There’s a discussion going on over on a boatbuilding board regarding laminating dimensional lumber for making masts, etc. One gentleman has made a comment that “it’s basic woodworking 101” to use an odd number of layers when laminating wood, in order to minimize warping, cupping, and twist. He says that he was taught this principle in early woodworking and that it’s been in every book he’s read on laminating boards.
Granted, my woodworking experience is mostly self-taught or learned from my father, but I’m also an avid reader (long-time FHB and FW subscriber), and I’ve never run across this principle. For the life of me, I can’t come up with anything that satisfies my engineer’s brain as to why an odd number of layers/plies is more stable than an even number. I can see that more is better, sure, but is 3 better than 4?
Most of what we’re talking about is 5 layers or less, not like 10+, and also not using veneers. I’d appreciate any input from the group out there.
-wte
Replies
when you look at how plywood is laminated, you will always see the two outside faces with their grain running parallel to one and other, and you will also see that the grain in plywood is oriented perpendicular to the ply beneath it. if you were to have an even number of plys, you would not be able to have the grain on the two outside plys running parallel, and you would create an unbalanced panel, as the forces of wood movement would be unbalanced. if you want proof, try veneering only one side of a piece of plywood and leave it for a while, and you'll see what can happen to an unbalanced panel.
Andrew - thank you for the quick reply!
I fully agree that plywood should have the grain on both outer faces running the same direction for that very reason. This is a somewhat different situation, though, because the grain direction is varied from layer to layer.
The backyard boatbuilding community is generally more involved in laminations where the grain is all running the same direction, as on masts, booms, tillers, curved cabin roof beams, etc. Is there ANY reason you can think of that would make an odd-numbered lamination more stable than an even-numbered one?
Thanks!
I've also heard of it, but again, only in the context of plywood.
Generally, I have to agree that balanced construction...odd number of laminations....is always the ideal. I know there is one type of a construction grade plywood out there that has six laminations, but if you check the definition of plywood that comes from the standards associations, it has to have an odd number of plies, so I've never figured out why this product is out there.
Anyway, in solid wood, glulams have a balanced construction (odd number of laminations), except when a curve is being built, and a tension/compression situation is being specifically engineered (according to the Wood Handbook: Wood as an Engineering Material I think is the titile....available on line at http://www.fpl.fs.fed.us/pubs.htm (Chap. 11 is on laminating) Every description of bricklaying curved furniture parts I've ever seen note that odd numbers of layers have to be used to equalise stresses. That's two examples off the top of my head. I'm not an engineer, but I am a woodworker, and I've seen the principle of balanced construction applied in many different parts of the industry, in many ways....I would only depart from using balanced construction as a first principle very warily.cabinetmaker/college instructor. Cape Breton, N.S
Adrian -
What a great publication! Lots of meat to chew on there! What I found in it on "balanced construction" never mentioned (that I could see) an odd number of laminations. Earlier in the document, I believe it even referred to laminated lumber as "two or more...."
What this document described as balanced construction of a laminated beam (glulam) is that both edges have the same (high) grade of wood on the outer layers, in order to carry tensile loads. If a beam is being supported only at the ends and has an "up" side and a "down" side, only the "down" side (which carries the tensile load) needs the higher grade of wood. The interior plies of the beam usually have lower-grade wood, and the side of the beam in compression can also be of lower grade (like the center plies).
If a glulam is supported and the ends AND at points in the middle, then both the top and bottom will carry tensile and compressive loads at different points, and both need the higher grade material on the edges.
If you've seen anything in there about odd numbers of layers, please point it out to me - I did read it pretty hastily this afternoon.
Thanks again ---
Under bending stress, a mast with an odd number of layers (one in the middle) might stand a slightly less chance of breaking. When you bend anything, the outer fibers/molecules are stretched (tensioned) and the inner ones are compacted (compressed). The center fibers stay nuetral. So if you laminate an even number of layers, with no center layer, then all layers will be under some stress (compressive or tensile) when the mast bends. But when an odd number of layers is used, the single center layer remains unstressed, and might be the only layer keeping the mast from breaking under severe loads.
