Glue lines

[Robert Chapman]

I have this annoying problem with maple glue ups. I use Titebond III, sand to 220 grit smoothness, stain with Minwax oil based stain, finish with 3 to 4 coats of a mixture of equal parts tung oil, polyurethane and minieral spirits and rub out with 600 grit paper. Here's the problem - after 3 or four days I get a roughness along the glue lines. I can rub these rough lines out but it's annoying and I don't understand why I'm getting them. Any ideas as to prevention/cause?

[Henry Ambrose]

Let it sit longer before finishing.

[Howard Acheson]

Here is some info that should help.

Creep has been known about for many years, perhaps even centuries. It's nothing new and has already been defined precisely, so no need to reinvent the wheel here. PVA is the classic and renowned creeper.

The tendency of the glue to ooze out of joints is one form of creep. A classic example is in a slab edge to edge glueup, such as a table top. When ambient humidity rises the timber and the the glue swell. When the ambient humidity goes down the timber shrinks again, and so does the glue, but the glue doesn't all shrink back into its original place resulting in a line of pimples disfiguring the finish. Actually, under sustained high humidity my anecdotal observations are that the glue keeps on absorbing moisture and creeps out of the joint without the need for the timber to shrink. The symptoms can also be seen sometimes at the shoulder line of other joints such as mortise and tenons.

Another cause that I've witnessed several times is to make a solid timber slab with edge joints in a fairly humid workshop without climate control using PVA as described before. Then right away get to planing and preparing the surface ready for polishing with scrapers, sanding, etc.. Right after that apply the polish, whether it be varnish, pre-cat, or another finish, and take the piece into a drier house, either heated or air conditioned. Three to five days later the rows of pimples will be apparent as the timber shrinks.

Somewhat related to this is using an aid for edge glue alignment, such as biscuits with PVA. These jobs need time to settle. If you are in too much of a hurry and prepare for and polish it all as soon as you get the piece out of the clamps, the chances are you'll see a row of rugby ball shaped depressions telegraphing the position of every biscuit a couple or three weeks later, although I've seen them appear as much as a year, or more, later. The moisture in the glue swells the wood either side of the biscuit, and it takes a week or two for this moisture to work its way out of the joint properly. My advice then is to not rush into the preparing for and polishing stages after this type of glueup, particularly table tops and the like which show off every flaw. I don't use PVA for this specific kind of glueup at all anymore here in hot, humid Houston. I tend to reach for the hide glue nowadays, either out of the pot, or the liquid stuff.

[Faust M. Ruggiero]

Titebond III is a bit dark for maple. If you need the waterproof qualities, you have no choice but if you are gluing up something that will not see the outdoor weather, you should stick to original Titebond.
faust

[Jamie Buxton]

Better jointing, or maybe better clamping. If you make really thin glue lines, they'll creep less, and be less perceptable when they do.

[Steve Schoene]

Carefully prepared joints, with hard clamping at pressures recommended by the manufacturer, will show no glue line.

[Robert Chapman]

Thanks for the responses - I will try some of your suggestions - pariticularly the hide glue idea.

[Steve Schoene]

I will add that opinions are mixed on such lines being caused by creep. Apparently Franklin attributes such lines to differential movement of the wood on either side of the joint. It would take very little relative change to be visible. Of course the only way to prevent that would be to ensure that your wood was all at the same equilibrium moisture, and that the boards were of very similar grain structures--ie. flat sawn against flatsawn, not flatsawn against quarter sawn or rift sawn.