Wood-fired!

Gluing to excess

I have referred elsewhere to the many systems woodworkers use to solve the problem of cleaning away excess glue in their builds. I am currently testing a method that I have eschewed in the past but seems ideal for the job in hand.

This system is based on pre-finishing the elements to be glued, allowing the glue to spread and dry wherever it will, and then carefully peeling it off the finished surface with a chisel.

These observations refer to water-soluble PVA-type glues but some of the techniques work for other glues.

For the uninitiated, a joint must have sufficient glue in it at glue-up to ensure that both surfaces are fully wetted with the glue. This means that some glue will be squeezed out of the joint when cramp pressure is applied. On the outside of joints this presents no problem because the surfaced can simply be planed clean before sanding and finishing. However glue on the inside corners of joints is not so easily removed. The bulk can be removed with a chisel or flat knife but some will enter the grain of the wood and dry there (if it didn’t then the glue would be ineffective in the joint). This can be sanded back (not always easy to do well on an inside corner) until it appears to be gone. That is until the finish is applied and the glue prevents the wood absorbing the finish resulting in obvious light patches where the glue has spread.

So woodworkers need a solution for this. My usual system is to protect the wood on the inside of joints with suitable masking tape. The excess glue is then removed with the tape and any that remains is easily removed with a fine chisel or a flat marking knife. This method is described and illustrated here.

Other practitioners allow the glue to squeeze out onto the wood but once the joint is cramped, the excess is washed away. For water soluble glues this usually requires quite a lot of water and scrubbing with a toothbrush or something similar. This can remove the glue ok but tends to raise fibres in the wood so that re-sanding is required. It also creates extra time pressure in a process that is usually very time-critical.

The method I am describing here is not new by any means but is one that I have mostly avoided because it requires pre-finishing, still requires some masking and just seems - well - wrong to me. It feels like you’re cheating a bit.

Because I have a lot of these joints to glue-up and clean - around 240 inside corners! - I experimented with all three methods.

Washing the glue away had the disadvantages I expected. It takes up a lot of time when you’re fighting the clock to cramp the joint true and get to the excess glue before it dries too much. And it left the wood feeling furry and in need of re-sanding. This grain lifting could have been minimised by pre-treating the wood in the same way. That is, you wet the wood, allow it to dry, sand it and repeat that process at least two more times. I sometimes use this technique for surfaces that must feel glass smooth but it wasn’t warranted in this case and would have taken ages. This method just didn’t suit my needs.

My usual method - masking with tape - worked ok but takes so much time! That’s 480 separate pieces of tape that must be very accurately applied.

I was also concerned about being able to apply the finish to the piece after it was assembled. I wanted to spray the finish on but the assembled piece has 68mm square compartments that are 58mm deep. Even though I could access them from both sides it was going to be tricky. So the method of pre-finishing the elements had additional benefits.

I tried it out. My finish is a quick-drying spray nitrocellulose. This is quite waterproof and should provide a good barrier when dry. However this method gives you the reverse masking problem - you can’t afford to get any finish on the gluing surfaces or the glue will definitely not hold. So I had to mask the areas of the element that would be glued as shown.

So I still have to mask - but nothing like the amount of work required for my usual method.

The excess glue is clearly visible against the finish when viewed in the right light.

Cleanup does require a great deal of care. Pushing a chisel across a sprayed finish is going to leave scratches unless:

• the chisel has a sharp fine edge - no micro pits, dings or nicks!
• the chisel has a very flat back - you need to be able to hold it flat and push it across the finish without marking it
• the touch is extremely light - in some cases the weight of the chisel is sufficient downforce.

Now it seems we are back at the very activity this whole discussion is trying to avoid, ie cleaning up glue by pushing a chisel across the grain. The difference is that the glue is barely bound to the finish so it separates easily, it is easy to see if your technique is marking the finish, and most importantly no glue is left in the grain to mar the finish.

After one light pass the bulk is removed leaving only a little waste in the corner.

A second pass is made across each face of the joint to remove the remaining glue. Make sure you rotate your piece to do this as holding the chisel vertically does not give the fine level of control that you need.

If all else fails you can still repair any nicks with a dab of finish.

And at the end of the 26 glue-ups all the inside faces are already finished!

The process of removing the waste with a chisel stills takes a bit of time but I have found it to be a very relaxing and strangely enjoyable task - so it seemed like no task at all. Perhaps this is because you effectively see the finished joint as you proceed - rather than just another stage on the way. But I suspect it’s also something about the fine nature of the task that appeals to me.

I have been reluctant to use this technique in the past but for this project the method has worked extremely well for me (so far - still more than a dozen glue-ups to go).

This table summarises my experiences with three different techniques for removing excess glue. My preference for this particular project is the third option. Other projects will have other requirements and a different method may suit you better.

2 comments May 20th, 2008