Towers B/C

[Random Human]

Balsa Man... I hope people appreciate the work you do! Your research shows hours of testing, stuff I always wanted to do but never had time to. Sometimes hard to follow but I love the conclusion. Because of years of hands on building with different sets of students I can see your data follows our experience of test and rebuild and test again.

In 2012 we used the 3/23 sq Bass for the base legs as we kept breaking the 1/8 sq balsa. In the top we used 3/32 balsa which worked very well. This year we used 3/32 sq bass for legs but the mass of the tower came in just over 13 grams, too heavy. I like to have new students build a little heavy to get some build experience and an ego rush from success . From there we then start to bring the mass down. Different cross section or density.

So from the lasted data posted, I see legs - 1/8 x 1/8 sq, 10-12 # density balsa legs with bracing intersections at about 12 cm (5 inches).. Also, I know from past posts about how important bracing is, I'm interested in your opinion about which pattern you like best. 'X' or 'V' bracing.

Thanks, Baker.
I always enjoy your insights. Its a pleasure to share the.....basic data we've accumulated; helps everybody up their game, and do some real engineering, and less shooting in the dark.

On legs, 10-12#/cf translates, for 1/8x1/8 to 36" stick weights between about 1.45 and 1.75gr. At 1/5 bracing intervals for a C tower meeting 29cm circle specs (12.27cm which is I assume your "about 12cm), you're definitely in the right ballpark. Specifically, a 1.52gr (to the light end of the 10-12#/cf range stick calculates out to 4572gr buckling strength at 12.27 braced intervals (which is force on each leg, +20% safety factor, for a 15kg tower load).

On bracing, see attachment on my post on October 21 (pg 10 of this thread); provides a pretty detailed discussion of the "Xs and ladders" we use- your middle figure. What's different from most in implementing this pattern is the use of 1/16th wide 1/64th" strips, put on pre-tensioned, for the Xs. I'm not 100% convinced its the very lightest way to go, but it works......

On the 13gr 3/32nds bass tower, that weight sounds about right. With 1.52gr/36" legs, ladders and Xs, weight calculates out to a tad under 10 grams. We'll see, one of these days, if reality aligns with all the calculation.

Len Joeris
Fort Collins, CO

Time for a Balsa Man reality check....
Built for bonus, 5 cm square at the top and 22 cm square at base
Legs -1/8 x 1/8 inch, 12#/cf balsa, 1.27 to 1.51 grams / leg (yeah, one was stripped a little heavier)
Ladders - 1/8 x 1/8 inch, 7#/cf balsa, 0.73 to 0.90 grams / side
X bracing - 1/64 x 1/16 inch, 8#/cf balsa, 0.35 to .53 grams / side
ladder intersections at 12 cm
After glue and some extra bracing at the top, final mass was 10.82 grams. Carried the full 15000 grams with no problems. With bonus final score 1571.16 I think being a little more careful and consistent with material weights we could have brought the mass to just below 10 grams. Will probable try 10#/cf for the legs next. All materials were stripped from 4 inch wide balsa and there was still variations in #/cf pieces. Darn close to Balsa Man's ball park guess... I'm just impressed that the legs held as well as they did.

When building each tower, is it extremely important to use the type of stick with the correct grain. (ie legs with c grain for comperession sticks) (b grain for bracing and ladders)

[BananaPirate]

Thanks, Baker.
I always enjoy your insights. Its a pleasure to share the.....basic data we've accumulated; helps everybody up their game, and do some real engineering, and less shooting in the dark.

On legs, 10-12#/cf translates, for 1/8x1/8 to 36" stick weights between about 1.45 and 1.75gr. At 1/5 bracing intervals for a C tower meeting 29cm circle specs (12.27cm which is I assume your "about 12cm), you're definitely in the right ballpark. Specifically, a 1.52gr (to the light end of the 10-12#/cf range stick calculates out to 4572gr buckling strength at 12.27 braced intervals (which is force on each leg, +20% safety factor, for a 15kg tower load).

On bracing, see attachment on my post on October 21 (pg 10 of this thread); provides a pretty detailed discussion of the "Xs and ladders" we use- your middle figure. What's different from most in implementing this pattern is the use of 1/16th wide 1/64th" strips, put on pre-tensioned, for the Xs. I'm not 100% convinced its the very lightest way to go, but it works......

