We value your privacy.
We use cookies to enhance your browsing experience, analyse site traffic, and personalise content.
By clicking “Accept all”, you consent to our use of cookies. You can also choose
“Essential only” or learn more.
Re: Can't use PVA glues for load bearing applicati
Hugh Evans
Load bearing, in a structural context, refers to beams that will be subjected to continuous stresses of potentially several tons. They are however very well suited to furniture making. Just don't try to frame a house using PVA instead of mechanical fasteners and you'll be fine.
Re: Can't use PVA glues for load bearing applicati
Barry Irby
Wilbur, I had to pause at that statement as well, but I thought it through. Creep is a movement of deformation of the material (glue) caused by a load or force exerted on it over a period of time. In this case, a rather heavy load of a long period of time. Think of taffy, if you warm it and pull it very slowly, you can stretch it a long ways, if you pull it quickly, it will snap, in effect, resisting the sudden pull until you reached its (pretty low) breaking strength. A beam or structural component might have a fairly high load (dead load) that was constant nearly forever and then some monetary loads (live loads) that were short lived. So, if you glued up a beam with PVA glue and use it to roll a wheel barrow of concrete across it would hold fine, but if you piled bricks on it and left them there, it might creep and sag.
Most furniture has very little if any dead or long term load. A chair sitting empty has very little strain on its joints. Neither does a table. The glue joint between to boards in a top should have very little strain on it. So, if you sit in the chair for an hour or two or pile the Christmas feast on the table, those loads are not long term in terms of Creep. I suppose it you leaned something really heavy against the end of a table and left it there, it might be fine at first but fail after a few months or years.
On the other hand, a bent lamination does apply loading to the glue lines that is long term. So, if you make a severe powerful bend and leave it, the joints might slip or creep over time. (My son made my X a table with bent laminated arched legs. Wonder how bad the explosion will be when they turn loose? )
If a phenomena is glue moving, or glue allowing something to move, in response to stress, you can justifiably call the phenomena "creep". Movement in response to stress is the requirement. But, I am bewildered as to what you are observing. What may have moved and what generated the stress, if either happened?
I did a joint survey about the house on a couple dozen PVA joints and I don't feel any of them. I don't expect to be able to feel a joint in a panel. So I am at a loss to understand the observation.
If one glues a higher moisture content board next to a lower moisture content board with PVA glue, after moisture equilibration a step at the joint will be observed( according to the literature I have read). One could argue that creep in the cured glue enabled the one board to shrink past the other. The stress was provided by differential shrinkage. Depending on one's perspective one could argue that the step is a short coming of the glue(creep), or a short coming of the cabinet maker , or a short coming of lumber supplier.
Will hot hide glue prevent this joint's tactile defect at the expense of capturing the resulting stress of the shrinking board in the joint? If so , and the locked in stress at the glue line doesn't cause some problem in the future, then one could say that hot hide is superior in this situation.
Creep, from the perspective of material science, is quantifiable. I have no doubt that Franklin has creep data on every wood glue commonly used. It would be of interest to know the creep values of PVA vs hide. Again, one could easily get the data in a home shop experiment.
As I mentioned before, a glue ridge is indicative of less than optimal bond. It is unlikely that this could be replicated with hot hide glue as the joint would probably fail outright.
Regarding empirical data on creep, we do occassionally perform ASTM D3535 "Resistance to Deformation Under Static Loading for Structural Wood Laminating Adhesives Used Under Exterior (Wet Use) Exposure Conditions." Due to intellectual property restrictions I can't post the procedure, but if there's interest I could post a picture of the test apparatus. Our test data on creep resistance from highest to lowest is: Hide Glue, Titebond Extend, Titebond Original, Titebond II, and Titebond III. Aside from difficulty of use, limited open time, and poor water resistance hot hide glue has fantastic properties.
Re: Can't use PVA glues for load bearing applicati
FredD
Hugh ,
thank you for taking the time to explain a little understood subject.
Also for explaining your take on structural stress.
I started using Franklin hide glue in 1952 in shop class when your company called it Franklins Horsehide Glue in the cold form.
Used your PVA's since you developed them.
As a retired builder In California I was always aware that Glu-Lams did not use Pva glue in assembly, but having been in stair lofts all over Northern California and one actually doing business in the building I had a Cabinet shop in, I observed many times Curved Stair horses being laminated with PVA glue and even in the field if there was a existing wall to clamp the curve to.
Never saw a failure or heard of one and as you know a stair combines a dead load with a ever changing live load.
For one brief moment I thought I had this thing nailed down, but I'm not so sure again and I can't even assess my own level of confusion, heh.
Anyway, what I was trying to get at, what I meant by sculpted lam furniture, for example, is shown in the work of Michael Coffey (a former teacher and a major influence on me). In his words:
"Sculpture has three dimensions. My work has a fourth: the sensation of touch, and yet a fifth: the pleasure of use. If an object is conceived as a piece of sculpture, but feels good and also has a function, it is a richer experience for both maker and user."
When he says touch he actually means the sensory experience of continuous silky, almost soft surfaces with no discernable tactile breaks. He also extends touch seamlessly to include all body contacts with furniture, from surface glides to pressure sensations to ease of shifting positions, and so on. And in this respect, his work is masterful and flawless. fluttering on the liminal edge in the most pleasing manner. Any tactile anomaly would shatter the spell. The feel of his finished surfaces are rooted deeply in my consciousness and inform my own workmanship.
If I can't easily let this alone, it's because it really really matters to me. And what we call things is important. Whether it's small art objects, jewelry, a smoking pipe, lighting or even cabinet hardware, I mix materials and I particularly love seamless transitions visually and tactilely, and while I employ sound mechanical construction methods, glues often play a strong role. I simply can't abide my fingers detecting transitions from one material to another.
I suspect that discernable ridges at joints is indeed due to differential shrinkage, and that the PVA glue is creeping and letting the wood reach its natural thickness while harder glues are absorbing the stress internally and keeping the glue line flush.
I have an Ikea solid-wood table laminated of many strips of birch, each roughly 2" wide. Looking at it in raking light I can see that the surface is not perfectly planar but rather there is some variation that corresponds with the laminations. I suspect this is due to differential variation due to humidity loss in my dry climate. However, the joints themselves are flush, likely due to the use of a glue with low creep.
you let the wood fully stabilize after gluing. The times I've gotten ridges are due to planing or sanding nearly immediately after glue-up. If I wait, I don't have that problem.