When glueing several boards to form a table top I use titebond glue. After the glue sets, 24-48 hours, I sand smooth and finish. After a couple weeks I notice that at the joints a little ridge forms leaving a rough finish. I suspect moisture form the glue has raised the grain at the joint. Is there another type of glue that I could use to prevent this form happening?
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
You probably should wait longer to allow the moisture from the glue to dissipate. I have seen a suggestion that using a hair dryer might speed the effect. The reverse effect can be seen when using biscuits. If sanded too soon, then eventually you will see dimples in the plank where the wood has shrunk back and is lower than the rest of the top due to the sanding.
I've experienced the same thing with well seasoned wood and well dried joints. It could be glue creep.
A polyurethane glue (like gorilla glue) will not creep and is excellent for tight joints like those between table top boards.
Thanks for the info. I'll get some and try it. Dick
Actually the problem is not with the glue at all, but a flaw in the gluing technique. Titebond is a really great product, and it is definitely not the problem.
Whenever gluing, always clean off the squeeze-out right away.
If you leave the squeeze-out glue on the joint while it is drying, the surface of the squeeze-out glue will harden, then the moisture inside of that little (or big) blob of glue will be absorbed into the wood along the glue joint. This causes the wood to swell.
So while it is swollen, you come along, assuming that the glue is dry, and sand the wood and glue. Then the wood shrinks, leaving a little ridge of glue (since it does not shrink).
It is best to clean up your squeeze-out with a sharp flexible putty knife so that you scoop it up rather than rubbing it in. I sharpen mine to a polish on my buffing wheel, just like I do my carving chisels, and clean it as well as I can, as soon as I can.
I also don't recommend wiping the excess with a damp cloth, which we have all seen on TV.
Hal
http://www.rivercitywoodworks.com
"Actually the problem is not with the glue at all, but a flaw in the gluing technique.... it is definitely not the problem."Then why, after years, does PVA continue to rise above a glue line that was cut back well below the original surface long after any moisture equilibrium problems should have disappeared? I believe the answer to be that it creeps. It probably will never quit creeping. It is plastic, after all. I am instinctively skeptical of a manufacturer rep's information anyway.Cadiddlehopper
You are correct Cadiddlehopper. This type of "creep" is a recurring problem with PVA adhesives. In the shop I used to be involved with, we had the technical folks from our adhesive manufacture trying to help us. While there was some impovement with re-formulation, we never fully resolved the problem. What seems to happen is the when boards are glued up during a time of higher humidity and the boards later contract, the glueline remains raised to the original height of the seam. This will leave a raised seam that is easily felt and seen. We had a number of call backs for the problem.The depressed glueline is an entirely different issue. When a waterbased adhesive is applied to the edges, it causes the edge area to expand. If you do not let the glueline fully dry out before leveling the glueline, you will have a depresion when the area later shrinks to it's normal thickness. This can happen with any type of water based adhesive.As I said earlier, using an adhesive that dries harder will eliminate the glueline creep problem.Howie.........
"using an adhesive that dries harder will eliminate the glueline creep problem."That is what I found that I had to do. The PVA extrudes from the joint years later even if sanded away after years of curing.Cadiddlehopper
In my experience, this doesn't happen if the wood is glued correctly. But then I have only been using Titebond glue since about 1974. Check your technique.
It is true that Titebond wood glue is somewhat thermoplastic. Type I is fairly thermoplastic, Type II is much less so. This is why Type II is better for structural work like curved laminations. In my experience, Type III stays gooey, and should never be used for structural work. But if two boards being glued together have similar structure, and the same moisture content, there shouldn't be a problem.
Improperly prepared surfaces are more likely to have something to do with the problem. For a while I worked for a guy that would sand his gluing surfaces and he had this problem, often.
I learned a long time ago that a properly prepared gluing surface will have a machined surface (jointed with a sharp jointer) with between 25 and 29 cuts per inch. A sawn surface, no matter what saw blade is used, is not good enough.
Hal
http://www.rivercitywoodworks.com
I haven't used enough Type II to tell that there is a difference, but I will attempt to observe. There have been some interesting suggestions on this thread. My solution has been to use more polyurethane glue if I must avoid telegraphing the joint. It has other advantages such as longer open joint time as well.Cadiddlehopper
Caddlehopper,
Reading this thread makes me feel as though I am in the Twilight Zone. I have used Types I, II and III and I have never noticed the problem. Maybe my eyes don't see so well or my fingers don't feel so well. I just looked at all of the furniture in my house for glueline problems. I checked over two pieces that I have in the final stages of construction. I still don't see and can't feel any glueline problems.What did I do wrong?
I feel embarrassed and naive to ask such a question, but what the heck, but what other way is there to learn? The best quip I have ever heard about someone who just realized that they were missing something obvious before is the 11 year old boy who blurted out in English class, "Ohmygosh, I never realized that I have been speaking in paragraphs."Why didn't I get the glue creap?
Thanks,
MelMeasure your output in smiles per board foot.
Well, the thing that comes to my mind is that well-made joints have very, very thin glue lines because the joints have no gaps to fill with glue. They are then well clamped--which also contributes to thin lines. Firm clamping pressure is called for--and its hard to see how anyone could overclamp boards over 3/4" thick.
If you want a non-creeping glue without the hassles of polyurethane glue, try hide glue. Even the liquid glue is virtually as strong as titebond, and stronger than the poly. glues. The hot hide glue is even stronger.
Steve,
Thank you for the info on glue joints. Since I have not had any problems with them, I'll just continue as usual.You did bring up an issue that I have been having a hard time getting good info on - the liquid hide glue from Titebond that I buy at Woodcraft. Woodcraft takes the date off, so you cant tell how old they are. Titebond told me that I shouldn't buy it if it hasn't got the date on it. Others tell me just to try it out before using it on a project. If it does its job on a test piece, all is well. Do you go with that rule of thumb, or do you suggest that I do what the Titebond rep said -- only use a bottle with the date on it, and don't keep it more than a year? Secondly, I am happy to hear that you feel the liquid hide glue is pretty strong. I would like to use something with greater open time than Titebond I,II or III. The liquid hide glue really takes a long time to set up, which is fine with me. I am not "in the business". Would you recommend it for most carcase construction? Are there any places where you wouldn't recommend it?I suppose that I should try real hot hide glue. Given my tiny workshop, I have shied away from anything that will take up more space. I do appreciate your advice. Given the many threads on Knots that have little to do with woodworking or only have to do with which tool to buy, I really enjoy and profit from a thread such as this which gets down to "how to". Again, thank you,
Mel
Measure your output in smiles per board foot.