On the 13gr 3/32nds bass tower, that weight sounds about right. With 1.52gr/36" legs, ladders and Xs, weight calculates out to a tad under 10 grams. We'll see, one of these days, if reality aligns with all the calculation.

Len Joeris
Fort Collins, CO

Time for a Balsa Man reality check....
Built for bonus, 5 cm square at the top and 22 cm square at base
Legs -1/8 x 1/8 inch, 12#/cf balsa, 1.27 to 1.51 grams / leg (yeah, one was stripped a little heavier)
Ladders - 1/8 x 1/8 inch, 7#/cf balsa, 0.73 to 0.90 grams / side
X bracing - 1/64 x 1/16 inch, 8#/cf balsa, 0.35 to .53 grams / side
ladder intersections at 12 cm
After glue and some extra bracing at the top, final mass was 10.82 grams. Carried the full 15000 grams with no problems. With bonus final score 1571.16 I think being a little more careful and consistent with material weights we could have brought the mass to just below 10 grams. Will probable try 10#/cf for the legs next. All materials were stripped from 4 inch wide balsa and there was still variations in #/cf pieces. Darn close to Balsa Man's ball park guess... I'm just impressed that the legs held as well as they did.

When building each tower, is it extremely important to use the type of stick with the correct grain. (ie legs with c grain for comperession sticks) (b grain for bracing and ladders)

If I test for something like buckling strength, don't I also test for whether the grain will be good enough for the legs of a tower since I'm applying a compression force anyways (like the C grain pieces will naturally test higher)? I'm not that good with identifying grains, especially A and b..If someone has any tips that would be great!

Also, I'm a little confused on what #/cf means?

[bernard]

Also, I'm a little confused on what #/cf means?

Pounds per cubic foot.

[baker]

Time for a Balsa Man reality check....
Built for bonus, 5 cm square at the top and 22 cm square at base
Legs -1/8 x 1/8 inch, 12#/cf balsa, 1.27 to 1.51 grams / leg (yeah, one was stripped a little heavier)
Ladders - 1/8 x 1/8 inch, 7#/cf balsa, 0.73 to 0.90 grams / side
X bracing - 1/64 x 1/16 inch, 8#/cf balsa, 0.35 to .53 grams / side
ladder intersections at 12 cm
After glue and some extra bracing at the top, final mass was 10.82 grams. Carried the full 15000 grams with no problems. With bonus final score 1571.16 I think being a little more careful and consistent with material weights we could have brought the mass to just below 10 grams. Will probable try 10#/cf for the legs next. All materials were stripped from 4 inch wide balsa and there was still variations in #/cf pieces. Darn close to Balsa Man's ball park guess... I'm just impressed that the legs held as well as they did.

When building each tower, is it extremely important to use the type of stick with the correct grain. (ie legs with c grain for comperession sticks) (b grain for bracing and ladders)

If I test for something like buckling strength, don't I also test for whether the grain will be good enough for the legs of a tower since I'm applying a compression force anyways (like the C grain pieces will naturally test higher)? I'm not that good with identifying grains, especially A and b..If someone has any tips that would be great!

Also, I'm a little confused on what #/cf means?

More than you may want to know about Balsa... http://www.go-cl.se/balsa.html

[BananaPirate]

Ah thanks, so it's just another form of density then, makes sense.

[Random Human]

Is it absolutely necessary to use the right type of grain balsa for your tower? Obviously it will be very difficult to get because there are no sellers offering a selection of both density and grain? Will having the right grain of your tower (c grain for legs) make a huge difference in the results? Is it very necessary?

[Balsa Man]

Time for a Balsa Man reality check....
Built for bonus, 5 cm square at the top and 22 cm square at base
Legs -1/8 x 1/8 inch, 12#/cf balsa, 1.27 to 1.51 grams / leg (yeah, one was stripped a little heavier)
Ladders - 1/8 x 1/8 inch, 7#/cf balsa, 0.73 to 0.90 grams / side
X bracing - 1/64 x 1/16 inch, 8#/cf balsa, 0.35 to .53 grams / side
ladder intersections at 12 cm
After glue and some extra bracing at the top, final mass was 10.82 grams. Carried the full 15000 grams with no problems. With bonus final score 1571.16 I think being a little more careful and consistent with material weights we could have brought the mass to just below 10 grams. Will probable try 10#/cf for the legs next. All materials were stripped from 4 inch wide balsa and there was still variations in #/cf pieces. Darn close to Balsa Man's ball park guess... I'm just impressed that the legs held as well as they did.

