Structural question 3

[Grawschbags]

Hi folks,

I'm new to the forum. I previously worked in the defence industry as a mechanical engineer. I have recently made the switch to a smaller engineering firm who deal with tanks and structures.

Now, my memories a bit rusty and I need a quick solution to a problem I haven't come across since my uni days. Hopefully you can help.

I am designing a structure that is to support a 60m^3 storage tank, that when full could weigh up to 90 tonnes.

This cylindrical tank is to be supported by 4 legs (with additional cross members) equally spaced around the tank diameter.

It has been proposed that the legs of the tank be made from 180X180X8mm square hollow section, mild steel.

The height of these legs will be roughly 4m tall, with the tank sitting on top.

The cross sectional area of the beams is 54.4cm^2, and the moment of inertia is 2661cm^4.

Could one of you kind people please confirm if these sections are man enough to do the job?

Thanks for your time/help.

G.

 

[rb1957]

ok, the answer'll depend on where you are, ie what building code applies (I'm guessing Oz).

generally you're looking at a column in compression, google/wiki "euler column load". simply put Pcr, the load at which the column will buckle is pi^2*E*I/L^2 and since you've got all the terms you can calc that (FWIW i get 3.4E6N).

then you'll have 4 legs, 4*Pcr = 13.6E6N.

Your load, 90 tonnes - 90000kg = 883000N.

then your safety factor = 13.6E6/0.883E6 = 15.5 ... sounds ok.

 

[Grawschbags]

Hi,

Thank you very much for your quick response.

At the risk of sounding daft, what does the term Pcr mean?

I'm guessing that the fact the factor of safety came out at 15.5, I could probably go to a smaller square hollow section that would still be adequate, say with a FOS of 5?

Thanks again.

G.

 

[rb1957]

Pcr = critical buckling load (determined by Euler) ... the load at which the column buckles ... if you're going to do structural work, you should know the terms ... wiki/google "Euler column load".

this is for a perfect column. your code might tell you that you need to consider an imperfect column, or a certain out-of-straightness might be typical. I'd investigate this first before going to a smaller section.

Another factor to consider is how you're supporting the column with the "other members". you might be able to say that the supports prevent column buckling deflections and create nodes (which you'll understand if you read up on the subject).

 

[Grawschbags]

Thanks again,

I've just looked out dusted down my copy of Roark's Formulas and will bury my head in that.

So that I can replicate the calulation you carried out for my own piece of mind, can you tell me what values you used for 'E' and 'I'?

The structure will be tied together as it's height increases by various walkways. I imagine these will go some way to support the structure.

 

[rb1957]

E for steel = 30E6psi = 208E3 MPa
I, you gave me, = 2661cm4 = 2661E4mm4 ...

becareful with your units ... i recommend N, mm

is there any lateral load to consider ? (we've only been talking about dead weight)

"walkways" = people ... any additional SF ??

 

[izax1]

Are the four legs only loaded in pure conpression, or is there any bending involved? (Are the legs only vertically loaded, i.e. only in the direction of gravity?) I would be careful with too small FoS if we are talking euler buckling. Any small variations in geometry of the legs (which you will certainly have) will reduce the buckling load drastically. Recognized design codes will take care of that. And remember, Roark&Young is not a design code!!!!

 

[Grawschbags]

Thanks rb1957.

These tanks/structures will be arranged equispaced around an 8m diameter.

They will all be tied together on three different floors via walkways.

Turns out the unsupported length of square hollow section will be 8616mm (not the 4000mm I stated previously). Total height of the tank support will be 10840mm, with 2224mm supported by welding to the tank.

The overall height to the top of the tank, which is 3000mm in dia, is 18000mm.

Cheers,
G.

 

[msquared48]

Grawsa:

RB asked if there were any lateral loads to consider, such as wind or seismic...

Mike McCann
MMC Engineering

 

[rb1957]

that'll reduce your SF by a factor >4, so now you're marginal if you want a SF > 5.

to reiterate three points i've made ...

1) this is for a perfectly straight column, research "euler imperfect column" ... you're looking for something that'll address a column that is not perfectly straight (and i think you're code defines how out-of-straight you need to consider).

2) a saving grace in your design may be the lateral bracing, but it needs to support the section in both lateral planes to be fully effective (so i doubt it's much of a saving grace).

3) look into your loadpath for lateral loads. maybe you've got diagonal members for the tank to the ground (not necessarily directly). if you're relying on the columns, they'll work but not very well, and they'll have a significant effect on the allowable column load.

i assume that you've got someone there looking over your shoulder ?

 

[rb1957]

seriously, reasearch "euler imperfect columns". i quickly found a link to an answer for a column that is initially not straight.

for a column that's initially not straight, the answer is quite a bit more complicated.
1) maximum deviation = c; the column is assumed to be slightly curved so that it deviates from a straight line by c at the mid-length. your code might express this as c/L = 0.4% (or something like).

2) calculate Pcr, the euler allowable for a perfect column, like we've been doing.