The liquid hide glue that Titebond makes is nothing like real hot hide glue. Well maybe it has hide glue in it, but in my experience it has none of the wonderful charactistics of real hide glue. It stays sort of rubbery (similar to Titebond III) and takes a long time to dry.
Another thing that probably many don't know is that Titebond III, the "Ultimate Wood Glue" is not really the ultimate wood glue. I find that it does not really dry very hard, although it sticks well and is easy to use.
Polyurethane glue is also not all it is cracked up to be. In my experience I found it to be the worst of all glues for woodwork. True, it will stick to anything, but it also creeps. Many think that since it will fill cracks, it will glue those cracks, but in reality it has no strength in the filled void like epoxy does. Also, it does not stick well to resinous wood.
Hal
Hal,
Thank you. You essentially said to stay away from liquid hide glue and from Titebond III. What do you recommend that I use?
Thanks,
MelMeasure your output in smiles per board foot.
Hi, Mel -- I've been reading this thread with a lot of interest because Hal Taylor has done a lot of research (which I've seen) on all types of glue and his conclusion for furniture is to use Titebond II and III. Since the color varies after curing, that's what governs his choice. At no time has he experienced the problems mentioned above, but he preps the glue joint extremely carefully and uses lots of pressure. He's never had a callback on any of his cold laminations or joints. I'm not sure of the physics or chemistry behind it, but I know there's barely any glue left in the joint after clamping and the work pieces start at the same MC. The test he's run show glue penetration into the wood fiber such that the failures during testing are not glue line failure but wood fiber fractures -- both butt joints and parallel grain joints. He says he's never had glue creep as there's not enough glue left in the joints to cause it. Good enough for me!
Jimma is correct. I think that it may be that you have too much glue in your joint. Clamp it tight.
It is interesting that PVA glue, like Titebond I & II, don't really glue like many people think they do. It isn't a case of the glue sticking to the wood on one side, and then sticking to the wood on the other side, with the glue in the middle holding it all together. The glue must imbed itself into the wood on both sides of the joint without the glue in the middle.
I don't know what could make your glue continue to rise out of a glued joint unless you are applying heat to it. But it takes a lot of heat. Is this an ironing board by chance?
Clamp tightly.
Hal
Yup -- like eight 1/8 inch lams over 30 inches with 15 "C" clamps screwed really tight on a heavy form.
I've just glued up an 8'long by 35" 12/4 hard maple workbench top. It was well prepared through the jointer and planer. It is clamped very tightly. How long should it remain clamped? How long before I run it through the wide belt sander? I used Titebond II.George
Edited 11/1/2006 7:46 pm ET by geodown
I don't think you need to keep it clamped more than an hour unless the joints are stressed. I usually unclamp with Titebond II, if I need to reuse the clamps, in 45 minutes.
In this case though, it is a big piece involving a lot of work and strong wood so I would keep it clamped overnight, or at least 8 hours. Then let it relax prior to sanding.
I would think you could sand it after 24 hours if your shop is warm. Leave it so the air can circulate all around it.
Hal
Hal, geodown,I disagree that the top can be sanded within 24 hours of gluing. The wood adjacent to the glue lines has absorbed moisture from the glue and will be raised above the surface. Sanding will bring the swollen wood fibers down to the rest of the surface, but over the next several days, the swollen areas will shrink as they give up the extra moisture and will sink below the surrounding surface, leaving a depression along and adjacent to the glue joints.I would give the top at least 5-7 days before sanding.Rich
Thanks. I really appreciate the advice.George
George,The ideal time to get the beads (or gobs) of squeeze-out removed is several hours after clamping when the glue is firm but rubbery, not wet. If you wait too long (days), it gets harder and harder.I use a very sharp plane iron, laid flat on the glued-up piece, bevel up, and just slide it along to slice off the excess glue. This doesn't take off any of the raised wood fibers - just glue.Rich
I totally agree with this. The worste advice I ever had was to scrape the gobs of glue off with a paint scraper... It took chunks out of the wood. It's one of those things you only do once.
Leaving the glue squeeze-out on the surface to be scraped off later is a big mistake. It should be cleaned with a sharp flexible putty knife immediately after gluing.
The surface of the blobs will seal over, then all of the moisture inside of the blob will be absorbed into the wood. This also is a bit like putting a lid on the glue in the joint near the surfaces, resulting in making it take longer for the glue to dry.
If the glue is lifted off right after gluing, you significantly reduce the amount of moisture that is added to the wood, thus reducing your drying time, and the time you need to wait prior to sanding.
I would also recommend against using a damp cloth to wipe the glue off. More moisture, plus the thinned glue penetrates into the wood.
Hal
http://www.rivercitywoodworks.com
Hal,Waiting an hour, even a few to remove the rubbery glue squeeze out is not a big mistake it's a very good way to do it.No argument that if you can cleanly get the glue off the wood immediately with a putty knife, do it. The problem is that it's not so easy to do when the glue is fresh. It's very easy to dislodge and smear blobs (or other squeeze-out) that would otherwise stay nicely in place and get firm for easy removal in a little while, long before any hypothetical re-entry of the blob's moisture into the glue line. And with clamps in place, it's hard to reach all the glue lines. Besides, I'd rather not have to be careful with a putty knife wet with glue. With my luck, I'd drip it on something else.I've gotten to the point that I get very little squeeze-out, using my methods of glue application. Still, glue-up, as a process is anxiety-provoking enough. After I get a piece clamped, I like to just leave it alone and come back later when it is "calm" and all I have to do is remove those harmless, soft, "dry" beads.Again, if scraping clean when wet works for you, it's a good method to use. Otherwise, slicing rubbery glue off a few hours later is the way to go. Waiting too long (days) makes the job harder (but not impossible).Rich
I congratulate you for being able to put on just the right amount of glue.
