Replies

1. 
   bandit9497 | Dec 05, 2008 01:04am | #1

   Citation,

   There is a discussion going on this in the Joinery section.  Some very good suggestions on this topic.

   Bandit

2. 
   thumbnailed | Dec 05, 2008 01:05am | #2

   For me, the best method is to apply just enough glue to get a small bead of squeeze-out, and scrape it off when rubber hard. It takes practice to get the right amount, and you don't want to let the glue get fully hard before scraping. Don't over clamp, either, to avoid a glue starved joint. I do finish before assembly when possible, as it does avoid the whole issue. There are a number of ways of dealing with glue joints, and it pays to learn as many as possible, as different ones will be better solutions for different problems.

1. 
   bandit9497 | Dec 05, 2008 01:44am | #3

   Citation,

   How do you determine the correct clamp pressue?  Is this just by feel or is there some guidance that is important to know and follow?

   Bandit

1. 
   user-83153 | Dec 05, 2008 03:17am | #4

   Bandit,

   The correct clamp pressure is however tight you can get them. You couldn't squeeze a joint hard enough to glue starve it if you tried.

       Rob

2. 
   thumbnailed | Dec 05, 2008 04:11am | #5

   Normally I wouldn't be strident, but the post after yours couldn't be more wrong. Except in special circumstances, your pieces should fairly accurately fit dry, and clamping pressure should be the equivalent of 'hand tight'. It matters less if the glued piece isn't structural, but you might as well learn to do it properly so you won't have something important fail on you. There's a risk asking for advice in a forum, since you have no way to determine if either I or the other guy know what we're taking about, so take everything you read with a grain of salt. As an example, do a forum search on 'thumbnailed' and find the thread on alternate ways of building a door. Good luck, and have fun.

1. 
   SteveSchoene | Dec 05, 2008 07:42am | #8

   Yes, thumbnailed is right--there is risk in asking for advise of a forum.  But, he (and Jeff) are completely wrong about using only relatively light hand pressure.  Manufacturers and wood scientists have shown conclusively that ideal clamping pressures are quite high.  For example, Bruce Hoadley in Understanding Wood speaks of dense woods such as maple and birch calling for 200 lbs. per square inch.  Tropical hardwoods up to 300 lbs per square inch.  For relatively light hardwoods, perhaps as low as 100 lbs per square inch.  Titebond calls for 100-150 psi for softwoods, and mediuim woods, 125 -175 psi; with hard woods 175 - 250 psi. See, for example,  http://www.titebond.com/IntroPageTB.ASP?UserType=1&ProdSel=ProductCategoryTB.asp?prodcat=1  At least one other study found that optimal clamp pressures were even higher.

   Those pressures,  can only be achieved with firm clamping pressure.  For example, Bessey K body clamps can only achieve a bit more than 1,000 pounds ---call it 1100 lbs.of pressure, the new K-body Revo clamps perhaps 1,500 lbs.  Three quarter inch pipe clamps perhaps 1,100 lbouts.  So if you have a joint that is three feet long, gluing up a panel of just 3/4" thick boards, you have 27 square inches of glue surface, and at 175 psi, would require 4725 lbs of force.  That is beyond the capabilities of 4 classic K-body clamps turned as tight as you can get them.  So to be sure, you need 5 clamps (spaced a little over 7 " apart)  tightened to about 85% of their design maximums if you have sufficient grip to accomplish that.  Increase the thickness of the panel just a little, to 7/8" and those 5 clamps will almost achieve the 175 psi required at full 1,100 lbs. of force.  Make the wood maple, and you would need at least 6 K-bodies for that 3 foot panel.   

   Edited 12/4/2008 11:44 pm ET by SteveSchoene

1. 
   thumbnailed | Dec 05, 2008 11:34pm | #10

   Steve,

   I don't remember which cynic said 'you can prove anything with statisics' , and I can't say there is any specific thing wrong with your info, but from a practical 'doing it every day for 35 years' point of view, it's just not that hard to clamp wood together and get a good glue joint. I have no idea where the author derived those clamping pressure numbers, but you can get a very strong joint just rubbing two jointed blocks together with glue and leaving them to dry. I'll allow that there are many valid ways to accomplish the same task, but IMHO that kind of clamping force is completely unneccessary. 

2. 
   SteveSchoene | Dec 06, 2008 06:45am | #11

   It's true that there is a wide range of clamping pressures that will hold two boards together under most all circumstances.  That's why the "I did this for 35 years and nothing has fallen apart", is valid, but the science does contradict the idea that you really are likely to over clamp.  It also is a guide to those few occasions where the glue joint might really be tested. 