Of course the wind and force on the mast may come from any direction, so if the mast must act like a single homoginized mass, responding the same from all directions, the gain from odd versus even layers is probably irrelivant.
Dave B.
Edited 6/25/2002 12:02:30 PM ET by 4DTHINKER
4DTHINKER -
You make the best point I think I've heard yet on this topic, but I'm not quite yet convinced. One thing I'd offer in rebuttal is that with most glues we'd be using, the glue joint is going to be stronger than the wood itself, so I don't see a problem having a glue joint in the middle instead of that middle, odd-numbered ply. Besides, most of the load is carried in the outer 10-15%(?) of the thickness of the beam, and the exact middle is neutral - neither compressive nor tensile. That's why you can saw an 8" hole in the middle of a 12" wooden I-joist. The outer 2" (top & bottom) carry the real load.
As you mention, this only matters if the wind is blowing against the face of your laminations - perpendicular to the plane of the plies. Hmmm.... the search continues.
PS - The claim, also, is that it simply prevents warping, cupping and (in particular), twist in a way that an even number cannot prevent as well.
WTE,
The reason that your engineer's brain can't find a reason that an odd number of plys is better than an even number is that you are applying logical thinking, not yet contaminated by the "accepted wisdom" that an odd number is better. There is no real justification for an odd number.
The (incorrect) justification goes something like this:
Starting with a single sheet of material (wood), laminating another sheet (usually a thin veneer) to one surface causes warping of the pair. The warping is the result of many stresses introduced by the lamination - usually moisture from the glue. If a contact cement is used, no warping occurs.
The warping is the result of penetration of moisture into the thicker sheet causing an imbalance of moisture across the thickness, swelling of the higher moisture content tissue and the inevitable warp away from the wet wood. Such a sheet is "unbalanced" for many reasons. If the glue is flexible enough, the sandwich will come back to near-straight as the moisture distributes itself through the entire thickness. However, most glues have set very tight by that time, and the sandwich stays warped, unable to completely return to the initial state.
The solution is to simultaneously glue a second veneer, similar to the other veneer, using the same glue, to the other side of the board, resulting in a "balanced" sheet. The warping stresses counteract each other and the 3-ply sandwich remains straight. The fact that the veneers are glued with their grains at right angles to the middle sheet have nothing to do with the story.
The conclusion, of course, is that since 3 ply construction corrects the above situation, that odd-ply construction is always needed.
Au contraire mon frere!
If two sheets of similar thickness and composition are glued together, the moisture-induced stresses will still be balanced and the sandwich will not move. Subsequent plys can be added, 2 at a time (one to each outside surface, of course) and the multiple, even- plied structure will not warp.
I believe that you can find plenty Baltic Birch plywood with an even number of plys.
Rich
Edited 6/25/2002 6:21:59 PM ET by Rich Rose
Rich,
That's sort of what I hope to discover out of this little quest of mine.
I'm not about to make such a declarative statment myself yet, but I haven't gotten anything back yet that really explains *why*; just that, like you say, it's the conventional wisdom.
It almost reminds me of the story of the newlywed couple in the kitchen. The wife is making her hubby a pot roast and cuts the end off of the roast before placing it the pan. The husband asks why, and she doesn't know. Her mother taught her to do it that way.
Long story short... they finally get in touch with grandma, who also cut the end off her roast. She replies, "Oh, honey - it's just that my pan was too small for a whole roast!" :-)
-wte
Ahh, the gentleman is correct, WTE, and so are your instincts that an even number would be stronger than an odd number. The point you're missing is even number of what? The correct answer is glue lines, not laminations. To achieve an even number of glue lines it takes an odd number of laminations.
The strength in a same grain lamination comes from the sheer strength of the glue line, the inability of the two joined pieces to slide when glued. The limit of the strength of the whole is the sheer strength of the wood used and the sheer strength of the glue line. No doubt your gentleman also suggested a rigid glue because of (knowingly or not) increased sheer strength.