3) calculate Mmax = c*P/(1-P/Pcr) ... as P approaches Pcr, the moment approaches infinity

4) calculate bending stress = My/I

5) compare with Fty (or Fcy) for SF

i've been assuming a simple pinned column. if your column is fixed at the top (sounds like it if it is welded to the tank for something like 2m) then Pcr increases by a factor of 2.05

 

[Grawschbags]

rb1957, thank you once again for your perseverance. Thank you to the others for the comments.

I've been getting drip fed more and more information on this from the guys at work as the day went on. Finished work for the day now though.

I am still getting to grips as to what the company works too. They have told me to work to a FoS of 5. We are a UK based company.

It turns out that the first 6m of unsupported beam have a load cell on top, then a further 4m of the structure is placed on top of this. Sorry I can't remember the exact figures as I don't have the drawings in front of me.

With regards to your post earlier where you mentioned FoS had dropped to 4, I have upped the beam cross section to 200X200X10mm square hollow section. I think this gave me a FoS of 4.6 if my calculations aree correct.

With regards lateral loads, the tanks will be situated outside and will be exposed to the elements. We are going to tie the base structures of the silo's together, hopefully to combat the lateral forces. For reference, the tanks are equispaced around an 8m diameter.

As for if I have someone watching over my shoulder, there is a gent there with 40+ years experience in the game, he doesn't claim to be an engineer though and relies on previous experiences. We are a small engineering organisation that makes use of an independent structural engineer when required.

I want to get to grips with how it's done before I pass the drawings under the nose of a structural engineer, more for my own piece of mind more than anything.

I will do the research in to the imperfect columns.

Thanks,
G.

 

[rb1957]

don't take this personally, but geez am i scared. maybe your "gent" with 40+ years can look at something and size it by eye, and maybe your c/- is trying to get someone else "up to speed" and designing structures with calcs. but you're designing a structure supporting 8*90 tonnes without much experience, and without much support (in-house).

i'm sure you're doing your best, but geez am i scared.

 

[Grawschbags]

Ha ha, I appreciate your concern. Once I've been in the door a bit longer and am a bit more offay with how all this stuff comes together, I will have a lot more confidence with future designs.

Like I said, we have a chartered structural engineer that we call upon to scrutinise major load bearing designs, and I'm sure he will get sight of this one I'm working on before it goes out the door!

Thanks again though for your help, it's definately given me food for thought. Forums like this are a breath of fresh air.

 

[msquared48]

Glad to see you know and are willing to admit to your structural limitations, asking for help when required. That in itself instills confidence.

Cheers...

Mike McCann
MMC Engineering

 

[dhengr]

Grawschbags:

It is incredibly dangerous to be trying to design this kind of support structure for a storage silo or bin the way you are going at it, and Wiki, Google or even Roark aren’t going help keep you out of trouble, given where you are starting from. And, you won’t get the kind of help and guidance and supervision you need from this distance. You moved to a smaller design firm, hopefully you didn’t over sell yourself; but you shouldn’t be doing this kind of problem by yourself, or even with our help. Go to one of the senior engineers, your supervisor for example, and be honest about your experience level on these kinds of design problems, and show that you really want to learn and will work hard at it. Ask him to be your mentor, most bosses and supervisors are willing because they want to see you grow for the good of the company, and if for no other reason they don’t want you BS’ing them about what you know and getting them in trouble for what you don’t know. You may have to be a bit humble, but your boss most likely expects you to have questions in a new area of engineering, after all none of us started out with 30 years of experience.

Do come back to Eng-Tips, but don’t come here with the kind of fundamental questions you are asking on this complex a structure. You should have gotten your basic engineering education at the Uni, not here, and in all fairness your boss has a right, no, the necessity of knowing what you don’t know.

 

[nuche1973]

Grawschbags:
I recommend you purchase a copy of Bendar's Pressure Vessel Design handbook. There is an example similar to what you have posted. Also, I agree with the others, make certain you have all the lateral loads accounted for in your design. And remember, to note all your questions so that your structural engineer can properly review your work. Good Luck.


There are days when I wake up feeling like the dumbest man on the planet, then there are days when I confirm it.

 

[Grawschbags]

@ dhengr

I appreciate what you are saying. I was hired mainly from my background in product design, and my ability to produce technical drawings and documents. I certainly did not put myself forward for the role as having much structural experience, that is a discipline in itself! It's certainly a field I want to get experience in though.

I know I am not going to teach myself the skills from google and sites like this, but they are a good place to gather ideas from, and the calculations I have been carrying out have evolved due to the comments of other posters. I have a knowledge of this subject area, it's just went a bit rusty having spent the past 10 years designing electronic boxes for military aircraft.

I have stressed, pardon the pun, to my boss that I am no structural expert and that they should continue to make use of the structural engineer they call upon for complex tasks such as this. In the meantime I will work to improve my skills in the field to a point where I have confidence in the designs I am putting together.

 

[Grawschbags]

@ nuche1973,

Thank you for the input, I will give it a look. I am looking to build up a library of good literature based on the subject.

Rest assured I will no doubt have a list of questions as long as my arm when I take my proposals to the structural engineer.

 

[MiketheEngineer]

When was the last time you sat on a four legged chair and all hit the ground at the same time.

You might want to rethink your leg loads in case this is a possibility.