However, cleaning it off later, whether you do it or not, is not in the best interest of most woodworkers, including yourself.
This is not just an opinion, but a scientific fact, that the surface of the glue blobs will cure quickly, sealing in the moisture in the blobs, with the only escape route being the wood. I am not making this up. You can see it for yourself by putting a big blob of glue on a piece of melamine or plastic, then the next day cleaning it off. The surface will be dry, the inside will be uncured and wet. If you do the same on a piece of wood, the moisture goes into the wood, and the curing of the glue in the joint is slowed considerably.
I feel that the people reading this post will be interested in knowing what is the correct and best method rather than knowing what you are doing in your own shop just because it seems easier for you and less risky. Then the others can make their own choices based on the facts rather than opinions.
Cleaning the glue off with a very sharp and polished putty knife is not really very hard to do. The glue is simply scooped up, a bit at a time. Then later no scraping will be necessary. It works well. It's the right thing to do for a better glue joint.
Hal
http://www.rivercitywoodworks.com
Hal,I don't know where you are getting your "scientific facts" about all this. The technique I have recommended is hardly my own. Removing excess glue after it has firmed up by a few hours is very commonplace. It's not my invention.Now, please tell me, if examination of next-day glue blobs on a piece of melamine reveals them to be dry-walled sacs containing wet glue inside, how that wet glue can get into the glue line of wood? Is there a channel from the inside of the blobs to the glue line? Of course not. The wet glue (if there is any of significance) is trapped inside the outer wall of the sacs. It's safe in there. Can't harm anything.Continue to use your wet glue cleanup method if that works for you. Cleaning up partially dried beads and blobs of glue a few hours after gluing is a good way for me and for most other workers who I have ever discussed this with or who themselves have written about this. And despite your claims of "scientific" evidence to the contrary, results in no excess moisture introduced to the wood.Despite your sarcasm about my claim of being able to put the right amount of glue down to achieve minimal squeeze out, I'll share it. It's simple, and I claim no particular ownership of the technique.It's nothing more than a small-scale version of the method used by tile layers who use a notched trowel to meter a very uniform layer of adhesive, mortar, mastic, whatever. Without that technique, their tile floor or wall would have an uneven surface and they could not be sure of optimum adhesion or absence of voids. If you've ever laid tile, you've spread the mortar, holding the trowel almost vertical to the floor and created the swirl pattern of ridges and valleys of mortar that tile-laying requires.I learned the technique from David Finck's, "Making & Mastering Wood Planes," Sterling Publishing, 2000, pp 98-99 (great book). The method uses strips of scrap about the thickness and twice the width of a tongue depressor. I notch the squared-off end on the band saw, about 3/32" deep about every 3/32". This notched "trowel" is used to spread glue on the gluing surfaces. Easily made in bunches, discarded after use.Squeeze glue out of the bottle and spread it with the trowel, wiping right off the ends and sides of the glue surfaces, with the trowel vertical to the surface. The glue will not run down the face of the boards. The trowel will leave ridges and valleys of glue and the surface will be coated with a very uniform amount of glue - just enough to create small beads all along the glue line when the boards are clamped. Far, far superior to spreading the glue with the bottle's applicator tip or even with a glue brush.It's possibe to adjust one's technique with practice to be able to expect nothing more than evenly-spaced, tiny glue beads all along the glue lines at maximun clamp pressure. Very nice.As I said before, gluing is typically an anxious time. Knowing that the right amount of glue has been applied, that it has been evenly distributed as a thin film on the gluing surfaces, that there is not enough to cause problems of running and dripping as the wood is handled and aligned in the clamps and that a very manageable amount of squeeze-out is going to occur, gives a MUCH more secure and less anxious feel to the entire process. One feels very much in control, not rushed or extended by glue running everywhere.Rich
Regarding: "It's not my invention."
I knew that. I saw the same technique used my my ancient old shop teacher back in high school.
"Now, please tell me, if examination of next-day glue blobs on a piece of melamine reveals them to be dry-walled sacs containing wet glue inside, how that wet glue can get into the glue line of wood?"
Rich, if you had done the requested demonstration that I suggested, you would see that the bottom of the blob is not sealed. Only the outer surface. The inside and underside is uncured glue, moisture and all. The wood that it sits on absorbs the moisture. The glue line that it is capping also doesn't cure well because it is essentially inside of the blob too.
"Despite your sarcasm about my claim of being able to put the right amount of glue down to achieve minimal squeeze out, I'll share it. It's simple, and I claim no particular ownership of the technique."
Okay, maybe I was a little sarcastic. Sorry about that. I also can put the proper amount of glue on a board most of the time. But really, you are the one that made the point of it.
Your gluing technique though, sounds like a lot of work just to glue a couple of boards together. As long as you are already going through all of that, it doesn't seem like too big a deal to spend a couple more minutes to scoop up most of the squeeze out, and end up with a better product that can be sanded sooner because not as much moisture has been absorbed by the wood near the joints. Isn't that the issue? How long to wait to sand?
Hal
Hal, you are exaggerating a problem that just doesn't exist. Glue blobs squeezed out of the glue line do not contribute moisture to the walls of the glue joint. At any rate they should not be large globs, just small beads. But even large globs are not the problem you describe at all.The swollen wood tissue on either side of the joint is from the glue in the joint. The swelling is real, and it needs several days at least to reslove. Sanding too soon can result in sunken areas at or adjacent to joints if sanding is done too soon and the wood then further dries and shrinks. I don't sand glue-ups for at least several days, usually about a week. I don't understand the need you describe to get it sanded immediately.No one is telling you not to use your method. If it works for you, that's great. Others can certainly do as they wish. Most will find it best to follow the advice of many other knowledgeable wood workers and teachers to allow the wood to dry for several days, at least before sanding a glue joint.Roch
Well I believe that the bigger the blob/ line of beads/gobules /whathaveyou the longer both it and the glueing wood below it take to dry, since the usual wood glues skin over- which slows down things.