Let's make a strong mini-mast. If I glue two pieces together and flex the resulting product where are the stresses placed? In the glue line? Nope, on the wood itself and the stress is tensile and compressive, not sheer, and that's not the strongest place. To get the sheer strength of the glue involved you need two glue lines separated by the wood. Flex a piece made with three laminations...stronger than the same thickness made from two? Yup... Why? Because now the sheer strength of the glue is where the stress is placed.
Simply build out from there and you see why the magic number is even, not odd...why trusses generally have two chords...why torsion boxes have two glue lines...why incredibly strong airplane wings have a skin on either side attached to a rib ...all employ sheer strength which is considerably stronger than tensile strength.
Clem
Well, as long as I'm spewing drivel...
A mast, a bow, a ski, or any same grain lamination, and plywoods are designed to meet very different requirements. A mast needs strength and plywood needs stability. Plywood is very wimpy compared to a torsion box or same grain lamination but a mast is wildy unstable by the standards we place on plywood
So, why the crossgrain lams in plywood? Because it exploits the wonderful stability of long grain. Wood is really weak across the grain and if I slice it thin enough cross grain laminations will not have the force to compress or lengthen long grain, the cross grain fibers will either compress or spread. Cross grain lams will have adequate strength to bend the long grain but not enough to actually shorten or lengthen it. So, by making plywood as it's made the sheet moves as long grain in both directions, length and width, and the necessity of balance is paramount and balance is achieved through even numbers of glue lines. This stability enables us to glue a plywood piece to a solid long grain piece (such as a plywood panel in solid wood frame) and since both move quite the same we need'nt worry about failure.
Rich Rose, yours is one of the most misguided pieces of logic I've ever heard to explain this stuff.
Clem
"Rich Rose, yours is one of the most misguided pieces of logic I've ever heard to explain this stuff."
Gee, Clem. How YOU doin?
No niceties of conversation here, boy. No, "I disagree with you sir for the following reasons . . . Here are my observations and conclusion, would you be so kind as to comment?"
Nope! Just come right out and insult someone with whom you find disagreement. Did I do something to offend you or are you always this pleasant?
Don't understand this stuff, huh? Then hang on, maybe you'll just learn something . . . or not.
When fact and theory don't agree, you throw out the theory. Have you ever laminated the kinds of structures you've described? Actually put the layers together under pressure?
I've worked with polyester resins, glues and other materials. Under pressure with correct control of the geometry (keeping the pieces flat at all points of the sandwhich and assuring equal penetration of the "glue"), it doesn't matter whether there are an even or odd number of laminations, the structure will not cup. That's experience, Clem.
So much for that behavior of plywood.
Agreed on the principle of controlling movement in cross-laminated plywood. (Actually such behavior was not a point of discussion in the original question, but that's ok, you get extra-credit points.)
As to the flexing behavior of a laminated mast. You've only got it partly right and you're failing to completely vizualize the very structure that you've described.
A two ply mast has little resistance to flexing from a force applied to the "face" of the structure. But then the two-ply is simply plywood, with all its weaknesses in such an application. The structure does not benefit from the resistance to shear (and it's SHEAR, Clem, not SHEER) as you've pointed out because the wood in compression and tension is not the equal of the glue line in shear. However, once we add additional laminations to that original two ply, one glue line "core," we have exactly the same ultimate structure with an even number of shear-opposing glue lines stepping out from the single glue line at the center.
We have an even number of laminations and an even number of glue lines where it matters. Once again, it makes no difference whether there are an even or odd number of laminations.
The ball's in your court, son.
Rich
Rich, I'm not going to patronize you and I'm not here to play tennis.
If you feel that after centuries of practice we have it wrong and that you're the first to actually consider an alternative I suggest you continue doing precisely what you are doing. Be prepared to hear stuff like ""Rich Rose, yours is one of the most misguided pieces of logic I've ever heard to explain this stuff." Oops, I said it again.
Don't sweat it, many genius' have had to endure the same.
And yes, thank-you...it is shear.
