Gusset Renovation - Shoe Cream ?

[Glug]

Hiya folks,
I've been taking a closer look at a leaky bellows and discovered some pinhole leaks and what look like thin spots.
(using a torch inside the bellows there are light spots and at least one definite hole).

It really needs replacement, but I thought I'd see what can be done: so patching the pinholes or similar.

Anyway, long story short, apparently gussets become less airtight with age and will leak through the leather even if there's no hole.
And one fix is to apply a leather treatment: Meltonian neutral shoe cream is recommended.

Here's the source: https://www.concertina.net/forums/index.php?/topic/16772-hands-free-bellows-testing/
It's a very interesting read.

I've already replaced 2 gussets on this bellows but I've got 4 new ones in the cupboard so if I ruin an existing one it's not a problem.
Currently I'm thinking of putting a smear of Bostik Leather Adhesive on the pinhole and the thin areas and just letting it dry (it's flexible).

 

[Gonk]

Definitely interesting. One potential concern: while concertina bellows are entirely covered in leather, accordion bellows corners are leather adhered to card and paper with rather light glue. I'd be a little worried about the oils from a shoe cream spreading into the paper bits, and potentially limiting the ability of glues to get a good fix (now or later). I'd use very small quantities if I were trying it, and perhaps keep the bellows spread until it cures.

I've been fixing similar issues (pinhole leaks and thin crispy bellows corners) by shaving very thin pieces of flexible top-grain leather with a leather skiver, and gluing them to the inside of the corners with flexible fabri-tac. So far so good, although if I go too thin on the skiving, the patches risk having pores as well.

 

[Glug]

Yep, concertina bellows are a bit different, but one of the comments does mention bellows papers and the theory should still apply.
They also say they're applying shoe cream with a small artists paint brush.

He found doping the leather with PVA worked and shoe glue got "much better results".
But it was tricky glueing anything to the leather with shoe cream in it.

On that basis I'm going to try a smear of leather glue - I can still glue a patch to it if that doesn't work.
But it does raise the question of whether we need a leather sealant to apply to gussets every few years.

It would be nice to be able to set up a test rig the way he has, but accordions are a bit larger than concertinas

 

[debra]

If the leather is so worn that it isn't airtight any more from wear over many years, rather than trying to "patch it up" with something like glue, I would suggest to just replace it by new leather (goat skin if I'm not mistaken). Not that I have experience with it. I tend to maintain more recent accordions mostly, not old enough to have the bellows leather parts to be completely worn out.

 

[Glug]

Yes I could replace the leather where required, I'm almost proficient at doing that.

The interesting bit is that the original article finds that leather starts leaking air as it dries out even if it isn't worn.

 

[Gonk]

Paul, I agree that replacing the gusset is optimal. It's just a huge job, if you want to do it right. The gusset's over the card but under the bellows paper, which is nigh impossible to remove without damaging it. So it would look like: remove bellows tape, remove metal corner, remove paper, remove gusset, replace gusset, repaper (with new paper/cloth), replace metal corner, replace bellows tape.

 

[Glug]

If you're careful you can peel back the paper - just dampen it a bit: seems to be held on with tape glue.

I've got all the tape off anyway, and some of the corners. I'm not going to retape it until I'm certain it doesn't need more fixing.
But as you say it's quite a big job.

 

[Dingo40]

I found this thread and the links interesting reading (thanks Glug!).
Personally, I'd be loathe to apply anything oily or waxy to my bellows for fear of creating worse problems down the line, particularly as regards glueing.
On the other hand, I was most surprised by the existence of the problem at all, as I haven't come across it myself.
Even the gussets ( original) in my ancient 1940s I. Busilacchio Challenge PA (surprisingly) appear to be as good as ever, as do those in my various newer accordions, the newest being approximately 30 years old.

 

[Glug]

Me too, I'm not (currently) seriously thinking of putting shoe cream on my bellows

But it would be interesting to be able to test an accordion bellows the same way that guy tests his concertinas and to find out just how much the leather leaks air.

 

[Dingo40]

While we're on the subject, I wonder if this product would work?
It's certainly effective with leaky roofs!

Rust-Oleum 340g Black LeakSeal Flexible Rubber Coating Spray

Find Rust-Oleum 340g Black LeakSeal Flexible Rubber Coating Spray at Bunnings Warehouse. Visit your local store for the widest range of bathroom & plumbing products.

www.bunnings.com.au

 

[Gonk]

Dingo, you first

My main concern with that kind of approach is the effect of repeated flexing. Weltmeister and Hohner both experimented with synthetic / rubberized corners instead of leather, and abandoned it because they tend to fail within 10 years. The toy 10-button plastic melodeons from China still use a kind of plasticized cloth, and it leaks badly.

But if we're talking about getting a little extra time out of some bellows that will eventually need full teardown and repair (or just teardown) ...I don't see why it couldn't help.

 

[Dingo40]

Personally, I wouldn't try it either!

