Arvins 62" osage design

[willie]

Arvin mentioned some specs for a new osage flight bow. and as I am learning to work with Virtualbow along with learning about what makes a bow good for flight, I thought it would make for a  "design along" to post as I go.  Comments and questions and insights welcome.

Arvins design:
Single piece selfbow
62" NTN 
50# @ 25" 
7" brace height
1.5" flipped  tips   
8" between fades

Virtualbow can be downloaded from https://www.virtualbow.org/
https://www.virtualbow.org/files/user-manual/   explains the program in detail

When Virtualbow first opens, it presents with a multi pane window. The upper right pane shows the bow in Back view, side view, single limb view or with both limbs,  as selected by the buttons in the lower part of the pane. One can also zoom or rotate. The lower right pane will show plots when Items are highlighted/ selected in the left sidepane.

Below are some typical views when working with the program functions.

units_moe.jpg  Two features are demonstrated in the first screenshot.

1. Selecting the "options" dropdown in the upper left pops up the units dialog where you can choose your default units, either metric (SI) or (US).

2. In the upper half of the left hand sidebar, by highlighting materiels > new materiels, the lower section in the side bar presents text entry fields to set Rho  (materiel density) and E (stiffness).  Having the correct Rho value is not required for simple "static" design but will need to be set properly when generating "dynamic" outputs.

The stiffness of the materiel is neccessary for knowing the draweight of a design. It is set on the E line by clicking on and resetting the desired MOE (modulus of elasticity). It is has been unchanged in the example below, as it is close to the book value for osage. Actual bend testing of a ripping from the stave will determine a closer value and will be done before the Arvins design is finalized.

MOE is explained in more detail at https://www.wood-database.com/wood-articles/modulus-of-elasticity/  and book values for various woods can be found by searching in the database. From the wood database, one might find an potential Elastic Modulus or MOE of 1,350,000 lbf/in2. Clicking the existing E in the sidebar allows one to enter 1,350,000 directly. The program will display  in scientific notation once the value is entered.

The values from the database are obtained by averaging across many samples of various densities at 12% MC.  If one has a primo stave with good rings at a 6% MC and applies heat treatment on the belly, then using a book value in the program could likely indicate the need for a wider stave then neccesary.

dimensions.jpg  brace height and draw length are set in the left hand sidebar by
highlighting "dimensions". Enter your preferences in the lower sidebar.

thickness.jpg  Virtualbow is optimized for designing  laminate bows. Each laminate materiel can be defined as a "layer" with its own MOE and its own thicknesses.

In this designalong, Arvins selfbow being from a single materiel, the thickness profile will include the handle, fades and limbs all in a single layer

Highlight layer > newlayer. A table will appear in the lower sidebar. Entries in the lower table for a 31" limb are explained line by line as follows.

1. 0 is center of bow.  2 specs the depth of handle at center of bow in inches.

2. In the left hand column of this table, the distance out from the center of bow is given in percentage.  making the percentage entries in the tables to need some calculation. Keeping in mind that we are designing a 62" bow, and each limb will be 31", the second row of the table is calculated thus. For a 4-5/8" handle on the bow, the half length would be 2-5/16" on each limb. Divide 2-5/16"(2.31) by 31" to get .075 which is entered as 7.5  in the left box and the right box keeps the handle depth at 2".

13%  at .87" thick is an arbitrary point I chose for Arvins proposed handle ramp.  It helps set the  slope and shape of the ramp where the handle thichness transitions into the limb.
Choose a percentage that works best for you if needed. Raise or lower the thickness (.87") in the second box with the small arrows that appear when you double click inside the box. 
The 13 in the first box on the line can also be raised or lowered also with up/down arrows in it's box. Shape the ramp to your liking by adjusting both values.

22%  is where the thickness taper becomes fully working. 
Again, divide the distance out on the limb that you desire the limb to start fully working by the overall limb length. 6.8" / 31" = .22  or 22 in the table when entered as a percentage. .42" in the second box is a thickness you estimate for the beginning of the limb thickness taper.

The last row defines the the trial thickness at the nock.  100%  of  .31"

The actual limb length of 31" will be specified innthe table and plot whan  the "profile" line is highlighted as explained in the next post below.

f1width(1).jpg Illustrates setting limb widths. Highlighting "width" in the sidebar brings up a width plot and tables in the lowers sidebar for setting the width values in inches. Again,  points along the limb are again shown as percentages of the total limb length.  Arvins design calls for a 1.5" wide handle,  8" (13%) from the widest part of the fades to the" widest part of the fades, then straight for just under 7" (35%),  then into a willow leaf taper to 5/16" at the nock. Arvin mentioned 4" straight section before tapering, and that can be done also. A straight taper can be had too. It just makes for small changes of the location of where the maximum stress occurs in the limb.

