Tower Base

[Jaimie]

Can you have a triangular base on top of the square base for division C? Example:https://i.stack.imgur.com/QtANH.jpg [img]<img%20src="https://i.stack.imgur.com/QtANH.jpg"%20 ... a%20square"/>[/img]
Thanks!

[Unome]

I guess you could, but it would be very unstable. Two squares is probably better.

[cool hand luke]

could you? yes, it would be legal.

Is it a good idea? No.

[Balsa Man]

I guess you could, but it would be very unstable. Two squares is probably better.

could you? yes, it would be legal.

Is it a good idea? No.

Its not a question of 'stability', and its beyond not being a good idea; its a bad idea.

First, let me say welcome to the towers forum. If you take some time to read through what's been posted (this year and last), you will learn/understand a lot, and as the season goes on, there will be much more useful info. Feel free to then ask questions, and you'll get useful answers.

That said, back to your question.....

Why is it a bad idea?
Because of the most basic fundamentals of how a loaded tower works.

Let's go back to last year's 'one part' towers; square base, much smaller square top; straight legs tapering in from wide base to narrow top (or. for discussion purposes, could be wide triangular base, small triangle at the top). Load block sits on the top square/triangle. When you load the bucket, the force put on it by gravity pulling the bucket down goes down through the legs to the test base surface. If the tower is symmetrical and the load block centered, in a 4 legged configuration, 1/4 of the load force goes down through each leg (in a 3 legged configuration, 1/3 of the force). This force, going down through/along the long axis of each leg is called an an axial force. At some level of axial force/loading, a stick/a leg will fail by buckling; it'll bow outward or inward, and break. If you put sufficient bracing between the legs, you will increase the buckling strength of the legs (by breaking them down into a set of shorter and much stronger 'stacked columns'). With sufficiently strong legs, and a sufficient number of bracing intervals, it will be strong enough to carry the full load.

In the 'two part' configuration required under this year's rules, the distribution, carrying, transmission of forces from load applied works exactly the same way- load applied on the top of each leg is carried down to/applied to the base. Yeah, the legs have a point at which they angle out more (at the 8cm circle plane (<20cm above the base), but this doesn't change the... basic physics at work.

What happens with the ... arrangement you show?
The upper triangular/3 legged section works just like a 'one piece' tower; load force is carried down the axis of the legs. That force is about 1/3 of the load on each leg- ~5kg at a 15kg bucket load. In the place in your drawing where one leg connects with/sits on top of the base section leg, it will be carried down the base section leg to the base. If the base section legs are properly braced, that leg will carry the load it sees. The other two upper section legs, each carrying a force of about 5kg, are sitting on a piece bridging between two legs- a 'ladder' between each pair of adjacent legs. Loaded that way, the ladder is much, much weaker than an axially loaded leg. One of the two ladders carrying the the two remaining upper section legs will snap at the point the leg meets it, under very little tower loading; depending on the density/strength, maybe 1 or 2kg, instead of 15kg.

Try this to see/feel: take a balsa stick, cut a piece off of it say 20cm long. Put it vertically on a solid surface, push straight down. As you push harder, at some point it will buckle and break. Cut another piece; bridge it between two supports. Push down on the top side toward one end (like where two of the legs in your drawing would be hitting/loading it), using a short piece of leg wood, held vertically - to create the 'point loading' of a leg (like in your drawing). With a small fraction of the force it took to break the axially loaded piece, the ladder/bridged piece will snap, If you use a scale in the first test, and 2 scales in the second test, you could measure the difference in force, but you can easily see/feel the difference; and its a BIG difference.

There are other.....secondary problems with the configuration you show, too. The load on the upper section is not centered on the lower section, so a disproportional amount of the top load will be put on the one lower leg where top section leg meets bottom section leg. Even if the ladders supporting the other two legs were strong enough to support them (i.e. really heavy/dense/strong), the disproportionally loaded leg would have to be much heavier/stronger than the other 3 legs to hold. Another inherent problem with a 3 leg configuration comes in when you try to brace them. With a 4 sided tower/tower section, the outside faces of two adjacent legs are parallel/in the same plane; a straight bracing piece will fully contact the faces of both legs, giving you a good, strong 'lap joint.' When you go to put a brace piece between two adjacent legs in a 3 leg/triangular tower/tower section, the brace piece will only contact edges/corners of the legs- with that tiny contact area, the joint will be massively weaker than the joint in a 4 leg setup.

Hope this helps

[Raleway]

Just a question for Balsa Man regarding this year's tower design. To address the bracing issue (and ignoring cost), would getting equilateral triangle style cut sticks solve the issue? I am a little confused about the leg loading issue you have presented as well. Past all of these woes that I can potentially see, my team had some success with a triangle tower. It was not terrible (averaging at around 1700 with bonus), but definitely not competitive. I was curious as to ask how loading block slippage would work if it ever maxed out as a square block does not fit well on an equilateral triangle top. Thanks for all of your amazing advice on these forums

[Unome]

Just a question for Balsa Man regarding this year's tower design. To address the bracing issue (and ignoring cost), would getting equilateral triangle style cut sticks solve the issue? I am a little confused about the leg loading issue you have presented as well. Past all of these woes that I can potentially see, my team had some success with a triangle tower. It was not terrible (averaging at around 1700 with bonus), but definitely not competitive. I was curious as to ask how loading block slippage would work if it ever maxed out as a square block does not fit well on an equilateral triangle top. Thanks for all of your amazing advice on these forums

I believe in the past Balsa Man has talked about some other issues with triangular towers - for example, since the sides aren't joined at 90 degree angles, there's significant shear in the braces (even if you used triangular cross-section legs).

