Am putting together a bookcase and have started gluing it up. I started using titebond, however, I remember from past experience that the polyurethane glues seem to be a great deal stronger. Also I could use a little expansion in a couple of joints. Is one really better than the other, or should I stay with the Titebond, which has not failed me yet? All comments welcomed.
Discussion Forum
Get It All!
UNLIMITED Membership is like taking a master class in woodworking for less than $10 a month.
Start Your Free TrialCategories
Discussion Forum
Digital Plans Library
Member exclusive! – Plans for everyone – from beginners to experts – right at your fingertips.
Highlights
-
Shape Your Skills
when you sign up for our emails
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply. -
Shop Talk Live Podcast -
Our favorite articles and videos -
E-Learning Courses from Fine Woodworking
-
-
Replies
Poly glue works fine. It satins and finishes better than PVA. I does tend to foam out of the joints but is easy to scrape off after dry. As Far as filling gaps, not to be offensive, but, Try to make your joints a little tighter. Nothing will really make up for ill fitting joints. I mean they will hold most likely hold together, but might not look so good. If you have a gap of bigger than 1/32 nd I would use an epoxy.
Good luck
Polyurethanes will never "starve" a joint, which means that PVA's can soak into end grain areas leaving none in the joint (plywood edges, etc), where PUR's will soak a little, yet expand to keep from starving the joint. The best plywood PUR glue I've found for case construction is PL Premium Polyurethane construction adhesive (in a caulk tube). Its as strong as Gorilla glue, but waayyy less expensive - 2.99 a tube at Lowe's or Home Depot.
"The furniture designer is an architect." - Maurice DuFrenes (French Art Deco furniture designer, contemporary of Ruhlmann)
http://www.pbase.com/dr_dichro http://www.johnblazydesigns.com
Due to the cost, I reserve my use of Poly glue to applications which will be exposed to the weather. I find nothing at all wrong with PVA for all my "dry" work.
As far as filling gaps, ... assuming it is only aesthetic and not structural, I think it depends on the type of finish you will use. Consider, for example, a thick gel stain... these stains will easily build up and hide wood filler.
Both are stronger than the wood itself, so either in a proper joint would be fine. Polyurethanes are completely waterproof. They have longer working time, which is a benefit in a complicated glue-up. Titebond can be cleaned off the hands, whereas polyurethanes cannot. It will also launder out of clothes unless you use one of the water resistant formulations. Titebond is less expensive and has a longer shelf life than urethanes.
Neither is too strong when there is a gap. One remedy if you have, say, a tenon that is too thin for a mortise and is fairly uniform is to make shims of paper or cardboard. You can use various materials, from kraft paper bags, to cereal boxes, to thick writing pad backs. You can shim one or both sides, whatever makes the surfaces come out flush. Dry fit then take the joint apart and glue it up with glue on both sides of the shims. If the joint is really loose, make wooden shims.
If the joint involves personal safety, such as in a chair, it would be better to make another piece.
I use poly glue all the time; it's great for some things, with all the advantages you list, but I have to disagree about it being stronger than wood. Edge-joints in hardwoods will break right along the glue line with no tearout.
In shear I think PU glue is plenty strong, and even in tension if there is enough surface area. But it does not make a bond that is stronger than wood, in my experience."Everything should be made as simple as possible, but no simpler." A. Einstein
http://www.albionworks.net
I have never had an edge to edge joint fail when I put enough Titebond on to squeeze out when clamped. Before glue application I make sure the edges fit. That is; only a slight spring joint. Each edge is square to its' respective face. Too much clamping pressure is as bad as not enough on glued edges. I have gone to using biscuits because it is easier to get one surface flat.
Whenever I have tested the joint of a cutoff piece the wood tears instead of the glue line failing! I don't glue any wood when the room temperature is less than 55 F or will be less than 55 within 12 hours.
Hatchet
The current issue of Wood Magazine has and excellent comprehensive test of woodworking adhesives. They tested a number of parameters from glue line strength to stength after being wet. Also, a number of old "truths" are investigated. For example, traditional hide glue was far inferior to today's premixed glue. The pre-mixed hide glue was far stonger.
The simplistic bottom line is that all the adhesives tested were stronger than the wood itself for edge to edge joints. Strength beyond that is basicly irrelevent. Therefore, ALL were more than adequate for woodworking purposes. There were difference between manufacturers and between types of adhesives there are relevent to woodworking.
Guess I'll have to get a copy and check into that; I wonder what wood they used?