Clem
WTE;
The info in the Wood Handbook is Chap 11, page 4 (under Bending Members); it specifically mentions a core piece, then building out from each face to create a balanced construction....the exception, when unbalanced construction is used, is when additional laminations are applied to the tension side. Doesn't apply to a mast, I don't think, if that is the application you have in mind. The appropriate standard is ANSI/AITC A190.1, but I'm not going to buy it for a debate on the internet.
As far as Rich Roses's comment that "The warping is a result of many stresses introduced by the lamination-usually moisture from the glue. If a contact cement is used, no warping occurs." That's not correct, unbalanced construction will warp, and the standards that apply where contact cement might be used (AWI and AWMAC, the architectural millwork orgs), specifically require balanced construction in all sheetgoods and countertops. For solid wood,if you face glue two pieces of identical dimension, you'll probably be okay, but as you add laminations, you should be trying for a balanced construction.
Also, I've never seen and am not aware of any Baltic birch ply that is made with an even number of plies. It may be available, but I've never seen it, not in North America or in Europe (including at the Interzum show in Germany last year, where there was a lot on display). You occassionaly see references to a part made with 26 ply BB, but that's just people gluing two pieces of 13-ply together. Do a quick internet search; see how much even ply plywood shows up.
Personally, I'm sticking with the established standards as my guide, and when I break the rules, it's with the knowledge that that's what I'm doing.cabinetmaker/college instructor. Cape Breton, N.S
Adrian - I'll have to go back and look at that reference again tonight. Thanks for checking.
Part of my question here would have to be with what constitutes "balanced construction." Yesterday's reading of that document - I thought - indicated that it means having the same thing on both sides of the centerline (simply put). Unbalanced construction would have something different on one side compared to the other, like the additional laminations on the tension side of the beam.
So, does it really matter that much what is *ON* the center line, be it a glue line or the middle of that odd-numbered layer? If that's the zero-compression/zero-tension line, WHY does it matter?
===
For what it's worth regarding plywood, here's a quote from the boatbuilding message board from a different gentleman in Indiana. If he's correct, there's apparently a good bit of even-layered plywood out there besides 13-layer BB. Still, my main concern is with same-grain laminations.
<<<BEGIN QUOTE>>>
Okay, I went out to the garage and checked. I'm using BCX plywood that I bought at Lowes.
The 1/4" and 3/8" plywood sheets are 3 plies. The center ply is twice as thick as either of the outside plies, so it balances.
My 7/16" and 1/2" sheets have 4 plies. The two inside plies run the same direction, perpendicular to the two outer plies.
My 5/8" and 3/4" sheets have 5 plies. Three plies run one direction and two run perpendicular.
Some plywood has more plies than these. At any rate, they do make plywood with an even number of plies.
<<<END QUOTE>>>
I'm sitting here looking at a chair in my living room. It was made in Denmark. It is entirely of laminated oak. The arms have 12 laminations. The legs have 11, the back has 14.
It's so easy to discuss these kind of things with reality on one's side.
Rich
Well, that certainly confirms what I have noticed about cheap mass produced Scandanavian furniture. Clearly, Rich, if you're going to hold up a contemporary piece of cheap mass produced furniture as your pinnacle of engineering, longevity and aesthetics I'm poorly equipped to continue.
Rich, you're a jackal with little understanding of the subject you're discussing.
Gawd, this forum has deteriorated.
Clem
The number of laminations and the number of glue lines is totally irrelevant in a glulam beam. The number of laminations in a glulam beam range from 2 up to 50+, including both the even and the odd numbers. The beam is selected based on the depth and width required.
As for those of you who stated that a glue line at the neutral axis has no stress, this is only true at the center of the span. At the beam ends the horizontal shear at the neutral axis is a maximum.
Some references that may clarify the issue are the National Design Specification for Wood Construction ("Structural Glued Laminated Timber"), and the American Institute for Timber Construction ("Timber Construction Manual").
