Failed Glue Joint - help me diagnose please

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Molynoox

Established Member
Joined
17 Jan 2021
Messages
975
Reaction score
672
Location
Billericay
So I'm making a cutting board for a friend (still) and two of the glue joints have failed.
Probably safe to assume its a general problem because it was clamped up very firmly and the joints seemed to be pretty flat and well mated in dry fit.

(Ignore the oil on the surface - I put a bit of tung oil on to see how it would look once finished, might have got a bit carried away with it)

1699182943736.png

1699182982897.png


Looking for help with the WHY DID IT HAPPEN and HOW TO FIX

WHY
The donor wood was from an old desk (20 years old) which I chopped up to make the laminations, the top had been oiled originally, and because its an end grain board the top surface of the desk is now the surface which is glued together in the chopping board as the glue joint. Due to my woodworking genius I didn't think to sand the (previously oiled) top of the desk before cutting up the laminations:
-is this likely the reason for the failure? that the oil in the mating surface stopped the glue working?
-do the pictures reveal anything about what has failed?
-could it be the glue? (the glue isn't old and hasn't been colder than 5 or 10 degrees)

HOW TO FIX
I assume the best way is to run my track saw down every single joint and essentially re-cut all the laminations?
Should I do any other prep to prevent it failing second time around?

thanks in advance
Martin
 
What glue are you using ?

Were all the mating surfaces clean and free from anything like oil or maybe from something like polish used on the desk that had soaked in ?

Then what finish are the mating surfaces, a tracksaw may not be ideal and if I am jointing boards like this then I either put them across my P/T or through the router table with offset fence to give a really clean finish.
 
Hmm, It definitely isn't over clamping as there is definitely dried glue residue there. I t could be that the wood itself is too oily/waxy. The two main culprits in this regard are usually Iroko and Teak. You have to degrease the surface of these before gluing up.

I could also be that you are picking up silicone or wax residues off the beds and fences of your machines.

Failing that, it could be that your glue has got too cold in storage and has been compromised
 
to the best of my knowledge the surfaces were clean
good idea to use a P/T or router table, but I don't have either hence my weird track saw solution :)

Glue is Everbuild D4 premium
 
Yeap I think you have answered your own question. The glue joint is contaminated by the original finish. Cutting down the joints would work but alot if effect. I would probably discard and put it down to experience!
 
glue hasn't been subjected to frost and its less than a year old
seems like the culprit is the potentially oily wood surfaces - doh!
 
.
..... Due to my woodworking genius I didn't think to sand the (previously oiled) top of the desk before cutting up the laminations:
-is this likely the reason for the failure? that the oil in the mating surface stopped the glue working?
Certainly looks like it, assuming that was normal PVA or similar
-do the pictures reveal anything about what has failed?
Glue hasn't stuck to wood. What else is there to say about it! :unsure:
.

HOW TO FIX
I assume the best way is to run my track saw down every single joint and essentially re-cut all the laminations?
Should I do any other prep to prevent it failing second time around?
Could clean them off again and just apply a few trial dabs of glue and see how they stick without putting pieces together. Should be like s**t to a blanket!
 
Hand plane the surfaces true and square. Dry clap to check. De-oil the naturally oily surfaces with meths. Add some biscuits or dowels to the joint and then re-glue.

Colin
I think the biscuits will only give me alignment and not really any extra strength (from what I've read) so probably wont bother, but that's a useful tip on the meths, thanks
 
well, after an hour in the workshop with the track saw and power planer I'm back where I started:

1699206074985.png

1699206095996.png


It's sitting together pretty nicely in a dry fit so i will hit it with some meths to de-oil and then try again.

Martin
 
I haven't used PVA in as long as I remember. Did try Gorilla once with Iroko but had failed joints. Cascamite? Not sure if it is food-friendly. The new manufacturers, Eureka seem to have gone back to a recipe that works. I'd be surprised if it was a finish applied 20 years ago and how many people keep their desks well polished? Unless it was silicon and Mr Sheen?
 
I've made a fair few bits with Iroko. It is oily and you are advised to give a wipe down with meths to surfaces before glue up.
I must now confess I've never done this. The photo of the chairs with footstools and tables was taken during a lockdown project of 2020. These have been left outside over winter with a cover since being made and have now gone a nice grey colour and no joints have failed.
The photo of the gable end of our bungalow shows a garden table and two benches. I made these about 1997. They have been left outside every winter under a cover. A few of the M&T joints did open up slightly but did not fail. As a precaution I cross doweled those effected through the joints a few years ago and added glue blocks as an extra. These have held tight since.

I have now used Titebond 3 on all my interior and exterior woodwork , softwood, hardwood and manufactured boards for about 30 years. It is a bit more expensive but only accounts for a small % of a project cost. I think these projects reflect how good the adhesive is. It has a good initial tack yet still also has a fairly long open time. On wide M&T joints of door rails it has demonstrated the ability to allow for expansion and contraction without joint failure. I think it is excellent and worth considering.

Colin
 

Attachments

  • P1040259-1600.jpg
    P1040259-1600.jpg
    222.5 KB · Views: 0
  • P1050792-1600.jpg
    P1050792-1600.jpg
    152.5 KB · Views: 0
Hmm, It definitely isn't over clamping as there is definitely dried glue residue there. I t could be that the wood itself is too oily/waxy. The two main culprits in this regard are usually Iroko and Teak. You have to degrease the surface of these before gluing up.

I could also be that you are picking up silicone or wax residues off the beds and fences of your machines.

Failing that, it could be that your glue has got too cold in storage and has been compromised

All good suggestions. My neighbour recently took up his decking (ipe) and let me have it. Similar to teak and iroko, it’s a pain to get strong glue joints. I had one fail, similar To the photo. On that occasion , I’d used old pva and my workshop was pretty chilly. iI’ve since only used Titebond 3 and made sure my workshop was pretty warm when working with ipe, whether the project is for external or internal use. Had no problems so far…fingers crossed
 
it might just need a good wipe with meths before you glue it up, I've found epoxy never fails with difficult woods, never had any problems with wudcare PVA but have had problems with titebond original in one case on a bookshelf where two joints failed, aerolite would be a good choice as well, I've heard cascamite has gone downhill and wouldn't trust it unless they've improved it again.
 
I think the biscuits will only give me alignment and not really any extra strength (from what I've read) so probably wont bother, but that's a useful tip on the meths, thanks
Biscuits can give quite a lot of extra glue area and strength but probably not the best option for this grain orientation. Using a couple of dowels would be better
Wiping some naturally oily timbers such as teak and rosewoods with meths is also a good general approach
 
Thanks. Seems pricey. I was hoping I could get some without using mail order but I guess not.
 
Back
Top