Interpretation of the result of the buckling simulation.

[Malenki]

Hi,

Could someone advise me on how to interpret the result of the buckling simulation for its first mode (see first attachment). Is the result safe? Can the buckling be twisted like a first screenshot at the bottom? 

   

Regards

[Malenki]

And archived model to post as above ...

[John_Holtz]

Hi @Malenki

The first mode is showing a rotational mode. It is like a torsion spring since your long bar is held on one end but twist along the length. (See the article Model "blows-up" instead of rotating when viewing displaced shape of a simulation for an explanation.)

I think that the part will not buckle in torsion, but I would like to know what other people think.

The results may be okay, but I suggest that you delete the constraints on the inside diameter of parts 2, 3, 4, 5. Put the constraints on the outside diameter of these parts. (For everyone's information, these parts are like bushings that allow the long bar to slide inside the bushings.) My concern is that the constraints on the same nodes as the sliding/no separation contact could cause some type of problem.

[Malenki]

Thank you for your answer.

As practice, such forms of excitement (straight bar) I have not seen. I doubt that this rotary buckling mode is normal and correct. That mode strongly resembles one of its form natural frequencies - not one of the forms of buckling .

> delete the constraints on the inside diameter of parts 2, 3, 4, 5...

Look at the "scenario 2" (attachment .arch at .zip file). The value of the buckling multiplier is rapidly decreasing, which is obvious in that case.

> Put the constraints on the outside diameter of these parts...

Note that the bar diameter is slightly smaller than the inner diameter of the guide bushes (D=0.2mm) for the stability of the calculation. A large clearance leads to a calculation error. Now I fixed all the surfaces of the sleeve, not just the inside. The shape of each mode has not changed.

I'm still not sure: is it safe?

[John_Holtz · Accepted solution]

Hi @Malenki

Actually, buckling and modal analysis are both forms of eigenvalue solutions, so the displaced shape is the same in both analyses. In design scenario 1, the lowest natural frequency is the rotary motion, and so the lowest "buckling" shape is the same!

But I agree that the rotary buckling is just a result of the analysis and not a real-life issue in the case of your model. Mode 2 is the first real buckling mode. With a buckling load multiplier of 7.8, that design is probably safe.

In design scenario 2, the buckling load multiplier is 1.12. I think that is too small of a factor of safety to be safe. There are a lot of factors that can affect the buckling load, such as

• slight variations in the magnitude or direction of the load
• parts are never straight as in the theoretical model
• "large" displacement effects are not considered in the analysis. If something causes the part to bend to the side, it is no longer straight and will buckle at a lower magnitude.
• constraints are never rigid.

The other thing you need to check is the slenderness ratio. The buckling analysis is valid for long columns only. As it the slenderness ratio gets smaller, real columns start to deviate from the long column formula. See Figure 2 on the documentation page Critical Buckling Load for an example of what I mean.

[Malenki]

Hi John,

Thank you for the feedback and suggestions.I agree with them in principle.

BTW. There are few people willing to share their knowledge. Well, it's a pity.

Regards

Malenki