How To Check A Simple Structure Design?

[abrogard]

I'd like to know how you check that some simple structure's design is adequate.

I know that's bound to be a baby question for you engineers but, believe me, the answer isn't easy to get out there in the world for us non-engineers.

If I, for instance, draw something like the frame of a small garden shed - four uprights and four bars joining them at the top - and then specify it is all built out of, say steel angle iron, how do I check it out for things such as .will it support itself , .will it stand up against a wind, . will it support a roof of such and such a weight?

This I realize is incredibly complicated if you once begin on any kind of a complicated structure.

Similarly I guess it is incredibly simple if you stick to force x distance high school calculations and you have some tables showing what forces can be supported by these materials.

But I'm thinking there should be an area in between where perhaps there's some software that will do these calcs for you on not too complicated structures...

 

[Lok]

It takes a lot of math ( simple though, with a bit of trig) just to know what a rough estimate is. Don't bother with it for a simple garden shed, if you take beam dimensions already somewhere used you'll surely get an overkill structure. Contact someone who has build something like this instead, years of experience are sometimes better then 2 hours of math.

btw. Computer aided design with finite analisys helps too. But it seems a bit much for a garden shed.

 

[xxChrisxx]

I was going to post basically the above earlier.

This isn't the tye of job you use calculations on, just make it stout.

 

[abrogard]

Thank you for your responses, guys.

I take your point - a bit much for a garden shed.

But it is not really a garden shed I'm asking about. Because whatever I draw may never, ever be built..

What I'm really trying to do is learn something. So that I can show it to my children, in the interests of given them an interest and an insight - and a tool, a power, that I never had.

I ought to tell you where I'm at so's you know what you're talking to in terms of maths ability, etc.

Well I've been educated to undergraduate degree level - my degree is in Business Computing, which isn't strong on maths, just a bit of statistics there for a while.

In fact I have no calculus. The closest I came was TAFE maths for engineering surveying which I did by correspondence for a while back when they taught that (they don't any more).

Their course was excellent and for a brief while I understood something of calculus and was able to do some quite remarkable things with it - prompted by their test papers, asking questions I'd never had dreamed of for solving by calculus. About the only one of which I remember now was solving roots with calculus. I remember the question. I don't remember how to do it.

I've lost all my knowledge of calculus. The little that was there for that brief time has gone.

But, if you see what I mean, I know there's something there that I don't know and that it is needed to make serious progress on complicated structures and analysis. So I'm not kidding myself about what I could or couldn't do, could or couldn't understand.

I still retain most of the trig from the surveying. That's fairly simple stuff, really, isn't it.

Computers are my field, I was a programmer.

So, you see, I don't expect to be able to do anything complicated, really, but I do expect to be able to do many things at a lower level.

What would be lovely is if I could show the kid (eldest one, only 7 as yet, don't need anything complicated) simple by hand calculations - 10kg acting at 2 metres = 20kg or something like that - and some basic software that does that same thing to perhaps a higher level of sophistication.

Yep. Like that. It's not really about a shed at all. We've got two. Probably won't ever build any more.

And I'd like to thank you for your interest and hope I'm not wasting anyone's time...

:)

 

[poor mystic]

I think everybody would like a fail-safe design method, not just for garden sheds but for everything we design. As far as I know, nobody has one.
The builder whom you might have called into construct a shed calls on his experience and understands that triangular structures have strength advantages over "un-triangled" rectangular structures.
More detailed observations are collated in handbooks specifying methods of construction based on what works.
As far as I can tell, there can never be a general design method for anything. You've just got to check, and build models, and check again, and ask someone else to check... and with luck and experience you'll end up with something worth keeping, but probably not the first time you try.

 

[abrogard]

Hi, and thanks for that.

When an engineer gets a drawing of a structure maybe designed by you or me according to those principles you've outlined, done in the way you mention, he checks it against the appropriate Australian Standards and if it passes he stamps it as passed for these standards and then we can put it to Council and they'll approve it (all other things being equal) and we can build it.

That's how it goes.

What I want to know is: How does the engineer do that checking?

And then there's architects and engineers designing everything themselves. Do they do it this same way or do they have a different way.

What I want to know is: How do they do it?

And then there's software that designs and checks structures - I know because I've read the blurb about somesuch but couldn't afford to buy it.

What I want to know is: Is there any such software I can afford?

And then there's Universities full of student engineers that probably have a course module amongst all their modules called 'Structural Design 101' or somesuch.

What I want to know is: What's the content of that course? And I'll find some way to ask at various universities.


So all of that is what I want, if I can get it.

But I kinda thought there might be some 'rules of thumb', some general rules than one could use to check a structure. I thought they might be widely known and widely used and almost axioms within the business.

Something people used to look at designs that would enable them to straightaway pronounce on the viability of that design.

