Grit & Gluing?

[Nahil.R]

I have seen it debated before whether a finer or rougher grit is better for gluing surfaces. The argument almost always seems to lead to as to whether glue joints are more mechanical or chemical in nature.

I am no scientist (nor do I play one on the internet) so I was just wondering from all of your first hand experiences which method do you subscribe to, and what is your reasoning behind it?

Personally intuition tells me that glue joints have both a mechanical and chemical action at play and that is what bonds the surfaces, but that still doesn't let me know which is better functionally. Rough grit or fine grit (hand plane)?

Thanks.

Nahil.

[redir]

It's chemical bonding so the closer the pieces fit the better.

What I thought you were going to say by the title was this, something that someone mentioned here once and I thought it was interesting... That an old wood working trick to keep pieces from swimming away when you clamp them up is to add a few grains of sand. THe sand bites into the two work pieces and holds them from slipping... But I digress

The old school way was to scratch the bridge up. You can see this when repairing very old guitars. But all it really does is make the joint weaker. Voids filled glue have no strength.

[D. Churchland]

Around WWI there was a tremendous amount of research done on what glue joints hold better with the rough vs. smooth. The research concluded that a cleanly planed surface is the strongest glue bond at least for hide glue. Check out how they scarf joint ship masts for a good example of solid gluing technique.

I'll see if I can find you the link. But I'll also say from experience that smoother surfaces bond better with hide glue every single time. I've had too many destroyed guitars with chipped up tops under the bridge because of a "scored" surface done during the reset.

[Nahil.R]

Thanks guys, that's really interesting about the research done, I'll be sure to check that out. I have heard about the scoring of bridge plates in the past, good to know from your first hand experience that it was not the right thing to do.

Very cool how methods change and evolve over time.

[HodgdonExtreme]

I build force sensors, that requires me to glue thin sheets of phenolic resin onto blocks of steel. The performance of the end product relies HEAVILY on achieving an excellent glue joint. There are lots of different adhesives for the application, and NONE of them work if the steel block has been polished smooth. All require a ~220grit surface finish to create a good bond.

[Rudy4]

Originally Posted by redir

It's chemical bonding so the closer the pieces fit the better.

What I thought you were going to say by the title was this, something that someone mentioned here once and I thought it was interesting... That an old wood working trick to keep pieces from swimming away when you clamp them up is to add a few grains of sand. THe sand bites into the two work pieces and holds them from slipping... But I digress

The old school way was to scratch the bridge up. You can see this when repairing very old guitars. But all it really does is make the joint weaker. Voids filled glue have no strength.

I've kept a SALT shaker close by for slip-free glue ups as long as I've been woodworking. It's a trick I learned from my grandfather.

A few grains of salt won't do any damage and stay intact long enough to do a glue up.

Even a FEW grains of sand will destroy a cutting edge on hand or power tools, so I'd advise against trapping any sand within your work.

The important thing for glue ups is not the finish, but the fit. It won't make any practical difference between sanding to 100 grit or 220, but it's important to mate your joint as gap-free as possible. The proper thing to aim for is a gap-free gluing surface before clamping force is applied. In the real world we usually end up closing things up when clamping force is applied, but make sure you aren't counting on your glue to fill a poorly fitted joint.

Note that luthiers who build classical guitars or viol family instruments sometimes simply glue plates together using a simple rub joint. No clamps involved, but don't try it if you can't achive a light-tight joint with your tools and/or skill level.

[Nahil.R]

Oh wow, now I am even more confused!

Contradicting views, I would love to hear more luthiers here weigh in

A question for you Rudy. You mention that the most important aspect is achieving a gapless fit between the surfaces, but also say that sanding to 100 or 220 won't make any practical difference ... technically wouldn't a finer grit provide more gluing surface are?

Thanks.

Nahil.

[charles Tauber]

I think it is important to distinguish between "polished" surfaces and "sanded" surfaces. HodgdonExtreme's experience speaks to that.

I have found that ebony that is sanded to 600 grit doesn't produce a very strong epoxy joint. The epoxy tends to peel off of the ebony surface.

I agree with Rudy that the difference in surface texture between 100 and 220 grit isn't sufficient to make a practical difference in glue joint strength.

Fine Woodworking magazine did a number of glue strength tests. They found that some glues lost less strength than others as the joint tightness decreased. For example, hot hide glue lost about half of its strength in a "loose" fit joint, while other glues lost relatively little strength. As Rudy stated, in general, the mating of the glue surfaces matters more than the grit to which it has been sanded, with the range of typical woodworking sanding. (Fine Woodworking's test did not examine surface texture, only relative tightness/mating of glued surfaces.)

Technically, a surface with lots of hills and valleys will provide a greater gluing surface. However, it will not create a greater portion of the surfaces being in contact with each other.

