Mesh Improving

Dear Francesc,

It is quite difficult to estimate the problem without the project.
Below you can find my assumptions:

1. 4.5×10^−3 is a bad mesh quality, but not so critical. I think a minimum value of 0.0001 or less is more critical for the calculation. Moreover, I can assume that the volumes inside the pit will be deactivated, which means that bad elements will also be deactivated. Probably, you are not interested in the results in this area.

2. The yellow bar on the legend is at 0.1. It's more interesting to know where the bad red elements with mesh quality less than 0.001 are.

3. The average size of a element, and therefore the local size of a element, depends on the size of the entire model. This means that if you have a large model (large model diagonal), the average element size will also be large. I understand that the boundary conditions should not affect the results, but perhaps you can reduce the size of the model a bit.

4. Is it possible to change the volume of the "wedging" soil? Is this layer so important in terms of its parameters? Maybe you can combine it with neighboring soils.

5. I might suggest adding some vertical surfaces that intersect the bad elements. Alternatively, you can add some volume to this area and then refine the mesh in it.

I hope some of this helps.
Good luck!

Dear Francesc,

It is quite difficult to estimate the problem without the project.
Below you can find my assumptions:

1. 4.5×10^−3 is a bad mesh quality, but not so critical. I think a minimum value of 0.0001 or less is more critical for the calculation. Moreover, I can assume that the volumes inside the pit will be deactivated, which means that bad elements will also be deactivated. Probably, you are not interested in the results in this area.

2. The yellow bar on the legend is at 0.1. It's more interesting to know where the bad red elements with mesh quality less than 0.001 are.

3. The average size of a element, and therefore the local size of a element, depends on the size of the entire model. This means that if you have a large model (large model diagonal), the average element size will also be large. I understand that the boundary conditions should not affect the results, but perhaps you can reduce the size of the model a bit.

4. Is it possible to change the volume of the "wedging" soil? Is this layer so important in terms of its parameters? Maybe you can combine it with neighboring soils.

5. I might suggest adding some vertical surfaces that intersect the bad elements. Alternatively, you can add some volume to this area and then refine the mesh in it.

I hope some of this helps.
Good luck!

Dear Rushan,

1. 4.5×10^−3 is a bad mesh quality, but not so critical. I think a minimum value of 0.0001 or less is more critical for the calculation. Moreover, I can assume that the volumes inside the pit will be deactivated, which means that bad elements will also be deactivated. Probably, you are not interested in the results in this area.

Actually there are no elements with values less than 0.001 so it does not seem critical as you say. Regarding its location some of them are behind diaphragm walls so tthis is why we are trying to improve it.

2. The yellow bar on the legend is at 0.1. It's more interesting to know where the bad red elements with mesh quality less than 0.001 are.

Yes, we usually modify legend setting to find the worst quality elements.

3. The average size of a element, and therefore the local size of a element, depends on the size of the entire model. This means that if you have a large model (large model diagonal), the average element size will also be large. I understand that the boundary conditions should not affect the results, but perhaps you can reduce the size of the model a bit.

As you say we set the minimum model boundaries in order to not be afected by them.

4. Is it possible to change the volume of the "wedging" soil? Is this layer so important in terms of its parameters? Maybe you can combine it with neighboring soils.

We realize that thin layers do not get along with mesh quality. This is an option if we get to issues during calculation.

5. I might suggest adding some vertical surfaces that intersect the bad elements. Alternatively, you can add some volume to this area and then refine the mesh in it.

Actually, we defined a local soil volume and an intersection line to refine this area. We will try your suggestion.

Thanks Rushan!

Dear Francesc,

Thank you for the feedback.

I wish you to solve this issue as soon as possible.