Actually it'd be something drawn from the experience and the tables/handbooks you mention. Something that somehow formalises that observation.

Like anyone can see that a little shed, say 2m x 2m, massively built of, say, four corner 'I' beams to support a corrugated iron roof, fixed to purlins of 150mm x 2mm 'C' section at 500mm centres fixed at every valley with tek screws - with the columns 2metres in the ground into concrete footings of .5metres cube would be virtually immovable.

We can see that. But what's the formula we apply to it to show that in figures?

Well it looks like it doesn't exist else I would have been swamped by identical suggestions.

"Use UFFDS..." or somesuch and when I checked I'd find 'UFFDS ' was the 'Universal Formulae For Designing Structures', obtainable at my local hardware store.

Nonsuch. Doesn't exist.

So I'm left with the questions I've listed above. And an intent to probe the Universities and see what they have to say.

I hope I haven't hassled anyone with what looks like dumb questions or, worse, an apparent inability to be told.

I did see a thread somewhere, maybe this forum, maybe another, where someone was guiding someone else through all the deficiencies of their design and I'll go look for that one and read it carefully this time. I just glanced at it before because it referred entirely to wooden structures, the thread was all about some bloke's pergola or somesuch.

But the discussion was all about design from an engineering point of view, strength, structural integrity.

I better go back there, make up a design and put it before those blokes and see if they'll look at a steel design and see how they go about critisizing it.


Christ, I waste so many words. I'm just not a good communicator.

poor mystic: yes: "... you've just got to check... and check again..."

What I'm asking is: How do you 'check' ?

 

[sardonicus]

Google: structural analysis software free download. You can get a trial copy and have at it. Many will even apply code prescibed loads. Use the help feature and examples to learn it. Looking forward to seeing the next Golden Gate (be sure to check the harmonics so it's not the next Tacoma Narrows).

 

[nvn]

Is there any such software I can afford?

Perhaps try Calculix (free), if you have a Linux or Windows operating system.

 

[abrogard]

Thank you for that. I've downloaded and installed Calculix and I found and downloaded a thing called 'Analysis' which has a GUI.

I also found a page: http://www.structural-engineering.fsnet.co.uk/free.htm which seems to have a heap of stuff on it. I never found any of this before, must have been googling on the wrong words.

Well now I seem to have enough here to keep me busy for yonks.

That Calculix certainly seems to be impressive. That analysis of the femur, was it?, and the turbine seems to indicate it can certainly do man size stuff. I'll buy it for the GUI if it has one, it is not crystal clear from the blurb whether that is included in the extended version or not.

Thanks for your help.

:)

 

[nvn]

The GUI is free. Upon further investigation, it appears Calculix surprisingly currently does not support regular 1-D beam finite elements. Calculix tries to create beam (and shell) elements by expanding them into volumetric (solid) elements. Therefore, although Calculix is excellent for 3-D solid elements, it currently appears inadequate for truss, beam, or framework structures. For shell elements, it automatically "inflates" them into solid elements.

Although I have not tried them, and do not know which free program would be best, someone suggested perhaps try Frame3DD or Wolsink Framework (?).

 

[abrogard]

There is a GUI and it is free? Where do I get it? Could you provide the URL, please? I've not found it. Or is it a GUI for the purchased version only?

I'll try both suggestions you've made - I think I already downloaded Framework but didn't unwrap it yet.

The URL I provided for the long list of structural analysis freeware hasn't been updated for years and I found prog after prog after prog no longer available, homepage gone or changed or software now in a later version only for sale.

Still I've got heaps to carry on with. Just even the wikipedia entry for structural analysis and related stuff is fascinating...

In fact my danger is getting lost in the fascination... I don't have the time... I need to keep focused on something simple for a simple task for my simple children instead of going off wandering down another primrose path... I do enough of that already...

regards,

ab :)

 

[nvn]

The Calculix GUI is included with Calculix, for free. If you are referring to the MS Windows version, you go to the link I posted above, then go to the download link on that page, then go to the "Windows executable" link on that page, which will then bring you to the Convergent page. On the Convergent page, the first link under the heading "Download basic version" includes (for free) the Calculix graphical user interface, called Calculix Graphix (CGX).

 

[abrogard]

It's okay. I've found it. I had only noticed no mention in the list of installed progs in the calculix folder on windows programs menu.

I find one evokes it by writing the command in calculix command.

Now I better go read all the doco.

regards,

ab

 

[nvn]

As mentioned in post 10, I was surprised that Calculix currently does not support regular 1-D beam finite elements, which are needed for a frame structure. Hopefully, Frame3DD or Wolsink Framework will work better for you, for a frame with no plate finite elements.