So I prefer to wipe the line with shavings or similar asp, then after a suitable period, use a scraper or putty knife to scrape the glue line- depends on factors like glue type , temperature, degree of panic/neurosis in some cases(,)- but this needs to happen way before that glue line has got too hard. When I say glue line I mean the now thin line that has formed after one has wiped the initial squeeze out: squeeze out keeps happening for some time, but it is now either a fine line or just a line of very small beads.
Reference the swollen wood tissue on either side: here I have found that the glue type and the wood species also come into the picture, and give two examples of the procedures it was necessary to adopt in a factory that used American Red Oak and Muninga (Pterocarpus Angolensis): for the former all tops or panels required to have waited at least 24 hours before they went through a wide belt sander, or we would see the dreaded "glue-line dips" when the finish was applied. However with the Muninga we could get away with sanding within two hours. But waiting several days or weeks is going overboard, I think.Reference to Uncle Bruces Fine Book entitled Understanding Wood will further prove my point.
So that is my six pence worth.Philip Marcou
Edited 11/3/2006 8:41 pm by philip
Aren't we just talking about a workbench top?
Hal
Dang, I can't member-were we? Anyway , I don't want no stinkin glue line dips in my bench top.Philip Marcou
Here are a few words from others on the subject - From R. Bruce Hoadley's, "understanding Wood," Taunton, 2000, Chapter 11, "Adhesives and Gluing," p 197(The same information was printed in Taunton's "FWW, on Boxes Carcases and Drawers," 1985 and in a previous edition of FWW, I don't remember the issue):Figure 11.5, a drawing, shows 4 stages of a glue joint:A shows the joint just glued with beads of glue squeezed out at the glue line.B shows the joint swollen and raised at the glue line.C shows the suface uniformly leveled.D shows the joint sunkenThe caption reads, "When an edge-glued panel (A) is surfaced while the glueline is still swollen with moisture (B, C), a sunken joint (D) results."The last paragraph on the page reads, "Finally, cured joints need conditioning periods to allow moisture added at the glueline to be distributed evenly through the wood. Ignoring this can result in sunken joints (Figure 11.5). When edge-gluing pieces to make panels, moisture is added to the gluelines, especially at the panel surfaces where squeeze-out contributes extra moisture (see Figure 11.5A). If the panel is surfaced while the glueline is still swollen (see Figure 11.5B, C), the glueline will shrink when the moisture is finally distributed (see Figure 11.5D), leaving a joint that is sunken."Hoadley gives no information at all as to when to remove squeeze out, nor how long to wait for "moisture added at the glueline to be distributed through the wood," before surfacing a glued up panel.From FWW, No. 166, Nov/Dec 2003, "Gluing up Tabletops," Gary Rogowsky, pp 46-51, exerpt from p 51:"Invariably, the pressure from the clamps is going to squeeze some glue from the glue joints. Once that glue hardens, it's a chore to remove. To make life easier, I allow enough time, usually about 15 minutes, for the glue to turn rubbery. Then using a putty knife, the glue will scrape off with realtive ease."He advises to wait overnight before removing the clamps. Rogowsky does not address the issue of swollen glue lines, nor how long to wait after removing the clamps before post-glue up surfacing.(I always wonder what the instructions mean, "to wait overnight" for some process to complete. For gluing, if I clamp my table top at 8 am, then wait overnight and unclamp at 8 am, the process has gone 24 hours. But if I clamp at 11 pm, overnight can result in about 9 hours. Why don't authors say "wait 24 hours" rather than "overnight?")There is not a whole lot the experts say that will resolve all the differences we have expressed. I think Hoadley's advice that glue squeeze-out can contribute moisture to the joint gives some justification to Hal's advice to scrape glue immediately.However, I feel that if one can limit squeeze-out to small beads, rather than large globs, the amount that the squeeze-out adds to the glue line moisture is minimal if not insignificant.On the other hand, Rogowsky advises to wait until the glue has become rubbery to scrape it off. He says rubbery happens in 15 minutes. My experience is that 15 minutes is a bit short, as "rubbery" doesn't happen for a little longer than that - about an hour. At 15 minutes glue has skinned over a bit but is messy wet if disturbed. Your mileage will vary and shop conditions obviously are a big factor. I lived in Hawaii for the last 14 years and I'm now in Arizona. I haven't been able to set up my shop here yet. In Hawaii, ambient humidity was never below 55-60%, and was often higher. In Arizona, I may find that glue squeeze-out becomes rubbery in 5 minutes. (Arizona does not have low humidity, it has negative humidity. If you put a glass of water out on the table, the air just sucks the water out of the glass. That's how dry it is here.)I still feel that at least several days is necesary to wait before surfacing a glued panel to avoid shrinking at the glue joints. Maybe 24 hours is enough, but "overnight" is too short.Rich
Edited 11/4/2006 10:37 am ET by Rich14
Well there you go then-and I gave you two different timbers to prove my point.So what differences are we talking about now? In fact what are we talking about?
"overnight" means a minimum of 12 hours ,I believe.Philip Marcou
Edited 11/4/2006 1:17 am by philip
philip,"In fact what are we talking about?"Well, since it started as an argument about gluing, shall we say the long thread is . . . "male bonding?"(running and ducking)rich
Jim,
Thanks a lot. If it is good enough for Hal, it certainly is good enough for me. I don't want my heavy curly maple bathroom wall cabinet to fall apart and fall and surprise someone! :-)
Good to hear from you.
MelMeasure your output in smiles per board foot.
I know you're joking about that, but a couple of months ago a friend had a bathroom cabinet at work fall off the wall on her when she opened it and wound up with a concussion,broken neck and lots of complications. The irony is she is head of an industrial healthcare coalition! I guess it was a case of bad construction work, not glue failure.
