Learning Forum

[asade010]

Hello everyone,

I have posted previously about issues with EREINF command and REINF264 elements, unfortunately, I did not get any reply.

The issue is that all the reinforcement in the Z direction gives erroneous results. I have made a simple model of a cube with reinforcement in 3 directions and this can clearly be seen in the results.

Does anyone have any suggestions or know where the problem is?

Many thanks,

[Sheldon Imaoka]

Hi Amir,

Thanks for taking the time to demonstrate the issue with a simple model.

It looks like this may be a graphical problem with REINF264 and /ESHAPE,ON.  The results seem to be fine, but the display of REINF264 with /ESHAPE,ON seems incorrect.  The conditions seem to be (a) when rebar lies on edge of underlying solid element and (b) rebar is pointed in global z-axis, although the combination of (a) and (b) do not always give this graphical error.  (I have filed a bug report on this issue you reported.)

I believe this is a graphical error since the displacement plots of the underlying solid look symmetric and correct.  Note that the REINF264 elements share the same nodes as the SOLID185 elements - you can confirm this with ELIST.  Thus, although REINF264 plot with /ESHAPE,1 shows a displacement of 16.3953, this is due to a graphical issue with /ESHAPE,1 and REINF264.  (REINF264 is actually a hexahedral element, just like the underlying SOLID185 element, but Mechanical APDL manipulates the display just to show the rebar, and that is where this issue is arising, I believe.)

In your actual model, please confirm the displacement solution of the homogenous materials - I think that you will find it to be OK.  Unfortunately, I don't have a good workaround for the graphical issue of the REINF264 plots.  One option is to change the mesh density so as not to have rebar along element edges (in your cube model, change element size to 100mm or 40mm - the rebar doesn't pass through the element edge, and plots seem to be fine).  What is also interesting is that if you change the alignment of the model so as not to be exactly parallel to global z-axis, it should work fine - however, I understand that these two workarounds may not always be practical. Otherwise, you may wish to plot rebar stress with /DSCALE,1,OFF (no deformed shape) since, as noted above, the results seem to be correct - it is just a graphical issue of plotting REINF264 displacement results.

Regards,
Sheldon

 

 

[asade010]

Thanks for the reply Sheldon,

That was my guess too. The results in my bigger models seem to be close to the experiments. I just wanted to make sure I am not making any mistake in the definitions and this is just a graphical error. Anyways, I will be looking forward to the result of the bug report you mentioned, if you please share with me when you get feedback.

Regards,

Amir

[Sheldon Imaoka]

Hi Amir,

I just wanted to let you know that this graphical problem will be resolved in Release 2021 R1.  Thanks for reporting this issue to us.

As noted previously, it's an issue with just REINF264 when rebars are parallel to z-axis and rebars go through the underlying element edge - the solution for non-REINF264 should still be correct.

Regards,
Sheldon