Okoume shear strength test

In engineering terms, there is no reason for the maximum bending stress should be less for increasing thickness unless designed otherwise.
It is just a case of ply thickness/layup design.

The basis of the F-Stress grade system is that this is the case. It is designed to be thickness independent.
Making life easy to design boat scantlings. I don't expect things to be perfect which is why the F-Stress system is a characteristic strength.

As you suggest, if Okoume is designed as a "bending" plywood then my key assumption is wrong (it happens from time to time).
In this case you would need to have specifications for each thickness if you want to use it for "structural" purposes.

Alan

 

Thanks again!
I'm doing the testing mainly to better understand the failure modes in plywood at various orientations, and also to better understand the data presented. I'm starting to build the spars for a wing for my boat, and am double checking some design decisions.

 

The modulus of elasticity in bending Em (in N/mm2 ), described in EN 326-1 of each test piece, is calculated from the formula:

Em= l1^3 (F2-F1) / 4 bt^3 (a2-a1)

Where:

l1 is the distance between the centres of the supports, in millimetres b is the width of the test piece, in millimetres t is the thickness of the test piece, in millimetres F2 – F1 is the increment of load on the straight line portion of the load-deflection curve. F1 shall be approximately 10 % and F2 shall be approximately 40 % of the maximum load a2 – a1 is the increment of deflection at the mid-length of the test piece (corresponding to F2 – F1 )

I see your concern now as increasing thickness will greatly affect the “stiffness” or resistance to bending. But as I have explained above, the directional strength is dependent on how the fibers/veneers are arranged. Maybe it has less fibers oriented in the longitudinal direction as it gets thicker, hence the result of the test.

 

Is that a box beam?

 

I essentially agree with your first statement. However, plywoods can be (and are in practice) "designed otherwise". For many uses, it is nice and convenient to think of plywood as being anisotropic and amorphous. But it is not designed that way. There are thin plywoods with thick cores and very thin outer skins (such as that sold as "underlayment"). There are also thin plywoods with three equal-thickness layers, with the two outer layers running with the grain lengthwise. If you think of a narrow strip of plywood as being like an I beam, with a tension flange and a compression flange and a shear web, then three equal plies can be quite efficient... much more efficient than 5 plies, in which the outer third of thickness is only partly functional as a flange because a large portion of the fibers are oriented in the wrong direction.

If plywoods consisted of a very large number of very thin laminations (say 50) then we'd expect that bending strength and stiffness numbers (expressed in MPa) would be essentially the same for different thicknesses. It is convenient, but not accurate to consider a 3ply material to have the same properties as a 7 ply material. Nearly all plywood has different bending stiffness numbers in the longitudinal and transverse directions, I think you would agree. The number of layers has a similar effect regarding how effectively the fibers resist bending.

Only 3mm Okoume is frequently called "bending plywood." All plywood bends... some more easily than others. 3mm is unusually easy to bend in the transverse direction, because 2/3 of the thickness is poor at resisting bending forces. If one is interested in a plywood that is good for in-plane shear, then there is aircraft plywood with 45 degree fiber orientation.
Birch and Poplar Plywood Aircraft Grade MIL-P-6070 - Wicks Aircraft Supply Company https://www.wicksaircraft.com/shop/birch-and-poplar-plywood-aircraft-grade-mil-p-6070/

 

Okay, I have been doing some research on Okoume/Gaboon (the name used in Australia).
I do see that one of its main attributes is "bending" plywood (i.e. low modulus of elasticity).
The nearest recommended F stress grade is F8.

Are you going to post your spar/beam design?

AlanX

 

Almost. It is actually two I beams, but the wing skins join those into a box that is effective in torsional resistance. The spar caps (I beam flanges) are carbon reinforced, with the reinforcement level reducing from base to wing tip.

 

How did you estimate the wing loads?

AlanX

 

Looking at the photo, it looks like a coupon for tensile strength, not shear. For shear strength, the material gets clamped on two jaws that are close together. They move normally (at 90 degrees) to the surface and measure force until it breaks. Wood failure is brittle.

 

Broke stuff today.
In honor of Newton, I will use pounds and inches, which is what he used.
I used my instrumented shop crane to apply loads in 10 pound increments, and measured displacement with feeler gauges inserted between faces of the jig that, when not under load, appear to meet -- but are in fact .004" apart. My guess was that the coupon would fail at 150#. The measuring faces remained essentially parallel through the test, and there was no sign that the coupon was doing anything other than lozenging.

Deflection was a little under .002" per 10# increment, and surprisingly consistent from increment to increment. (The stress strain curve line was quite straight.) The last check was at 190#. There were no clicks, creaks, or groans until the final, satisfying bang. Failure occurred just after 190 lb. ( at which point the deflection was .030") on the way to 200.