Jimma,
That is one for the record books. I have been discussing how to hang the heavy bathroom cabinet with my wife. I will relate your story to her right now. After hearing that story, I will take extra precautions. Thank you.
MelMeasure your output in smiles per board foot.
If the cabinet fell off the wall, I doubt glue had anything to do with it.
Hal
I've been following this thread and I suspect that humidity may be a major factor. I'm in CA (low humidity) and have never had a problem with Titebond II - or any other PVA glue. I usually wipe off squeeze out right away with a damp sponge or rag and (if I'm in a hurry) have been known to take off the clamps and sand after only 3-4 hours. I've never had the problems I've seen described here.
I use a moderate amount of glue and spread it fairly evenly (usually with a finger) then clamp with just enough pressure to give me an even amount of squeeze out which gets wiped off as soon as I can get it. Most of my glueups sit overnight, but not always.
I'm betting that humidity is causing a lot of the problems I've been reading about.
I do the same thing, sanding within an hour, often.
I just don't think the glue is the issue here.
Hal
Dear 9619,
Aren't you the one that was saying that all of your gluing has been fine until you read this?
Keep doing what you are doing.
Hal
Hal,
Yup. I am the one who never had a problem with Titebond glue creeping. I was wondering why others were having the problem and I wasn't. I plan to keep using it. Steve Shoen did mention both hot and liquid hide glue. I have used the liquid but not the hot. I guess that I still have that adventure to try. I would like to learn the real good and bad aspects of the liquid hide glue, just so that I know.Meanwhile, I am happy as a clam that my glueups are working fine. I wish that I had more time, but so does everyone who doesnt use hide glue.Have fun.
MelMeasure your output in smiles per board foot.
Mel,
<!----><!----> <!---->
Hot hide glue can indeed be an adventure. We used it almost exclusively in the pipe organ business for leather and fabric to wood applications. Hot glue joints are easily steamed open so it is frequently used when a joint or repair must be reversible. Check out this link for more info:
http://en.wikipedia.org/wiki/Hide_glue
Nazard,
Thanks for the pep talk. I am expanding the set of things that I have tried as I near retirement. Hide glue is in the "to do" list. Thanks for the Wikipedia link. I already read it. Very useful. Thanks for letting me know about the applications of hot hide glue for leather to wood applications.
Enjoy,
MelMeasure your output in smiles per board foot.
I just finished running my hand over a piece I made in 2001 using a PVA glue. The glue lines have telegraphed again. They were sanded down once during the 5 years that passed. Maybe the glue was Elmer's or Titebond I, I can't be sure of the species. Perhaps I put them all in the same class mistakenly. A very similar piece glued with polyurethane exhibits no glue lines. It is newer by about one year, made of same lot of wood.I read all the previous posts. Perhaps humidity is a contributor to the condition since I live in a hot, humid climate. But why not the polyurethane-glued piece? The glue lines aren't very visible, but they can be felt with the hand.I usually err toward excessive glue in a joint, but use plenty of clamping pressure. Considering that I use a jointer/planer to prepare surfaces, there is little chance of getting a dry glue line with adhesion resulting from surface penetration only (if that is possible).Notice that I wrote in paragraphs!Cadiddlehopper
Cadid,
These are the type of problems to have. Your paragraphs are superb!
Have a great week. I am off to California for a week. Back in the saddle next weekend.
Have fun.
MelMeasure your output in smiles per board foot.
Telegraphing PVA glue lines have been happening, to some of us, for a long time. Some of us have gone back to hot hide glue for this reason. In my experience the creep resistance of hide glue is superior to the PVA's. Also, hot hide glue is less likely to show up through some finishes. I like Polyurethane(PU), gorilla glue and similar, when it has to be weather proof. I've not had creep problems with PU's in curving stair rails, rocking chair rockers, or other laminations.
I think PVA's have their place and I will continue to use Titebond I, II, and III, but just as I don't use hot hide glue for a project that will be exposed to weather, I don't use PVA's when I'm concerned about creep.
Ditto what Pondfish wrote.
Cadiddlehopper
Dick,
I called Titebond a while back to get their opinion. In the absence of a moisture meter, they suggested you wait two weeks after glue up before you sand/plane and apply finish to insure the top is in moisture equilibrium...
Titebond glue is a fine glue. The suggestion of using a putty knife is a good idea, because you don't want excess moisture added to the joint with a damp cloth. Something else that might cause this problem is clamping pressure. A good rule of thumb is to have enough pressure to get a uniform bead of glue at the glue joint and then stop clamping. Allow this to dry after using the putty knife to get the bead cleaned up, for at least 24 hrs. This is best option I can think of. If you don't use the putty knife, allow the glue to cure and then use a cabinet scraper to clean the bead. This does two things. First you can get the joint clean, second the scraper flattens the joint so it is uniform with the rest of the surface. Hope this helps!
Hmmm... How do you prepare your boards for glue up, specifically the edges? Do you buy rough sawn lumber, or dimensioned stock? What type of clamps do you use for glue up (pipe, bessey...)? Finally how much squeeze out do you get?
Wow!!! I just got back from a couple days on the river and am really suprised at the interest in this topic. Guess I'm not the only one interested in this issue. I have just checked furniture project I have made dating back to the middle 90's and some have glue line telegraphing and some don't. I can't tell you why.
I buy rough lumber, the project in question is a quartersawn white oak table, and work it down. I don't have a jointer but use a 6' carpenters level clamped to the fence on my table saw and use a sharp Forrest blade to prep the glue edges. No visual voids on the edges. I biscuit joint every 8 or 9 inches, apply a liberial amount of Titebond II or III and clamp with pipe clamps every 12" or so. I tighten the clamps down moderatrely tight untill there is a glue line squeeze the full length of the joint and wipe it off with a damp cloth. Usually clamp over night, let it sit for 3 or 4 days and then sand and finish.
My shop is in Central Oregon and we have low humidity both summer and winter.
Titebond's clamping recommendations call for oak to be joined with 200 to 300 lb. per square inch of pressure. Just to reach the lower pressure calls for a Bessy clamp every 6 inches each tightened nearly to its maximum. Not moderately tight, but really tight.
