thanks for this incredible video i have a question , why there is second mode and third mode of buckling while it's supposed to collapse at first mode of buckling ?
In Linear Buckling it's just a "math operation" - you can calculate something akin to "what would be a second mode, if the first one would not exist" etc. In nonlinear FEA, you can't do that - the structure will fail into the first mode, and there isn't much you can do to obtain a second one (apart from applying load very fast in some cases, but that is beyond static analysis :) )
The video was very insightful but it would be great if you could please tell us about the boundary conditions and the type of element that you used for this particular example.
Hey Mian! Sorry I missed your comment - I rarely visit my won youtube channel :) It is far more effective to comment and ask on the blog: www.enterfea.com :) To be perfectly honest I don't remember what BC I applied as I have made this model so long ago (it was nothing particular - just an example). But as I know myself that would most likely be: Top - radial direction supported Bot - radial direction supported, on the place where the vertical support is additionally vertical direction supported. Symmetry on vertical edges. Elements were QUAD4 if I recall, less likely they were QUAD8. I hope this helps you. If you have any more questions feel free to ask :) All the best! Ł
Well... I don't know... because I don't know what you mean by "post buckling" - I've heard this name a few times, but I was never able to pin point what folks mean precisely by this term... However, there is a chance this is the same - let's say it this way if I would have to guess, I would guess that it's the same... the thing is, I don't like guessing, especially in engineering!
I haven't used Abaqus in years... but back in the day, it was simply called "Riks" (like the Riks algorithm). Perhaps they changed it to something else in the last decade or so - I have no clue... This may be called "arc-length", Riks, Modified Riks, or Crisfield... I would guess that would be the most common "names" out there. All the best! Ł
Yes, i.e. in plate bending LBA will see a lot of "fake" buckling (since it cannot see membrane state), and it isn't there - so technically the nonlinear analysis in such case would give a better outcome than LBA. In shells, that would be rare, and I would be careful - it is possible to "miss" buckling and calculate your model in an unstable equilibrium in FEA if you are not sure what to do. All the best, and good luck!
Hey Tony! I try not to say anything definite about the things I know nothing about :) I only used Adina, and it's ok (some things are iritating with how the arc-length analysis is defined and how to "steer" the arc during analysis. As for other two, I have no idea really... I guess that MSC is almost the same as NX Nastran, so you will end up with SOL 106 most likely. It is super easy to set up, but it's not as "powerful" as SOL 601 (Adina). I would also guess that Ansys is at least as good as Adina in solving nonlinear stuff (maybe better), but I have no knowledge to back this up - just a hunch really...
I modeled the same geometry and same load with zero displacment in the x and y direction for the upper edge und fix support for the small feets and I got 0.62 from Linear buckling (LBA). What could be the reason ?
Hey Basem! First of all I'm glad that you were inspired enough to try the example :) I think (in all honesty I can't remember...) that I used zero displacements on horizontal directions (yours x and y I assume) at the bottom edge as well, with the support additionally with zero vertical (your z) deformation. However, I think I allowed for all of the rotations. See if that will work for you, and let me know! All the best! Ł
@@basem506saeed Hmmm... that is interesting for sure. I think that there is a video on my channel when I do a nonlinear shell analysis - I think that it was the same shell (I hope!). There is a step by step guide on how I support and define the model there - maybe this will help...
@@oguzhanakarsu6703 No, no... this is the outcome of a SINGLE nonlinear analysis. I mean, you do a nonlinear analysis, and the outcome is, that the model fails under 0.612 of the load - there is no need to guess anything, nor do the analysis several times to get the GNA outcome - the nonlinear analysis gives you such an answer in a single run. Otherwise, if guessing would be involved... design would take forever, and wouldn't be very practical :)
Be careful with this Mate... while in many cases the 15% will be decent, you can encounter a difference that is substantially different. And if yielding takes part in the failure (i.e nonlinear material) the difference can be even bigger!
You mean between linear and nonlinear buckling? That would depend on the case I guess, as I still use LBA in some simple occasions... but if I need to make a "blanket answer" then nonlinear buckling all the way :)
@@TonyA-ex1sj, To be honest, I don't see your first question... so I'm not sure what this is about. I'm using Femap with NX Nastran if this helps you... Please write your question again so I can answer it... since apparently, it disappeared...
This was a very good explanation. Thank you.
You're very welcome!
thanks for this incredible video
i have a question , why there is second mode and third mode of buckling while it's supposed to collapse at first mode of buckling ?
In Linear Buckling it's just a "math operation" - you can calculate something akin to "what would be a second mode, if the first one would not exist" etc. In nonlinear FEA, you can't do that - the structure will fail into the first mode, and there isn't much you can do to obtain a second one (apart from applying load very fast in some cases, but that is beyond static analysis :) )
The video was very insightful but it would be great if you could please tell us about the boundary conditions and the type of element that you used for this particular example.
