Home Articles Forums Gallery Marketplace Videos

WoodCentral Forums

Knowledge You Can Build On — Est. 1998

Super Resolution Thermal Imaging Camera

Starving Glue Joint

Posts

Starving Glue Joint

#1

Starving Glue Joint

Warren in Lancaster, PA

>There was some discussion earlier about clamping hide glue and the risk of "starving" the joint with too much pressure. Since I had not seen this myself I did an experiment this morning to see if I could cause it to happen. I glued a block of walnut 1 inch wide and 2 inches long to another piece of walnut and clamped immediately with two Record sash clamps. The wood had been carefully planed slick and smooth and no light could be seen under my straightedge. Two bar clamps on two square inches of surface is near the upper extreme pressure I can produce. After two hours (much longer than usual), I removed the clamps and the join was sound with a very fine glue line and it could be worked like solid.

Gluing long edge joints with hide glue is difficult; there is not much room for error. Nicholson recommends using three men for large pieces. In this one case I think it wise to consider other glues.

Re: Starving Glue Joint

#2

Moses Yoder

Re: Starving Glue Joint

Moses Yoder

>Warren, do you use a traditional glue pot over a heater of some sort? Would be interesting to know what kind of set up you have for hide glue. What would you recommend for a guy who would only use it about once a month or maybe just on the weekends?

Re: Starving Glue Joint

#3

Re: Starving Glue Joint

Warren in Lancaster, PA

>Mose, I use an electric glue pot and sometimes just put the inner pot on the woodstove in winter. When I was just working weekends I put the glue in a yogurt cup. I would boil some water in an old pot and bring pot and glue out to the shop. Sometimes I mixed up as little as a teaspoon of glue. It may seem like a lot of trouble, but you don't have to wait hours for the glue to dry and there is a lot less clamping needed.

Re: Starving Glue Joint

#4

Go to Wally World

Neal in San Jose

>Wal Mart has a small (one quart, maybe) crock pot for under $10. I bought mine on sale a few years ago and it was one of the best investments I've made in my shop. I just use a small glass bottle, like a small mayo bottle, and mix up what I think I'll need. Put the mayo bottle in the crock pot and I'm good to go. Low heat on the crock pot is just right with an inch or two of water to transfer the heat, but I sometimes use high to melt the glue a little faster or re-warm an old batch. At the end of the day I take the bottle out, put on the lid and stick it in the fridge. I try and date the bottle, because I don't always get back to using the glue in a reasonable amount of time, and it is so cheap to mix up a new batch, I don't feel bad throwing away an old batch

Re: Starving Glue Joint

#5

thanks for the data

Bill Tindall, E.TN

>I know an expert in adhesives and he claims that the thinner the glue line the stronger the joint. To that end it would seem impossible to squeeze out too much glue with hand clamp pressure, as you have shown. Franklin recommends up to 300 psi with the yellow glue.

Re: Starving Glue Joint

#6

why.....

Bill Tindall, E.TN

>Upon thinking about the situation........the viscosity of the adhesive rockets as the water is removed. Water removal by absorption into the wood from a thin glue line will be very rapid. It will be impossible to squeeze the resulting high viscosity glue because it won't flow under readily obtainable pressures.

Re: Starving Glue Joint

#7

Moses, It's Ok to use . . .

Tom Colligan

>a microwave, though in very short spurts, i.e. 15 seconds, then stir. I always have a batch kept in a Smucker's jar in the refrigerator. I heat the entire jar over a 2 minute period while stirring with a thermometer commonly found in the utensil isle at the grocery store. After you make your first batch, i.e. 50% water and 50% hide granules let it mix and then put it in the refrigerator for a night before using it. I learned here that the strength of the glue improves greatly if it is first cooled for a night.