   I'd bet the reason for joints that appear to be glue starved isn't over clamping, but having glue dry before the joint is clamped together.  This is more of a problem if you try to get the absolute minimum of glue.  I'd rather deal with the squeeze out.

   One other point should be made, relative to the OPs questions.  That is that there is never any need to discover glue when stain is applied.  I always wet down the surfaces with naptha that reveals any glue problems so they can be corrected while it is easy, not after stain has been applied and the fix up has become difficult. 

    

3. 
   JeffHeath | Dec 07, 2008 07:03pm | #20

   It always amazes me how every thread that you are involved in becomes a pissing contest about "Steve is right and everybody else is wrong".

   You must be some sort of engineer because you're always citing statistical mumbo jumbo.  It's been a long time since I was (your words) completely wrong about anything, and I especially appreciate the delicate way at which you consistantly put things around here.  Do you really think you need to read a bunch of articles and scientific papers to understand how to avoid glue squeeze out????

   In 25 years of building furniture, the last 20 professionally, I've never, not once, had a glue joint come apart in any of the woods that I have worked in.  BTW, that includes almost all of the North American hardwoods, and a small percentage of imports.  I've been using rub joints to glue up raised panels doors without clamps with no failures, and I can tell you from practical experience that if your woodworking skills are good, and the joinery is prepared properly, you don't need alot of glue or clamping pressure to make the joint structurally sound.  Enough pressure to close the joint is all that is required.  Any more is overkill.

   Why don't you put the books and encylopedia's down, and actually build something.  And, try to be a little more courteous with your responses, especially if you're going to include my name in them.

   I said to use enough clamping pressure to close the joint.  I never said anything about light handed pressure.  Don't misquote or misinterpret what I say to someone else again, you forum troll.

   Jeff

   Edited 12/7/2008 11:07 am ET by JeffHeath

   Edited 12/7/2008 4:41 pm ET by JeffHeath

4. 
   user-83153 | Dec 07, 2008 08:50pm | #21

   "You must be some sort of engineer because you're always siting statistical mumble jumble."

   Jeff,

   I believe the correct term is "mumbo jumbo", not "mumble jumble". You are wrong again. :^) (Ducks head and runs)

   This discussion has become largely theoretical and has probably ceased to be of any interest to the guy that asked the question. I say we should all just mill our parts accurately, apply a light even coat of glue to all surfaces to be joined, clamp them up nice and snug and don't worry to much about glue starvation. I don't believe you will encounter much difficulty with this method.

   Oh yeah, if you are going to do rub joints, hide glue works great for them.

       Rob

   Edited 12/7/2008 2:40 pm ET by Rob A.

5. 
   JeffHeath | Dec 08, 2008 12:38am | #22

   Hey, Rob

   You're right.  Mumbo jumbo it is.

   Have a great holiday!

   Jeff  (no head ducking required)

   PS  Analytical paperback experts who are always all-knowing (aka Steve Schoene) drive me nuts, and also drive people, myself included, away from this site.  Main reason for my previous 6 month absense.

6. 
   SteveSchoene | Dec 08, 2008 04:45am | #23

   And, all I said was that you can use just about as much pressure as you like without starving the joint.  The advise that one should use only the minimuim pressure needed to pull the joint together is wrong.  Factually wrong.  But it's not the same as saying that using such light pressure will lead to failing joints. That's the informatiion that you are well qualified to provide.  And, I never challenged your opinion on that. But it isn't the only way that works, and that's what was said, and what I challenged.  

   What I said was that it appears a very wide range of pressures work.  That, therefore, one need not generally worry about clamping too much.  The numbers only indicate that there are a lot of people, including the glue manufacturers who also suggest relatively high clamp pressures.  That is the accepted viewpoint of people who have studied the matter systematically.  It's not my opinion, but that of a lot of very knowledgeable workers, who have based it upon scientific study. 

   I didn't have the advantage of a mentor for much of my learning about woodworking, I've alway found reading as my best source of information, just as others find hearing things verbally to work better, and a lot of folks, particularly who do work with their hands, find hands on methods of learning to be best.  No particualar method is the best, only the best for me, or the best for him or for her.