Clem, its nice to meet you. I tried to tell them before but I didn't really hit it off to well either. It seems logical that the less glue lines you have, the stronger your material will be. In my experience thow, for most projects, I go for the 7-ply every time for stability. Poplar cores are seemingly the best and very user friendly. The de-laminating of plys happens a lot more frequent but the end result is sufficient. Hope to see you around more. ~RFW~
About your chair.....what part of what I wrote are you having trouble with? I stated that, according to the Wood Handbook, you can have an unbalanced construction, and that (to use a beam as an example), extra laminations are applied to the tension side.....there are examples all over the place, if you want to look, but hopefully thay are in fact engineered to meet a particular requirement, and you have to be able to predict the anticipated loads (or if something is going to used in a particular way, like construction grade sheetgoods that are going to be fastened down, that gets factored in.) In the case of a chair, that's pretty easy. In the case of a mast, if in fact that's what we are talking about, I don't know if it is easily predictable, but I would assume marine architects have some sort of reference standard to go by. I don't doubt that the Danish designers took all that stuff into account, they're not just going by a gut feeling.
Now, you came in to this discussion all full of certainties, implying you have the whole truth....in spite of the fact that your opinion seems to be in contradiction to the results of decades of research by scientists in research labs and codified in the standards used by the manufacturers of products like this. I'm no research scientist, but my little bit of practical experience jibes with 'the rules'. You got anything to back up your assertion that the conventional thinking is all wet? Anything? Or are you just 'sure', deep down, that all that stuff is wrong?
I'm up to my ears in reality , by the way...cabinetmaker/college instructor. Cape Breton, N.S
Ahh, but Clem, I never said that I thought an even number of laminations was better than an odd number. I said that, as far as I can tell, more is better.
I congratulate you on bringing up a new (and presumably valid) characteristic - that the glue line is stronger in shear than the wood itself (with the right glue). But let's look at the three-layer mini-mast you propose. Yep, that third layer has pushed the glue lines from the neutral-stress center out to where they can do some good.
But, whoops! Somebody sneaks in and adds a fourth layer. What has this done to the picture? It's just added a third glue line in the middle where it does no additional good, right? Wrong. It has also pushed the other two glue lines farther to the outside where they can do even more to handle the shear stresses.
By your own logic, it sounds like four layers are better than three, and that, yes, more is better. Nothing I read in your post would begin to indicate that an odd number of layers is inherently better than an even number.
As for your extrapolation in the last paragraph, it seems that none of it really has to do with justifying an odd number of layers in a same-grain lamination. If you'd care to elaborate more on any of those examples, I'll keep reading.
Thanks,
-wte
(...and no, the gentleman never suggested a rigid glue, FWIW.)
WTE,
No, you never asserted an even number of laminations is better, it's not. I said, and I'll repeat since you missed it that the objective is not lamintations, it's glue lines...count the glue lines...the strength is in the glue lines...laminations are secondary...glue lines...glue lines, glue lines.
Now, from your own post..."But let's look at the three-layer mini- mast you propose. Yep, that third layer has pushed the glue lines from the neutral-stress center out to where they can do some good."
Now, you've acknowledged that three lams and two glue lines are better and stronger than two lams with one glue line and I'll assume you will accept the notion that balanced construction is better than unbalanced construction and that balanced construction would entail adding a lam and glue line to both sides.
Unless you abandon the notions that two glue lines (3 lams) are stronger than one (2 lams) or that balanced construction is better than unbalanced construction you have to be lead to the reason why odd numbers of laminations are used. To be lead to the conclusion that even lams are fine you'd have to abandon one notion or the other or both. Do you?
So, what's the problem?? Are you simply debating for the sake of debating or are you looking for the actual reason...the reason is well stated.
I've never seen even laminations in plywood with the exception of certain bending plywoods which are designed to be inherently wacky. Please direct me to a product not intended for bending that has an even number of plys.
Clem
Clem,
First, I presented the original topic in terms of number of laminations, not number of glue lines. Unless you're somehow laminating without actually gluing the laminations together, I don't think any person would argue that the number of glue lines...glue lines... glue lines = number of laminations minus one. Try the math - it works! :-) Bottom line, it's semantics, and it seems to be a red herring purely for the sake of argument. With that issue hopefully set aside, let's move on.