When building each tower, is it extremely important to use the type of stick with the correct grain. (ie legs with c grain for comperession sticks) (b grain for bracing and ladders)

If I test for something like buckling strength, don't I also test for whether the grain will be good enough for the legs of a tower since I'm applying a compression force anyways (like the C grain pieces will naturally test higher)? I'm not that good with identifying grains, especially A and b..If someone has any tips that would be great!

Also, I'm a little confused on what #/cf means?

Thanks for the feedback, Baker.
Ain’t engineering cool?

I do believe from our analysis that for a C-Div tower, bracing at 1/5th intervals (as you’re doing) is the optimal configuration (tradeoff between stiffer legs & longer bracing interval, and floppier legs with shorter bracing interval). Moving to a bit lighter leg wood is indeed the path to a better score. I think you understand, but just to make sure; the guiding parameter in selecting leg wood going forward should be the buckling strength. Density numbers you’ve established/verified get you in the ballpark- from this test result, from just under 1.3 to 1.5 gr/36” (which translates to 8.80 – 10.16 #/cf - BTW, BananaPirate, #/cf is pounds per cubic foot, one way of expressing density). Now the game is looking at sticks in that range (focused on the lighter end of the range) and finding the lightest sticks that have sufficient buckling strength.

I find it very interesting that the (at least one) leg at 1.27gr (which I assume is 1.27gr 36” stick weight, which would yield a leg weight of about 0.85gr = total leg weight of 3.4gr) held to full load. With 4 legs at this density, and with the same buckling strength of that one, you’ll see a significant cut to tower weight.

This result provides data that speaks to an important open question about measuring buckling strength of a 36” stick, and applying the results to wood selection (see the “Measuring/using buckling strength-new information” thread).

The open question is about what is the appropriate factor to multiply the buckling strength measurement you get doing a “single finger push down test” on a 36” stick by, to get a 36” buckling strength value you then calculate braced interval/braced leg buckling strength from. As discussed in the “measuring/using” thread, we’ve narrowed down the factor to between 2.0 and 2.3. Simplified just a bit from the full discussion, it depends on the “end conditions” of the stick undergoing buckling strength testing (and what the end conditions for each of the shorter ‘stacked columns’ you end up with in a leg when you brace at points along the legs). The end conditions define an “effective length factor”, which gets us to the 2.0 to 2.3 times factor.

Doing single finger push down, you have “pinned”, also referred to sometimes as “hinged” end conditions at both ends (‘pinned/pinned’). If a stick with a measured pinned/pinned 36” buckling strength of “X” were to be tested with both ends “fixed” (which means the ends can’t move, can’t rotate- like you glued one end of the stick onto the scale, and glued the other/upper end onto a stiff plate, parallel to the surface the scale is sitting on) and you were able to do the pushdown on that upper plate in a way it stayed parallel to the base surface, the math/the theory says you’d see a buckling strength about 2.3 times “X.” If you did one end fixed, and the other end pinned, math/theory says you’d see a buckling strength of about 2.0 x “X” The open question is, what end conditions (in each of the braced segments of the legs) are actually getting created by/at the braced points along the leg?

To see what this means in terms of how light can we go on leg wood and still have a tower hold close to full load, let’s look at some numbers:

The ‘design load’ (which is the force on each of the 4 legs- in a C-div tower meeting 29cm circle bonus, at 15kg tower loading), is 3,810gr. Looking at that same C-div/29cm circle tower, if you’re using a 1/5th bracing interval, which it sounds like you are- “ladder intersections at 12cm” – leg length should be about 61.34cm; 1/5th of that is about 12.27cm (61.43/5); 12.27 is 0.134 of 36”(which is 91.6cm) (12.27/91.6=0.134); from our ‘inverse square table’, 1 over 0.134 squared = 55.74. That tells us that if a 36” stick has a buckling strength of 68.33gr, a 12.27 piece (braced section) of that leg will have a buckling strength of 3,810gr (68.33 x 55.74)- just barely able to carry our design load (if everything were perfect).