[Balsa Man]

Just a question for Balsa Man regarding this year's tower design. To address the bracing issue (and ignoring cost), would getting equilateral triangle style cut sticks solve the issue? I am a little confused about the leg loading issue you have presented as well. Past all of these woes that I can potentially see, my team had some success with a triangle tower. It was not terrible (averaging at around 1700 with bonus), but definitely not competitive. I was curious as to ask how loading block slippage would work if it ever maxed out as a square block does not fit well on an equilateral triangle top. Thanks for all of your amazing advice on these forums

I believe in the past Balsa Man has talked about some other issues with triangular towers - for example, since the sides aren't joined at 90 degree angles, there's significant shear in the braces (even if you used triangular cross-section legs).

Thanks, Raleway,
On equilateral triangle cross section solving the issue - the face orientation part of the issue, yes, but that's not the whole 'issue.' If you orient the triangular legs so that one of the edges/apexes is pointing away from the tower centerline, when you look at any two adjacent legs, their outer faces will be in the same plane; a bracing piece spanning between them will be in full contact with the face on both. Solves the glue joint weakness issue nicely. But.....

The bracing pieces will be inherently weaker than they would be in a 4 leg configuration. In other words, if you look at a pair of adjacent legs, triangular cross section in a 3 leg config, and square cross section in a 4 leg config; same size, assume identical wood (as in identical "E" - modulus of elasticity/stiffness); assume identical brace pieces (same cross section and same "E"), the 3 leg setup will fail under less load than the load at which the the 4 leg setup will fail.

This is because of the direction the legs be trying to flex/deflect (as they come under axial loading and start/try to buckle). With a square cross section, deflection will be toward (and perpendicular to) one of the faces. If one, or the other, or both of an adjacent leg pair are trying to deflect toward each other, that deflection will be directly toward the opposite leg. With a ladder brace between them, that incipient deflection will put an axial load on the ladder. With a ladder brace lap jointed on the outside leg faces, the force put on it will be...almost axial. But when we look at two adjacent triangular legs, they will not be trying to deflect/buckle toward each other, hence not trying to deflect along the axis of a brace piece between them. The direction of incipient deflection will be at a significant angle. This is because of the location of the 'centroid', and orientation of the 'neutral plane' So, the non-axial loading of the brace piece will be trying to bend it as soon as any force develops. Because of this additional bending force, the brace will fail under less load than it will under axial loading. So, it is not "shear" in the braces that Unome recalls, it is non-axial loading.

On load block 'slippage' and square block 'not fitting well' on a triangle of leg tops - once there's any significant load on the block (assuming nice flat, all in one plane leg top ends), the block will not be slipping/trying to slip; it will just be putting force axially down the legs. The tricky issue/problem, challenge is getting that force to equally load all three legs. With a square leg top configuration, it is easy to position the load block so that it's centroid lines up with the vertical centerline of the tower. With that lineup, force is equally distributed on the legs. On a three leg setup, the centroid of the load block square has to be lined up with the centroid of the triangle formed by the three legs. If it is not perfectly lined up, one or two legs will disproportionate load, and fail prematurely. That lineup is... how do I say this, not visually intuitive. If you draw it out to scale, it... doesn't look right. When we ran a 3 leg tower in 2011, we ended up solving this problem with a load block setup plate- a rectangular piece of plastic with a 60 degree V cut into one end, and 'guide strips' at 90 degrees glued on the top side forming an L. To set up the load block, right at the top pf the tower, you'd move the plate in so the apex of the V cut was up against one leg, then holding that contact, the load block was moved into contact with the strips,

Last on any confusion on the leg loading issue. All the words were simply saying load force is carried down from the top, axially, along the legs. Having two of the upper section leg ends not ending up on top of the lower section leg ends (so the force could be carried down those leg segments onto the testing base), but, rather, onto the top of two (very weak) bridged pieces is a bad idea.

[Jaimie]

Thank you so much for all the replies! This is my first year in Science Olympiad, and as I looked over the tower requirements, I thought that a triangular tower would cut down on the amount of material I would have to use, and I didn't see anything in the rules about putting a triangular tower on top of a square base, so I thought why not put it on top of very, very thin square base so that it was technically complying to the rules. I had no idea that it would cause problems, so thank you so for responding before I started building.

[Balsa Man]

Thank you so much for all the replies! This is my first year in Science Olympiad, and as I looked over the tower requirements, I thought that a triangular tower would cut down on the amount of material I would have to use, and I didn't see anything in the rules about putting a triangular tower on top of a square base, so I thought why not put it on top of very, very thin square base so that it was technically complying to the rules. I had no idea that it would cause problems, so thank you so for responding before I started building.

Glad the insights were helpful.
Yeah, there's nothing in the rules about how many legs. You could run triangular base and triangular upper/chimney (but it wouldn't beat a 4 leg tower). The dimensional rules drive the shape - leg ends outside 29cm circle at the base, legs fit within an 8cm diameter circle at 20(C)/25(B)cm above the base, and leg ends are fully within a 5cm square (the load block) at the top.

Do yourself a favor, and go back and read/study the posts in this forum, and strongly suggest some review of the 2017 archive.

[Snarknado]

I work primarily on the construction side of the tower, and am fairly confident in my team's ability to construct an adequately balanced and effective three-sided base, which we will probably continue to work with for now until we create a new clean-sheet design for a four sided tower.

That being said, I've lightly looked through the forums from this year and last, finding that people have had success in the past with three-sided towers. My main question is how did you build the top section (this is assuming a pyramidal bottom section with a top piece more or less stacked on top)? I'm concerned by the gluing at 60 degree angles that would need to occur.

Or, are successful three-sided towers just a single pyramid shape as opposed to a pyramid topped by a triangular prism of sorts?