Frame buckling anaylsis

[Anonymous]

Hi guys,

I tried to run a simple test on a Frame buckling analysis today. 

Just confused witht he results from advanced----------->critical forces.

When I add all the columns to the table (I want to get critical force for each element and also effective length for each element), it looks strange.  The bar with the load on it (Bar 1) has N/A for all of the parameters except the first column.

A lot of the other bars also have this value (n/a).  Why is this?  Also, why is the critical force reported as such a small number (ie 0.13kN), when how is the member effective length so long (300m?)

Any help appreciated.  File attached.

[Artur.Kosakowski · Accepted solution]

Tony,

Please check the 2nd (and perhaps 7th) post from http://forums.autodesk.com/t5/Autodesk-Robot-Structural/Buckling-Analysis/m-p/3280269/

The effective lengths is not shown when bar is under tension or is divided into calculation elements (have nodes along its length).

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Anonymous]

Thanks Artur,

So the function described here: http://docs.autodesk.com/RSA/2012/ENU/filesROBOT/GUID-07FF6757-9008-47EC-B497-A4CD862BCFE-1040.htm has nothing to do with the critical buckling load?

Is there any way to implement effective length calculated by buckling load into AS4100 steel design module?

Thanks

[Artur.Kosakowski]

The automatic buckling length in the Steel Design code is based on calculations of stiffness of the column itself and elements at its top and bottom nodes. This is the typical approach that I believe is described in any steel design code.

[Anonymous]

Artur,

There was some debate in our office about a clause in the AS4100 code that says;

So essentially you can gain a bit of capacity in buckling by running an Elastic Critical Buckling Analysis and justifying a slightly reduced ke factor.  Some programs available here allow a 'ke' calculated by critial load analysis to be automatically assigned to the member in steel member design module. 

My original question is now kind of redundant as I realise that;

To achive a truly refined effective length for EVERY member in a frame you would need to run a seperate critical buckling analysis for EVERY member.  Then get a refined 'ke' for each member individually.  In a large frame this is obviously not effeicient.  But for certain applications it may be useful. 

Anyway thanks for your help as usual!

Tony

[Artur.Kosakowski]

Tony,

Looking at AS4100 I can see that the automatic procedure of calculating buckling length as available in the Steel Design module of Robot seems to be what the code describes in 4.6.3.3 and 4.6.3.4. I was sure that this approach has to be common for all steel codes

This is based on stiffness of the column and the adjoining beams rather than buckling analysis itself.

My assumption about the point you refer to is that it is for structural roof (build from number of pyramids/triangles forming a space truss) rather than rectangular frames the Robot option is intended for.

[Anonymous]

Artur

May I please raise again the topic of discussion outlined by Tony.

(Tony and I are work colleagues)

Your latest response in the discussion chain references two clauses of section 4.6 of AS4100, which is the Member Buckling Analysis design approach. 

The design approach using Frame Buckling (section 4.7) is alluded to in the design standard.

• There are numerous research papers available that discuss this design approach.
• Sydney University research report number R891 - Frame Buckling Analysis (2008) is one of these.
• The conclusion drawn in paper R891 is that there are design efficiencies that can be achieved by designing using the Frame Buckling approach, as compared to the "traditional" approach (AS4100 Section 4.6) of calculating the effective lengths from the member end conditions.

The Frame Bucklimg method of design is automated in other design software packages. Namely, the user can make the software package automatically calculate member effective lengths and from these the software automatically calculates the member buckling loads. The member buckling capacities are affected by stiffnesses and also the load distribution within the frame. ie The member buckling capacities vary with each load case.

Clause 8.4.2.2 (refer one of Tony's emails above) allows the designer to design braced and sway compression members with an effective length factor of 1.0 (using results from a 2nd order analysis), provided that the member is also checked for buckling under the axial load alone. It is this latter condition that the automated software packages are checking for when the user requests the software to calculate effective lengths from a frame buckling analysis.

I have designed many structures using the Frame Buckling approach.

In some cases the design by Member Buckling would lead to the same result, however in structures with complicated geometry and/or partially braced columns, design by Member Buckling would result in heavier designs.

Design by Frame Buckling is impractical however, unless the software automates it.

From what I understand of Robot, the software does calculate the effective lengths of the frame members based on a Frame Buckling analysis, and these can be extracted from the model. However, can this be progressed to enable automation of the member design using effective lengths calculated from the Frame Buckling analysis run by the software?

Thanks

Tom Rose

[Artur.Kosakowski]

I'm sorry but the buckling analysis in Robot is not intended for calculations of buckling lengths of particular elements of a model and there is no automatic way of extracting such information from the model so that it could be used for the steel design part.

Do you know where I could possibly find a step by step example of the approach described in the section 4.7 so that I could better understand it? 

I can see that there is a growing interest in alternative approaches to the 'classic' solution described in the Chapter 4.6 which is adopted in Robot e.g. the Direct Analysis Method used in the US but that approach seems to be not the same as what is written in the chapter 4.7 of AS4100.

[Anonymous]

Artur

I dont have a guide to show the step by step process of design using this method.

I can say that it takes some practice to develop a good understanding of the method, and to be able to pick up modelling errors which frequently occur in 3D models. The design method comes into its own when designing geometrically complex structures, in which the traditional effective lengths are not evident.

On googling the subject, the following book came up, and seems to me to present a clear description of the process in part - Steel Structures Design Manual to AS4100, Brian Kirke (Senior lecturer in Civil Engineering, Griffith University) and Iyad Hassan Al-Jamel (Managing Director, ADG Engineers, Jordan)

In terms of how the software calculates the mathematical solution - I cant help there!

Once the program has calculated the effective lengths for the compression members, it is a simple step to then calculate Ncx and Ncy. This is the latter check in the process defined in clause 8.4.2.2 where it states - "provided clause 6.1 is satisfied for Nc calculated using .........."

Tom