For veneering with a hammer I'm not sure you have to use hot glue on the top surface while hammering. You can instead continually spray water as your hammering. You must use something on the top though it doesn't have to be HHG. Hope this helps; just my understanding of using HHG. Tom

Re: Starving Glue Joint

#8

Re: Starving Glue Joint

Adam Cherubini, NJ

>I have starved joints. Warren, you may not have gotten enough pressure to do it. With two clean faces and a short sample, you could only get so much force. The cleanliness of the samples meant you probably distributed that force over the entire area (and threfore may have been low).

I attached a leg of a Wm & Mary piece using hide glue. The leg had a dowel turned on the top of it which fit into a hole in a glue block inside the case, I turned the leg so I had a force fit. I didn't want the leg sloppy. When I later had to remove that leg, which, with the addition of glue forced me to mallet the leg all the way in, there was no glue on the dowel surface, only the shoulder.

Also seen this with M&Ts. Starving a long joint would be harder to do physics-wise. Harder to produce enough pressure which means it's also hard to produce ENOUGH pressure for PVA. This is one more reason I prefer a sprung joint with PVA. But that pressure overload at the ends of a sprung joint might just starve the joint of hide glue.

Glue joint starvation is real. Let's not pretend it's not because Warren or someone else offered some anecdote. (And I know Warren and I know that's not what he's saying). It's something to keep in the back of your mind. If you are hammering together dt joints dry, you may not get good HG strength in there when you glue. BTW, it's the same with epoxy, but not PVA. And teh opposite is also true. If you can't get good bondline pressure, that's when you want to use HG.

Adam

Re: Starving Glue Joint

#9

Different definition of "starved joint"?

Bill Tindall, E.TN

>If I understood your situation correctly your glue was scrapped off upon assembly, not squeezed out by clamp pressure.?.

When edge gluing, rapid loss of water next to the wood is going to make the viscosity of the adhesive so high in this area that squeezing a joint "dry" will be nearly impossible.

When we took the class at Headley's one of the most useful things I learned was proper joint fitting. If the joint (M&T or dowel) had to be hammered together it was too tight. Their target was a joint that could be pressed together by hand only. Because this shop has several generations of experience I took this advice to heart.

Re: Starving Glue Joint

#10

What glue did they use?

Adam Cherubini, NJ

>

Re: Starving Glue Joint

#11

Re: Starving Glue Joint

Warren in Lancaster, PA

>I design an experiment to deliver extreme pressure (one bar clamp per square inch of surface) and you call this anecdotal evidence? Not enough pressure? Do you actually know anyone who would use 30 bar clamps on 30 inch edge joint with one inch stuff? I deliberately produce a very flat surface to avoid the possibility that glue in the center is not being squeezed and then you suggest that the pressure is too well distributed? If you want to talk about anecdotal evidence, just look at your own leg dowel example. And as Bill pointed out, this example is complicated by other factors and was not an experiment designed to test the effects of pressure.

This morning I repeated the experiment with four times the pressure: one bar clamp on 0.25 square inches. This is equivalent to 120 bar clamps on a 30 inch long joint. I might have crushed the walnut a bit, but the joint is sound.

Re: Starving Glue Joint

#12

Re: Different definition of "starved joint"?

Adam Cherubini, NJ

>Starvation is just a lack of glue, however that happens. The glue can extrude out due to being scraped or being forced out.

As far as viscosity goes, HG can/will have different viscosities in different situations. Manipulating the viscosity of r a given application is part of the art of using HG and the reason why it's good to have these discussions with experienced guys (like Warren or Jeff Headley) in person, as you have found out, Bill.

Pat Edwards likes really watery runny glue. In a cold shop that will become thick really fast- probably too fast. Some old English workshop folklore mentions the warming of baords over the shop's heat source prior to glue up to reduce the globulization (I made that word up just for you Bill).

I think when the viscousity increases, it's almost easier to press the glue out of the joint. Unlike PVA, I don't think HG penetrates the wood much. When you scrape it of, it doesn't leave a mark etc etc. It doesn't seem to stick more to the wood. Like Pat I like watery glue that flows off teh brush in one continuous strand. In fact, I recently wrote about thisin PW and even got pictures of the glue flowing off my home made brush (thanks to John Aniano's hair!)

