[BUG] Serious bug related to shear design of quad elements

Dear SOFiSTiK team,

I am writing to inform you about a rather serious bug related to shear design of quad elements. I am currently using SOFiSTiK version 2025-4

I have created a simple example to illustrate the problem. In the example I have modeled a wall that is loaded with lateral earth pressure:

I have modeled the wall with two different support conditions. In one case the supports are described with springs and in other case with fixed lateral supports.
The spring stiffness is chosen such that the internal force distribution in both walls would be the same.

Despite having equal internal forces, the shear reinforcement requirements are different:

In case of spring supports I have noticed that for many nodes and elements SOFiSTiK incorrectly takes the shear force in shear design as VEd = 0. But also values other than 0 are sometimes incorrect. A comparison between shear forces in quad elements for group 11 and 12 is given below, where group 12 has correct values:

The example model:
Bemess shear bug.sofistik (109.3 KB)

I discovered this bug when checking results for a similar case where SOFiSTiK gave shear reinforcement requirements only on top of the wall, where shear forces were significantly lower than at the bottom of the wall (this is without vertical loading so normal forces are not the cause):

With best regards,
Martin

Hello Martin,

I checked the behavior. It seems related to “width of support” BFIX which is defined in SOFiMSHC. In supports, command SLNS, SFIX is default with 1 m. If you change it to e.g. 0 m you get quite similar results. I wanted to give you a fast feedback. I have to check the behavior more detailed and will add information here later.

Br, Adrian

Hello Adrian,

Thank you for the fast response! It indeed seems that BFIX is used to determine the width of support for shear design. Maybe it would be safer if the default value for support width is set to zero?

I often use SOFiPLUS to create model geometry and in case of Structural Line there is an option to select support width that is used for shear design. But it seems that the settings in bedding overwrite this choice (BFIX). Maybe it could be explicitly stated in the structural line window?

To be honest, based on SOFiMSHC manual I also did not realize that BFIX had such an important role in shear design. Maybe the BFIX function could also be given a bit more in-depth description?

With best regards,
Martin