You already get similar ( like from RyanY ) answer from Avcase in your thread "Is the early wood the weak link to cause set?" on the bottom of the page 20 : "The amount of energy that a material can store in bending is a product of the elastic modulus and amount of strain or deformation it can handle before taking a set. It can be taken a step further by throwing in the density in order to get a property that rates materials as ft-lb of energy per ounce of material.", which is probably more understandable :-) .
IMHO - for rapid and good repeatable results - one have to start with such measurements of the appropriate sample of prepared wood stave/billets, then using these values to design your bow in some software ( Supertiller, VirtualBow ... ) to achieve desired properties and continue with shaping the bow by this model measurements ( with reserve ). As wood is not very homogeneous material, you cannot rely completely on theoretical dimensions and have to continue in tillering by bendmetering, to be sure you are keeping the strains in optimal ranges. ( Good description of this method is in this book :
http://www.broadheads.de/docs/literatur/disputation/Elmer/Elmer_TargetArchery.html - please, see the pictures slightly above the bottom of the page - BTW I advice to rather buy - more handy - digital dial indicator for this purpose . Free quote of homebody's contribution somewhere : “Tillering Gizmo is only simplification of a bow-making bendmeter, first devised by Russell Willcox in the 1920’s. Russ and others refined it over the years. He used it in tillering his Duoflex bow … " )
Bendmeter can be very precise, hence you can determine exact strain values at every “station” on the bending limb and “tiller” it to be precisely equally stressed ( or what stress distribution you choose ) along its length in accordance with the diminishing thickness ( elliptical tiller ). At the beginning you have to measure initial offset ( on each station along the limbs ) on non stressed bow, then the deflection and the thickness on actual station on ( partially ) stressed bow and then compute the actual strain value ( in percent ) at this station. And so on for all stations and tillering stages .
I used bendmetering several years ago on my wooden bows . I measured the bow only at about half-draw to eliminate string follow caused by long time spent at full-draw position, when measuring it by bendmeter ( shape of limb bend and strains at this half-drawn position can be resolved through mentioned software ).
With true bendmeter it is possible ( I hope ) to tiller the bow by amount of set/string follow in each station along the working limbs ( and by this way eliminate long time measuring on drawn bow ) , because it is so precise ( I never did it ).
Or, maybe ( only my fantasy ), if you want to stay on primitive way, you could observe the chrysals on intentionally highly arched belly of the bow and when they are placed everywhere evenly at full draw, you could remove the damaged wood ( slightly more ) without changing the drawn shape of the bow and you will obtain "evenly" strained bow.
Bendmetering is quite “boring” and “longlegged” method of tillering ( nowadays facilitated by spreadsheets on computers ) , but gives very reliable results ( even for novices ) and should be used at least to enhance flightbow making if too laborious for common bows.
Guess, I was not very boring ( sorry for my English and lengthy contribution ) and hope, that more contemporary bowmakers will be interrested in this “old masters technique” again.