As I say, my own testing shows conclusively that PU glue joints are weaker than white oak. Edge joints without splines, given a sharp tap on the bench, will break every time along the glue line with no tearout of the wood. Same thing with Type II is harder to break and there is usually some wood tearout."Everything should be made as simple as possible, but no simpler." A. Einstein
http://www.albionworks.net
They used maple which is the wood used by the Forest Products Lab to evaluate adhesive strength.
BTW, poly adhesives take a long time to develop strength. They require tight clamping for a number of hours where the PVA's typically required about half an hour. We found in the shop I was involved with that poly adhesives took about 24 hours to develop full strength. Once fully cured, they were stronger than the wood itself.Howie.........
I'm familiar with the strength gain over time (the samples I mentioned had been glued for days), but I thought you could strip the clamps after the glue hardened (about an hour here). That's what the instructions claim, anyway, and I haven't noticed any strength difference between joints I clamped for an hour and those left in clamps overnight."Everything should be made as simple as possible, but no simpler." A. Einstein
http://www.albionworks.net
According to the instructions published on the Gorilla Glue site, parts should be tightly clamped 3-4 hours and they go on to recommend "24 hours for best results". That's what I have found too.
Personally, I don't use PU adhesives much but I have never found them to be any less strong than other adhesives.
You need lots more clamps when you use polyurethane glues.Howie.........
Albion,
It seems to be repeated a lot that since glues are stronger than the wood itself, [fill in the blank] doesn't matter.
My simple-minded brain wonders if penetration shouldn't be a factor. A glue that just sits on the surface might not be as strong as one that penetrated several pore thicknesses into the wood. The joint would have to break more material in order to fail.
PVA glues recommend applying glue to both sides in order to "wet" them, whereas poly glues generally recommend one side. I wonder if factors such as wood porosity and humidity might prevent a good bond to the other side.
Can any experts comment?
Dan
Hi Dan and all - you mentioned: "PVA glues recommend applying glue to both sides in order to "wet" them, whereas poly glues generally recommend one side"
That would make perfect sense for these reasons:
PVA glues dry "too" fast for adequate 'soak-in' properties - notice how fast they skin over and grab with initial tack. Wetting both surfaces gets "fresh-out-of-the-bottle" glue to contact and soak the most onto the glue surface, but once on the surface, the air in the wood (pores and fibers) itself begins polymerizing the glue. Not saying PVA's are bad here - other adhesion properties make up for this, like cohesive strength.
PUR's on the other hand, cure slower allowing deeper penetration and "wet out" of the wood fibers, which makes PURs superior for porous joints like MDF edges, plywood edges, end grain etc.
This is precisely why pre-heating (with a heat gun) the joint like done with epoxy is a very effective method of multiplying the joint strength.
Epoxy is superior in porous joint strength due to deep fiber penetration (as long as the joints are prewet for 10 - 15 min to prevent joint "starvation")
I glued five sections of mahogony cove moulding to create a half-round section of compound curve cove moulding (pic below) for my boat and I pre-heated the end grain before applying the epoxy, and the wood just sucked the epoxy in. After cure, the section was strong enough to clamp in my workbench vise without cracking the joints. Niether PUR or PVA glue would have been good here, but the penetration is the moral of the story here.
"The furniture designer is an architect." - Maurice DuFrenes (French Art Deco furniture designer, contemporary of Ruhlmann)
http://www.pbase.com/dr_dichro http://www.johnblazydesigns.com
RE: PL Premium Polyurethane construction adhesive
You have been pushing this stuff in several threads, so I finally bought some. Tried it out today (rebuilding exterior wood stairway), and you were right -- it is way better, and easier to work with than conventional construction adhesive. Squeeze out seemed to wipe up very easily with a paper towel and paint thinner.
I guess only time will tell how it stands up.
My question is, if I have occasion to use it on a piece that will be stained and varnished, what kind of care do I have to exercise in cleaning up any squeeze-out?
Also, the directions advised that you should use gloves when working with this stuff. I can tell you they weren't kidding; my hands are full of dark stains that no amount of handcleaner will remove.
Thanks again for promoting this..................
Came across the Wood magazine with the glue testing article the other day. Didn't feel like paying $7 just for that article, so I read it in the grocery store... ($7 for a magazine???)
Anyway, I didn't see where it claimed that all the glues produced joints stronger than the wood. In fact, it looked to me like the weakest glue - Franklin PUR, the stuff I have been using - produced joints that failed on the glue line, just as I have experienced.