The biscuits don't help pull the joint together or add any to strength. If they aid alignment one every 18" should be plenty.
You also need to evaluate how tight "no visual" gaps really is. If you stand the boards on edge in a darkened room and shine a light behind the boards would you see any light between the boards? If that light can be seen with more than light hand pressure pulling the boards together then it isn't really tight. I sure can't get that tight on the table saw.
Steve
Thanks for your information. Maybe I'm not using enough clamp pressure. That could very well be the variable I was missing. May have to invest in a jointer. Maybe if I'm real good Santa will find my workshop. Dick
One of the problems with bisquits is that each one becomes a pocket full of moist glue that takes a really long time to dry. All of that moisture is absorbed by the wood.
On an edge to edge joint, there really is not a significant amount of strength added to justify using them, unless the glue joint is not done correctly.
Hal
http://www.rivercitywoodworks.com
William Tandy Young’s “The Glue Book,” Taunton Press (out of print—but see Amazon.com) covers many of these issues of glue line creep etc. He is very much an advocate of hot glue, (and other glues as appropriate). I used hot hide glue on a few projects, and have been very happy with the results. However, like many of us, I also use Titebond, liquid hide glue, and fish glue. I became interested in the ‘correct’ application parameters for Titebond, and found the application notes at Titebond.com. As Steve Schoene notes in his post, Titebond calls for incredible clamping pressure. I was also interested in the recommended spread rate. Here are <!----><!----><!---->Franklin<!----><!---->’s guidelines about these two parameters:
<!----><!----> <!---->
Application Guidelines: <!----><!---->
Minimum required spread: 40 lbs. per 1,000 square feet<!----><!---->
Required clamping pressure: Enough to bring joints tightly together (generally, 100-150 psi for softwoods, 125-175 psi for medium woods and 175-250 psi for hardwoods)<!----><!---->
<!----> <!---->
The spread rate works out to .007 inches thick (7 mils) wet film. The hardwood clamp rate is 12 to 17 atmospheres (try that with a vacuum press), or, about 25,000 to 36,000 pounds per square foot. How big is that veneer piece you’re clamping? I talked to a tech guy at <!----><!---->Franklin<!----><!----> to try to make sense of these numbers. He said not to use so much pressure that you starve the joint. I pointed out that most of us can’t develop that type of clamping pressure, and he fell back to “as long as the surfaces are well-machined …”
<!----> <!---->
I’m an amateur, and have lots to learn, but as a result of my research, when I use Titebond, I lay a very small amount of glue on the wood, usually do quick rub of a straight joint to distribute the glue, and then clamp tightly to close the line (i.e. squeeze the glue). If I see a lot of squeeze out, I know I screwed up. Seven mils ain’t that thick.
<!----> <!---->
About six months ago, I contacted Mr. Young (The Glue Book author) to suggest that FWW might be ready for an article that covers some of these aspects of gluing—spread, pressure, creep, etc. He didn’t think there was a need, but this thread might indicate there is interest.
<!----> <!---->
Thanks for your insight on the ues of biscuits. My learning curve keeps going up. Dick
Dick,I agree with Hal about the limited usefulness of bicuits. I'll go much further - I have no use for them at all. They are among the most misunderstood "tools" in a woodworker's arsenal. They were developed by the furniture industry to help in alignment. Like dowels, they add nothing at all to the strength of an edge to edge joint, and little if any strength in any other kind of joint.In mass production, time is one of the most expensive commodities. Dowels and biscuits allow rapid, accurate alignment for glue-up. But a buscuit in an edge joint actually robs the joint of strength because the buscuit slot removes potential long grain to long grain wood contact in that area. And long grain to long grain contact is the only thing that gives a glue joint strength.Biscuit jointers are a phenomenon unto themselves. Everyone's gotta have one, because "that's what the big boys use."There is nothing wrong in using alignment devices to get your glue-ups properly oriented. But understand what's happening in there. Use the smallest biscuits you can, and the fewest that will suffice. It actually would be better to use dowels (again the fewest that suffice) becasue less gluing surface is removed from the joint. Don't fill the biscuit hole with glue or get glue on the bicuit at all. Just get glue on the surrounding wood joint surface. Better yet, learn to use battens to keep the pieces aligned.Rich
Rich,
Biscuit joiners were developed as a method to join chipboard, not as an alignement tool. See the history of Lamelo for a brief history of the tool. When used as intended they are a very usefull tool. I assure you that the strength of a but joint would be greatly increased by the use of a dowel. That said, I think there is smarter joinery.
I find the tool usefull in the quick shop jobs, and those that don't approach 'fine' woodworking status. Primarily anything that is to stay in the shop or goes into a garage.
Rich has anyone actually done a test regarding the strength of edge joining with and without biscuits? I've never seen one. It's clear the joint is plenty strong without biscuits, so I don't see the need to add them, however I doubt your claim that the joint is made weaker. Your claim that the only thing that give sthe joint strength is long grain to long grain discounts any mechanical strength given by the biscuit. Regardless I think that any strength is inconsequential since the joint is already very strong.