[willie]

Defining the limb length and side profile

Virtualbow can use multiple segments to build out a limb. One can define each segment seperatly, both by length and shape. Once all segments of the limb are specified, the overall profile and total length of the limb are viewable. Segments can be straight sections, arcs, spirals or splines. Lengths entered in the lower sidebar tables and shown in the plot graphs are in inches.

length1.jpg  19" straight limb

Select profile > line in the left sidebar. The entry in the lower section of the sidebar has been set to a new length of 19"

length2.jpg  12" spiral section added to the 19" straight section.

To add a spiral segment, go to the upper left corner next to "comments" and click on the  left-most of four boxes and select"spiral" in the dropdown. Under the highlighted "profile" in the sidebar, a spiral entry will be added and a table shows in the lower sidebar.
The first row  defines the spiral segment as 12" long.
The second row defines a longer radius as the spiral slowly begins it's bend.
The third row defines a much shorter radius that ends the spiral.

Arvin wants a 1.5" flipped tip, so playing with the values of the two radii creates what I think his caul produces.  Left clicking on the plot area will bring up a context menu that lets one select a photo to overlay the plot on when building out the segments.

[willie]

Basic program outputs.

If you have gotten this far, its time to look at some basic outputs Virtualbow can generate. Clicking on the left hand green arrow at the top will generate various tables and graphs for your design.

Virtualbow uses the stiffness of the materiel entered earlier to calculate the draweight of your design.  If you can brace and pull a 2x4 stud to 36", you are a better man than I, and Virtualbow will output some fantastic draweight. However, the program does not attempt to inform you if your design will actually work out in pratice.
All materiels have a stress limit. In the case of a self bow we generally want to avoid stress above the level that causes set, and will need to obtain that info some other way.

f1width(2).jpg shows some basic characteristics of our design as numerical values. Of note is 51+# of pull @ our design draw of 25" and occuring at a maximum stress of 16,407 pounds per square inch. or roughly 16.5K psi.

f1width(3).jpg  selecting the shape tab displays the bend shape at  brace height. Not the  slider along the bottom that allows the output to display limb shape at any drawlength up to our stated full draw.

f1width(4).jpg the same design shown at full draw


Selecting the stress tab brings up a plot showing the stresses along the limb from the handle at the left to the tip at the right.
Moving the slider to full draw will let you see where along the limb  the maximum stresses occur and the amount of stress which can be read on the vertical scale at the left. 

f1width(.jpg shows the distribution and maximum stress at full draw. not sure how that smily crept into the file name

[willie]

Arvin has some pretty good ideas about limb width profiles. 
Different design limbs have their stress distrubuted  in different parts of the limb.
Virtualbow allows us to try different thicknesses, profiles and widths to allow us to anyalyze where different conditions occurs.

Does the width or thickness profile matter when shooting flight? What should we be looking for, presuming we can adjust thicknesses approiately?

[Selfbowman]

If I understand you correctly there is no width requirement in flight rules.

[Selfbowman]

This is going to be interesting. I see all this information and understand some. I will look at the pictures real close! The challenge for the designers is not only the design but getting the builder to understand it. 🤠🤠🤠request for information ( RFI) will be present I’m sure.🤠🤠

[willie]

If I understand you correctly there is no width requirement in flight rules.

No width requirements that I know of, but I have not looked at the rules lately.  Your stave might limit the width tho, and dictate some design decisions.

Most people have a back profile in mind and cut the width to a pyramid, semi-pyramid or more of a straight limb when they set out to build the bow and then tiller to bend profile by removing  thickness.

Edit: setting widths has been added to the first post.. 

With this program we can define a back shape and adjust the thicknesses to make it work, or vice versa, define a thickness taper and adjust a back proifile to suit.  I can post examples of any other design offered.

With Virtualbow, once a profile, either a back shape or a  thickness taper is tried, the other can be adjusted and the resulting stress plots drawieghts and bend profiles can be generated by the program.

So long as the stresses shown in the stress plots do not exceed the limits of the materiels, any configuration of limb can be built without set.

[sleek]

Something i didnt see mentioned is are the tips stiff or working? If stiff, how long is the stiff portion?

[willie]

Something i didnt see mentioned is are the tips stiff or working? If stiff, how long is the stiff portion?