This relationship (that if the 91.6cm (36”) buckling strength is “X”, then the buckling strength of a piece of the same material 12.27cm long is 55.74 x “X”) is correct if the end conditions (hence ‘effective length’ are the same in the long (full 36” stick) and shorter (12.27cm segment). Our 1 finger push-down testing is giving us buckling strength for pinned/pinned end conditions. We know that the end conditions of braced segments are somewhere between fixed/fixed, and fixed/pinned. Applying effective length factors, if the end conditions of leg segments in the tower are truly fixed/fixed, then their buckling strength will be about 2.3x what it would be for a segment with pinned/pinned end conditions. If the end conditions of leg segments in a tower are pinned/fixed, then their buckling strength will be about 2.0x what it would be for a segment with pinned/pinned end conditions.

So, finally getting the end point- what matters. You test a 36” stick (1 finger push-down = pinned/pinned end conditions); you get a measurement of, let’s say, 32gr. If the 2.0 factor is correct (pinned/fixed end conditions in leg segments), its 36” buckling strength under pinned/fixed end conditions is 64gr (32 x 2.0 = 64) which is only 94% of our design strength of 68.33); however, if the 2.3 factor is correct (fixed/fixed end conditions in leg segments), its 36” buckling strength under fixed/fixed end conditions is 73.6gr (32 x 2.3) which is 108% of our design strength. So, ignoring possible variations within the stick (which are actually likely to some extent), if it tests at 32gr, and the 2.0 factor is correct, it should not be able to carry the design load, hence the tower won’t quite be able to carry 15kg. But, if the 2.3 factor is correct, it should carry design load, and has a +8% safety factor, which may very well cover variations in strength in the wood.

Looking at a plot of our data (for about 150 sticks at this point), showing stick weights vs measured 1 finger (=pinned/pinned end conditions) buckling strength, the trend line shows, at a stick weight of 1.26, a buckling strength of 32 gr. We don’t (yet) have, and haven’t tested, any sticks between 1.1 and 1.33gr, but looking at the variation around the trend line for stick weights we have tested suggests that finding sticks as light as 1.2, maybe as little as 1.15/1.16, with a 36” pinned/pinned buckling strength of 32gr is quite possible, and that suggests a tower weight a bit under 9 grams is quite doable.

Given all this, I’m… not sure what to make of Clueless Builder’s post in the “Pictures, Videos and Scores” thread, reporting a 6.5gr C-Div tower holding full load. On the one hand, I’m always amazed each year at what the very best, one or two best in the Country, come up with; I have no idea how almost 2 ½ grams could be cut from a 9gr tower; I’m really curious what leg stick weights, and bracing intervals are being used to get to this level; I must be missing some important technical insight/understanding, because I can’t see or figure out an engineering explanation.

One last interesting thing in the data Baker provides- the #/cf numbers vs stick weights. I put together a table years ago converting pounds per cubic feet to 36” stick weights for various stick sizes. For 1/8” sticks, it comes out as follows:
0.7gr stick = 4.74#/cf
0.8gr stick = 5.42#/cf
0.9gr stick = 6.10#/cf
1.0gr stick = 6.77#/cf
1.1gr stick = 7.45#/cf
1.2gr stick = 8.13#/cf
1.3gr stick = 8.80#/cf
1.4gr stick = 9.48#/cf
1.5gr stick = 10.16#/cf
1.6gr stick = 10.84#/cf
1.7gr stick = 11.51#/cf

The stick weights you’re seeing/reporting at various #/cf densities are significantly lighter (e.g., 7#/cf w/ stick weights 0.73-0.90) than the calculated stick weight in gr from density in #/cf in my table. I’ve double-checked my calcs/conversion factors, and have no idea where this discrepancy is coming from, so a question. Using the 7#/cf example, a 36” stick at 7#/cf should weigh about 1.02gr, and in a sheet 4” x 36”, there would be 32 1/8 x 1/8 sticks; and 32 x 1.02=32.64gr sheet weight. With reported stick weights 0.73 to 0.9, using an average of 0.81gr, 0.81 x 32 = 25.92gr sheet weight. Is your sheet weight closer to 33gr or 26gr??

Last, re: Random Human, and BananaPirate’s comments/questions on grain types. A few comments to help clarify this subject. There is a good explanation of A, B, and C grain types/patterns on Specialized Balsa’s website:
http://www.specializedbalsa.com/balsa_g ... cation.php

A few key insights-
Grain type really is about/has to do with/applies to balsa sheets, not square cross section sticks.

