takedown sleeve test

[Richard B]

I am in the process of building a takedown bow http://www.primitivearcher.com/smf/index.php/topic,45456.msg665225.html#msg665225 and have seen bows made using different joints between the wood and the takedown sleeves. This post describes a test I carried out to try to determine which approach was stronger. The two approaches tested were:

1. The approach described in TBB3 with the wood fairing into the inside of the sleeve (the right hand piece in the first picture).

2. A shoulder on the belly and half way up the slides to the depth of the sleeve thickness (the left hand piece in the picture).

I made up the pieces shown in the picture from pine (20 x 18mm rectangular section) using a 50mm length of 16mm (ID) stainless tube. The ends were fastened into the tube using Smootheon MT13 epoxy.

[Richard B]

I then mounted the test piece on my tillering stick and loaded it progressively (in 5lb increments, exercising 50 times at each increment) until it failed.

The results of the trial were not as clear as I expected:

At 30lb the unshouldered end (method 1) started to develop some set, suggesting that the wood at the join with the metal tube was starting to crush. There was no set at all on the shouldered joint, suggesting that this join had not reached its proportional limit at this load.

However, at 40 lb the shouldered joint fractured (I think, by tensile failure on the back) - see picture below. I think the cause of this may have been the stiffness of the tube bearing on the shoulder at the belly overstressed the back. A bit odd, because I would have thought it would crush at the belly first (as it did on the unshouldered joint).

It is only one test, with only one wood. I probably should have used yew for the test as the relative strength of the belly vs the back would have been more representative.

I expected the unshouldered joint to fail first, because it is a shallower section. Indeed, it did start to develop set first. However, it seems that the stiffness of the tube bearing on the shoulder may overstress the back.

I think that I may get rid of the shoulders on my yew takedown bow and go for the approach described in TBB3.

[bubby]

Why are you bending it sideways instead of to the belly?

[lebhuntfish]

Very cool Richard,  thanks for posting. I want to make a take down soon myself. Patrick

[TimBo]

I have the same question about bending it sideways, but it's great that you are testing it. 

[J05H]

He's not bending it sideways. The test pieces are square. I assume this is so that the weakest point is at the joint being tested.

[Marc St Louis]

Not quite sure what the point of this test was, Pine is not Yew.  Try it with a piece of hardwood and see how much weight you need before it breaks, you'll see that your results will differ quite a bit

[Jim Davis]

Couple of things.

It is almost certain that the "tension" failure was preceded by crushing of the belly wood. In  bending, the belly always crushes first, if the back growth ring is intact (and we're not talking about cherry or red cedar. Only in a sap-laden green stick would the belly compress and wrinkle enough to be obvious without the back breaking. In dry wood, when the belly compresses, the neutral plane of forces moves closer to the back of the bow and that's when  the back lets go.

Thinking of the shouldered limb, the step down from the shoulder to the inside diameter of the tube creates a spot where the crushing force is concentrated.

OK, three things. When you make limbs instead of square test pieces, the handle area is plenty strong enough to be the area that will not break if the bow is overdrawn.

This works great:

[Richard B]

J05H you are right . I was not bending it sideways. The "limbs"of the test piece are deep (18mm) to make sure that the test piece fails at the wood to metal joint, so that I could compare the two approaches. Naturally one would never build a bow like that. The idea of the test was to try to identify the most robust way of designing the joint so that it can support the maximum draw weight. Clearly, one would design an actual bow with fades in the limbs so that bending is kept away from the joint as much as possible.

I also agree with asharrow, most wood will fail in compression before it fails in tension. However, it looks to me that a tension fracture occurred on the shouldered side, with similar symptoms to those  described by Tim Baker in TBB3. One reason for this may be that the shoulder, only going up half the sides makes the back narrower than the belly. This has the effect of moving the neutral plane towards that belly with the result that the tensile stress will increase at the back and reduce at the belly (compared with a symmetrical section). It is the same effect as trapping a bow to reduce compressive stresses.

Mark St Lewis, yes again, yew would have been a more relevant test, as I think the sapwood is more resistant to tensile failure than pine is. However I was not really interested in what load it failed at, just what the relative performance of the two approaches was.

At least, doing the test has convinced me to go with the non shouldered approach for my bow (i,e. as described in TBB3) so it has been useful for something!