The best shape for a reflexed bow?

[willie]

I also like the curvature graph—you can see exactly where the bow is actually bending.

Yes quite helpful with reflex/deflex and other profiles that are hard to judge when looking at the bow on a tiller tree.

[simk]

Thanks Tuomo!
I'd say the results are intersting to get a general idea about these designs. The setback handle being worst and the d/r best will will most likely match with reality.
However: Having the front profiles and tapers not optimized will distort the outcome relevant. Especially having the same front profiles on all bows neglects specific advantages/disadvantages of these designs. The recurves need a lot more mass on the outers to stabilize than a self-stabilizing longbow. In my experience the theoretical advantages of recurves and other complex designs over longbows mostly get consumed with the additional mass required. Therefore I would cast into doubt  the "placements" 3-5 and also I think the gap between 1 and 6 is not that big.
 

[bjrogg]

Very intresting Tuomo.

I think I shoot that bow Bob showed and it was a very nice shooting bow.

What would work best scaled down to a 24” draw.

I would be interested in the best performing length and reflex shape for a 24” @ 50 lb natural material Selfbow.

I’ve only done one deflex handle with static recurves . It was more draw length than I could comfortably draw at 28” .

Would the static tips in my bow add to much mass ?

It’s kinda hard to wrap my head around all the mechanics of a bow. Something so simple. Yet so complex.

Thanks for sharing

Bjrogg
I’d like to know the best front profile for it too.

[RyanY]

If we isolate the variables and assume that these bows are all evenly strained and 100% efficient then wouldn’t the difference come down to energy storage alone? I’m not sure what would make the difference just based on the unbraced profile. It’s hard to separate the real world effects that were used to making assumptions about such as strain, limb mass, hysteresis, limb vibration, etc.

[Tuomo]

Simk: How do you optimize the front profile, and especially the tapers? The front profile is more of a practical problem, because in the model you can make it as narrow as you want, but in real life stability becomes a real concern. That’s the main reason I wanted to use the same front profile for every bow model.
As for the taper rate, I optimized it with strain distribution in mind. What would be a better approach? The 0.008 taper I used seemed to work well for every model.

Ryan: Yes, energy storage is mainly determined by the unbraced profile. That’s exactly what I wanted to examine, and of course the results of that difference also. All the “bows” had the same net reflex, yet they still behaved very differently.

[Selfbowman]

The design I’ve been using for years is closest to the third design. The speeds are close to what I get most of the time. I have had upward to 186 ft per second. Not often though. Interesting topic thanks for sharing. Notice the farther the tips get in front of the back of the bow the more speed. That’s with the deflex in the handle to help the stress on the inner limb.

[willie]

As for the taper rate, I optimized it with strain distribution in mind. What would be a better approach?

Model exceptional performing bows in virtualbow to find what the best strain distribution looks like.
Use this distribution model to test different degrees of deflex/reflex, different tapers and lengths etc.

[Tuomo]

Nice idea, Willie, and definitely worth trying. But I see many problems as well. How can I be sure that strain distribution is the main feature of an exceptional bow? Why not mass or curvature distribution, or the side profile combined with some other parameter?

I don’t have an exceptional bow—just some good ones. Their common features are that they have minimal set and they are deflex-reflexed. Set is a material-related parameter, and when making wooden bows, minimizing set is only possible when strain is distributed as evenly as possible. I’m going to make some bamboo-horn bows in the near future, so with horn the set parameter is almost eliminated. Simk has made a lot of very fast horn-belly bows and has taken very good and reliable measurements of their performance (see Traditionell Bogenschiessen 117), so there is potential. But again, this is how we implement theory in a real-life bow.

So, I’m going to make deflex-reflex (or even deflex-reflex-recurve) bows with evenly distributed strain and test them. A long to-do list…