Arvin said he wanted the last 3-4 inches to not bend. I dont see much bending in the screenshots just now added to reply2 

should they be stiffer?

To help see the tip stiffness, I modified the design to let the progran show a straight tip. They are kept at the same stiffness. The bow lost 3.5 lbs when the tips were straightened, so the inner limb was thickened slightly to bring the bow back up to the same draweight .

The maximum stress without the flipped tips is, 8% lower than with. The 8% reduction will also allow us to draw the bow 2 inches more if one wants to bring the bow up to the original stress level. Straightening the tips makes a noticable change in the force draw curve. It shows as a lower "hump" in the early draw portion, and a slightly noticable difference all the way out to 25"

f1width(9).jpg shows the force draw curve with 1.5" recurved tips.

f1width(14).jpg shows the force draw curve for the modified limb without the recurved tips.
 
f1width(12).jpg is the bend shape of the straightened limb,   

Note: I have edited this post and some posts above for completeness

[sleek]

How wide will it be out the fade, and at the last part of the bending limb before it gets stiff?

[willie]

sleek, I have added a screenshot and instructions for widths in the first post.  The yellow bow design  is used to generate the stress curves and bend shape posted in reply 2,

one can have the backshape either more pryamid or parrellel requiring less width, and by changing thicknesses, still be able to control where the limb will bend and consequently, be stressed the most. 

bending limb before it gets stiff

the transition from bending to stiff is gradual with most all bows, unless one sets out to deliberatly design a lever bow. 

A bigger question is how should the best part of the limb work? should one try to keep most of the bend  toward the inners? towards the mids? keep the inners and the mids working as much as possible? etc.
the screenshots shows the stress plot of a limb that works

mostly in the inner
inner and mid equally
mostly mid

[sleek]

I ran my formula with the numbers you gave me and the dimensions are exactly what they need to be for the poundage.

I have opinions on the profile of a bow, but being uncertain of their accuracy, ill only say what I firmly believe to be true. The more you stress the inner limb, the more torque you get, the more you bend the outer limb, the more horsepower you get. The limbs profile AND length needs to be built according to those goals. Longer gives more torque shorter more hp. Which, is a large reason my shirt bows have a hard time keeping up with Arvins long bows in shooting broadhead, but also why his long bows struggle against my short in flight.

[Selfbowman]

I ran my formula with the numbers you gave me and the dimensions are exactly what they need to be for the poundage.

I have opinions on the profile of a bow, but being uncertain of their accuracy, ill only say what I firmly believe to be true. The more you stress the inner limb, the more torque you get, the more you bend the outer limb, the more horsepower you get. The limbs profile AND length needs to be built according to those goals. Longer gives more torque shorter more hp. Which, is a large reason my shirt bows have a hard time keeping up with Arvins long bows in shooting broadhead, but also why his long bows struggle against my short in flight.

If agree Kevin.

[willie]

I ran my formula with the numbers you gave me and the dimensions are exactly what they need to be for the poundage.

I have opinions on the profile of a bow, but being uncertain of their accuracy, ill only say what I firmly believe to be true. The more you stress the inner limb, the more torque you get, the more you bend the outer limb, the more horsepower you get. The limbs profile AND length needs to be built according to those goals. Longer gives more torque shorter more hp. Which, is a large reason my shirt bows have a hard time keeping up with Arvins long bows in shooting broadhead, but also why his long bows struggle against my short in flight.

I agree about an agressive side profile stressing the limbs more, especially recurves that are getting the tips out ahead of the handle.

From playing with Virtualbow with various reflex and  deflex arrangements,  I am seeing more energy being stored in the inner halfs of the limbs and not as much as I had previously thought being added by bending the outers more.  That said I have just been playing with 62" bows and havent really spent a lot of time with longers designs - yet.

below are two graphs available in virtualbow that show the amount of curvature in the limb:
the first at brace and the second at full draw.  you can move the slider and watch the curvature change as you "draw" the bow.
Looking at the stress graphs sort of tells you the same thing,   Maybe the dedicated curvature graphs with its exaggerated scale helps some visualize better.  I can see the usefullness for someone who mainly builds glass laminates and tillers by tapering lams before glueup and doesnt often do a lot of sanding and trying to change the bend by eye after the fact
Using the shape view to see the changes in curvature after making a tweak to your model is not that easy, (for me anyways), especially if there is a lot of reflex or deflex



 

[Selfbowman]

🤠🤠🤠just a guess but less mass in thickness he fades from  ramps  to mid limb. Thin the tapper