The grain patterns seen on sheet surfaces come from the orientation/alignment of the sheet relative to the concentric rings of grain structure you see looking at the end of a balsa log.

The orientation of a sheet, and its resultant grain type, does affect the stiffness of a sheet, across the width of a sheet; e.g, how easily it can be rolled into a tube.

There is no such thing as A, B, and C sticks. You can have A-C sticks (where 2 opposite faces show A-grain, and the other2 opposite faces show C-grain), and you can have B-B sticks (where all faces show B-grain. There is no material/significant difference in buckling strength between A-C and B-B square cross section sticks.

Both legs and ladders, in a loaded tower….work under compression loading- actually axial compression loading (loading/force along their long axis). To work/not fail, they have to have sufficient buckling strength, because thin columns fail under axial compression loading by buckling. The legs are under axial compression load because of the load block sitting on their upper ends; the ladders are under axial compression loading when the legs they are in between start/try to start bowing/buckling inward toward the end of a ladder. X braces, in a ladder and Xs bracing configuration/system work under tensile loading- when the leg, where they’re attached tries/starts to bow outward, away from the end of a ladder/end of X brace.

[baker]

Balsa Man,
Sorry for the mis-leading info on the leg weights. The 1.2 to 1.5 gm was the cut length of the legs, 60.5 cm long. The 36 inch weights varied from 1.7 to 2.3 grams. The heavier one was the one that was 1.52 gm cut to 60.5 cm. Cut lengths of legs were 1.20, 1.27, 1.23, 1.51 grams. Like I said we were cutting from sheeted material and each stick came out varied so we picked the 4 that were closest to each other and had no flaws. Now that we feel good about the design, next one will be a little more careful on wood selection.

[Balsa Man]

Balsa Man,
Sorry for the mis-leading info on the leg weights. The 1.2 to 1.5 gm was the cut length of the legs, 60.5 cm long. The 36 inch weights varied from 1.7 to 2.3 grams. The heavier one was the one that was 1.52 gm cut to 60.5 cm. Cut lengths of legs were 1.20, 1.27, 1.23, 1.51 grams. Like I said we were cutting from sheeted material and each stick came out varied so we picked the 4 that were closest to each other and had no flaws. Now that we feel good about the design, next one will be a little more careful on wood selection.

Thanks for the clarification, Baker.
So, total leg weight of 5.21 grams....
In our estimating sheet, we're currently running total leg weight at 4.07gr (based on 1.52 gr/36" sticks), so you're +1.14gr. With 4.07gr in legs, we have total tower (legs+ladders+Xs+glue) estimated at 8.85gr. Our ladders and Xs weights are close to your #s. So, 8.85 + 1.14 =9.99 (VERY close to your "just below 10 grams"; add a bit for extra glue and bracing, your 10.82 lines up pretty closely.

I do believe that legs with sufficient buckling strength from 1.27-1.3gr/36" sticks can be found (based on the belief that the 2.3 factor is correct. That means looking for sticks with a one finger push-down reading at/above 32gr. If so, that would reduce leg weight to 3.4gr (0.67 lighter than current 4.07gr). 8.85 - 0.67 = 8.18. We'll be doing a build to test this after the holidays, and I'll report results.Looking forward to hear how your weight reduction goes.
Merry Christmas!

[baker]

Balsa Man, thanks for your data. What glue are you using? We've been using Titebond II yellow wood glue. Students apply it with a tooth pick to both pieces to be joined and remove excess. It really soaks into the fiber of the wood and when there is a failure it rips the wood. But I wonder if it might be a little too heavy. I always felt that CA glue makes wood hard and brittle. I know it's a son of a gun to sand for final adjustments.

You know one thing I don't think we have ever addressed here... Final Adjustments. After Tower build we place a loading block with about 50 cm of chain on top of the tower and check how level the tower is.
- Does the block sit flat on top of the tower with no rocking. Figure out which leg is slightly longer and lightly sand with a flat sanding block. Do this a little at a time so as not to do too much because then you'll have to sand 3 legs.
- Does the Tower sit flat on the table without rocking. Here we use the chain to help to see if the center line is straight down the center. Again with the level on top of the loading block, make the necessary adjustment to make the tower level and firm to the table.
- This slight sanding on the top and bottom helps to make sure that the full surface of each leg is making full contact to the block and to the table.