Designs

[Balsa Man]

Ah, at any level of detail, when discussing some concept, there is almost always the next, deeper level of detail. I totally agree with SLM that splinting (hence significantly stiffening) a leg that breaks in the testing scenario I discussed could result in that leg seeing, and carrying more load/force in re-test. And that could change the forces on the other legs to some extent. The extent to which that would be a.....problem (as in really mis-lead you as to what weight/strength leg material will work) would, to some extent, depend on how wide the range of weight/strength was in the 4 leg weights you used for testing- if the lightest was really floppy (as in not close to strong enough), and the heaviest really stiff (as in way too much), stiffening the floppy one (after it broke) with one or two lamination strips so it was now way stiff would throw off the .....overall structural dynamics somewhat. If the range of stiffness/strength is.....reasonable, the effect, I believe, will be fairly small. Is the technique perfect; going to get you within....2.47% of minimum weight that will hold, no. But is it a lot better than shooting in the dark and guessing- believe it is. Its a tool, not a magic bullet.

Question, though, on the result; the effect, would be toward over-engineering, or under-engineering? If the broken, now seriously stifened leg did see/carry (slightly)more force, then the other three would be seeing a bit less - yes? So if the next (stiffest) leg was going to break at ....a 13.5kg tower load, it might take 13.6...or 13.7- i.e., you'd think that it was slightly stronger than it would be without the one stiff leg effect, so if you went with that weight, you'd be under-engineering, I think.

[fishman100]

Would having a tower with a chimney with slanted legs be better than a chimney with straight legs?

[SLM]

Question, though, on the result; the effect, would be toward over-engineering, or under-engineering?

I assumed that once you are done re-testing the tower with the stiffened member (say leg), then you will built a new (competition) tower using the revised section for all the legs. This does result in an over-designed tower since the legs, now having the same stiffness, carry the same force which would be less than the force that was being carried by the re-built member.

[Balsa Man]

Question, though, on the result; the effect, would be toward over-engineering, or under-engineering?

I assumed that once you are done re-testing the tower with the stiffened member (say leg), then you will built a new (competition) tower using the revised section for all the legs. This does result in an over-designed tower since the legs, now having the same stiffness, carry the same force which would be less than the force that was being carried by the re-built member.

Sorry, but I’m still not following your conclusion. Maybe I’m missing something. Let me run through what I’m thinking in more detail.

What I’m suggesting, to put some made-up numbers to it, is – using the same (cross) section fpr legs in a test build; just varying the weight/density of the legs. Let’s use the example of a C-tower chimney-55cm long. 4 legs, with, let’s say a degree or 2 of lean-in. At full tower load (15kg), each will see (assuming symmetrical construction) 3.75kg. Its actually a tiny bit more than 3.75, but even at a 2 degree lean the additional force is really small; like 0.000x more. Say you use one leg at 1.3gr, one at 1.4gr, one at 1.5gr, and one at 1.6gr. Bracing interval (between ladders) = 11cm. So you load, and at 11.6kg, the 1.3gr leg (the lightest/weakest) buckles between two sets of ladders. This is telling you that at 11cm exposed length, you shouldn’t expect 1.3gr legs to carry full load. You laminate splint strips on it- fixing the break – those strips should be most of the length of the exposed section-oh, 8 or 9cm long. You also put them on the other four exposed sections. That leg is, of course, now much heavier and much stiffer than it was; with the lamination I would presume its significantly stiffer than the 1.6gr leg. Now, you re-load the tower. If it carries 15kg, that’s indicating that 1.4gr legs are enough- or more than enough - to carry full. You could load beyond 15kg to get a read on how much over needed strength 1.4s would be, or you could “go conservative” and build the next tower with 1.4s and have pretty good confidence that those 1.4s will carry full. Say, though, on second loading it goes to 13.9kg, and the 1.4gr leg fails. You do the same sort of fix the break and stiffen the rest of the leg gig as before, and run a 3rd test. This time it goes full load. Now you’ve got an indication that 1.5s are enough, or more than enough – again, you could overload beyond 15kg to get a read on how far over needed strength you. On the next tower, you could go conservative and use 1.5s, you could go to 1.45s and see.

In your initial comment, I understood you to be saying that because the repaired and reinforced leg is so much stiffer, it carries significantly, or at least measurably, more than 3.75kg – yes? I’m not saying it won’t, but on the other hand, I don’t understand at this point why it would. I do believe/think that if this is the case, the amount is pretty darn small. So, for discussion, let’s put a conservative, made-up number to it; say it carries 3.78 (instead of 3.75) in the second test. That would mean the others would be carrying an average of 3.74 ea. Or, let’s say because it is the weakest, the 1.4gr leg sees 3.72 (though, again, I’m not sure why that would be the case). Let’s look at both possibilities.

