Home Articles Forums Gallery Marketplace Videos

WoodCentral Forums

Knowledge You Can Build On — Est. 1998

Build it. Rock it. Rockler has your woodworking tools.

Yellow glue creep data

Posts

Yellow glue creep data

#1

Yellow glue creep data

Bill Tindall

Creep explained and defined: Creep is an inherent property of any polymer. All adhesives are polymers so all adhesives can/will creep more or less. The more or less can be sensitive to ambient conditions of temperature and humidity, stress and time under stress.

While a piece of dried glue (or any polymer) may appear to be a rigid solid, down at the molecule level the extremely long and tangled together polymer molecules are wiggling, more or less. A good analogy is at the molecule level the dried blob of glue looks like a blob of worms. If the worms are cold they aren't wiggling much. If they are frozen they aren't wiggling at all. If they are warm and wet they are wiggling enough that the mass of worms will conform to the shape of the container, they can be stretched, or even pulled apart without breaking any of the worms. Ditto for the glue blob molecules. If we grab each end of the worm blob and pull the worm blob will extend. This extension of the wiggly worm blob in response to the stretching force is creep. For the more technically inclined, creep is the slow deformation of a polymer(material) in response to stress.


Creep is measured for a polymer by molding a strip of the polymer, clamping one end rigid and hanging a weight off the other end. The lengthening of the strip is measured over time. Creep is this lengthening. No molecules are broken, in fact the material is not broken, in a creep test. Adhesive creep is measured by gluing two pieces of non-creepy material together, clamping one end and hanging a weight off the other. The lengthening (elongation) of the assembly over time is creep.

Yellow glue creep: There is reason to fret over yellow glue creep because the people that make the stuff say you should. They have lawyers and worry about law suits if you glue your house together with yellow glue. But do people gluing furniture need to worry about yellow glue creep? Well, here is some data and some caveats to go with it.

The experiment and result: Two maple sticks were glued together. Total glue surface is 1.5 sq. in. A scribe mark was made through the joint to enable movement of one strip relative to the other to be detected in the creep test. The joint was cured for one week. Then one end of the assembly was clamped to the rafters and a twenty pound weight was hung from the other end (see picture). The assembly was hung for 60 days. During this time the temperature was between 65 and 70 degrees and the humidity was between 45 and 55%. The scribe line was examined after 60 days. I could detect no movement of one strip relative to the other and I think I could easily see 0.002" of movement if it occurred.

Conclusion: A well cured joint made with Titebond Regular glue does not significantly creep under quite a large load in 1. 60 days, 2. at 65 to 70 degrees F, and 3. at 45 to 55% relative humidity.

Caveats: I might have been able to detect creep if the joint had not cured well before the test, or the temperature or humidity was greater or the stress was applied for a longer time.

Yellow (or white) glue is formulated so that the polymer molecules (worms) are essentially "frozen" by "hydrogen bonding" (chemistry term) at typical ambient conditions so it was no surprise that I saw no creep under the conditions tested. At 150 degrees or thereabouts I would have seen significant creep and at wet 200 degrees total failure of the bond.

Stressed applied perpendicular to the long grain of wood will cause the wood molecules to creep. This creep will lessen the stress on a bent lamination over time.

Re: Yellow glue creep data

#2

Re: Yellow glue creep data

Dave Bair, Eaton Rapids, MI

In 1985, I made my own 24' trusses for my workshop. They are in the style known as a "king post" truss. I made them from 2x4s, 3/4" plywood gussets, and yellow carpenter's glue. I used 1-1/2" box nails to hold the gussets until the glue set. My shop is still standing so it was with some interest that I read the disclaimer on my latest bottle of Titebond 3 that said it wasn't to be used for "load-bearing" applications. I felt uneasy for a second 'till I realized that I have 25 years of actual proof that yellow glue is perfectly acceptable for structural applications.

Re: Yellow glue creep data

#3

Re: Yellow glue creep data

Thomas S Stockton

Interesting stuff Bill

Most of the cold creep problems I have seen are in laminations, sometimes you can just feel the different layers. I have a pine lamp I made in high school out of 4 layers glued with white glue and you can defiantly feel the layers, but I think some of this can be caused by pieces that had different moisture content when glued.

On a funny note, someone has a web site that claims that anything built with yellow glue will eventually fall apart because of cold creep.

Bottom line use the correct glue for the application and your problems will be minor.

Tom

Re: Yellow glue creep data

#4

Re: Yellow glue creep data

Jesse Cloud

Thanks Bill, very informative.

I would add that "creep" isn't all bad. If you look at a cross grain application like mortise and tenon, that tenon is going to grow and shrink in reference to the piece that has the mortise. Creep allows the glue bond to flex a bit and the joint to survive movement.

Tom is absolutely right though, creep in a lamination or veneer is creepy. Sorry, just couldn't resist the pun.

Re: Yellow glue creep data

#5

me too

Bill Tindall

I too have home made trusses for a lumber storage barn that haven't creeped to failure yet as well as several home made laminated house beams made from glued together 2 x 12's.

If I lived where it was very hot and soggy maybe it would be a different story.

Re: Yellow glue creep data

#6

creep in laminations

Bill Tindall

It seems common to attribute a glue line that can be felt (proud glue line or what ever) to creep. It is not clear to me how creep as it is defined in my discussion (and in Material Science) would lead to a glue line that could be felt. Franklin claims that glue lines that can be felt in a panel are a result of differential shrinkage of the wood on each side of the joint. It seems likely that some similar phenomena is in play in laminations. In thin strip laminations the moisture content of the wood will become very high as a result of the water in the glue migrating to the wood.

If creep occurred in laminations the result would be a slow change in shape over time as the stress caused the laminations to slide past one another. But, in a lamination there is a huge glue surface to distribute the stress over so the typical stress seems to be less that I subjected my test piece to.

If the phenomena that is attributed to creep does not continue to occur over time then the cause is definitely not a result of creep. Creep goes on forever in a linear fashion so long as the stress is applied. So a question.....the phenomena that people are calling creep, does it continue over time, ie. continue get worse? If what ever happens happens in the first week and then doesn't get worse I would look for some moisture effect on the wood to be the source of the problem.

Re: Yellow glue creep data

#7

As usual Bill...

John in NM

You come up with interesting and well thought out experiments :D

One thing I've noticed on the internet.... lots of people (including myself of course) talk about things with which they have little or no direct experience. I think a lot of the "creep" discussions confuse creep with springback when making bent laminations. PVA's are nice flexible glues, the traditional lamination glue I've used (plastic resin) is not. Never tried resourcinol or similar glues myself.

I'm interested in your (and any other's) truss building. Did you do any calculating or just rule of thumb in the building of them? I'm not sure I would have trusted PVA, but you certainly know the chemistry better than I do.

Re: Yellow glue creep data

#8

Truss-Building

Dave Bair, Eaton Rapids, MI

Back then (1985), pre-internet, I used a wood-engineering manual from the library. As I remember, several truss types were described along with tables giving dimensions, angles, materials, loads, etc. Probably all that stuff is online now. I laid out the truss template, in chalk, on the garage floor, cut the pieces to length and glued up each, one at a time.