Tips for this structure I am designing

[JM_10]

Hello all,

I need to design an structure which is able to support a cylinder loaded with 10 tons of weight distributed around its external face

The current model is the one you can see in the pictures. A FEM analysis has been carried out using "Simscale" and this is the result.

I would appreciate any advice or comment in order to improve this structure (made of steel, initially).

On the other hand, you can see a picture of a 2D model, in which a vertical load plus a moment is applied (exactly the ones caused for the cylinder plus external weight), and you can see the reactions in red. In green, I have written the reactions that the FEM simulation is calculating. I wonder which could be the discrepancy in the left support, in which 2D and 3D simulation are differing in the force direction

PICTURES:

https://imgur.com/a/yxEs0

Thank you in advance
Regards,

Javier

 

[drawoh]

JM_10,

If you are not confident about what you are doing, maybe it is time to contract this out to a structural engineer. However informative the FEA is, there is soil mechanics, all sorts of QA[ ]issues, especially if this thing is welded, plus building codes, and occupational health and safety. We don't want to lose control of ten ton (tonnes) of stuff.

--
JHG

 

[rb1957]

the biggest difference in your reactions is FEA is predicting a redundant self-reacting pair of loads (66), which you couldn't calculate by hand. You can relieve with a roller. Odd that element 1 is reacting almost as much normal (bending) load as element 5, but 1 is braced and 5 has to carry this in bending; but maybe you have to (possibly to access one side of the structure). Have you considered load cases that'd out the cable stays in compression ?

agree with the previous comment ... be very careful with a load of 10 tons. I'd suggest having some "learned" input.

another day in paradise, or is paradise one day closer ?

 

[JM_10]

@drawoh @rb1957

Thank you for your answers

About contracting this out, initially the company is not considering it. However, we have time to learn some inputs and develop this carefully

So,

Yes, 1 is braced and 5 not, but not for particular reason, we just thought it would not be needed, as putting two cables in the upper side is more effective

Cables in compression have not been considered. But the structure will not turn to the cables side, center of mass is so far on the other side, even further than the first pinned support

So then, using a roller should be considered? Actually, even with this setup, both FEM and analytical calculations show conservative values

Thank you

 

[drawoh]

JM_10,

I have done some safety calculations. I made very limited use of FEA. I went through the structure joint by joint looking for weakness. Most of your failure points are bolts, flanges and welds. Let's just pick one element in your structure.

You have two cables. I can work out the loads on your cables using statics. I may have to make some assumptions about the rigidity of your base. Your cable has a load rating which must not be exceeded. You have cable fittings at each end with load ratings that must not be exceeded. You have some sort of brackets fixing your cable fittings to the floor and to your structure, that must be strong enough. You have bolts and/or studs holding the brackets to your floor and structure whose load ratings must not be exceeded.

It is very easy to make FEA show structural strain under load. Working out the stresses of a complex structure is an order of magnitude more complicated, and it requires a thorough understanding of the loads and how your reactions work.

Since I am not a civil engineer, I do not understand studs in concrete floors. How flat is your floor? How flat will your welded base be? I wonder if your mounting system is imposing strain on the structure holding your ten ton weight.

--
JHG

 

[Settingsun]

Without knowing all the story and constraints, it's difficult to propose improvements. Perhaps though a single cable would be simpler? And taking them down onto the base member so there are only two base connections to design rather than the current three connections?

Maybe just a problem with my reading of the numbers but the free body diagram reports vertical equilibrium whereas I read 54+51-53-57=-5. I really don't like results that aren't in static equilibrium so would check this.

rb1957, consider the vertical (axial) force in the post member. It is applied near midspan of the base member and equilibrium requires essentially equal reactions at either end of the base regardless of the brace to one side.

 

[JM_10]

Thank you all very much for your help.