   (I retired from a career analysing fixed income (bond) investments, particularly of companies in "distress", and making recommendations to institutional investors (mutual funds, hedge fund, insurance companies, and the like) on whether to buy or sell those securities. And, no I never recommended mortage backed securities. Incidentally, my formal training is as an economist not engineering.)  But since my "thing" is reading, I naturally I read about woodworking, just like I read about everything.  Personally, I learn from reading, remember what I read, and can often put that into practice.   My copy of Hoadley's Understanding Wood is from 1989, though the first printing of that edition was in 1980.  That's how long knowledge about using higher clamp pressures has been available to amateur and small commercial woodworkers. I probably read it first on a train commuting to work in the City.    

   Is it necessary to have a scientific study to figure out out to glue a joint?  Perhaps not, but the people who really do lots of it--furniture manufacturers-- have obviously decided that it is a question worth spending real money to have answers they can rely on. They likely do as many panel glue ups in a year than those working in small commercial shops do in a lifetime and must feel that the anecdotal evidence from their defect rate wasn't sufficient to tell them how to adjust their manufacturing process to either reduce defects, or to produce more cheaply without increasing defects uneconomically.  

   And, as a general matter, without statistical analysis (mumbo jumbo) properly done, you can say almost nothing very authoritatively.  Statistics and its biostatistics and econometrics sub-fields) is a discipline essential to sorting out sense from non-sense.  Dismissing statistical analysis out of hand is just anti-intellectual snobbery.  I do know (or once did) enough to do it, and likely still know enough to understand the limitations, and to take it apart when it is done poorly.  Challenges aside it is still an essential tools of science. 

   In a basic sense, science never proves the truth of anything, it only demonstrates the absense of falsity on a specific issue.  (Think of the difference between being innocent of a crime or being not guilty.)  But, you can't argue that a scientific hypothesis is false, without providing the same kind of rigorous evidence--right down to the statistical analysis and proper experimental design that would have been used to provide support for the hypothesis.     

7. 
   JeffHeath | Dec 08, 2008 05:56pm | #24

   Steve

   My bad.  Get outside and go for a walk, will you??

   Happy Holidays,

   Jeff

8. 
   SteveSchoene | Dec 08, 2008 07:55pm | #25

   Heading out now. 

   Happy holidays,too. 

2. 
   user-83153 | Dec 07, 2008 12:43am | #14

   Actually Thunmbnailed, I could be much more wrong since what I said is basically the truth. I don't recall saying anything about using clamp pressure to force improperly prepared boards together. I will, however, say that it is virtually impossible to glue starve a joint without using mechanical or hydraulic force.

   If you wish to continue to disagree then you should probably do a little research. This article would be a good start.http://www.taunton.com/finewoodworking/SkillsAndTechniques/SkillsAndTechniquesPDF.aspx?id=29561 

   In the article the author, Roman Rabiej, states that using an average of the maximum amount of force that 4 different test subjects could generate for different types of clamps, achieving the recommended pressure to edge glue two 3/4" thick x 3' long pieces of red oak would require 9 heavy duty bar clamps, 12- 3/4" pipe clamps, or 26 quick grip clamps. He also says that the recommended pressure is half of the optimal pressure required for the strongest joint, which is just short of glue starvation. The pressure would double again if the edge being glued were flatsawn instead of quartersawn. So unless a man of above average strength is routinely using 24 pipe clamps (48 if the edge is flatsawn) tightened as tight as he can get them, to clamp two 3' boards together, he probably is not going to starve the joint.

      Rob

1. 
   thumbnailed | Dec 07, 2008 01:37am | #15

   OK, Rob, you have many stats, and creditable sounding references, and obviously you're not a crank just contradicting for its own sake, so I'll ask if you have done any of those things you described, or caculated the results you post from an academic point of view? I was going to throw an "only" in there, but realized that would be an unfair minimalization of potentionally accurate logic and calculation. I should have said that I can't say you're wrong, only that it doesn't correspond with what I've experienced. There's an implication that the concept of "glue starved" is a phantom phenomena, except perhaps when insufficient glue was applied in the first place. The recommended glue clamping force stats you use may well be for the assembly of massive gluelam beams in a factory setting by minimally trained employees using the lowest common denominator for the quality of the wood. Without the context and references for the stats, I have to give more validity to my life experience and what I've actually done than to even the best phrased argument based on numbers and stats. For instance, what does " the optimal pressure is double the recommended pressure" mean? Who's recommended pressure? Optimal for what, according to whom? If I read the whole article, which I admit I haven't yet, does the author do measured breakage tests to confirm his statements? I'm going to finish reading the article, and see if I find it consistent and logical, and I'll get back to you.

