Cracked dial glass - should I try a repair ?

[GrimJosef]

Well I've tried the repair and you can see the results in the attached pictures. While the crack has not disappeared altogether I think it's much improved and I'm going to declare the exercise a success . In fact much of what is still visible is damage at the two surfaces of the glass. Down in the bulk material the adhesive is hardly visible at all. I suspect that either when the crack formed, or in subsequent flexing perhaps, very tiny chips were flaked off the edges where the crack meets the surfaces and it's the effects of these that can now be seen. My justification for this is that in the fourth picture, where I partially repaired a crack in a test piece of glass, the repaired section really is invisible. The crack here was tighter than the one in the PE80 glass and, furthermore, was never flexed significantly. So its edges were never damaged.

In a bit more detail: I started out with the test piece of glass, shown after glueing in pictures 4 and 5. This had a 20mm long 'running' crack coming in from one corner. I initially put a drop of Loctite Glass Bond acrylate onto this and attempted to get it to penetrate the crack. I tried applying the glue to both surfaces, pushing it around with a cocktail stick and also gently flexing the glass (an earlier attempt with a much larger piece of glass and slightly less gentle flexing resulted in a lot of shattered glass on my kitchen floor - not a good start). Despite my efforts though I believe that I completely failed to get any Glass Bond into the crack. Now in fact this was a very tight crack indeed. The crack in the PE80 glass, while 'tight', could be opened a little by flexing because the screwhole stopped it propagating any further. So maybe this wasn't an absolutely fair test of the Glass Bond. But at this stage it was the best I could do.

Having failed with the Glass Bond acrylate I turned to the Opti-Tec epoxy. This was much, much runnier than the acrylate. But still, to my disappointment, a drop placed on the surface did not immediately flood into the test crack. However it did seep in a little at the 'open' end of the crack, even without flexing, and with a bit of effort (pushing around and gentle flexing) over a couple of minutes I managed to get full penetration from the top to the bottom surface over perhaps a 2-3mm length of the crack. At this point it was clear that even if it wasn't great it was a lot better at penetrating than the Glass Bond had been. So I turned to the PE80 glass and set about glueing that with the epoxy. The batch was now perhaps 10 minutes old. It was still, as far as I could tell, as runny as when it came out of its mixing bag.

Getting the epoxy into the crack in the PE80 glass was not difficult. However filling the crack completely was a little more tricky. With a continuous bead of glue overlaying the top of the whole crack length I was left only with air bubbles towards the bottom. I could move these with flexing but I couldn't seem to squeeze them out completely. After a couple minutes trying I ran a second bead of glue over the back surface of the crack and this did the trick. With the glass turned over I could encourage almost all the bubbles out. The only ones which remained were right at the screwhole end of the crack where it was hardest to flex. These are visible as two dark blobs in the third picture (which was taken without flash, unlike the second one) very close to the screwhole. They serve as a reminder of what the whole crack looked like before it was glued. Perhaps with more work I could have eliminated them too. But turning and flexing the glass was causing the glue to run away from the crack and occasionally letting more air back in. In the end I decided to quit while I was ahead. I pushed the stray glue back over the crack, put a blob on each end to seal the ends, and left the whole lot to dry. Checking the unused glue in the pot I reckoned it had thickened a bit after an hour, it was really quite thick after 2 hours and it had become a gel after 3-4 hours.

Turning back to the test piece I noticed that in fact I had managed to get some epoxy penetration down into the crack along more-or-less its whole length. The fifth picture shows the crack itself and also its reflection in the bottom surface of the test glass. There is a clear gap between the crack and its reflection which means that the crack must now no longer reach the reflective surface (you can confirm this if you like by touching a mirror - where your finger actually touches it there will be no gap between the finger and its reflection). So perhaps I had been a bit too impatient with the test piece. If I had left a bead of glue on the crack it may, over time, have wicked in. In this case putting glue onto both surfaces might not have helped since it would inevitably have trapped an air bubble in the middle.

So my conclusion is that for a tight crack like the one in my glass the runny epoxy is definitely the best stuff. Its penetrating properties make all the difference. However if I had been trying to join two separate pieces of glass then it may not have been the best option. The fact that it won't really stay put and that unlike the Glass Bond it can't be persuaded to set quickly may make it harder to achieve a sound, invisible joint than would be the case with the acrylate.

One final consideration is cost. The epoxy is a specialist glue. Its storage life is limited by atmospheric moisture so it comes in a hermetic outer bag containing five separate 4-gramme sachets along with some dessicant. The best price I could find for a 5-sachet bag was about £20, which I paid. The Glass Bond price varies from about £3.50 to £5 depending on the source. I'm mentioning this partly because one sachet of epoxy was much more than I needed. So I'm putting the other four up for sale over on the Sets and Parts Offered forum in case anyone wants to try it without paying for 5 sachets.

And just let me thank everyone again for your help and advice.

Cheers,

GJ

[Nickthedentist]

A very impressive result, and interesting write-up, thanks for sharing it with us!

Comparing the first pics in posts 1 and 21, you can see how well it's worked.

I look forward to seeing the finished set.

N.

[GrimJosef]

Thanks Nick. I tried to take the first two pics in today's post in exactly the same lighting conditions as I used in post 1. If anything I think the crack looks better in reality now than the latest pics suggest.

Cheers,

GJ

[Patrick Dixon]

That looks like a really good result!

[emeritus]

A really impressive result!

Re the Loctite, something that might warrant investigation is whether it gets significantly more runny if heated. I know Araldite certainly does, having often heated the items to be joined (such as ceramics) to promote rapid setting of standard Araldite. I can see that there might be a problem with radio glassware, as heating the glass to a high enough temperature to make the adhesive runny could damage the paint.

[GrimJosef]

I did think about heating, both for the acrylate and also for the epoxy. I have used it in the past with epoxy to very good effect and I didn't need to raise the temeprature far at all. I seem to recall that 50-60 deg C made the stuff I was using then (Varian 'Torr Seal') much runnier. The downside was that it also accelerated the curing. So if we were going to heat the epoxy then we had to work very fast indeed before it started to set. I don't know if it would have the same effect on the Loctite though.

Cheers,

GJ