The place where your argument falls apart is exactly where you left off quoting my previous message. Compare a 4-layer mast with a 3-layer mast: which one has glue lines farther from the center where they can do more good? You answer; I don't want to presume. Granted, the 4-layer mast has an *extra* glue line which doesn't add much, but that's beside the point.
You also state that "balanced construction would entail adding a lam and glue line to both sides" from this odd-layered starting point. Do you assert that this holds true for laminations where all lams are glued together at the same time?
I was taught early on that you can prove almost anything if you start with a false assumption. I don't have to abandon either notion that 3 lams are better than 2, or that "balanced construction" is better than "unbalanced" based on anything I've said. What I have trouble with is your starting point. Your assumption is that you have to start a laminated object with a layer in the middle, after which you add equal layers on each side. This progression (1+2+2+...) will always yield an odd number of lams, so obviously, odd numbers are better, right?
Question: can a person not start by gluing two lams to each other, then add two, two more, etc.? This gives a progression of (2+2+2+...) which always gives an even number. One question that I just put to Adrian is whether this, also, constitutes "balanced construction." Your opinion is as welcome here as anyone else's on that matter.
Given a bunch of children of equal weight and a seesaw, I can put the same number of kids on either side, or I can put one kid over the fulcrum and any equal numbers on either side. It's balanced either way. Does the same principle apply in laminating wood? If not, WHY NOT?
I'm really NOT debating just for the fun of it, but I also don't see that the reason is at all well stated or, necessarily, well-founded.
O-kay, this is my last stab at this...
WTE, I stressed the glue lines because that's where the strength come from, not the wood. Shear strength at a glue line is far greater than within the wood itself. Glue lines, glue lines, glue lines
I'll use the terms you're more comfortable with...
Three laminations are stronger than two. You seemed to accept this but you turn your back on it when you go on about the 2+2+2 stuff. If you don't accept this there's no point in my continuing and I'd suggest a high school physics book to you for more information. If you accept that three are stronger than two and if you accept the concept of balanced construction you have to accept these two concepts as the reason for odd numbered laminations.
Your idea of 2+2+2 ignores the concept that three laminations are stronger than two. The strength of a laminated structure begins at the core and the strongest core is made with three lams then adding balanced lams ad infinitum. Three lams transfers the stress to the glue line, two does not. Five lams transfers the stress to the glue lines, four does not, etc., etc, etc. Accept this or you'll never get it. Yes, your 2+2+2 is balanced but is it as strong as boatbuilders want their masts to be?? No, no, no, it's inferior and if one is concerned about building the best of whatever why would you even consider an inferior structure. Why? Beacuse many applications are done for economy, building a curved part by the thousands is cheaper done by lams than sawing solid, less waste. Mast builders want strength and that was the point of your query.
I have neither the time nor the inclination to find an engineering source to define this further for you but if you really need it I'd suggest you purchase Timber Construction Manual. Go out to your shop and make a two lam ply, observe it's lousy stability, break it, make a three lam ply, observe it's stability, break it, observe the difference in strength. Then, feeling secure that you've figured out which is stronger apply the concept of balanced construction to make your mast however big you want to. Adding lams will not negate the forces that make three lams stronger than two.
Now, there are instances where liberties can be taken...glue lams, in which the lams are generally stressed with a crown or stressed with the application, chair parts that are restrained, too short to sweat or bent so they are prestressed and others.
The original query was concerning masts. Built purely for strength odd numbers are better.
Some will get it and some won't...
Clem
Some will get it and some won't...
At least we agree on something. I don't think you got what I was saying. You certainly didn't even make a stab at answering the direct questions I asked which might call your position into question.
I haven't yet said that an odd-layered lamination is not stronger than an even-layered one, but the only justifications you give are stated as assumptions with nothing backing them up. I'm not looking for declarations from you (like "it's inferior") that I have to accept on faith.
