler, do your best to remove oil residue in the vicinity of the back side of the weld. Perhaps pour in solvent and swish it around, dump out, repeat as needed. You also need to probably clean out the solvent as much as you can. (Something like a heat gun or hair dryer blown through the intercooler should help evaporate the solvent) |
|