View Single Post
Old 03-31-2021, 11:32 PM   #8
Registered User
MP&C's Avatar
Join Date: Jan 2013
Location: Leonardtown, MD
Posts: 1,525
Re: Roof patch causing warp / oil canning

Not that it helps in the center of a roof, but the following should help in understanding warpage.

I hear people all the time suggesting to only use the bare minimum amount of a patch panel needed to repair the rust. I prefer to look at it another way, For example, most rocker panels have an outward crown that, if we cut out a patch for the front 1/3 only, we are left with a vertical seam. The shrinking that will indeed take place is going to give us a nice valley right at the vertical weld seam. So IMHO, it makes more sense to install a FULL rocker panel, where most are installed with spot (plug) welds at top and bottom flanges. Without using a vertical butt weld seam that you'd need for a partial patch, there is virtually no warpage at all in installing the full panel.

For any other patches, (quarter panel, door corners, etc.), I worry less about using the least amount of patch possible, and more about putting a weld seam where you do have access to planishing. Additionally, a half height quarter patch puts your weld seam in the middle of a low crown area for a guaranteed warpfest. Using a taller quarter (even if not the OEM full height) puts the weld seam up higher where the profile usually has a higher crown, which helps to control the warping a bit better. So if given the choice here, I am using the tallest quarter available and putting the seam:

1) where I have access for planishing

2) in a higher crown area to help control warping

3) near body crease to help control warping (keeping enough distance for dolly placement).

Other body panels would follow same 1-3 considerations in that same order.

For trimming, I try to keep absolute tight joints and if using MIG, set the welder hotter, adjust wire feed faster if it tends to blow holes, and control heat with shorter length of time on the trigger pull. first and foremost, we need to insure full weld penetration, so some practice on scraps of same material thickness and clamped in mid air to simulate what is on the car. (metal welding table is a heat sink) To further explain gap vs no gap, anything welded is going to shrink when the weld cools. If you have a gap, the weld will pull the panels together as it cools, and each subsequent weld will pull it that much closer together. So if your panel had been trimmed correctly, now you have low areas being put in as the panel is moving. If the new patch is trimmed for a tight joint, then it will still shrink, but you will only need to planish to overcome weld shrinkage, and should not have panels moving together creating low areas. In essence, it will still need planishing as would a gapped joint, but with no gap it should remain more consistent where any planishing required is also consistent. With panels pulling together in a gapped situation, you will have differing planishing requirements based on how much the panel pulled in that particular area. Hope my rambling makes sense.

I had done some test welds a few years back and I think the pictures taken will help out in understanding the weld location and shrinking. The tacks were done using the TIG and NO filler for minimal warpage. This also means we need absolutely tight joints... Here's the tacking process, and as said in video, amperage is set at 70. Based on 18 gauge thickness this should have been about 45, but as we also do with MIG "dot" welding, higher amperage and less elapsed time on trigger pull = flatter welds, less HAZ.

Note minimal weld size, minimal HAZ with the higher amperage, shorter burst...

Patches started out flat and for the most part remained so..

Adding a weld pass we are quick to see some distortion...

Examining this further, even though we have absolutely tight gaps for less instance of the panels pulling together, we still see distortion.. This is your typical weld shrinkage as the weld cools. Note in the next picture the panel is still fairly flat along the edges (red line), some shrinking at the weld (yellow arrows) and show a dramatic pucker between the two. Note that the weld has yet to be planished, so the weld shrinkage is pulling the metal alongside it together, the areas unaffected by heat remain largely unchanged (red line) and the area between the two are forming a bulge due to these differing forces. Here we address the problem, not the result. Planish out the weld to stretch it in length and the bulge will disappear. Don't make a habit of chasing the result, a shrinking disc on the bulge is not the correct resolution; if this were a crowned panel that action would be causing a severe low area.

Referring back to an earlier statement I made on weld location:

......let's try this same scenario using a crowned sample near a body crease so we can take advantage of all 3 choices...

Weld pass....

Here we can see how the weld location and panel features (crown, body crease) helped to control and limit any warping effects. The weld will still need planishing to restore the crown of the center bead, as no doubt it has pulled in slightly, but this is hands down a dramatic improvement over the flat "patches" we did the first time. This shows how these features in your body panels can help out in controlling weld distortion, so take advantage of these in weld location and leave the limiting of panel size as your absolute last consideration.

Last edited by MP&C; 04-01-2021 at 06:23 AM.
MP&C is offline   Reply With Quote