100ft span Hollow Steel ARCH design 6

[cad1009]

Gents please help me out.
i am designing a Steel hollow Arch Gate structure. Arch is 100ft support to support with 30ft height.
Arch is serving the purpose of Entry gate with no dead load on top except facade glass pannels.

I am modeling the Arch in STAAD. to design the pipe section what should be LZ definition for Arch.
Support to support distance is 100 ft and Arch curved length is 130ft.

in my Opinion LZ (bending length) for design should be 130/2 = 65 ft.
Please see the pic attached for reference.

https://files.engineering.com/getfi...470b-4a30-8020-3ce283568e65&file=Untitled.jpg

 

[BAretired]

What is your definition of bending length? Do you mean laterally unsupported length? Is the arch laterally braced at its mid point? Are the ends of the arch fixed against rotation and translation?

Why do you think LZ should be 65'?

BA

 

[JStephen]

I'm not clear on the question.
Are you looking for an effective length for column analysis? In that case, the column analysis doesn't apply to a curved member like that.
Or is LZ an unsupported length for lateral buckling? There again, questionable applicability to an arch or to hollow structural tubes.
Or is LZ the span for determining bending moment? In that case, bending moment would come from your analysis, not just from a length.

 

[WARose]

You can't do a code check of a arch using STAAD. It's good for analysis (i.e. to get your loads). But the allowable buckling values need to come from elsewhere. There are several references that can do that. Roark's is one. 'Guide to Stability Design Criteria for Metal Structures' is another. For the latter reference, I've got the 3rd edition, and this is covered (for both in-plane and out of plane loads) on pp. 455-487.

 

[JLNJ]

Your shape appears to be neither circular or parabolic. Weird.

You might want to consider hiring a good structural engineer.

 

[fahedjavaid]

Guys thanks for your responses. i am attaching STAAD model diagram for more clarity.
i want to design the hollow section from STAAD considering it as beam by taking unsupported length = half curved length.

deflections and other analysis results i am taking from STAAD.

https://files.engineering.com/getfi...c8-43bf-9b11-5a334c829cf6&file=Untitled-2.png

 

[fahedjavaid]

@WARose can you upload that reference? "Guide to Stability Design Criteria for Metal Structures"

 

[ukbridge]

I would do an eigenvalue buckling analysis using STAAD etc to work out what your Pcr for the arch is. There should be a worked example in the manual showing how to do this, although I'm not familiar with STAAD myself. It should be P(factored Fx in arch) * Eigenvalue = Pcr, taking care to examine the buckling shape.

There are also some simplified methods in EN 1993-2 Appendix E that calculate effective lengths for arch bridges (for in-plane and out-of plane buckling lengths) so you can calculate your Pcr Pi^2*EI/L^2, although clearly the the grillage steel will provide some degree of lateral restraint. Any other decent code will likely have these as well, they are probably in some other structural mechanics textbook.

It'll probably be in the order of L/2 - L/3 with the steel restraint, but thats why its helpful to do a buckling analysis to give you a (more) "real" answer that you can compare to more simple hand calculations to give you more confidence.

edit: just noticed the middle arches wont have any lateral restraint clearly....

 

[BAretired]

I don't know how anyone could have guessed the configuration of the latest model from the original post. Is cad1009(OP) acting on behalf of fahedjavaid?

BA

 

[ukbridge]

Another few thoughts:

1. Look at the design of the connections early on to size the arch. If you have a very nasty wind load, and the connection transmits a moment you will likely need a very thick wall for your arch cross-section. Time and time again I find the connections of these type of structures to control the design.

2. Have a clear idea of how the connection is formed between the CHS and the steel members. There are many different ways (varying in ugligness and fiddliness) to do this, which should be agreed with the architect and client.

3. Design the arch using hand-calculations i.e. using Pcr, then allowing for imperfections and so forth as per your required code. Dont just input LZ and assume the software will do all the right checks for you. I don't like using structural design tools in FEA packages for novel situations unless I know exactly what they're doing.

 

[SAIL3]

ukbridge...all good and pertinent points....also check snap-thru, although, if only the dead load of the arch itself is involved , it may not be a problem.....

 

[ukbridge]

Snap through need to be checked as well, good point there. I would be surprised if STAAD deals with that specifically. Terrifying really isn't it...

 

[WARose]

[blue](fahedjavaid)[/blue]

@WARose can you upload that reference? "Guide to Stability Design Criteria for Metal Structures"

I don't have it in PDF form. However, in every design office I've ever worked in it has been on someone's shelf.

 

[ukbridge]

If I can believe the slightly dubious copyright claims of the first page, page96 gives you all the formulae you need. Id be surprised if it were any different from the ones in in Stability... or Roark, no (or not many) arch bridges in europe fallen down yet.

http://www.phd.eng.br/wp-content/uploads/2015/12/en.1993.2.2006.pdf

gl

 

[fahedjavaid]

Unsupported length suggestion isnt solved and snap-through problem is raised
@SAIL3 i would really appreciate if u can share literature on snap through. and 4 out of 6 arches will be under lateral load due to their orientation in structure.

 

[ukbridge]

@fahedjavaid

Everything you need to design the arch for axial compression is in that document on pg96. If the arch is significantly out of plane you will need to do some form of elastic buckling analysis, see my previous post.

I'm not sure about the term "unsupported length", but fundamentally to design an arch (except the connections as mentioned above, dont forget them they will come back to haunt you):

1. Calculate Ncr (using the formulae provided for an arch) checking for in-plane, out of plane buckling and snap through cases
2. Use your code to calculate Nb,Rd (i.e. the buckling capacity of the arch allowing for member imperfections, or whatever method is prescribed in your code).
3. Check it for combined axial compression and bi-axial bending and shear i.e. N/Nb,Rd + My/My,Rd + Mz,Rd. The eurocode allows for shear typically by reducing the bending capacity if V/Vr > 0.5 your code may vary.

thats it.....

 

[SAIL3]

another indespensible resource in understanding the theory of elastic stability can be found in Timoshenko and Gere "Theory of Elastic Stability".....invest some time in understanding the fundamentals of elastic stability and it will serve you well going forward not only in this problem but in just about every design you will encounter in the future.....

 

[BAretired]

Snap through is a problem only for very shallow arches where the compressive strain is sufficiently large to reduce the rise of the arch to zero. It will not be an issue for this arch.

For this arch, an inflection point can be taken at the center of the arch. When the arch is laterally braced, as it is in the model, at about the third points, the unsupported length for buckling about a horizontal axis can be taken as L/2 where L is the curved length of the arch.

The above would not be true for an arch without lateral bracing as described in the original post. An example of that type of arch is the Gateway Arch in the City of St. Louis which deflects 18" laterally under design wind loading.


BA

 

[cad1009]

i ran the model on staad & performed buckling analysis buckling

i am getting very large -ve buckling factors .... what do they mean,
i know that buckling factor <1 means that member will buckle before the designed load,

the snap shot attached below is for Self weight. (Plz See the Pic attached)

https://files.engineering.com/getfi...-f680-4a1d-9a82-fcbd22d6f40e&file=Capture.JPG

Secondly how can i know the buckling factors of any other member of my choice .
@ukbridge
@SAIL3
@BAretired

 

[Compositepro]

An arch that is close to buckling due to self-weight alone would be pretty useless. Using a constant cross-section for the whole length would also be sub-optimal. Again, look at the St. Louis Arch. You do not want dead weight at the apex that does nothing structurally.