It also looked to me like there were some serious flaws in the testing procedure. For one thing, they don't seem to have allowed enough time to reach the ultimate strength. Take a look at the graphs of strength vs time - all the lines are still going up steeply at the end of the test, meaning the glue is nowhere near done curing. They only allowed a couple of hours, IIRC.
The joints could not have been stronger than the wood, or all the joints would have the same failure strength, equal to the strength of the wood adjacent to the joint. The very fact that there were differences in strength between the different glues means that the joints were not stronger than the wood.
I'd like to redo this analysis but carry it out to the ultimate strength, include all the glues that are available on the Ace Hardware shelf (there are at least 5 PURs, for example), and present the results a little better. But I need to find someone with the clamping and testing equipment."Everything should be made as simple as possible, but no simpler." A. Einstein
http://www.albionworks.net
A reasonably well made joint in a competently designed piece of furniture shouldn't create stresses on the glue line that even get close to the failure point of the glue. There are a lot of things that are important to know about the performance of wood glues, but their failure strength is at the very bottom of the list of a dozen characteristics.
Testing glues to their failure point would be like testing ordinary car tires at 200 mph, the results would be interesting, and give everyone something to talk about, but pointless.
John W.
I suppose that is true in principle, but I don't think it is quite that simple in practice. For one thing, furniture is frequently subjected to stresses that were not anticipated by the designer, because users are always unpredictable. For another, good design often involves getting the most performance out of the materials - using high strength materials in order to lighten the structure. In such cases the strength of the glue might well be as important as the strength of the wood.
If we could be confident that all reasonably well-made glue joints were indeed stronger than the wood, then we could ignore glue strength. But I think this paradigm is unproven and contradicted by experience. Specifically, I claim that Franklin's PUR glue does not always give an edge-to-edge joint that is as strong as the wood I'm using (white oak). So when one of my customers dropped his chair in a certain way, the back split right along the glue joint. (That made it easier to repair, but still cause for concern.)"Everything should be made as simple as possible, but no simpler." A. Einstein
http://www.albionworks.net
Reading magazines in supermarket lines is always fraught with danger. For example:
>>Anyway, I didn't see where it claimed that all the glues produced joints stronger than the wood.
Under "Strength of bond" (Pg 82) "Our first test was joining edge grain to edge grain and with virtually every glue, the wood broke before the glue joint."
>>In fact, it looked to me like the weakest glue - Franklin PUR, the stuff I have been using - produced joints that failed on the glue line, just as I have experienced.
The chart you were looking at was presenting the results of EDGE grain to END grain joints, not EDGE to EDGE.
>>It also looked to me like there were some serious flaws in the testing procedure. For one thing, they don't seem to have allowed enough time to reach the ultimate strength.
From the description of the testing process on page 81
"All joints were clamped in a pneumatic vise set to clamp with exactly 150psi of pressure, the pressure recommended by glue manufacturers." "... the sample joints were removed from the vise after two hours. (Polys set slow, so those were clamped for four hours). All joints were then allowed to cure for 72 hours before we tried to break them."
>>Take a look at the graphs of strength vs time - all the lines are still going up steeply at the end of the test, meaning the glue is nowhere near done curing. They only allowed a couple of hours, IIRC.
That chart is not showing ultimate break strength. It is showing that joint strength increases with time as the adhesive cures. The chart shows the strength at 5 minute intervals up to 30 minutes. It has nothing to do with ultimate strength which was tested after 72 hours as indicated above.
>>The joints could not have been stronger than the wood, or all the joints would have the same failure strength, equal to the strength of the wood adjacent to the joint. The very fact that there were differences in strength between the different glues means that the joints were not stronger than the wood
Again, the chart is not showing ultimate break strength.
You may want to spend the money for the magazine and take another look at the article. I think the tests were well designed and performed consistently. Howie.........
Reading magazines in supermarket lines is always fraught with danger.
I guess you are right about that... apparently I missed a lot. Wish I knew somebody with a subscription; I still just can't see paying $7 for one article of interest. Especially if the conclusion is that all the testing is irrelevant anyway because all the glues are stronger than the wood...
Still, I note they said "virtually every glue," which is not quite the same thing as every glue. That says to me that at least one glue produced joints that were weaker than the wood. Since they did not choose to present the actual failure stresses in a table (unless I missed that too), but rather gave a qualitative rating, it's hard to know for sure. The Franklin PUR seemed to be rated significantly weaker than the others, so I suspect it was the cause of the word "virtually."
"Everything should be made as simple as possible, but no simpler." A. Einstein
http://www.albionworks.net
>>at least one glue produced joints that were weaker than the wood
Yes, the quoted statement went on to say "The only exception was Titebond Polyurethane, which failed under a very acceptable average of 1,650 psi of shearing pressure".
