Overestimated peak strength for cube compression. Abaqus/Explicit

[MrusTgr88]

Hi,

I have done compression test for concrete cube using Concrete damaged plasticity model. I assigned BC in Reference point which I coupled with a surf of the cube.

I have got stress-strain curve that looks nice, but Peak strength is 2-3 times higher than it should be.

Ia have uploaded pic of force-displacement.

Can anybody help me?

Thanks

 

[bkal]

It is not clear if the stress you are showing in your diagram is tension or compression. You need to consider that concrete model shows higher stress in multiaxial compressions, and the model you used is not appropriate for a hydrostatic (compression) stress state.

 

[MrusTgr88]

Bkal, thanks for your answer!

Curve is for compression. It is simple uniaxial compression test. There is no Hydrostatic compression state.

I have also checked energy. Allae is small, also kinetic energy is small. So the model should work fine.

If you have some suggesstions I would appreciate it!

 

[bkal]

Could you please send a sketch of your model with BC and loads, as well asd material data you have used, or alternatively submit a cae file. You mention your boundary conditions; are you sure that they are not causing triaxial compression in the sample?

 

[bkal]

In a hindsight, the curve looks very much like what you would expect for tensile stresses. This would also explain stress level if expressed you used Pa (N/m^2).

 

[MrusTgr88]

I am sure that I am not using triaxial compression in the sample. I have just assigned BC, U2 in Reference point which I have coupled with upper surface of the cube. And down surface of the cube I have assigned encastre BC.

This picture I have attached is Force-displacement curve. So, when I divide force of 2.3*10^6 N with Area of 150mmx150mm, I get peak stress of around 100MPa, which is huge!

This is curve for compression for sure, and I am sending you curve for tension.

Also I have just attached sketch with BC from ABAQUS.

 

[MrusTgr88]

BC for compression.

 

[bkal]

Can you upload the CAE file if it is not too big or confidential?

 

[MrusTgr88]

Sure!

Here it is.

I hope you will have some suggestions!

 

[bkal]

I've had a quick look at your model and some initial observations are:
- the cross sectional area is 0.2m * 0.2m (not 0.15 * 0.15),
- you have fully restrained all the nodes at the bottom and the top of your model, this has restrained lateral deformation and caused more complicated stress state compared to a pure uniaxial state (poisson's ratio effect),
- you have attempted to apply 25% deformation, which is excessive, but does not make difference apart from skewing your deformation/displacement axis.
- I have tried to put sliding (in plane sliding) BCs to the top and bottom surfaces (providing only a minimum BC to encourage a uniaxial stress state) and the results seem to match better what you would expect.

I will spend some more time over the weekend to look into this.

 

[MrusTgr88]

Sorry, I have sent you cube 0.2x0.2m. I have also used cube 0.15x0.15m, but the peak strength is again 2 times bigger.

How can I put in plane sliding BC to the top and bottom surfaces? And can you sent me the results you have got?

Thanks for your time and effort!

 

[MrusTgr88]

Hi bkal,

is there any progress?

 

[bkal]

To put sliding BC at the bottom surface you need to restrain all the nodes at the bottom surface only in the Y direction (perpendicular to the bottom surface). In addition, you need to restrain one of the corner nodes at the bottom surface in all three translational direction. Also, chose one of the neighbouring corner (bottom surface) nodes, if it is along the X axis, restrain that node additionally in Z direction (to prevent rotation about Y axis) or if it as along the Z axis, restrain it additionally in X direction.

For the top surface, when you tie your reference point to the top surface, you can define which degrees of freedom do you want to tie.

 

[MrusTgr88]

I have just done all of this what you have said, but the results arent good.

Can you send me your CAE file?

Also, have you tried tension test? Just to change the direction of the BC? I have, and it doesnt work well.