Interesting info on clamp pressures

[dyne]

I think the numbers they're reporting are really lbs-force. They'd have to calculate the pressure using the clamping face area and if they put the effort into that they'd specify in the article. Not to mention that if they were actually reporting psi, 370 psi for a parallel clamp would be absurdly high.

Also using a bathroom scale is fine, you wouldn't need to be super precise for this type of comparative test.

[kscott]

exactly.
Ken

[Doug_H]

Guys

I find the results on achievable clamp pressure from various clamp designs very hard to believe. As others have suggested, maybe it's the "instrument" used (Bathroom scales). BTW who has a bathroom scale that weights over 500 lb? (If there were one I should need it after Thanksgiving.)

Necessary clamp pressure is controversial. I strongly prefer research results to expert opinion, but both kinds of information need to be taken in the context of their unstated assumptions.

A glue joint should entirely cover both mating surfaces and should be .0004" thick or less. A thicker PVA or polyurethane glue film is weaker. (With the possible exception of slow-setting epoxy, glues do not fill gaps.) Clamp pressure is needed to force glue into the wood, to overcome swelling caused by the water in the glue, to hold the surfaces tight against each other and to immobilize them while the glue dries.

According to Roman Rabiej, starved joints are not a worry with hand operated clamps. On the contrary, it is difficult to get enough force with hand clamps. Necessary clamping pressure depends on the wood species and whether the mating surfaces are radial (quartersawn) or tangential (flatsawn). The pressures he recommends are difficult to achieve in a home or small commercial workshop.

Consider a 48" long panel of 7/8" thick red oak showing a quartersawn face. According to his research, the flatsawn glue faces require a minimum of 900 psi for a strong glue joint. Each glue joint has 48 x 7/8 = 42 sq. in. of glue surface, needing a total force of 37,800 pounds. A pipe clamp, operated by a strong man, can exert about 1400 pounds of force. Therefore, to receive the recommended clamp pressure, this panel requires about 27 pipe clamps along its length, that is, one clamp every 1¾". You could barely fit them.

A flatsawn show face (quartersawn glue face) would require about half as many clamps. Softwood would require somewhat lower pressures, about 300 for quartersawn glue face and 150 for a flatsawn glue face.

I do not doubt the accuracy of Dr. Rabiej’s engineering studies. Perhaps his recommended clamping pressures would be necessary to make a cantilevered panel or one that would be supported only along its edges. Set against this, however, is years of experience by many respected cabinet makers. Also, Bruce Hoadley (p. 197) recommends much lower clamping pressures, about 100-300 psi. Evidently, for a panel such as a table top that is supported across its width by an apron, joints made with much less pressure seem strong enough.

R. Bruce Hoadley. Understanding Wood: a craftsman's guide to wood technology. Newtown, CT. Taunton Press 2000.

Roman Rabiej. Get Serious About Clamping: Most Woodworkers Are Underclamping Their Joints. Fine Woodworking Nov/Dec 2007.

Best

Doug

[danceswithticks]

The proof is in the pudding as the saying goes. According to the article, I've probably been under clamping for many years, but AFAIK, my projects have stayed glued together. I would guess most here have not had significant glue up failures. And sorry, bathroom scale or not, my quick clamps do not exert more force than my parallels.

[Robert Adams]

Like others I don't believe the numbers either. On the subject of how tight is tight enough there have been lots of comparison tests over the years and it's better to be too wide of a glue joint instead of too thin. And the strength of a perfect glue joint isn't much different than one loose with a thick glue line.


Overall it's not a big deal just glue it and clamp it and go on. Just don't crank it so tight all the glue comes out. Seen that happen many times where the joint has no strength at all IE break it by hand.

[AHill]

The most useless part of the article was using the formula to calculate how many clamps you need. Seriously? So where's the data for how many psi of clamping pressure you need for each kind of glue? I'm pretty sure it also varies by species of wood, moisture content, etc. Their formula used 450 psi of clamping pressure. I looked up Titebond's recommended clamping pressure, and the highest recommended clamping pressure was 175-250 psi for hardwoods using Titebond I (original). It was 100-150 psi for softwoods.

[Robert Adams]

AHill said:

---

The most useless part of the article was using the formula to calculate how many clamps you need. Seriously? So where's the data for how many psi of clamping pressure you need for each kind of glue? I'm pretty sure it also varies by species of wood, moisture content, etc. Their formula used 450 psi of clamping pressure. I looked up Titebond's recommended clamping pressure, and the highest recommended clamping pressure was 175-250 psi for hardwoods using Titebond I (original). It was 100-150 psi for softwoods.

---

All the more reason to glue it clamp it and forget it.

[barnowl]

kscott said:

---

... I simply cannot believe that the quickgrip provide more force than the parallels.
Ken

---

Neither can I.

Ask Joe Grout to post his clamping pressure test photos.

[handi]

First, as stated above, PSI is a function (obviously) of square inches, so a K-Body clamp may register fewer PSI while applying more actual pressure. Sort of like a 40 ton tank exerting less ground pressure than a 5 ton truck. Those tracks spread the weight over a LOT of square inches.

Second, in my experience, clamping a joint too much and "starving" a joint by squeezing all the glue out is bogus. The only way to "starve" a joint is to not apply enough glue in the first place. I made thousands of stair treads over the years clamping them up in an industrial clamp rack. 4-6 pressure heads each exerting a couple thousand PSI, and never once had a joint fail due to "starvation".

Thirdly, clamping force figures are highly over-rated. If you need to exert hundreds or thousands of PSI to get glued faces to meet up, you are not only doing it wrong, you are actually building in stress that will likely fail in the future. The mating sides of a glue joint should meet without pressure, the clamps are just there to hold them until the glue cures. If there are gaps when brought together, then there is more work to do before clamping.

Ralph

[barnowl]

Excellent points Ralph.