I'd rounded the belly a bit and cut second nocks so I could get a shorter string on it.
With the winch pulling the long string I had to take it to 110# just to get the short string on at a very low brace.
I winched it back to about 70# and it looked ok. I thought I'd take it further. At 100# I heard a 'tic' the death knell of a bow (probably the back giving way).
Turn the sound up and you'll hear it, a split second later it exploded.
No point me pontificating as to why, if I have to guess I'd say the back wasn't perfect. It could even have been damaged when the up and over garage door jammed against the bow as I opened it... I neglected to check the back. Of course there are marks on the back now where it smashed into the underside of the door.
The good thing is the splice held solid and the lower limb remained in one piece.
This illustrates the point about a short string applying greater load than a long one.... I'd pulled it to 110# on the long string several times, yet it blew at just over 100# at low brace.
Checking the original videos I can see the tip deflection on the right limb with the long string was 3 bricks. On the short string it just about hits 4 bricks and then explodes.
The sound doesn't appear to be there but watch closely! (Maybe it's a UFO ?? ;) and Aliens caused it.
The video isn't hi res because it's compressed. I've studied the original and the break definitely starts at the grip. It probably is the discontinuity in the sapwood at the splice. This confirms my suspicion that a sapwood overlay over a splice is a good safety measure.
You could say I was working the bow too hard in the middle, but don't blame me! I left it slightly over size compared with the Mary Rose bow dimensions... nuffin' to do wiv me mate ;)
You can see from the stills that the crack has propagated along the length of the limb through the heartwood.
Studying the video closely I think the sapwood gives way in the centre, this removes the protection from the tension-weak heartwood which gives way and splits along the right limb, as it splits the limb flexes more and actually breaks at a second point about 2/3 of the way along that limb.
In terms of actual tip deflection it wasn't all that far from a full draw, which on a similar bow is only one more bricks worth of tip deflection.
I'm a bit disappointed, but I think the careful study of it has taught me a few lessons.
I think I've found the culprit!
I reassembled the centre section and sure enough there is a sliver missing. It's not the actual join that was the problem, more the discontinuity where the sapwood has been rasped to match either side of the join, you can see the crack going from the missing sliver to the edge of the bow.
A thin sapwood overlay would probably have held it all together. Mind I couldn't overlay any sapwood because of the nice feature knot...catch 22! c'est la vie.