I'll try to state this very simply. Please stick to the topic if you choose to reply even after your "last stab."
I ask again: are four layers stronger than three? (see earlier post) I've shown that the glue lines on a 4-lam are closer to the sides than on a 3-lam. If it's not stronger, why not?
You agree that an even-layered (2+2+2) beam is balanced in construction. Do you agree that two layers are stronger than one?
If "the strength of a laminated structure begins at the core," then wouldn't the strongest core be a glue line, not a middle lam?
===
One of us has a big mental block about this. It might be me, but if so, I'd like to get past it, and I'm trying to come at it with an open mind.
Maybe you're right, but maybe you've forgotten or never really understood the principles behind the (supposed) facts. Be that as it may, arrogance and condescension are not highly effective tools for convincing me.
-wte
Edited 6/26/2002 2:53:02 PM ET by wte
WTE;
You asked me about the 2+2+2 thing....I can't give you a definitive answer: I'm not an expert, and I simply don't know precisely how it relates to the 1 + 2 +2. My strong feeling, based on everything I've ever seen or read, is the latter is stronger....by how much a margin, I don't know, and whether that margin is critical or not, again, depends on the situation. When I'm not an expert on something, I try to find the best reource I can....in this case, I think the Wood Handbook is very solid, and it states that balanced construction....with a single member at the core, not a glue line.....is the strongest situation. Unless you know where the load is coming from, and you want to beef it up on the tension side, as noted before. The reference standards Ben Malamed provided are or use the same information as the ANSI/AITC standard I provided earlier, and I'm assuming the Wood Handbook and the research centres that put that together provided the data that the ANSI/AITC standards are based on. As to whether a particular glulam has an odd or even number of plies, they are bought by the dimension, but they're engineered according to the principles we're discussing....as long as they are the right dimension and can carry the desired load including the desired safety margin, the actual number of laminations may not be that important; lots of variables to play with there. I agree with Clem, though, a mast is a different story.
Why don't you forward your question to the Forest Products Laboratory, or a school for naval architecture, or the AITC? If this question is that important for you, go to the source.cabinetmaker/college instructor. Cape Breton, N.S
Well guys, it's really not that important to me, or at least it wasn't. Now my curiosity is piqued, and I may have to take the question to a lab such as the FPL, like you suggest. Maybe a trip to the local library would yield something. I thought I'd give this board a shot.
To summarize, here's what I've learned or confirmed:
The term "Balanced Construction" is thrown around a whole lot. I need to get an authoritative definition on what that is.
Ben Malamed reminded me that the center line on a beam is only zero-shear at the center. The ends are at maximum shear.
Glue lines are stronger than the wood itself.
Two lams are (probably) stronger than one (no one refuted this)
Three lams are stronger than two (I never refuted this)
Four may or may not be stronger than three.
Given those last three or four items, and having no prior "odd is better" bias, I would tend to extrapolate, saying "more is better." Clearly, lots of people have been taught that "odd is better" even if the underlying "whys" or "hows" are long forgotten.
Clem's need for "why" is satisfied with the answer that, somehow, an odd number of lams pushes stresses to the glue lines. I'm chasing the goal of explaining the "somehow."
I'd certainly agree with Clem and Adrian both that a mast is a different animal. It is often pre-tensioned in some direction, has supports in multiple places and carries loads from many directions. Plywood is yet another animal, because of cross-grain laminations.
In case anyone is still interested, I'll post my findings at a later date. Further comments are still welcome here, though!
-wte
Edited 6/26/2002 4:18:13 PM ET by wte
Your question...
"I ask again: are four layers stronger than three? (see earlier post) I've shown that the glue lines on a 4-lam are closer to the sides than on a 3-lam. If it's not stronger, why not?"
I can't give you an absolute answer but if the overall size of the pieces are identical and the only fifference is the number of glue joints I'd bet my dough on the three lam piece. Why? Because the stresses are pushed to the glue lines, not the wood.
I'm with Adrian, go to a laboratory that specializes in this sort of research.
Clem
This forum post is now archived. Commenting has been disabled