From one of the charts, the wood failure point averaged about 1,800 psi.
The relevent point of the chart showing edge-grain to end-grain performance was that some PVA adhesives produced joints stronger than the wood and the poorest failed at about 1,200 psi. All of the poly adhesives failed at less than 800 psi.
Of interest was how poorly the PVA's did in the water resistance tests. Even the new "waterproof" Titebond III did less well than their older "water resistant" Titebond II.Howie.........
Tim, I don't have a definitive study, but two points to consider...
1. The only glue joint that I have had fail (other than fixing something while saying "I wonder if this will hold") was the corner of a fairly large mirror frame, done with polyurethane glue. I cleaned the joint out well, went back to the good old fashioned yellow glue, and it has been in service now for a couple years.
2. When I do a glue up, and later trim the length, I save the scraps. Whenever a visitor questions the strength of a "yellow" glue joint, I give them a couple scraps to break. Only once has the break occurred near the glue line.
Today I put some crown moulding on an already-finished bookcase. I used PUR for better adhesion to the already finished surface. An hour later I was still trying to keep the crown moulding tight and clean up the foam-out. So I spent a few minutes scratching the finish on the second bookcase, used PVA, and had a perfect job in far less time.
Did I mention that I don't like PUR glue?________________________Charlie Plesums Austin, Texashttp://www.plesums.com/wood
I'm not liking it as much as I used to. It is easy to use and sets up fast here in cold, humid Albion; but the strength is just not there. I do much the same thing with offcut scraps, and they nearly always break along the glue line with PUR. So I am going back to yellow glue, and will just have to heat my shop when I do glue-ups.
NB these are edge joints, clamped tight overnight. I do a lot of work in 1/2" white oak and typically use three bar clamps on 18 to 24 inch long panels; do I really need clamps closer than 9 inches?
As I'm sure you know, PUR has to be clamped or the foam pushes the joint apart. Clean-up is extremely easy, though - you just wait until it has hardened (gve it a few hours). The foamy squeeze-out is brittle and comes right off, especially from a finished surface (wax makes it even easier). Cleaning it up before it sets would be a nightmare."Everything should be made as simple as possible, but no simpler." A. Einstein
http://www.albionworks.net
Tried your tip on the PL adhesive in a tube. I am assembling a buffet, the case being veneered plywood. So far, so good. Thanks for the tip!
By the way, DOES Cleveland rock? I was born in Toledo, and went to a Browns game about 30 years ago, but don't recall anything specific.
Well, . . . Cleveland has to claim something.
Back when I built cabinets for a living, I used PL Premium butt joints with a glue cleat exclusively for all my case glue ups. You heard right - no dados, biscuits, screws, or dowels. Strongest cases around too.
Don't agree or don't believe me? Well, do some tests like I did. I made 4" x 4" x 4" sample "L" joints in all the joints listed above, and tested them against identical "L" butt joints of PL and put them in an arbor press with a scale to record threshold weight before breaking.
All broke around 35 to 40 lbs, even a PL butt joint without a glue cleat broke at 37 lbs. Dado being very weak because people don't realize that cutting a dado groove destroys the skin strength of plywood.
The PL butt joint with a 3/4 x 1" glue cleat for added glue surface broke at 125 lbs.
And I guaranFREAKINtee you that is the fastest case joinery method there is. Just glue and shoot a couple of pin nails and throw on a deep throat clamp or two.
I routinely did double oven cabinets and TV cabinets where the weight bearing shelf didn't have more than two finish nails in each side. You could still throw them down a set of stairs without joint breaks (substrate failure maybe).
PL is even better on melamine. Just about welds to thermoset surfaces.
"The furniture designer is an architect." - Maurice DuFrenes (French Art Deco furniture designer, contemporary of Ruhlmann)
http://www.pbase.com/dr_dichro http://www.johnblazydesigns.com
Since this discussion has focused somewhat on various glues, their relative strengths and the effect of curing time -- just thought I'd pass along the fact that the current Wood magazine features a "glue review" that included some tests that relate. Have just scanned it, so no details from me, but you might want to take a look.
forestgirl -- you can take the girl out of the forest, but you can't take the forest out of the girl ;-)
Another proud member of the "I Rocked With ToolDoc Club" .... :>)
Hate to plug the competition, but "Wood" magazine compared several kinds of glues in their most recent issue and concluded that polyurethane was stronger than any other, including Titebond.
This forum post is now archived. Commenting has been disabled