 

[Glug]

FYI: It's also available in white (and aluminium)

NEW Rustoleum Leakseal Roof Gutter Water Leak Repair Spray Paint - All Colours | eBay

NEW Rust-Oleum Leakseal Spray Paint 500ml. Application: Aerosol. Choice Of Colours. An easy to use, rubberized protective utility coating designed to fill and seal leaks and cracks. Coverage: 0,74 m² per spray can.

www.ebay.co.uk

That leather glue was horrible to work with and didn't seem to wet the gusset.
I switched to PVA (Titebond 3) and used a 00 artists brush to paint the pinholes and thin bits, the drop test time went up by 6 seconds.

This particular bellows has very thin, almost papery, gussets. What I really need is a way to test one of them on its own.

 

[Glug]

Ok I was bored and though I'd evaluate the options, so junior science lab time:



The syringe pumps air into the reservoir (Kilner jar) using the one-way valves (red bits). Then the air leaks out through the test sample held in the aluminium box. The air pressure is monitored by a 0-15 psi digital pressure gauge. The idea is to time how long it takes for the pressure to fall and use that as a measure of how leaky the sample is.

I estimated that the pressure in the bellows is going to peak around 3psi - thats 29 reeds sounding at around 10mbar each. So I'm using 4psi to 2psi as the pressure drop.

Here's the sample holder:



The sample is held between two slabs of 3mm neoprene and leaks air through a 5mm hole in the rubber and top plate.

And here's the results (so far):


'Relative leakage' is the plasticard avg / sample avg and is slightly easier to compare than the times.
Each row is 3 tests on the same sample spot - to check there's no error in measuring. So 3 rows in a section is 3 different sample spots.

Main conclusions so far:

Christmas cards don't leak air
Untreated gussets are actually not very airtight in absolute terms
PVA reduces leaking quite a lot and the gusset remains flexible, but it doesn't fix worn spots so much.

I've got a can of Rust-oleum Leak Seal (white) in the post, so I'll update the results when that turns up.

 

[Gonk]

Really interesting. One thing I wonder about with PVA/titebond/leakseal is whether in time it will stiffen and lead the gusset to crack. If the coat is thin enough, presumably small cracks would emerge in the glue itself, but not the leather. I imagine one could try to test that by doing repeated flexing. But volatiles might still be leaving the glues over a period of years, leading to eventual inflexibility. Of course, it might still be a net gain over untreated gussets!

This doesn't seem to matter as much on large boxes like piano accordions, but in my experience, getting really good compression with smaller boxes like the Liliput is difficult. I'm interested to see whether this could help.

 

[Glug]

That's definitely a consideration.
I'm trying to 'fix' a 1950's bellows which should really be replaced so if it fails after six months I haven't really lost much.
I also want to know which bits of a bellows cause these problems.

One big advantage of testing gussets in the new rig is I can see how flexible they are after painting with PVA etc. The answer so far is surprisingly flexible, but as you say, time will tell.

 

[Glug]

Ok, so Rust-oleum finally turned up after Fedex found the right door.

Quick hint: don't spray it - it comes out far too thick for anything other than construction use. You can spray it on a plate etc and paint it from there. And you'll need something stronger than soap to clean the brush. You can repaint after 20 mins.



And here's what the gussets look like after 'treatment':



Opinions so far:

• I think fixing worn spots needs 2 coats, but I dont have a lot of proof.
• LeakSeal isn't as airtight as PVA but is much better than untreated new gusset.
• I think 2 coats of PVA is slightly less flexible that 2 coats of LeakSeal
• LeakSeal gives a much better finish than PVA

So it's basically it's a toss up between 2 coats of PVA or 2 coats of LeakSeal (painted).

I could rip out another old gusset and do some more testing now I know not to spray LeakSeal.

But I'm probably more inclined to just do 2 coats of LeakSeal (painted) - it's good enough and it should improved the look of the old gussets quite a lot. I'm thinking if I use a long bellows stretcher I can peel back the cloth (apply water) without removing more corners and just paint it on fairly easily.

 

[Dingo40]

Interesting !

 

[Glug]

Finished painting, and it seems to have worked:

Drop test after putting PVA on pinholes: 77 secs
Drop test after 2 thin coats of Leak Seal: 104 secs

That drop test is done with 1 layer of masking tape on the treble soundboard (I've already checked the bass pallets and replaced the air button facing). But that masking tape does leak air so the numbers are only a relative indication.

The Leak Seal is surprisingly flexible and the bellows seem to be working normally, apart from some creaking which I'm putting down to being on a stretcher for several days.

Here's before and after on some gussets:

 

[Gonk]

Doing the Lord's work, Glug!

That looks really sharp. The big question remaining for me is about its longevity, which of course can't be answered yet. But since it doesn't make it any more difficult to eventually replace the gusset, I suppose it does nothing but expand their lifespan.

Did the creaking go away?