Buster,With the slightest push, this thread will go down that usual path of useless argument that discussions about the utility of biscuit joinery always goes. You already relegated biscuit projects to jobs that never leave the shop or go into a garage (whatever that all means - no need to elaborate).I could always start with, "If it's worth doing, then it's worth doing right . . ." But I won't. Yes biscuit joinery was developed for man-made sheet goods, in particular particle board. But you misunderstand a lot of the needs for that joinery if you don't realize that alignment was the principle factor in the biscuit development.No matter. How many woodworkers use biscuits for particle board joinery? Very few. Biscuits have become "standard practice" in all manner of softwood and hardwood joinery. Mostly in serious misunderstanding of what acually happens in a glue joint (properly or improperly made).The single most significant statement that can be made about the requirements of a wood/glue joint is that there must be long grain to long grain contact. I'm sorry if you don't appreciate that.In the case of edge joined boards, the long grain of each piece runs parallel to the other. In the case of a mortise and tenon joint, the long grain of the cheeks of the tenon runs at right angles to the long grain of the faces of the mortise. (No other wood to wood contact within that joint is of any significance to the glue bond.) There must be intimate contact of freshly-prepared long grain to long grain for bonding and adhesion to occur.In a properly made edge joint (such as a glued-up table top), the wood-glue bond is stronger than the (surrounding) wood itself. Place enough breaking force across any of the table top's joints, and the top will break AWAY from a glue line. Not AT the glue line. That is, the wood itself will break, but glue lines will hold, or the break will happen adjacent to a line, leaving wood fibers still adhered all along the line.The joint strength is proportional to the area of the joint. Anything that reduces the joint area (voids created for biscuits, dowels) reduces the joint strength. Since the joint is stronger than the surrounding wood, how can the overall structure be made stronger by placing some mechanical, "joint strengthening" device in the joint? If only a few dowels are used to align edge glued boards, there is not much gluing area sacrificed. But what about the case (not uncommon) where a novice places dowels every few inches to "strenghthen the joint?"In fact, dowels don't only weaken a joint by taking away potential gluing surface, but the dowel and its hole in the surrounding wood only share a tiny bit of long grain to long grain contact. All the rest of the dowel-board contact is long grain to short grain, or a transition from long to short grain.Biscuits remove a lot of potential gluing surface within the joint.You may go on thinking of biscuits as mechanical aids to joint strength (little "floating tenons" within the joint?), but even if you used actual wood tenons, oriented in the correct direction within the joint, you could not increase the strength of the joint, only weaken it by creating glue voids in it.It might be argued that a tongue and groove joint could increase the strength of edge gluing. Yes and no. Again T&G is principally for alignment purposes and makes clamp-up of long boards a lot easier than using many cross battens. But the typical T&G joint does not result in close proximity of the wood surfaces within the joint and since most of the mating surfaces do not touch within the joint, the actual glue bond is weaker than without any T&G at all. The tongue in the groove contributes very little, if any strength to the joint.The only argument that really works is that of a so-called "glue joint" or "lock miter" (shaper or router bit) joint. This effeftively doubles or tripples the glue area down the edge of the boards, exposing much more long grain to long grain contact.The "tongues" and "grooves" of this joint mate intimately with each other but by only 1/8" to 1/4" and are not a mechanical factor in the joint. The lock miter joint is a great way to absolutely align work pieces, and it looks great in exposed joinery if one can pull it off. The resulting glued-up piece has a look of great precision and integrity. But without a high-end, massive shaper assembly driven by power feeding equipment, all necessary to hold the work in precise alignement during the cut, it can't be made well for long edges.And what's the point? The plain edge joint is already stronger than the surrounding wood.I know of only one "study" done and published in a popular publication about this kind of thing. It was many years ago by John D. Wagner and published in FWW, I believe some time in the 80s. I remember reading it, but I don't have that issue any more. I do have Taunton's "Practical Design, Solutions and Strategies," published in 2000 and it has a reprint of the article, pp52-57, "Comparing the Strongest Joinery for Doors."The article compares lag bolts, T&G, dowels, loose tenons, mortise & tenon, two biscuits and three biscuits, in a stile to rail joint. As an "experiment" compaing the various techniques, it was so seriously flawed by uncontrolled variables and inconstistencies, as to make any conclusions invalid. Early in the article, the author stated, "It would be a stretch to call these tests a rigorous scientific examination."Yet, the study did exactly that. The stress at which the joints failed was "measured" and "compared" as if the values actually meant something. (The biscuits, for instance, were stronger joints than the mortise and tenon!). The article was published without any critical analysis of the methods and materials, (a serious omission by FWW) and I remember it's being quoted as a reference for the (high) strength of biscuit joints. I wonder how many door makers would stay in business if they used biscuits in their stile to rail joints?I am sure there are many industry publications with real data regarding glue joint strength. I stand by my original post.Rich
The point isn't to disagree with you. My argument isn't that biscuits are agood or bad, I believe in most cases specifically in hardwoods they are unnessesary. However to reject any joinery in whole based on loose 'facts' is silly. Biscuits have their place, I wouldn't put them in the must have catagory, but there is a place.
I think the discussion of making the joint weaker is silly. Unless adding biscuits reduce the strength significantly, or to the point of failure then srength is not an issue. I suppose academically this is important, but warning against using biscuits on this fact is pointless.
In your example of dowels, if you had a 3/4" X 10" glue surface, and drilled 1/4" holes ever two inches (starting two inches from the end for a total of 4 dowels) you'd end up decreasing the total long grain area by about 2.6%. If left as a void to you feel this will significantly reduce the strength of the joint? Of course they are filled with a dowel which reduces the glue void to a negligible amount. This adds an undetermined amount of strength directly to the joint. Finally due to the orientation of the dowel (longrain perpedicular to the long grain in the wood peice) the area aroud the wood will be supported against "bending" forces.
If I remember correctly, the article you are referring to, the testing of biscuits vs. mortise and tenon joints in Fine Woodworking Magazine (or maybe it was Fine Homebuilding), was based on using them for joinery a full sized cherry, frame & panel house door.
I think that at the time there were very few readers who actually bought into the results.
They are stronger than nothing in a butt joint though, and easier to use than dowels.
Hal
http://www.rivercitywoodworks.com
A biscuit jointer was not designed as an alignment tool. If it was, then it was one helluva lousy design.
If you visit a modern furniture factory that uses a lot of sheet goods you will see the huge pneumatic powered jigs that they use for workpiece alignment.
I do agree with you that dowels make a much better alignment aid and were I to use an alignment aid it would definitely be dowels and not biscuits.
The problem with using dowels for alignment is that it is nearly impossible to get the dowels aligned well unless industrial tools are used for the drilling and alignment.
I agree that biscuits are not for alignment. I use them when gluing plywood to plywood joints like in a frame for a fireplace mantle, or plywood framed wall paneling.
It really does add a lot of strength to a joint like this. I always build the frames in my shop, and have never broken one during transportation or installation. Without the biscuit, the glue joint is very delicate.