There appears to be a range of surface texture that is optimal for gluing two surfaces together. If the surfaces are too smooth, such as a high polish, they don't glue as well as a rougher surface. However, if the surface is too rough, less of the two surfaces being adhered are in contact. If the two surfaces don't have large portions of the surfaces in contact - due to gaps or coarse surface texture - the adhesion is weaker. In that case, the space between mating surfaces is filled with glue, which has less strength than the bond between the glue and the surfaces to be adhered.

[Nahil.R]

Thanks Charles, that helps clarify some of my confusion.

I normally try to aim for around 180 grit (spruce braces to spruce soundboard) with more emphasis placed on the fit of the X and the radiused braces against the soundboard.

Cheers.

Nahil.

[Rudy4]

Originally Posted by Nahil.R

Oh wow, now I am even more confused!

Contradicting views, I would love to hear more luthiers here weigh in

A question for you Rudy. You mention that the most important aspect is achieving a gapless fit between the surfaces, but also say that sanding to 100 or 220 won't make any practical difference ... technically wouldn't a finer grit provide more gluing surface are?

Thanks.

Nahil.

Charles posted excellent information, so I won't pontificate.

One hazard to excessive sanding is it's very easy to end up sanding your surfaces to a greater extent along the edges of a joint to be glued. This ends up causing a glue line that's much more visible, and sometimes inferior.

Sometimes we have a tendency to over-think a simple process, and in the case of glue ups it just isn't that important. If there's something about the strength of your glue joints that gives you pause you're probably better served by examining your joinery technique.

[redir]

Originally Posted by Rudy4

I've kept a SALT shaker close by for slip-free glue ups as long as I've been woodworking. It's a trick I learned from my grandfather.

A few grains of salt won't do any damage and stay intact long enough to do a glue up.

Even a FEW grains of sand will destroy a cutting edge on hand or power tools, so I'd advise against trapping any sand within your work.

The important thing for glue ups is not the finish, but the fit. It won't make any practical difference between sanding to 100 grit or 220, but it's important to mate your joint as gap-free as possible. The proper thing to aim for is a gap-free gluing surface before clamping force is applied. In the real world we usually end up closing things up when clamping force is applied, but make sure you aren't counting on your glue to fill a poorly fitted joint.

Note that luthiers who build classical guitars or viol family instruments sometimes simply glue plates together using a simple rub joint. No clamps involved, but don't try it if you can't achive a light-tight joint with your tools and/or skill level.

Oh that's right! Not sand. I remember now that it was indeed salt. Thanks for the reminder and correction.

[Nahil.R]

Thanks Rudy, this post has been very insightful ... and that trick with grains of salt is a technique I am definitely going to keep in mind for those slippery glue ups!

[tadol]

Some of the confusion comes from the old practice of using toothing planes to prepare surfaces for gluing, particularly for veneer work. Seemed to make people think a roughed-up surface made a better glue joint. As well stated above - the more matching surface area you can get, the stronger the glue joint. But the surfaces also have to have a certain degree of permeability, at least with water based adhesives, so over-sanding (polishing) can be counter-productive. A freshly hand-planed or hand-scraped surface is generally your best choice -

[nickv6]

Very interesting thread...thanks to all contributors.
Nick

[Alan Carruth]

Back when I was taking my violin making lessons I was told about two studies that the Forest Product Lab did back in WW 2 on gluing, related to wood aircraft structures. One found that freshly milled/worked surfaces glued better, and the other showed the planed surfaces glued better than sanded ones.

The planing/sanding study came from problems with laminated wood propellers that were coming unglued. The ones where the layers were planed to thickness held up better. This makes some sense. If you look at microphotographs of sanded surfaces they can look like a plowed field, particularly when coarse grits or dull paper are used. There can be loose 'whiskers' of wood, and dust in the pores, which would bond weakly, and could keep the glue from wetting the surface. Once you get up to about 220 grit the scratches from the sandpaper are about the same size as the pores in the wood, so going any finer doesn't help.

Freshly worked surfaces have higher 'surface energy'. When you remove material you can break chemical bonds, and it takes a while for these open bond sites to find something to latch onto. Glue can bond to those open sites.

The test for surface energy is to spritz the surface with water. Water is a polar molecule, and '+' and '-' ends of the molecules will be attracted to slight charge of open bond sites; the water spreads out into a film. If it beads up the surface energy is low, and glue won't stick as well. They actually use a microscope the look at the angle at the edges of the drops to measure the surface energy.

At any rate, the FPL study said that they got stronger joints when the pieces were glued within 15 minutes of working them. That argues that at least some of the strength of the bond is chemical. This would have been using Resorcinol glue, which, I believe, is hide glue, with Formalin to render it insoluble once it's cured. Until fairly recently that was the only glue that would pass the Navy boiling test. That glue also typically contains wood flour as a filler, I believe.