ABAQUS BUCKLING RIKS ANALISYS

[mikylangel]

Hi, I am doing a riks analysis of a cylindrical fuselage with windows. I have applied contact interactions between the different parts, and I have applied a unit value load to a reference point, joining the displacements of that point to my figure. However, the LPF comes out constant, and does not come out as it should, since I understand that it should come out a maximum and then decrease. Does anyone know what is the reason for this?

Attached is my LPF graph

 

[FEA way]

LPF vs arc length plots are rather unintuitive. It would be better to plot force vs displacement for a representative region (reference point). Also, have a closer look at the deformation of the model in different stages of the analysis (animation will help).

 

[mikylangel]

How can I graph force vs displacement?

 

[FEA way]

Either request those variables as history output for the reference point before running the analysis or use Create XY Data --> ODB field output to extract them from that point. Then use Create XY Data --> Operate on XY data to combine those two variables into a single force vs displacement plot.

 

[mikylangel]

I have graphed lpf vs displacemnet and I still get the same thing. The problem is that the LPF comes out constantly, and it shouldn't. Any idea why I get constant?

 

[FEA way]

I was thinking about applied force (or just reaction force) vs displacement. What does this plot look like ? Do you have any imperfections included in this analysis ?

 

[mikylangel]

I have not included imperfections, should i?. The applied force is of constant unit value at the reference point, shouldn't it continue to come out the same?

 

[FEA way]

Imperfections are almost always included in Riks analyses so that you can get past the bifurcation. That's a very likely source of the problem here.

Riks procedure scales the applied loads (hence the additional LPF output) but you can still use the reaction forces to track them directly.

 

[mikylangel]

How could I include imperfections? I have never done it, do you have any ideas or video tutorial?

 

[FEA way]

The easiest way is to add them as scaled mode shapes from linear buckling analysis preceding the nonlinear one. Check the documentation chapter "Introducing a Geometric Imperfection into a Model".