Perry reflex versus pre-bending cores and backing

[Eric Garza]

I am gearing up to make a fairly simple laminated wood bow. I will be using a hickory board for the belly wood, and a hickory strip for the backing. I plan on heat-treating the belly core before glue-up. My question is this: Would it be better to heat treat the core while it is straight and glue in Perry reflex, or would it be better to pre-bend the core while I heat treat it and not add much, or any, Perry reflex during glue-up?

I have made a few laminated bows, but not enough to be able to think through these alternatives in an informed way. Thanks in advance for your thoughts!

[Aussie Yeoman]

I haven't done what you're suggesting, but I've done a reasonable amount of engineering when it comes to bows.

My suggestion would be to heat the belly with a small amount of reflex, and then add more in the glue-up. The benefits of heating wood are different mechanical principles to those achieved through Perry-reflexing.

AY

[superdav95]

I second what Aussie said.  Hickory has a great memory if induced reflex when heat treated.  It also helps with compression strength too on the belly.   If it was me I’d pre reflex the hickory during heat treat.  Good luck

[bownarra]

Bend the belly into deflex (heat treating with dry heat). Then when doing your glue-up bend the deflexed belly and straight back into a reflexed form and clamp.   That's proper Perry reflex.....and it works There is a bit in the TBB that explains it.

[mmattockx]

My question is this: Would it be better to heat treat the core while it is straight and glue in Perry reflex, or would it be better to pre-bend the core while I heat treat it and not add much, or any, Perry reflex during glue-up?

Heat treating the belly wood increases its compression strength limit before set occurs while Perry reflex reduces the maximum stress the belly wood will see. Without being able to quantify how much the heat treating improves the compression strength it is not possible to calculate the best balance, but bownarra's combo is the way to achieve minimum set and high performance.

Bend the belly into deflex (heat treating with dry heat). Then when doing your glue-up bend the deflexed belly and straight back into a reflexed form and clamp.

This is the winner, with the only limit being a tension failure on the heat treated belly side from the tension it sees when you pull it into reflex for the glue up with the backing.


Mark

[superdav95]

Hmm.  Interesting.  It’s seems counter intuitive to me to heat into deflex then clamp and glue up into reflex.  Obviously it works. Hence pert reflex… I guess my limited understanding has me with more questions then answers.  I’ll have to research and experiment with this myself.  Good luck with the build Eric. 

[mmattockx]

Hmm.  Interesting.  It’s seems counter intuitive to me to heat into deflex then clamp and glue up into reflex. 

By deflexing the belly lam first you increase the tension preload the belly sees when you pull the lam forward into reflex. There are limits to how much Perry reflex you can have in a bow before the back fails. A lot of reflex also makes it very difficult to string and much harder to tiller. Deflexing first allows the tension preload on the belly to be achieved with a moderate amount of final reflex. This keeps stringing and tillering relatively easy but gives the performance benefits of lower stresses/strains on the belly wood that the high amount of preload gives.


Mark

[Yooper Bowyer]

If I understand it right, bownarrra has it right.

[Eric Garza]

Thanks everyone for your thoughts and suggestions. I will consider heat treating in some deflex, but am more likely going to heat treat the belly core so it is straight (no reflex or deflex), and then glue in some reflex. I am debating whether to make the bow deflexed in the handle as well, to add a bit more string tension once braced. The bow will be on the shorter side though, 58-60 inches, so not sure I want all that excess string tension.

[Aussie Yeoman]

On reflection, and reading the above, a deflexed belly then glued in Perry reflex is absolutely the right way to go.

A bow will take set, or fly apart in tension failure, when its strain (the change in length due to compression or extension) exceeds a threshold particular to that piece of wood. It makes sense then, for the belly to be deflexed before putting in the Perry reflex, because this means the belly is under less strain when the limb is at brace height (and in fact at all stages throughout the draw).

By adding in the Perry reflex, you increase the volume of wood doing an appreciable amount of work, which makes the limb stiffer per mass. So increasing the stiffness while reducing the strain on the outer surfaces can only be a good thing.

AY

[RyanY]

Thanks everyone for your thoughts and suggestions. I will consider heat treating in some deflex, but am more likely going to heat treat the belly core so it is straight (no reflex or deflex), and then glue in some reflex. I am debating whether to make the bow deflexed in the handle as well, to add a bit more string tension once braced. The bow will be on the shorter side though, 58-60 inches, so not sure I want all that excess string tension.

If at brace the string angle with the limbs isn’t super low the tension shouldn’t be too much higher. A mild design with a bit of deflex and limbs that still come around out of the reflex would be good for a short bow with the deflex taking a bit of stress of the limbs and keeping string angle favorable.

[scp]

I am not sure how much deflex and reflex people are talking about. But I would be concerned about reflexing the heat treated belly too much.

Deflex in the handle would make the braced string tension weaker.

[Morgan]

By heat bending in deflex, does this also help mitigate compression failures? It is my understanding (probably wrong) that compression failures are pretty much a result of too much wood in thickness in too tight of a bend radius. Does pre bending in the softened plastic like state help this?

[Marc St Louis]

I've found that heat-treating a good portion of the reflex in first and then gluing the rest of the reflex in to work much better especially with a white-wood.  The bow will keep more reflex once tillered

[mmattockx]

I am not sure how much deflex and reflex people are talking about. But I would be concerned about reflexing the heat treated belly too much.

How much is always the question, isn't it? I also think caution is wise regarding how much tension is put on the heat treated belly. It will be much more susceptible to failure after heating.

By heat bending in deflex, does this also help mitigate compression failures? It is my understanding (probably wrong) that compression failures are pretty much a result of too much wood in thickness in too tight of a bend radius. Does pre bending in the softened plastic like state help this?

It isn't the heat bending, it is the pulling into reflex of the belly lam that reduces the compression stresses on the belly. You can glue in the deflex as well when making a 3 lam bow and still get the same tension preload effect on the belly wood when the final glue up pulls the belly into reflex.

Your understanding of compression failure is correct, the wood is bent into too tight of a radius for its thickness and the bending stress exceeds the compression limit of the wood.


Mark