You go to full load in the second test; and you conclude that the 1.4gr leg is sufficient or more than sufficient; i.e., it will carry 3.750kg or more.. And let’s say, for argument’s sake, that you were right at the edge with the tower carrying 15kg- at 15.01kg, the 1.4gr leg fails. So, if in test 2 the non-stiffened legs – in this case, the 1.4gr leg - was seeing 3.72 kg at a 15kg load, it would be seeing 3.7225 at 15.01kg. (the extra 0.01kg tower load/4 = +0.0025kg/leg) Or, alternatively, let’s say it was seeing 3.74kg at a 15kg load, which would go to 3.7425 at a 15.01kg load. So, you build the next tower with all 1.4gr legs. At a 15kg tower load, they’re all seeing 3.750x kg. Depending on how big the “weight transfer” effect to the stiffest leg was in the second test, our 1.4s are under-strength; will fail at 3.7225, or 3.7425. A tiny amount, for sure (we’re into the range of discussing how many angels can dance on the head of a pin) but they’re under, not over-strength- under engineered, not over engineered. The bigger the “weight transfer” to the stiffest leg effect might actually be, the bigger the under-engineering result. If I’m missing something in this analysis (and I certainly could be- its early in the morning), I’d love to know what it is.

Most importantly, I believe, the range of difference – of over-or under-engineering – that could be driven by weight transfer to a stiffened/much stiffer leg in a test with a leg set covering a range of densities is very small (as in an order of magnitude smaller) compared to the range of actual strength/stiffness within a set of legs matched in density/weight. As has been discussed many times before, weight/density is simply a good (but imprecise) indicator of strength/stiffness- for a given cross-section, denser/heavier wood will on average be stronger-stiffer. The natural variability of wood means no two pieces will be identical in properties. In Euler’s buckling equation, it’s “E”- the modulus of elasticity that is the property that actually matters. Actually testing for E would require a very difficult set up, it would have to be taken to failure for precise results, and in the end would only show you that for a given cross-section, at a given density, there is variation by some percent around the average in individual pieces. That range – from the literature, and from discussions on this board - seems to be on the order of 10-15%. This is a lot (significantly) more than the 1-ish percent potential difference/bias that might come in with one or more “over-stiffened” legs in the testing scenario that started this discussion.

So, what I’m saying is simply that a) testing as I’ve been discussing provides an objective, measured way to get to a density value (something that you can easily measure) that will work, and b) because wood is wood you have to apply a safety factor to get that density value to work consistently. This approach will, I believe, get you a lot closer to “optimum”, and get you there a lot quicker and more easily, than guessing/shooting in the dark, and building and breaking a whole lot of towers.

Would having a tower with a chimney with slanted legs be better than a chimney with straight legs?

As alway, just IMHO. This has been discussed before - check back through earlier posts, this year and last. My take is, yes it would be "better." The value is in stability of the structure. The difference in leg loads is for all practical purposes - at least within 2 or even 3 degrees - is nothing. It is what happens when you get some bucket swing, or if the test base isn't perfectly level, or if your build is not perfectly symmetrical (there's some lean in the chimney), or your load block is not perfectly centered - or, most likely, a combination, to some extent of all of these factors, that some lean-in helps with. It gives you a safety factor. Each of these "off-center loading" possibilities put disproportionate load on one or more legs; if they're close to the edge/to the limit with a centered load, they get more load than they can carry, and they break. Rather than trying to explain the math, you can get a feel for this by Googling up the Johns Hopkins Bridge designer app - "jhu bridge designer" It's a little 2-D program that lets you see/calculate loads/forces. Do a triangle with load from the top; first with the load straight down, then with the load at an angle, but inside the angle of the legs, then with the load outside the angle of the legs. When that load is at more of an angle than the the leg is angled, force in the leg goes up quickly. Try the same with a straight-sided structure (you'll have to play with it to get it to analyze a pair of parallel legs) - any load other than straight down is "outside" the angle of the legs- is rapidly, as it gets off-center, increasing load on one leg.

[WOLFPACK]

WHAT DESIGN WON THE NATIONALS LAST YEAR? DOES ANYONE HAVE A PICTURE OF THE WINNING TOWER?

WOLFPACK

[Littleboy]

WHAT DESIGN WON THE NATIONALS LAST YEAR? DOES ANYONE HAVE A PICTURE OF THE WINNING TOWER?

WOLFPACK

yes sir http://scioly.org/wiki/The_Best_of_2011#Towers

[chalker]

WHAT DESIGN WON THE NATIONALS LAST YEAR? DOES ANYONE HAVE A PICTURE OF THE WINNING TOWER?

WOLFPACK

Please don't type in all caps.

[coolio12]

ok so for this year i have already began testing my towers and if i do say so myself they are turning out very well

Tower 1:
height: 68.4 cm
kg held: 15 kg
tower weight: 9.43 grams

Tower 2:
Height:69.6 cm
Kg held: 15 kg
Tower weight: 6.96 grams

[coolio12]

Would having a tower with a chimney with slanted legs be better than a chimney with straight legs?

in my experience i find the tower to be more efficient when i add a minor slant to the chimney
both of the results from my towers( posted above) had a chimney with about a 2 degree inward slant.
i find it to make the tower more stable and less likely to fall

[mrsteven]

ok so for this year i have already began testing my towers and if i do say so myself they are turning out very well

Tower 1:
height: 68.4 grams
kg held: 15 kg
tower weight: 9.43 grams

Tower 2:
Height:69.6 grams
Kg held: 15 kg
Tower weight: 6.96 grams

I love how height can be measured in grams teach me thy black magic lol
anyway, has anyone tried making curved legs? I'm trying them and I want to see if my results are typical or if I'm just a failure haha