Sidewall strength

[Vanq]

real world I don't think its that much different- I doubt my supports would ever be so perfectly verticly loaded as to fail by compression before bending... one side will always be in a bitt of a hole or bump turning the roof load into a partial bending load with the maximum possible leverage on that sidewall.

[Forrest747]

What I see in this picture is that you applied the load (and the supports) to the flat face of the plywood. If what I see is correct, this would test a bending load, which would be better indicator for the floor or roof.

I would agree with you it would be best for a floor or roof. In full disclosure I did this test "off book" and needed to complete it quick and dirty. THe test results were accurate for the set up.

It seems like the right test for a wall panel would be to - (this is the best way I can think to describe it) - would be to place a 4x8 sheet upright on the 8' edge and load it (on the thin edge) vertically until it fails. I know this wouldn't be practical due to the size and I don't think the test would scale well either since a small width of the material material size isn't braced the same by the rest of the material around it.

I would say a test with the 4x8 foot panel of plywood would be best with the load applied to the 4 foot section, sadly I can not do that test at this time. I also agree its a bending moment however i know if I were to do this test with a clamp set up at each of the sample and compress until fail, that the failure profile will look very simular to a three point setup. and that the loads would be very simular with in the margin of error. I have a 3/4 inch sample i will do after lunch in that setup. I will claim refrence measurement to the uppers.

[Forrest747]

Ok using a 3/4 inch baltic birch 48 inch by 3.5 inch sample, in a compressive load clamp setup. the max static load is 1712 lbs. Failure point was 3 plys from edge, at 28 inches from the load cell. MTS system used in clamp compression setup.

using the data from the 3 point test set up of 1658 lbs. this is a 3.2% diffrence from both data points. This is less than the out of safety specifications and with margin of error and acceptable linear range.

It is compareable to the other test and while other samples would allow a more linear collitaion to data, this intial set is a good refrence point and acceptable under industry practices.

Notes:
Equation ((1712-1658)/1712)*100 = 3.1542

Data was collected from two diffrent sources with both systems under current calibration. Both systems calibration data is with in the margin of error.

File picture of set up attached to notes. May not reflect actual test samples.
File picture of MTS system used in test attached to notes. May not reflect setup or location of test. See Engineer for any questions.

[Forrest747]

I wonder if anyone has an old teardrop that could be backed under the load cell.....

It took several peanuts and eventually a white lab mouse but we finally got the elephant on top. I think Mikes Benroy held up pretty good.

[angib]

2x4 2917 lbs
1 1/2 inch 2647 lbs

This is an interesting result. From what I can understand, these were two samples the same size (actual 1.5x3.5 ply = nominal 2x4 timber) and the home depot stud was stronger than the ply.

This is probably the opposite of what most forum members think, but exactly what would be expected - the grain of half the plies in the plywood are running across the sample so not providing much bending strength, whereas all the grain in the stud is running along the sample.