what have I learnt from this break? not sure...

[joachimM]

Started a shortish field elm (Ulmus minor) self bow, 52.5”, from the knot-free sister half of this log.
Pyramid design, decrowned, 1.15 cm thick upper limb, 1.05 cm lower limb, 5.2 cm wide (2”) tapering to 9 mm wide nocks, lower limb 1” shorter than upper limb.
I steamed in a tiny bit of handle deflex and 1” of reflex in the lower limb. Lightly heat-treated on the belly. Came exactly in at the expected weight according to the mass principle (351 g). Drew 48# at 24” Shot it about 50 times, clocked some 175 fps from a 500 gr carbon arrow, flight shot it with that same arrow 205 m. Didn't take appreciable set.

I was delighted. Calculated the strain level on the back, which was close to 1%. Close to the limit, but acceptable, so I thought. Perfectly circular tiller in the lower limb, upper limb a tiny bit stiffer in the outer part (see tiller pic).

Sanded it lightly a first time yesterday with 60 grit to remove tool marks, and wanted to shoot it again today.
And broke it at the first full draw in the lower limb, where the reflex bending was most pronounced. Clean tension failure. Not the first time that I break a bow after sanding it…

Bummer. At least I now have half a bow to match to a similar broken black locust bow. But what should I learn from this failure?

[Springbuck]

So, decrowned, like you leveled the back of the bow?  I kind of gave up decrowning as a bad job.  It's even harder to follow the grain perfectly than when marking out a snake stave from a bigger log.  I have ONLY had success doing it on kids' bows.  But, I also know it can be done, so.......

The full draw tiller pic has the lower limb bending a lot more than the other, as you noted, but what I noticed was the reflex on both limbs not only flattens out, but bends just as much as the rest of the limb.  When I put even gentle reflex in something, unless the whole limb is only slightly reflexed, but the same amount, you can still SEE the reflex there at full draw.  Does the make sense?  Like, it may flatten out, but you can still see where it WAS even if you can't see it directly.  I fear you may have reflexed it, then required the limb to bend enough there to not only flatten out,  but to continue to bend to the same arc as the rest of the limb.   That may leave that spot bending 1.5 X the same amount as the rest of the limb.

That's my guess, and if it then found some tiny weakness from the decrowning, that's that.

I know getting a nice 2" width is hard on a smaller elm sapling, but I totally just leave that crown.  I have made 55 lb bows out of 2.5" across elm saplings, and if you can't get that width, just get a tiny bit more length.

[willie]

joachim-
thats is a bummer, it looked like a nice tiller and on its way to being pretty bow.

where the reflex bending was most pronounced

most pronounced before the heat treatment?  in other words, did it break in the heat treated  spot?

just curious also, how do you make the tension strain calculation?

willie

edit. my question concerns the same thought as, springbucks ....

 

I fear you may have reflexed it, then required the limb to bend enough there to not only flatten out,  but to continue to bend to the same arc as the rest of the limb.

[Del the cat]

Bummer, it was looking good .
Was it the left limb in the pic? There looks to be a weak point?..  (see pic) It's difficult to see as the wood is similar colour to the backgound...
Del

[Pat B]

Will you post a pic of the break?

[bradsmith2010]

if you are going to decrown a stave that is pretty maxed out,, you should back it with rawhide or sinew,, that would be my suggestion,,also you need to know what the moisture content of the wood is,, if you are working inside in the heat,, the moisture content of the bow may be to low,, causing it to explode,, I have a feeling this was a contributing factor,, and why the bow shot so well when shooting, it was very dry,, 
also I think your design may be a bit more maxed out than you are thinking,, you may have to overbuild just a bit to keep the bow together,, it all looked good on paper,, but wood has to many variables to follow every calculation,,
but I still think the moisture content was the main cause,,

[PatM]

Looks like you exceeded 1% in that area. A formula is fine IF the bow is actually conforming to it.  Your bow was not.
 More limb width is needed in hard working areas. Going from reflex to a slight hinge is the greatest stress of all.