My testing came about because I had come across a very low for shear in Okoume plywood, but the test method was not described. (Maybe the coupon aspect ratio was too low, or the jig was not self aligning, or inadequately restrained, etc.) The test provides some confidence in the figures provided by rxcompomposites.

Doing a test like this -- or any other properties test -- does not produce a number that can be reliably used for design purposes. That's what numbers like those provided by rxcomposites are for.

On the other hand, people who make a living testing materials do not always agree on test methods, and even large organizations (DIN, EN, NZ, etc) disagree on test methods and results. It is convenient for someone designing a wing to first calculate tension and compression loads, and then shear loads, and to think of the shear web as handling just shear, and the caps handling just tension and compression, and to think of maximum shear along the longitudinal centerline of the shear web... but in real life things don't work like that, nor do they work as they do in a test jig. Notably, the caps in a wing spar are not constrained, and the wing will have bent substantially by the time a shear failure becomes a possibility... and by that time, there is no pure shear in play.

But it's fun to break stuff, nevertheless.

 

I use a technique called "profoundly wishful thinking". It results in very light structures.

Actually, I'll tell you how I have done it in the past, because some parts of this current wing and boat are things I'll patent.

The Windrocket is a boat I designed and built a while ago.



When sailed correctly, it planes on the center hull, and the skipper will have moved out to the windward float, which, with the boat kept flat, is about 2 feet above the water. The boat is 20 feet wide, so a tack entailed moving 20 feet from one side to the other, and then hooking into foot straps (which can be seen as black things in the pic). So in a steady state condition, the righting moment will be about 2000 lb ft. In the Windrocket, the vertical center of effort of the wing changed as the boat accelerated, for somewhat complicated reasons. But imagine it had a "Hershey Bar" wing, like an old Piper airplane. Then, a useful and safe simplification is that the center of effort is at the center of the wing plan. In the Windrocket, that would be 10 up (because the wing was 20 feet long). That means that at approaching capsize, the maximum lift that can be used is 200 lbs.

You can imagine dynamic conditions in which the wing load becomes higher: a gust comes along, the skipper fails to ease, and the boat rotates to leeward with some rotational acceleration that lifts the skipper further from the water. When, in addition, the leeward float hits the water, its little bit of buoyancy ads to the righting moment... etc. So perhaps better to design for twice the steady state condition. If the boat were bolted to the ground and a gale came along, the wing, if trimmed to say 12 degrees angle of attack, could generate perhaps 1000 lb lift or more, and something would break. But it would not make sense to design for that possibility.

In the current boat, (not the Windrocket) the maximum usable lift is that lift that causes the fully loaded boat to capsize, given a couple people on the rail. Unlike the Windrocket, the current boat has a quite even distribution of lift (from mast base to wing tip) and it works out that the wing can supply about 800 lb of lift as the boat capsizes. The skipper has complete control over the wing's coefficient of lift, but if the skipper is bent on capsizing the boat, about 25 knots of true wind is required to do so.

In the Windrocket pic, there is a black spot near the base of the wing. That is a window that exposes the needle of a mechanical strain gauge which measured strain in the wing spar cap (which was a group of carbon fiber pultrusions ). I never saw the needle move appreciably during those gusts I imagined while designing the wing. (The strain was rarely more than about .1%... the loud bang would occur at about 1%).

 

It may be hard to see in the picture, but the coupon is the little piece of thin okoume. Most of what you see it the jig. With the face ply grain oriented in the desired test orientation, and with the jig hanging vertically, the right side of the coupon is bonded to the part of the jig pulling up, and the left side of the coupon is bonded to the part of the jig resisting than force. As you can see from the broken pieces, the coupon failed along a line nearly perfectly parallel to the applied force.

 

At least you are on the ballpark and gives you a whole lot of self confidence. You did it right.

Those doing the Quality A testing are required to report in three decimals and do it 7X, throwing away the highest and lowest values and average only the 5. It is very impractical.

 

I've built small boats using both 3 equal layer plywood and with two thin skins and a thick core. The former bends very easily across the face grain and is stiff the other way. The thin skinned ply was closer to equally stiff in either direction.

Either ANC-18, 0r ANC-19, or both, discuss using plywood for shear webs in aircraft spars, including how far apart to put stiffeners to prevent buckling*. These were, at times, publications of the US military, and, I think, at other times, publications of the US FAA. (Federal Aviation Administration, not Fleet Air Arm!) I don't know if they discuss using +/- 45 degree grain orientation, but I know that's much stiffer in shear. Not sure what the difference in strength is.

*I've seen a plan for a rigid wing hang glider that used a shear web of 1/32" ply! I've also seen video of it flying. I'm sure the designer had to take a good hard look at possible buckling of the shear web.