torsion boxes, why no triangles?

[TheUnicorn]

more of theoretical question than for anything specific I'm planning to build.

I've seen a few torsion boxes made up (youtube) and I notice that the internal web always seems to be boxes and rectangles, no triangles, it seemed odd because triangles seem to be so common place everywhere else that you want to maximise strength and minimise weight (I know that is massive over simplification).

Have I missed something obvious?

 

[Jacob]

More difficult to make?
n.b. any box is a "torsion" box - it's a redundant term with no special meaning, but popular with woodworkers!
What you are talking about is a box with stiffening internal braces.

 

[TheTiddles]

Orthogrid is easier to make than isogrid especially in wood. Isogrid may have improved performance if the topology is optimised but you’re probably not going to get much of a cost/time benefit to make it worth it

 

[TheUnicorn]

More difficult to make?
n.b. any box is a "torsion" box - it's a redundant term with no special meaning

my understanding (possibly misunderstanding) of torsion box is two (normally thin) skins held apart by a framework. the purpose being to build a light, strong and stable (against torsion). I don't think of a box as a torsion box because it doesn't have the opposing side

 

[Jacob]

my understanding (possibly misunderstanding) of torsion box is two (normally thin) skins held apart by a framework. the purpose being to build a light, strong and stable (against torsion). I don't think of a box as a torsion box because it doesn't have the opposing side

One skin can be stiffened by bracing, two skins more so. A 6 sided box is stiff even without bracing, compared to the same box with side or sides missing i.e. the sides are the braces. It all depends on what is is being designed to do.
Most of what are called "torsion boxes" aren't designed to resist "torsion" at all, but merely to be stiffer/stronger etc in the application they are designed for. Ships a notable exception.

 

[TheUnicorn]

Orthogrid is easier to make than isogrid especially in wood. Isogrid may have improved performance if the topology is optimised but you’re probably not going to get much of a cost/time benefit to make it worth it

It's certainly much easier to make a grid of 90 degree joins than the complexities of triangles.

 

[Jacob]

It's certainly much easier to make a grid of 90 degree joins than the complexities of triangles.

Or no joins at all i.e. parallel partitions going the full length, will stiffen a box. A long load-bearing box becomes a "box beam".

 

[Adam W.]

It's certainly much easier to make a grid of 90 degree joins than the complexities of triangles.

Which is probably the reason why you're not seeing it on youtube.

 

[recipio]

Much easier to cut 90 degree half lap joints than any other. I do it that way for peace of mind but it certainly makes for a lot of work. I have seen some people simply staple the grid work together as it is the glue line that gives strength rather than the joinery.

 

[Ollie78]

I would agree probably dificult to make, but more to the point the extra strength provided by triangles is probably not worth the effort in a woodworking situation.
I could be wrong but I thought purpose of using a torsion box method ( a grid of equal thickness strips with a skin either side ) was more for ensuring flatness.

Many composite materials use hexagon patterns inside. I would imagine the strongest would be loads of thin skinned spheres between the skins but that would be trickier than triangles.

When I was a kid my Dads friend invented a construction method using some sort of triangles or maybe pyramids inside a skin ( he was a mathemetician I think ) he made a cardboard version you could drive a car on, but the concept was to build steel bridges and things with vastly reduced materials and costs, many were interested but he never sold the design as far as I know. I remember visiting him and he had all sorts of models and test parts around the place.

Ollie

 

[TheUnicorn]

Which is probably the reason why you're not seeing it on youtube.

I did also do a google image search for torsion box and 99% were squares and rectangles. I thought that at least on a factory / machine engineered level, triangles (as well as honeycomb structure) would be fairly widely used. it was just an assumption, and I was surprised to see it wasn't the case commonly.

 

[TheUnicorn]

I could be wrong but I thought purpose of using a torsion box method ( a grid of equal thickness strips with a skin either side ) was more for ensuring flatness.

