Board edge delamination from mounting stress

[helius]

I was working on this unit for over a week before noticing this...

It could have been my fault, I wasn't 100% careful because disassembling the unit is a giant pain. There is a lot of weight just under this area, with large VFDs.

Is there some material or technique that can reinforce the crack so it doesn't spread further? I suppose if necessary I can cut a slot and epoxy in a section of copper-clad FR4 material, but I really want to take the easy way out if possible. Would wrapping a piece of Kapton tape around the edge be good for stability, or just make it worse?

Thanks for any ideas you may have.

[Koray]

I would apply some epoxy, impregnate the cracked parts  and put pressure with some suitably sized clamps. It might be a good idea to cover the fresh glue with a thin paper or food wrap to avoid clamp adhesion. I don't think you need another piece of material on the existing PCB but if you do that would be even stronger.
K.

[T3sl4co1l]

VFDs aren't that heavy... unless you meant Variable Frequency Drives rather than Vacuum Fluorescent Displays.

It looks more like something sharp hit it in two diagonal whacks.

How deep is the gouge?  What does the rest of the board surface look like, is there evidence of cracking/delamination?  What traces  or vias are nearby, under threat?  What mounting is nearby to carry the stress?

Tim

[helius]

It looks more like something sharp hit it in two diagonal whacks.

That is entirely possible. Like I said, I wasn't 100% careful with this, because it was already a test of my patience. This board has a soldered-in ribbon cable that causes it to hang from the unit during disassembly, before it can be unplugged. That results in a certain amount of banging around. If I had it to do all over again, I would have wrapped the board in pink poly bubble film during the process, or started with desoldering that cable.

How deep is the gouge?  What does the rest of the board surface look like, is there evidence of cracking/delamination?  What traces  or vias are nearby, under threat?  What mounting is nearby to carry the stress?

It's two cracks that are formed from the center section of the edge flexing, not a gouge or scrape. Sorry if that's not clear in the image, it was difficult to focus on the narrow edge.
There are traces running horizontally along this edge (on the front and the rear, as pictured). You can see one of them in the upper right of the image, which has been lifted and twisted by the crack. Outside of this edge area, the board looks fine.
It is mounted with ordinary pan screws to two metal bosses in the top two corners, if the cracked edge is facing up. The opposite corners are unsupported, but there is a tact switch area near the center that is mounted with five sheetmetal screws to a PA6.6-GF stiffener. The screw mounts are basically compression only, they don't have precise registration fingers. Proper torque is kind of a mystery.
I can see a shadow through the board where delamination has spread inward from the two cracks, about 3 sqcm area. This shadow area crosses 10 traces on one side and a wide VCC trace on the other side. 7 PTH and 1 via are inside the shadow area. This is looking worse and worse, but not yet B.E.R. The board is 2-layer with low detail and fully through-hole.

Koray, is there a low-viscosity resin that you recommend for this? Most of the epoxies I found were 6,000 cPs and up, but something water-thin would seem to be able to soak in better to the glass fibers.

[T3sl4co1l]

Ah, so it's mounted at the edges, and something must've pulled back on it so fucking hard that it partially ripped the board, that's insane.  Like, obviously-too-hard prying with a screwdriver, kind of hard.

Lucky the VFDs aren't smashed, I guess?!

Epoxy can be thinned with solvent, or heat (as long as it soaks into final position before curing, which is obviously accelerated with temperature).  Uh, probably the solvent should be mostly evaporated by the time it's gelling up, since it'll shrink a lot as it evaporates.  So, apply a thinned blend, then heat?

Vacuum would be ideal, but maybe incompatible with some components (hmm, electrolytic capacitors especially, but maybe not much else, actually?), and obviously, you need a vacuum chamber first.

Get some stuff in there, then clamp it with something nonstick, say plastic film or teflon sheets, something like that?

Damaged vias can be drilled out and patched with solid wire; traces can be peeled back, scraped clean and patched with wire.  Or the enterprise-grade solution, new trace can even be epoxied down and UV-cure soldermasked over to make it literally good as new.  (Not sure if filling and bonding a delam is kosher in e.g. military repair practice, maybe they'd want it sectioned and replaced?  But whatever process they'd do, that's basically as far as you'd ever need to want to go.)

Tim

[helius]

Actually only one of the corner mounts was there, the other (cast metal) boss had sheared off. So only one corner held this board to its cast metal case. And that weight was then hanging from the ribbon cable. So it's not such a mystery at all.

The vias and PTHs seem okay for now, there is not very much spreading in that shadow area, only a peek of air from the delamination. I can section and replace if necessary, but I would like to take the easy way out and just stabilize the board to see if it works.

I'm not sure about the effect of solvent on epoxy strength. What would you use, DCM?
I noticed that there are UV-curing acrylic adhesives that are really thin, like 300 cPs, and claim to bond well to glass.

[T3sl4co1l]

I would probably just use acetone.

DCM, I'd be weary of; it swells and dissolves (such that it can be said to do so) cured epoxy as well.  Maybe that's actually a virtue: it's able to diffuse out of the laminate as the new resin cures.  But that swelling is even more risk to nearby vias, so, ehh...

It's something I'd maybe want to test first, but this isn't exactly a situation you have many redos on!

Tim