   OK, I've read the article, and am not impressed. Why should a joint be clamped just short of the wood's crush point? Author doesn't justify that. There is no other justification for doubling the clamp force for different grain orientation other than that original assumption. All his pressure #'s in his chart have to be based on the crush factor values, and again are only as valid as the first assumption. And his failure test is invalid. If you split a board and the grain runs off, it shows nothing about the glueline. The sample looks like it was split outside the glue line, no glue at all shows in the picture. Bench clamping one edge and forcing the other edge is a much more valid test. I feel it is likely he had an intellectual opinion, and sees what he wants to see to support it after the fact. Often opinion colors results, which is why undirected research is usually the best science.

   Edited 12/6/2008 6:09 pm ET by thumbnailed

2. 
   SteveSchoene | Dec 07, 2008 06:40am | #16

   I would bet that the original research, upon which the watered down summary article was based makes a number of the reasons for the choices. 

   Another thing to note is that the research was financed by furnitue industry sources, so they were looking for practical applications.  By the way, academic research seldom works the way you suppose.  There is likely to be a body of existing research that the author was working to confirm, or more likely, to disprove.  I am not willing to buy the original articles--though if you have access to an academic library in Michigan or North Carolina I'd bet the Journals in which such research is published might be available. 

   But the point is  that "life experience" proves nothing about the optimal pressures for strength, only that lower pressures don't fall apart.  That doesn't even address the issue.  What the science does prove is that higher pressures are not bad, and unless you have actually used the high pressures on a number of occasions, there is nothing that can be said that refutes the thesis of the research, which was originally published in peer reviewed journals if I recall correctly, where people who both understand the prior art, and have the capabilities of accessing the statistical analysis and full experimental design. 

   Edited 12/6/2008 10:45 pm ET by SteveSchoene

3. 
   thumbnailed | Dec 07, 2008 07:46am | #17

   I grant you good intent and intellectual integrity, but when you start out with " I would bet" then you're simply accepting what you've read without question. And, no, I don't have an idealised view of how modern research is conducted, but I contrast its general success and applicability when compared with non-directed, open ended classical research, where the intent was to discover what was true, not prove what you already thought was true. The author has stated that his base findings are scientifically derived, and that might very well be true, but just because it is in print doesn't make it true. I saw no footnotes or references listed with the article, perhaps I missed them, I'll go look again, but without them its just opinion, and so is my viewpoint. I have no references or footnotes, no scientific background, but a good deal of "visceral engineering", and I say with my experience, "what possible advantage can you demonstrate in clamping wood to near its crush point ?" Demonstrate where that is any better or even as good as firm even clamping. Show measured force tests that demonstate higher resistance to joint failure. This is diverging somewhat from the glue starvation issue per se, but the disagreement stems from the validity of the authors basic premise. Sometime I'd like to buy you a beer and discuss this more fully.

   Edited 12/6/2008 11:49 pm ET by thumbnailed

4. 
   SteveSchoene | Dec 07, 2008 04:06pm | #18

   You didn't see the article.  There is another article, at least, in a industry research journal, that I couldn't read, because the publisher charged to download the pdf.  But I don't quite understand your quibble.  The point is that the crush point is obvious the maximum clamping pressure that would be acceptable.  If "starving the joint" by excessive squeeze out were true, it would have to be happening at that maximum pressure.  But it wasn't.  Practically, it means that manufacturers don't have to design clamping systems to reach some precise modulated force, but just that they don't crush the wood. 

   You said "...non-directed, open ended classical research, where the intent was to discover what was true, not prove what you already thought was true."  I don't understand this.  Scientists have to have a hypothesis to test.  A successful experimental test doesn't prove anything to be true, it only proves it to be not false under the condidtions of the experiment.  Every experiment must have the possibility of not working--the hypothesis must be falsifiable. 

5. 
   thumbnailed | Dec 07, 2008 06:12pm | #19

   Let's agree that we disagree on the approach to determining proper clamping pressure. The description wasn't 'maximum', but 'optimum'. I'm not going to do research to clarify it, because it will make no difference to how I do my work, though it might make me a more thorough or consistent person.

   On the scientific method issue, of course an hypothesis is necessary to be validated or not, but that differs greatly from a natural human tendancy to find proof for a pre-existing belief and disregarding contrary evidence. Just an example, Percival Lowell, not a deluded fool but a well respected astronomer, built the then largest telescope in the world, and proceeded to discover a wide network of canals accross the surface of Mars. So far we have been unable to find them since. Unbiased science is incredibley hard to design, let alone perform. Phrenologists 'proved' the inferiority of ethnic people with 'scientific' skull measurements. Legal scholars defined the value of black humans as a percentage of a white man's. Einstein's professors told him to stay away from physics , because the underlying nature of the universe was fully understood by scientists. Science and general understanding are always colored by personal biases and cultural values, and sometimes corrupted by them. Skepticism is always a good first defence against  possibly unwarrented conclusions. "Believe a man who says he is searching for the truth, disbelieve a man who claims he has found it".