I'm wondering of there's enough interest in this subject for me to bring my HG to WiA. Are you going Bill? What do the rest of you think? I always hope I can meet some of you in person at these things. And if a picture is worth 1000 words, doing with somebody experienced or even, in my case, fitting together what I've been talking about with what I'm actually doing, has got to be worth a lot more than 1000 words.

Adam

Re: Starving Glue Joint

#13

Re: Starving Glue Joint

Adam Cherubini, NJ

>Bar clamp is not a pressure. Bar clamps/inch is not an engineering unit. Are you suggesting that it's impossible to extrude hide glue out of a joint? I've given a few examples from my shop.

It happens. It's an issue to consider. I've worked with Warren for years. He has a light craftsmanly touch. I've seen people who aren't like Warren. If you spring a long joint, the clamp pressure is magnified at the ends. This is the scenario that concerns me. If you are making long joints with hide glue, I'd be careful about "over clamping". But of course "over clamping" is a function of the clamp, the strength of the user, the size of the pad, the stiffness of the wood, the moment of inertia etc etc.

Adam

Re: Starving Glue Joint

#14

two vastly different situations

Bill Tindall, E.TN

>Case 1 edge gluing- By one means or another, before the clamps are set both wood edges will be covered by a layer of glue. Water from this glue will be rapidly penetrating the wood and drying the glue, thereby increasing its viscosity at the wood surface. It will not be possible to flow this thin layer of glue from the wood surface by clamp pressure, as Warren has amply demonstrated.

BTW, a properly made sprung joint can be closed with a piddly amount of clamp pressure so this factor is an insignificant contributor to joint pressure.

Case 2 M&T joint.- It is common to only apply adhesive to the mortise so glue doesn't gush out of the joint upon assembly. Upon inserting a "piston fit" tenon the leading edge of the tenon can scrape glue from the mortise sides leaving behind too little glue to wet the tenon surface. Because M&T joints are no where near the perfection of mating achieved in edge gluing, it may well be that the gap filling mechanical action of the glue is the most important factor in M&T joint strength. Therefore, it takes much more glue thickness to make this joint.

Re: Starving Glue Joint

#15

Re: Different definition of "starved joint"?

wilbur

>"Starvation is just a lack of glue, however that happens."

This is true. But to be fair, Adam, when I looked back on where this discussion came from, this is what you had said:

"It's true that high clamp pressure will starve HG joints."

So if we are talking about whether high clamp pressure can lead to glue starvation, I'm not sure how your example of a tight fitting M/T joint really applies to this discussion as you originally defined it. As has been mentioned, it seems more likely that the glue escaped the joint because of the scraping from the tight fit of the tenon into the mortise.

As to whether Warren's careful planing of his workpieces led to a decrease in pressure which saved the joint from starvation, that just doesn't seem reasonable to me. Firstly, the amount of pressure in his setup far exceeds what any reasonable woodworker would do, given that he had completely covered the joint with clamp pad surfaces with his REcord sash clamps. I also have no reason to believe that those clamps don't generate at least the same amount of clamp pressure as any other commonly used clamp today. From your descriptions of clamp pressure leading to glue starvation, it doesn't seem like you are talking about people who put a clamp along every inch of the joint, as Warren did in his test, and I certainly don't think that such an approach describes your clamping style.

Secondly, if I had made that joint, I certainly would not have done as good a job as Warren did, which would lead to smaller areas of contact due to a not precisely matched joint. This does lead to greater pressure at those areas, but it also leads to areas of less pressure in between where the sides of the joint don't touch. It is hard to see how glue starvation would occur in those spots where the joint doesn't match well, since the high clamp pressure that supposedly causes the glue starvation won't be happening in those areas.

Warren's test is only a piece of anecdotal evidence. But so are the other examples that you have given so far.

