Buckling Analysis for Alpha_cr Value

[Anonymous]

I have a couple of helpful attachments that will make the following discussion easier to understand.

The first is an output from the Steel Construction Institute's Frame Stability tool. It relies on the same basic principles of EC3 to calculate the Alpha_cr value of a simple braced frame. 

The second is an equivalent Robot model.

I've tried to produce the same basic steel frame in both the SCI tool and in Robot, with the same loading. Subsequently, I followed this Robot guidance (https://knowledge.autodesk.com/support/robot-structural-analysis-products/troubleshooting/caas/sfdca...) to get an alpha_cr value from Robot.

In the SCI output, I get three alpha_Cr values (one for each storey) and the lowest is at the lowest storey (as expected). In the Robot model, in the Buckling Analysis Results table, I only get one number (presumably for the whole building) and it's similar to the lowest output value from the Frame Stability tool. However, if the Robot output is only giving me one result, I would assume it's for the overall frame (from ground to roof level). Can I get some confirmation as to how Robot is getting this singular result, and how I might get a result for each storey -- the recommended check for sway.

[Anonymous]

*** I have editted the final paragraph of the original post after some further developments *** 

In the SCI output, I get three alpha_Cr values (one for buckling at each storey) and the lowest critical value is at the lowest storey (as expected). In the Robot Buckling Analysis Results table, I only get one number for each mode. Based on past forum posts, I gather that the number is related to a mode shape, so I check the mode shape of the first three modes (see image below) and I believe that the first three mode shapes are consistent with the collapse of the first three storeys, respectively. However, the critical values are nearly zero. Something seems to be incorrect here.

Can I get some confirmation as to how Robot is getting this singular result, and how I might get a result for each storey -- the recommended check for sway.

(NOTE: I have further edited the Robot model to include X-braces, rather than a single brace at each storey. This didn't really affect the result)

[Pawel.Pulak · Accepted solution]

Hello Mike,

answering and commenting various topics of your posts above:

1/ Results of buckling analysis are not related to any specific story - structure is treated as w whole

2/ The screen capture above shows 3 different buckling modes - but all of them are related to local buckling of bracings. It would be necessary to calculate and check higher modes to observe some global buckling in the model for which this screen capture was made

3/ As concerns critical buckling coeffcient Alpha_cr related to stories you probably mean this parameter calculated using the method similar to formula (5.2) from EN 1993-1-1:2005 (EC3) - " αcr.=(Hed)*h/(Ved*δh,ed)". It was already discussed in this and this forum threads.

It is available in the table of stories providing stories are defined in your model.

When comparing to SCI tools pay attention to the same conditions - like pinned connections in columns, beams and bracings.

---------------------------------------------
If this post answers your question please click the "ACCEPT AS SOLUTION" button (the button visible when logged to the forum). It will help everyone to find answer more quickly!

Regards,

[Anonymous]

Pawel, huge thank you! I'm glad that you've helped me understand what the Critical Buckling Factors were doing -- I was starting to understand that concept before we chatted earlier, but you've helped clarify.

As for the alpha_cr: I am glad to be on the right track now. I think I was just caught off guard by the other guidance I linked to in the first post, suggesting to check this buckling factor. I gather that the buckling factor I was looking at, would be equivalent to the alpha_cr value in the case of a portal frame, as there wouldn't be these lower buckling modes of the braces.

I will look at those threads you've linked to and follow up if necessary.

Cheers 🙂

[Anonymous]

Hi Pawel,

I've spent an embarrassingly long time trying to get some match between the SCI tool and the Robot alpha_CR value. I've learned a lot from the investigation into the values that both tools (Robot and SCI calculator) are putting out, and I've summarized the ongoing problem in an attached PDF with extensive mark-up. I've also included the Robot file. You might first look at the final image on page 2 of the attached PDF; if you are able to quickly figure out / fix the lateral deflection of the truss under gravity-only loads, then you might solve the whole problem. Otherwise, some further explanation may be necessary.

Cheers