 

[nvn]

Instead of analyzing a shed frame, what if you analyze a shelf? You could say all the weight of the shelf is carried by one simply-supported beam, the beam has a symmetric cross section, and all the weight on the shelf (on the beam) is a box (a point load) placed at the beam midspan. For this, you would want to study (1) vector mechanics, (2) statics, (3) how to draw and analyze a free-body diagram, to determine the bending moment on the beam midspan, (4) how to compute (or look up) area moment of inertia of the beam cross section, (5) how to compute applied bending stress on the beam, and (6) ensure applied bending stress does not exceed the beam material allowable bending stress. Unfortunately, I do not have good tutorial links to where someone (or some university) is trying to teach these subjects to beginners.

 

[abrogard]

Well you can see how ignorant I am, straight away - because it seems to me that if my shelf is supported on one beam then unless it lies along the beam it will be balancing and inherently unsafe for any structure I might have in mind.

But if it lying along the beam then why is it there at all?

So what I in fact (attempt to) analyse is a simple beam (or girder, or length of c section, or length of wood, or length of corrugated roofing material) supported at each end.

Firstly, by itself - when will it sag and fall under its own weight, and then under loads.

Currently I'm putting these scenarios into various freeware packages I've gotten hold of.

Your post seems to suggest that doing this is impossible without a great deal of arcane instruction - arcane indeed when we consider the portent of '... teach these subjects to beginners' which seems to suggest there's prior stuff required before one could even begin on this stuff.

Which I find very confusing because I've got progs that attempt to do this for me and you yourself have been directing me towards such progs.


But, all that aside, the simple question remains: How do people check simple designs?

And if the answer to that is something like this:

"People don't. Only incredibly highly trained engineers can do this. A training that you can't even begin. You'd have to start with the training for the training."

Then it has taken a lot of words to get to that point - well, in fact, we never got to it, I've hypothesized it.


And it makes the world look pretty stupid - and those engineers not helping much, either.

Because after decades of experience and decades of highly developed mathematics it surely should be a trivial exercise in this day and age to write a prog that will, for instance, look at a beam supported at each end and carrying a weight on itself and from the given properties of that beam decide whether or not it will snap and, failing that, how much it will bend.

A trivial exercise.

And given that simple knowledge one can design simple one story shed-type structures that are proven safe, inasmuch as they can be designed to support a point load at mid beam greater than any total distributed load (i.e. a roof).

Draw a little more on the widely available sophisticated mathematics and we could even put a figure on the amount of leeway we've built into the structure by designing in this deliberate 'over design' way.

And then we could answer questions such as mine by saying some like:

"Use the 100% Simple Structure Check" tool and it will provide you with a structure with a margin of safety of 100%.

Turn this simple mechanism on its side and it tells the story of what happens to a column planted in the ground when subject to wind loading.

Look up specifications for nails, screws, strips of tin and it should be equally trivial to decide if a roof is going to fly off in a gale.


That's how stupid I am. I see it like that.

:)

 

[nvn]

The beam underneath the shelf runs along the length of the shelf. To be more realistic, there can be two of these beams, each carrying one half of the shelf load. You can say you want to design the beams to carry the full shelf load, neglecting any additional strength contributed by the shelf plate. This simplifies the problem, for hand calculations. Post 15 indicates how we would check that structure.

For simple problems, you can use hand calculations, such as mentioned in post 15. For more complicated structures, you might run a finite element analysis. Structural analysis with no prerequisites might be difficult; we study statics, mechanics of materials, and structural analysis for months or years in university courses. But you have an adequate mathematics background to start studying the subjects. Unfortunately, I do not have good links to refer you to; but maybe someone else does.

 

[sardonicus]

Abrogard,

There are programs that will give you simple answers, vendor programs, try something like this, all this is from just one website:


http://www.ilevel.com/services/

Forte™ Software
Javelin® Software
Comparis® Software
TJ-Beam® Software

Those are for wood I joists and other wood products.

As for a simple program for a shed, use your noggin:tongue:

For your program start by listing all the variables that should take a day or two.

 

[rhit2013]

it's basic statics. look it up. easy stuff

 

[abrogard]

Thank you for that, rhit2013, you've placed my feet on the path, put me outside the right door.

One of the first googled hits I got was at http://www.fao.org/docrep/s1250e/S1250E0b.htm

and amongst the first paragraph was this extract:

" There are two distinct stages in structural design. First the structural engineer with his experience, intuition and knowledge makes an imaginative choice of preliminary design in terms of layout, materials and erection methods. Estimates of the various forms of loading are made and then the chosen design is subjected to detailed analysis based on the principles of statics." (my bold, my italics)

and that latter phrase is precisely what I wanted to know for starters.

It was a bit hard there for a while, not even knowing the name of the discipline I wanted to look at.

I'm looking forward to reading the rest of that document. It looks like it might provide me with all I want to know.

Thanks again.

ab :)