Re: Starving Glue Joint

#16

Yet another Starving Glue Joint theory

Bill Tindall, E.TN

>In assembling a panel, if glue is applied to only one edge, and this glue begins to dry before the joint is assembled, the partially dry glue will not wet the mating edge of the board next door. Upon disassembly this joint will appear "starved' as evidenced by the lack of glue on one of the surfaces.

Re: Starving Glue Joint

#17

Re: Starving Glue Joint

Patrick Gibbons

>I've been building a project that I'm using plans from Charles Hayward and Bernard Jones (they've both got almost the same instructions and dimensions) and have been putting together panels. I did not use hide glue so maybe I'm out of line here having no experience but the panels were originally done using hide glue and the joints were grooved, or feathered. The technique was to make one end of the insert thicker than the other and rub the joint til the glue tacked. These would be reasonably long joints (up to 36") that if done correctly would not require great clamping pressure and because of the joint would have much greater glue surface.

Re: Starving Glue Joint

#18

Re: Starving Glue Joint

Warren in Lancaster, PA

>Yes, Patrick, I routinely make rub joints with hide glue and would have done that with these small pieces of walnut also. In this case I used higher pressure than previously just to see what would happen.

For my second trial I used a Jorgensen heavy duty steel I-bar clamp (7200 series) for which the the manufacturer claims a load limit of 7000 lbs. The clamp has a 5.75 inch cast T handle and 5/8 diameter square threads with a pitch of .167 inches.

Re: Starving Glue Joint

#19

Re: Starving Glue Joint

Wiley Horne--So. Calif.

>Warren, I just want to thank you very much for the data. What you have done is not an anecdote--it's a repeatable experiment with N=1. Any of us could reproduce what you've done, closely enough for woodworking purposes. I consider this a strong demonstration, in that it eliminated all variables except clamping pressure, which was the original question.

I also think that Adam's experience has great value, and am glad he explained it in such detail. Because any one of us could replicate his experience, as well, had he not presented it in detail. The force-fit dowel joint and M&T's that he cited could with a bit more info be differentiated from your face-join/edge-join example, based on the scraping factor, combined with the time it takes to get the mating pieces fully together, and whether both of the mating faces were wet when joined.

In your initial post, you emphasized that join time was the critical variable, so one is not clamping a joint that's already gelling.

This has been a great discussion, all owing to the experiment you went to the trouble to perform.

Thanks again,

Wiley

Re: Starving Glue Joint

#20

Re: Different definition of "starved joint"?

Adam Cherubini

>You'd need to test the "overclamped" specimens to see if you were getting similar properties as unclamped pieces. That might be another definition. Glue starvation may involve not having enoough glue to achieve full strength. Industrial users of epoxy concern themselves with clamp pressure and bondline thickness. This may have influenced my perception of hide glue as I think epoxy and hide glue share a lot in common.

Reiterating, I don't really care what causes the lack of glue. The point is, HG is not like PVA. I think you can overclamp it. I think the glue will extrude out and you will loose bond strength. I think that (back to the original question) 18th c craftsmen didn't have a lot of clamps, didn't have great long clamps, and therefore didn't/couldn't starve joints. This may have contributed to their success. But the onus is on me to proove all this. And that's something I'm not working on. But here are my suspicions. I think mine is a consservative approach. Simply, don't assume HG will function like PVA. Wiping it on and pulling up any gaps or reducing the bondline with high clamp pressure may be a bad idea.

In the bigger picture, I think guys did stuff that effected the survivability of their furniture. We skip that stuff then wondder why our pieces behave differently or declare the original designs unsound. For example; I think they floated logs across the Delaware. I've heard tell that submerging logs in water is a great way to "dry" them. The water replaces the free water and eventually (?) the bound water preventing the cell walls from collapsing. I'm not sure about this, but this is what I heard. We typically don't do this so our white cedar for example, may not be the same as theirs (if you follow me).

adam