[bradsmith2010]

PatM that is a very good point about going from reflex to slight hinge,,  you have to leave that reflexed area a little stiff,, but it is still working the same as the other wood,, it is a fine point in tiller that is hard to read sometimes,,,

[joachimM]

Decrowning: I do that often, with very little issues. Helps me to get an even strain across the back. I've had more tension issues with crowned backs, and since then I often decrown to get a nice flat back.

Moisture: it's very moist here in winter. Rarely gets below 65% and 12% MC of the wood, I have the opposite problem of having too much set because of too high moisture if I don't heat-treat during tillering.

Pics: see below.
Yes I was in dubio for a backing but wanted to be straight-edge self bow...

The working reflex: good point about leaving it wider and a bit thinner. I may indeed have maxed out on that part, whereas the other limb was still a bit stiffer there.
Del: yep, that's pretty much where it broke, actually a few cm to the left of that.

[PEARL DRUMS]

Its hard for me to believe a micro tweak in your tip tiller would have saved it. To me, it looks like a clean break in tension. You can see right where the tear began. 

[Springbuck]

Decrowning: I do that often, with very little issues. Helps me to get an even strain across the back. I've had more tension issues with crowned backs, and since then I often decrown to get a nice flat back.

Funny how different experiences are. The only tension failures I've ever had with small diameter elm have been when 1. Bugs ate it and I didn't see. 2. I WAY over-heattreated it. 3. I cut wood really early in the spring and it had a paper thin summer ring over the winter ring. Or, 4. I de-crowned it.

Now, set, that may be different.

[willie]

Funny how different experiences are.

right you are there, Springbuck

I have been comparing notes with others on a different species.  In spite of the similarities we find, sometimes one ends up with more questions than answers.

[joachimM]

most pronounced before the heat treatment?  in other words, did it break in the heat treated  spot?
just curious also, how do you make the tension strain calculation?

I heat-treated the entire limb evenly but only reflexed the outer limb a tiny bit (see the first pics for the non-broken limb)

Tension strain calculation:
During tillering I take pics to compare how limbs bend and if they bend according to what it should for that profile. I paste these pics in Inkscape, a free vector-drawing software. I draw a circle till it matches the most bending section of the bow, and measure the length of that part of the bow. In this case, it's the entire bow limb out of the fades (58 cm). The circle I reduce to an arc, and inkscape tells you the size of the arc (top left: starts at 210° and ends at 270°, so 60° arc).
The maximum thickness of the bending limb is needed as well, which is measured with calipers (1.15 cm).
I added a simple excel sheet with a strain calculator that explains the rest. The other sheets tell you how to pike a bow to desired draw weight, or how to thickness tiller to desired draw weight.
Yes I know, this assumes a lot of things about wood that don't match up to real wood. But I find it handy for initial tillering to get in the ball park, and to check afterwards how far I'm stretching the design. This time, too much...

Thanks everyone for your thoughts. I'll keep the outers a bit stiffer for the sister stave, and will back it.

[willie]

Joachim

the spreadsheet is useful, I need to play around with inkscapesome more, to find angles, (I may have to download the full version, as my distro only has inkscapelite).

My question in quote should have been, and I apologize for asking twice

 Did it break in the spot where you

I steamed in................... 1” of reflex in the lower limb

the reason I ask is,  some have theorized,      that when when you steam or heat bend while bent into a reflex, that the fibers are consolidated or compacted on the tension side (belly).
In effect, increasing the density at the surface. A different effect than simply hardening the surface with heat tempering.
thanks
willie

[joachimM]

i

Did it break in the spot where you

I steamed in................... 1” of reflex in the lower limb

willie

It broke close to where the bend was strongest, but honestly look at the profile of the other limb (treated the same) that's not much of a bend is it? It's not that I added big hooks or did anything funky. Not like I heat-treated 6" of reflex like Marc St-Louis does. But I have to admit, I should have used an adapted caul to make it bend more evenly. At the time, I thought it was a bit too abrupt in one spot.
And I guess the entire tiller profile looks good for a straight stave, but easily fools the eye, even for an ever so slightly deflexed-reflexed limb. So in retrospect, the upper limb might have bent a tiny bit more and the lower a bit less.