3. 
   Quickstep | Dec 05, 2008 04:46am | #6

   I did some experimentation because I was a "slob" gluer. I also had the same problem you describe. In my experiments, I spread glue about as thinly as I possibly could on both surfaces to be glued. I discovered that even when I spread the glue very thin, I still got some squeeze out and in ideal situations the little bubbles that Thumbnail described. 

   I've also used blue tape right at the glue line.  I put the tape on so it overhangs the edge and then carefully trim it back with a razor blade. It works, but it's a bit of a hassle. Also if the glue drips past the area that's masked, you're right back where you started.

4. 
   JeffHeath | Dec 05, 2008 05:23am | #7

   Citation

   One of the biggest mistakes made on all DIY shows demonstrating woodworking is the amount of glue slobbered on.  I see guys like Norm Abrams really load up a glue joint, and have 2/3's of the bottle squeezing out of a joint.  YIKES!!

   The very best way to fix the problem you are having, forever more, is to avoid it in the first place.  With todays modern glues, you only need to 'wet' the long grain wood surfaces with the glue.  If your joints are properly cut and sized, and they mate properly,  just barely 'wetting' the long grain surfaces of the mating pieces will result in a joint, when dry, that is stronger than the wood itself.  You should put the glue on so thin that you can see the wood grain right through it.  Wet looking, and that's it.

   And, with full respect to Rob A in this thread, you most certainly can over clamp, and squeeze the glue completely out of a joint, resulting in a failure.  A good rule to follow when clamping is to clamp just enough to have the joint close.  That's it.  Proper joinery is required, of course.  If you are glueing up a tabletop, and trying to squeeze two boards together that are off by 1/4", you are asking for problems.

   Good luck.

   Jeff

5. 
   photonut | Dec 05, 2008 10:01pm | #9

   Since this is in the finishing section, not the joinery section, I will try to comment mostly about the finishing aspect.

   1.) Like others have said, use less glue. Squeeze out just isn't necessary.

   2.) Use Titebond II Fluorescent wood Glue. It fluoresces under black light. Buy a cheap black light at a home center, and examine your project before you apply stain. Scrape or sand it to remove it.

   3.) Just my opinion - Don't use stain. Use wood that is the color you like and finish with tung oil, danish oil, or poly. A little stray glue isn't as likely to show through a natural finish. I hate stain, and I also hate water based finishes, but that's just my opinion. The only kind of stain that I use at all is danish oil with stain in it.

6. 
   BG | Dec 06, 2008 02:37pm | #12

   Citation,

   You did say all suggestions, here is one that I experienced a couple of months ago.

   I wanted to experiment with French polishing and did so with a rather large piece of mahogany. Planed, scraped and sanded the wood perfectly flat before applying the finish protocol. It came out very nice.

   A few weeks later, being cheap, I recycled the mahogany into a small project...forming mortice and tenon joints from the polished mahogany. I was very impressed, not only was glue not a problem but the joinery was exceptional, perhaps because the stock was well perpared.

   Anyhow, under the 'all suggestuons' catagory.

1. 
   Citation | Dec 06, 2008 04:16pm | #13

   Sounds like you had it right.   Thanks.

7. 
   bones | Dec 08, 2008 07:59pm | #26

   Don't feel bad i've sanded thinking I've had it all till I stain and blink there it is.  I've used in the past a glue that had an addative that glows under a black light.  I used that on a project for my nieces wedding.  It was a cedar chest that had the box glued to the base.  I was really concerned about glue squeeze out where the two surfaces met at a 90 angle.  Maily because if I did not get all the glue out sanding would be across grain on one of the surfaces no matter which way you went.   It worked great!  I have pets from time to time have used the product urine-b-gone and it comes with a little portable battery florecenet black ligt.  I hit the joit and wiped away with wet sponge and not an issue.  It's great suff.  I only use it on tough locations, otherwise I simply flood the area with a soaked sponge and get the glue off while its clamped. Good luck. 

   Government's view of the economy could be summed up in a few short phrases:
   If it moves, tax it. If it keeps moving, regulate it.
   And if it stops moving, subsidize it.