I think you are right, I'm happy to admit that it could be something I've misunderstood and misused over the years but I've always thought that it was a way of making something flat and stable with minimal materials and therefore lightweight, hence its use in doors and (again could be wrong here) aeroplane wings

 

[Jameshow]

I think you are right, I'm happy to admit that it could be something I've misunderstood and misused over the years but I've always thought that it was a way of making something flat and stable with minimal materials and therefore lightweight, hence its use in doors and (again could be wrong here) aeroplane wings

Triangles will always produce the lightest structure apart from monocoque construction.

But I woodworking mft then weight isn't of primary importance.

 

[Jacob]

I think you are right, I'm happy to admit that it could be something I've misunderstood and misused over the years but I've always thought that it was a way of making something flat and stable with minimal materials and therefore lightweight, hence its use in doors and (again could be wrong here) aeroplane wings

Well yes - variations of the "box beam" etc.
The term "torsion box" seems to have been coined by JE Gordon "Structures or Why Things Don't Fall down" but he's talking about bi-plane wings and the struts and strings tying them together, which keep them from twisting, as a sort of box, rather than the wings themselves. Nothing to do with box beams or panels.
Excellent book by the way.

 

[Adam W.]

I did also do a google image search for torsion box and 99% were squares and rectangles. I thought that at least on a factory / machine engineered level, triangles (as well as honeycomb structure) would be fairly widely used. it was just an assumption, and I was surprised to see it wasn't the case commonly.

If you want triangles and curves, then waffle structures might be your thing.

This one's from Expo 2000 at Hanover.....It's big.

Still based on a square grid thoughbut.

 

[Rodpr]

Interesting (to me anyway) that nature doesn't seem to favour right angled structures. Hexagons are good, and Fibonacci spirals, but perhaps square forms are more difficult to grow or less adaptable to changing environments. Once you have a nice stable, flat environment square boxes work just fine (unless they move).

 

[TheUnicorn]

Interesting (to me anyway) that nature doesn't seem to favour right angled structures...

nature might not, but gravity seems to quite like going up and down, and water seems to quite like going straight across.

 

[rogxwhit]

Have I missed something obvious?

Yes. The triangulation is inherent in the web and skin components. Because they are continuous sheets, the triangulation isn't explicit, however it is there and fully functional.

 

[StephenArfer]

As an ex-aircraft maintenance engineer (A&C) I reckon you'd be hard pressed to find any modern aircraft without torsion boxes in primary structure including modern composite structures. There are many different implementations of a torsion box, some complex, some not so.

I made a "torsion box" MFT table a few months ago, basically to the Ron Paulk design principle. When I look at it now, I wondered why on earth I had to make it so complicated for no real purpose other than a lot of practice creating lightening holes in webs and longerons. Triangular internal bracing in a simple zig-zag pattern would have done just as well to ensure torsional rigidity IMHO. Bending stresses easily withstood by virtue of a significant spacing between well braced top and bottom skins and the side longerons.

However, a side benefit of the conventional MFT construction with lightening slots is being able to get your arm into the structure to pick up bits that dropped through a hole and magically discover tools and fixings that disappeared ever since you put them there (it's an age thing).

 

[Jameshow]

As an ex-aircraft maintenance engineer (A&C) I reckon you'd be hard pressed to find any modern aircraft without torsion boxes in primary structure including modern composite structures. There are many different implementations of a torsion box, some complex, some not so.

I made a "torsion box" MFT table a few months ago, basically to the Ron Paulk design principle. When I look at it now, I wondered why on earth I had to make it so complicated for no real purpose other than a lot of practice creating lightening holes in webs and longerons. Triangular internal bracing in a simple zig-zag pattern would have done just as well to ensure torsional rigidity IMHO. Bending stresses easily withstood by virtue of a significant spacing between well braced top and bottom skins and the side longerons.

However, a side benefit of the conventional MFT construction with lightening slots is being able to get your arm into the structure to pick up bits that dropped through a hole and magically discover tools and fixings that disappeared ever since you put them there (it's an age thing).

Does your mft fly?!