Hey Mian! Sorry I missed your comment - I rarely visit my won youtube channel :) It is far more effective to comment and ask on the blog: www.enterfea.com :)
To be perfectly honest I don't remember what BC I applied as I have made this model so long ago (it was nothing particular - just an example). But as I know myself that would most likely be:
Top - radial direction supported
Bot - radial direction supported, on the place where the vertical support is additionally vertical direction supported. Symmetry on vertical edges.
Elements were QUAD4 if I recall, less likely they were QUAD8.
I hope this helps you. If you have any more questions feel free to ask :)
All the best!
Ł
Thank you. Perfect video !
Thank you Eugene! :)
Is non linear buckling and post buckling same?
Well... I don't know... because I don't know what you mean by "post buckling" - I've heard this name a few times, but I was never able to pin point what folks mean precisely by this term... However, there is a chance this is the same - let's say it this way if I would have to guess, I would guess that it's the same... the thing is, I don't like guessing, especially in engineering!
Thank you
You're welcome Saad :)
Hello
Which step is used for non-linear buckling analysis in abaqus?
I haven't used Abaqus in years... but back in the day, it was simply called "Riks" (like the Riks algorithm). Perhaps they changed it to something else in the last decade or so - I have no clue... This may be called "arc-length", Riks, Modified Riks, or Crisfield... I would guess that would be the most common "names" out there.
All the best!
Ł
hi, pls let me known buckling load the bigger the better or the smaller the better. thanks
Hey Mate! It's better if the buckling strength is bigger - this means that the element can take more load before it fails due to buckling :)
Is It possible that the non linear buckling gives to me An higher load than linear?
Yes, i.e. in plate bending LBA will see a lot of "fake" buckling (since it cannot see membrane state), and it isn't there - so technically the nonlinear analysis in such case would give a better outcome than LBA.
In shells, that would be rare, and I would be careful - it is possible to "miss" buckling and calculate your model in an unstable equilibrium in FEA if you are not sure what to do.
All the best, and good luck!
which of these 3 programs do you think is the best MSC Nastra,ADINA o ANSYS?
Hey Tony!
I try not to say anything definite about the things I know nothing about :) I only used Adina, and it's ok (some things are iritating with how the arc-length analysis is defined and how to "steer" the arc during analysis. As for other two, I have no idea really...
I guess that MSC is almost the same as NX Nastran, so you will end up with SOL 106 most likely. It is super easy to set up, but it's not as "powerful" as SOL 601 (Adina).
I would also guess that Ansys is at least as good as Adina in solving nonlinear stuff (maybe better), but I have no knowledge to back this up - just a hunch really...
I modeled the same geometry and same load with zero displacment in the x and y direction for the upper edge und fix support for the small feets and I got 0.62 from Linear buckling (LBA). What could be the reason ?
Hey Basem! First of all I'm glad that you were inspired enough to try the example :)
I think (in all honesty I can't remember...) that I used zero displacements on horizontal directions (yours x and y I assume) at the bottom edge as well, with the support additionally with zero vertical (your z) deformation. However, I think I allowed for all of the rotations.
See if that will work for you, and let me know!
All the best!
Ł
@@Enterfea Actually, I tried both and still getting lower values of eingenvalues
@@basem506saeed Hmmm... that is interesting for sure. I think that there is a video on my channel when I do a nonlinear shell analysis - I think that it was the same shell (I hope!). There is a step by step guide on how I support and define the model there - maybe this will help...
Hello! How did we get 0.612 in GNA? Depends on what value?
Hey! There is a 50kN/m load on the top circumference... if you would apply 0.612 of that load in GNA analysis you will get failure.
Hope this helps!
Ł
@@Enterfea Thank you! But why 0.612, you found it as a result of the analysis or by trying?
@@oguzhanakarsu6703 No, no... this is the outcome of a SINGLE nonlinear analysis. I mean, you do a nonlinear analysis, and the outcome is, that the model fails under 0.612 of the load - there is no need to guess anything, nor do the analysis several times to get the GNA outcome - the nonlinear analysis gives you such an answer in a single run. Otherwise, if guessing would be involved... design would take forever, and wouldn't be very practical :)
What code did you use for this situation?
Hey! I'm using NxNastran in Femap mostly (this is why most of the model is in the "famous" magenta... but I'm switching to gray as the base color :) )
thank you!
You are most welcome Jose!
Thumb Rule of 15% is something I am going to remember till death
Be careful with this Mate... while in many cases the 15% will be decent, you can encounter a difference that is substantially different. And if yielding takes part in the failure (i.e nonlinear material) the difference can be even bigger!
what would you choose?
You mean between linear and nonlinear buckling? That would depend on the case I guess, as I still use LBA in some simple occasions... but if I need to make a "blanket answer" then nonlinear buckling all the way :)
@@Enterfea no dude, I meant, what software do you choose? Did you read my first question?
@@TonyA-ex1sj, To be honest, I don't see your first question... so I'm not sure what this is about. I'm using Femap with NX Nastran if this helps you... Please write your question again so I can answer it... since apparently, it disappeared...