14 Replies Latest reply on May 3, 2016 10:57 AM by Mjnb L. Kjhb

Difference in theoretically calculated and simulated buckling load?

Mjnb L. Kjhb Apr 29, 2016 8:19 AM

I have applied axially compressive load on a I-column and performed the buckling analysis keeping the material to be brass and found the buckling load factor 16. Now I found the Buckling load=16*applied load. Comes out to be 771742

But Theoretically Critical buckling load=pi^2 EI/L^2

If i calculate critical buckling load through this, i get 9817835. How is this possible?

Re: Difference in theoretically calculated and simulated buckling load?

James Riddell Apr 29, 2016 11:51 AM (in response to Mjnb L. Kjhb)

Have you refined your mesh size?
Like Show 0 Likes(0)

Actions
- Re: Difference in theoretically calculated and simulated buckling load?
  
  Mjnb L. Kjhb Apr 29, 2016 12:55 PM (in response to James Riddell)
  
  I just now decreased the element size from 12.5 to 8.5 and it took 20 minutes simulation to get buckling load factor 15.5 which is even lower than the earlier one(15.9). The error only increased. Isn't there some other explanation?
  
  Like Show 0 Likes(0)
  
  Actions
  - Re: Difference in theoretically calculated and simulated buckling load?
    
    James Riddell Apr 29, 2016 3:44 PM (in response to Mjnb L. Kjhb)
    
    What is your 'n' factor? Did you account for the different end conditions?
    
    Like Show 0 Likes(0)
    
    Actions
    - Re: Difference in theoretically calculated and simulated buckling load?
      
      Mjnb L. Kjhb May 2, 2016 7:37 AM (in response to James Riddell)
      
      Yes, I did account for end condition. For simply supported beam, n=1. Anyway, n factor cannot be as small as 0.08 as required by it to be consistent.
      
      Like Show 0 Likes(0)
      
      Actions
      - Re: Difference in theoretically calculated and simulated buckling load?
        
        James Riddell May 2, 2016 8:54 AM (in response to Mjnb L. Kjhb)
        
        I can't tell from your description and there is no picture but my first guess is that you are fixing the ends improperly in your model.
        
        And what is the answer to Seckin's question?
        
        Like Show 0 Likes(0)
        
        Actions
        
        Re: Difference in theoretically calculated and simulated buckling load?
        
        Mjnb L. Kjhb May 3, 2016 7:59 AM (in response to James Riddell)
        
        I fixed one end with roller/slider support and the other end with Ref Geometry (X=0,Y=0,Z=0) to give the hinge support.
        
        Like Show 0 Likes(0)
        
        Actions
        
        Re: Difference in theoretically calculated and simulated buckling load?
        
        James Riddell May 3, 2016 8:30 AM (in response to Mjnb L. Kjhb)
        
        Then you do NOT have the n=1 case obviously. Also, buckling equations that you used to check do not account for other than end-on loading. So, you have mis-applied that analysis.
        
        Like Show 0 Likes(0)
        
        Actions
        
        Re: Difference in theoretically calculated and simulated buckling load?
        
        Mjnb L. Kjhb May 3, 2016 9:10 AM (in response to James Riddell)
        
        http://www.imbm.bas.bg/tm/jtam/files/42-4/19-30.pdf In this paper it is derived for one end pinned (hinged) and the other end roller, n=1. I do not understand your answer
        
        Like Show 0 Likes(0)
        
        Actions
        
        Re: Difference in theoretically calculated and simulated buckling load?
        
        James Riddell May 3, 2016 9:49 AM (in response to Mjnb L. Kjhb)
        
        That is not how you have restrained & loaded this model.
        
        Like Show 0 Likes(0)
        
        Actions
        
        Re: Difference in theoretically calculated and simulated buckling load?
        
        James Riddell May 3, 2016 9:56 AM (in response to James Riddell)
        
        Also, the paper you cite develops into Euler's equation. The side load is not included in Euler's method, merely vertical loading.
        
        Like Show 0 Likes(0)
        
        Actions
        
        Re: Difference in theoretically calculated and simulated buckling load?
        
        Mjnb L. Kjhb May 3, 2016 10:57 AM (in response to James Riddell)
        
        I have not made this a case of simply supported beam? Have I not made this one end hinged (pinned) and other roller? What is wrong with it?
        
        Like Show 0 Likes(0)
        
        Actions
Re: Difference in theoretically calculated and simulated buckling load?

Seckin Uslu May 2, 2016 5:24 AM (in response to Mjnb L. Kjhb)

could you please share us some details. Beam type , force and boundry contidionals.
Also did you use solid or beam elements?
You can get good results with solid elements
Like Show 0 Likes(0)

Actions
- Re: Difference in theoretically calculated and simulated buckling load?
  
  Mjnb L. Kjhb May 3, 2016 7:59 AM (in response to Seckin Uslu)
  
  Beam is I-beam. Force axially 44482N both ends and bending load 22241N applied 30in away from hinged end on the 80 in long beam. Boundary condition is one end roller supported and the other end pinned (hinged) (Reference geometry x=0,y=0,z=0) . I used solid mesh. I don't understand how to check if solid or beam elements are used.
  
  Like Show 0 Likes(0)
  
  Actions
  - Re: Difference in theoretically calculated and simulated buckling load?
    
    James Riddell May 3, 2016 8:31 AM (in response to Mjnb L. Kjhb)
    
    Also, it does no good to apply a force to a fixed surface - nothing will happen.
    
    Like Show 0 Likes(0)
    
    Actions

Go to original post