• We invite you to join UKWorkshop.
    Members can turn off viewing Ads!

Polyurethane Glue - Interesting Report


Help Support UKworkshop.co.uk:



Thanks Noel, interesting. I'm sure Chris will have something to add to this based upon his own tests with these glues


Established Member
19 Mar 2003
Reaction score
West Dartmoor, Devon, UK
Thanks Noely - I've been recently playing around with some PU glue - so interesting to hear some facts behind it.

I got some from Axminster - pretty cheap actually. And one thing not mentioned in the article is it sets quickly. So I've been using it to build some workshop jigs etc. Still not sure I'd trust it for proper projects ...



Chris Knight

Established Member
14 Jan 2004
Reaction score
SE London - NW Kent
Thanks Noel, an interesting article.

As Tony mentions I have done a lot of testing with glues to ensure that they did what I needed in some unfamiliar joint configurations ( with plenty of end grain to long grain and even end grain to end grain configurations as well as the more generally desirable long grain to long grain joints).

My testing was fairly rough and ready inasmuch I made no attempt to quantify bond strength but simply counted a bond good if the wood around it failed, rather than the glue line.

The single most important factor I found in many tests was the benefit of roughing the joint surfaces with coarse (100 grit) sandpaper before gluing ( making sure I blew the dust off before applying the glue) . I had few really bad failures (except with Titebond poly) when I did this, no matter what glue I used. Titebond poly was the only glue to demonstrate total glue line failure, all other glues ranged from 100% wood failure to around 50% wood/50% glue line failure.

Generally the most successful glue (for its strength, price, and ease of use ) was Titebond II followed by Titebond Original and Titebond Extend. I was however seeking a poly glue at the time of these tests, because I needed its slippery nature for some extremely close fitting joints. The best I found was Balcotan which gave me 100% wood failure in most of my tests, whatever the joint configuration. It is very easy to use but rather expensive. It foams much less than other polys. Gorilla was OK but not as reliable as Balcotan.

Epoxies (three different makes) proved disappointing with partial glue line failures in all of them.

I did not try Cascamite/Extramite as being water based, I anticipated the same swelling problems in the joint that Titebond gave.

I have repeated some of these test with joints including ebony as I am including some ebony in my current rocking chair in the laminations and in the headrest. It sticks well with Titebond II - as long as it is roughed up with the 100 grit first.

The problem with all failures (except Titebond poly which I cannot account for) seems to me to have been inadequate wetting of the joint faces by whatever glue I was testing. I think that in most cases this is caused by oil from my fingers getting on the joint and making it impossible for the glue to penetrate properly. When you think about how often a joint is handled during its making, dry assembly and so forth, it does not seem unreasonable to assume this. I have noticed on some of the walnut components I am working with at present that my hands have produced a rather nice finish - simply by frequent handling, a bit like the leather handle of a briefcase goes after a while.

I had read that solvents would help gluing exotics and I tried this (acetone) with the ebony - it did not work, sanding did.


Established Member
22 Oct 2003
Reaction score
Up the proverbial creek
waterhead37":1dwly2zv said:
I have repeated some of these test with joints including ebony as I am including some ebony in my current rocking chair in the laminations and in the headrest.
Mmmm, can't wait for the pics... :D

Cheers, Alf