stick testing

[cool hand luke]

A question and a proposal

1. Is it better to test the sticks at the purchased length (36/48") and then calculate down to the needed length, or cut them to a uniform length that is very close to the end length (like 12") and test them. I think the second method will get you a better single finger push down test result that is more applicable to the way the piece will be used in the build. This dawned on me when I had a couple pieces with a slight bow to them at the full length, and got horrible results, but when paired down they got much better. (yes I understand that 12" should be 9x as stiff as 36", they were averaging 18 times as much)

The drawback would be weighing them at the cut down length would lose a lot of accuracy. I only have a scale down to .1 grams. Does anyone know where you can get a cheap one that is more accurate?
this could be remedied by weighing the whole stick and then diving by 3 or 4.

2. proposal, fill out the table Balsaman started for buckling strength vs. stick weight. we could all combine our data here to fill in the blanks. I'll show you mine if you show me yours!

[Balsa Man]

Give da Man a gold star!


If we can compile data for everyone to access, that would be so cool and helpful; a fantastic learning and teaching tool! (Isn't that one of the important things S.O., and the Scioly forums are all about?!) I have, among other things, a data set from 75 48" 5/32" sticks (in the density range from 6.2 to 8.1 lbs/cf), the lumber pile we bought last year to get sticks for Nationals (and more data than shows on the graph I posted on 1/8" x 36"). By making...compiled data available, it allows everyone to make informed/rational/scientific decisions on design. Beyond the data that's been shared on Scioly, the only real data set (as in values over a significant density range) on E vs density for balsa is the old Forest Service study I, a few years ago, (and Crtomir last season) posted a link to. As I've... preached about for a while, that last increment in competitiveness lies in finding/using the very lightest sticks that just provide your 'design strength'; a statistical problem. Having a database with ...a large "n" (# of samples) provides a way to....focus what you buy/have to buy to ...get good sticks to ...a reasonable number of sticks within a fairly narrow density range. And for those that re just looking for the path to...do pretty well, it provides a "safe" density " range to buy/work from. With just size, length, weight, and BS measurement on... what balsa exists in, nobody is...giving any secrets to their competition; but everyone gets real, scientific information to base design evaluations/decisions on. Everyone can use the data set to do their...own engineering.

Question is, how do we, collectively, make it work; make it into the valuable tool it would be? The data needed is cross section size, length tested at, SFPD scale measurement, and maybe a way to deal with/differentiate between ... pretty darn straight sticks, and sticks with... observable bow ( could be just a "B" or "S", could be reporting average of strength when it bows in the way its already bent and strength when you push the bow the opposite direction- what we do with visibly bowing sticks). Important that testing be done ...consistently- one finger on top, pushing straight down, measured weight/force in gr where its holding a slight bow. Yeah, there's going to be some error/variation from...individual technique, but the larger the data set, the better handle we get. (Cross sectional) sizes? I would argue anything you test at/with. For towers, 1/8 is...sort of the default standard, but/and as I've posted about, 5/32 will provide better structural efficiency than 1/8, and ...fairly light to really light 5/32 falls into the strength range needed for towers. BS of bracing pieces matters too- again, getting to needed strength with least weight, so 1/16 square, and 1/16 x 1/32 (maybe 1/16 x 1/64?) would...make sense/be useful.

The ...brute force, easiest to manage the development of a combined data set, but most cumbersome to access and use is just everyone posting a Google sheet with data to be shared, and everyone wanting to see/use plows through what gets posted. I'm pretty confident we can come up with a way to make posting data, and accessing the compiling dataset easy workable, but I don't have the technical knowledge to...help that happen; anyone have a handle on that; ideas? I'll check with my youngest son, an app developer/programmer for ideas, and provide any thoughts I get.

Oh, yeah, the answer to your question on length to test at, YES, testing at just over installed length IS the way to go. As you've seen, and as... one might think from the oft discussed variability in balsa, different sections of a stick are pretty darn likely to have different buckling strength/"E" (the modulus of elasticity). We went w/ 48" so we could cut 2 sticks from each stick, and data shows the differences within a longer stick. Cut to lengths just a bit longer than B/C legs. That allowed (successfully) reducing needed safety factor down to 6%.

I really hope we can make this happen!

[Balsa Man]

Oh, scales-

Numerous on line sources; to 0.01gr; I got a small one 0-400gr w/ 0.01 precision for ~$16

[cool hand luke]

I'm not tech savvy enough to know an automated way of doing it. However I've done stuff like this with car racing data (many many times more data)
By creating a spreadsheet template. Everyone can enter there data and then post it here then you can copy and pasting or we could have a sheet here we load it into.

Ill take a crack at the spreadsheet this weekend and then post it here to make sure it works correctly.