Hal
http://www.rivercitywoodworks.com
I've had dowels fail (break like a twig) when used in challenged joints, but never a biscuit. Biscuits make great splines for miter joints, and certainly add strength if it's a picture frame endgrain to endgrain miter.
I do recommend dowels to my students when absolute alignment is needed, as biscuits generally only help in one direction. I've got two students using threaded 3/8ths x16tpi metal rod with epoxy to assemble chair frames this semester, in two very different designs. The threads give the epoxy something to grab where a smooth rod would not, and steel in the joints adds stiffness and transfers forces across much better than a wood dowel or biscuit could.
It doesn't seem like a dedicated dowel hole drill machine made much like a plate jointer would be hard to engineer. You would have to be a bit more careful to put the alignment mark right on the line, but I can imagine a strategy to make alignment of multiple dowels reliable with such a machine.
4D
This has been an interesting thread to watch. Over the last three decades or more in the business I've witnessed and experienced various forms of creep with PVA glues. I find it interesting that Hal maintains the only reason a furniture make will experience creep in their work is due to faulty work practices. I believe he's only partly correct. The funny thing about this phenomenom in my experience is that its appearance (or not) seems to be somewhat unpredictable. I've glued up table tops and the like, been pretty sloppy about my procedures and the parts have never shown creepy pimples. On the other hand I've been very careful about the job and, 12 months later, there the pimples are. Strip the piece, clean the pimples off, refinish, and they show up again some months later.
Of the two primary forms of creep that exist the one that opened the thread related to the row of pimples sometimes seen in panel glue ups, such as table tops, so here's a story that might give pause for thought.
In the early 80's I needed to glue up a panel about 60" long in a hurry. I whacked matching edges straight over a surface planer, scooped a sprung joint quickly with a hand plane, slopped on PVA and pulled the lot tight with a single cramp in the middle. I made two panels roughly 19" wide like this.
I left the glue runs where they were and moved on to another job. Half an hour to 45 minutes later the panels came out of the cramps, the glue scraped off and smeared about and they were shoved through the thicknesser on both faces. These two 19" panels were then glued together using the same simple plain edge joint with three cramps to make a top approximately 38" wide. There was a bit of hand planing done, followed by power sanding in an oscillating thickness sander and hand sanding.
Within one work day (8 hours) the panel went from rough sawn timber to a completed top with two coats of pre-cat polish sprayed on it. Next morning the top was attached to the frame, the piece blanket wrapped and delivered to the gallery for exhibition.
I thought I'd probably need to make a new top when expected problems showed up. The funny thing is I still see the job from time to time. The couple that bought the piece friends of my parents. There has never been the slightest hint of a problem with that hastily made top. True, the polish is getting a bit past it and could do with some work, but that's a different matter.
My time living and working in Gulf Coast Texas was when I experienced the greatest number of problems with the 'row of pimples' creep. No matter how carefully I worked creep would almost always rear its ugly head. In this case I came to the conclusion the problem was not my work practices, but it was to do with the difference between my workshop conditions and typical interior conditions. A hot, humid workshop (typically 70-80%RH all year and often 95ºF+) is quite a way out of synch with the temperatures and RH values experienced in Houston houses and offices. Put the furniture in a house and the wood shrinks, but the glue doesn't shrink at the same rate as the wood-- therefore pimples. I couldn't change my workshop conditions for various reasons, but I could change my glue. I switched over to mostly using hide glue, a non creeper, and that fixed the problem. Slainte.Richard Jones Furniture
Thanxx!! This is very useful information.Cadiddlehopper
Richard,As always, good info. In all of this discusion, I assumed that PVA glue was the subject and resisted mentioning hot hide glue as a solution to many of the problems we were "sharing" here.Also interesting that your problems with gluing happened in Houston, and my frame of refernce was the humid environemnt of Hawaii. All the furniture and picture frames that I made there are now in the near-zero humidity of northern Arizona and they are all doing fine, except one picture frame whose miters have opened at the inside corners. It's the only frame without splines in the joints. All glued with Titebond II.The ones with splined joints are as tight as they were when first made. (in all probability, the frame members are curved now with the vast change in moisture in compensation for the tightly held joints, but I'm going to ignore that possibility, as that wood movement is not visibly obvious, while gapping miter joints are.)Rich
The siren song of hot hide glue!!! We used it in my high school shop. Veneer buffs really like to use it. Polyurethane glue is my chioce wherever I think that hot hide would be a better choice than PVA. That being my choice, I don't need a heated glue pot & stock of flakes. Whatever anyone else does for his own satisfaction is fine with me. It all adheres.Cadiddlehopper
Until the stationary-tool fairy visits your shop and drops off that 8" jointer, you may want to consider investing in a decent jointer plane, either a #7 or #8. With a little practice it's quite simple to create glue-ready jointed edges.Jim
"There are two spiritual dangers in not owning a farm. One is the danger of supposing that breakfast comes from the grocery, and the other is that heat comes from the furnace." - Aldo Leopold
After reading a lot about glue "creep" and other problems which I have never experienced, I have to conclude that the basic problem that some folks are having are two fold
First, it seems like they are not clamping tight enough. I know from my engineering studies that bonds are stronger the thinner they are. I believe it is a myth that you can get a glue starved joint by clamping too tightly. If both sides of the joint have been wetted with the glue, all tight clamping will do is force out excess glue.
Secondly, I believe it is very, very wise to let a glued panel sit for a few days since the moisture in the glue joint ends up in the wood. I try to wait five days or more before sanding and finishing the panel.
When I follow these two practices, tight clamps and waiting, I never have glue line problems no matter what glue I have used.
Bill
Bill
After reading all the comments here I arrived at the same conclusion. In the past I have been worrying about squeezing all the glue out of the joint so didn't clamp real tight. Wish I would have figured this out 10 years ago. Thanks Dick
Unless you are using epoxy, which is stronger, the thicker it is and the more that is left in the joint.
Hal
This forum post is now archived. Commenting has been disabled