Seismosoft | Earthquake Engineering Solutions

Hello everyone,

I'm modeling an unreinforced masonry wall using 'masonry' sections, but I'm not satisfied with the hysteresis response. I tried adjusting the deterioration factors, but they had little effect.

So, instead, I modeled the wall manually using separate elements, as suggested in the documentation: modeling shear with lateral springs at the element ends and flexure using a fiber section. This approach seemed promising, but I ran into issues:

1. Using MIMK_Pinched for lateral springs: It works fine without deterioration, but once deterioration is added, the spring behaves normally until it yields. After that, it loses stiffness completely and doesn't recover in subsequent cycles.
2. Using MIMK_PeakOriented: The same issue occurs, even without any deterioration defined. Once yielding happens, the spring no longer provides stiffness beyond the peak of the first cycle.

In both cases, the analysis runs without numerical divergence or errors, but the springs behave as if they have no residual stiffness once plasticity is triggered.

• Has anyone faced this before? Any ideas on the cause or possible solutions?
• In the case of 'masonry' element, how can I reduce the pinching effect? The hysteresis is too pinched.

Thanks in advance!
Alireza

Both MIMK curves have several parameters that define the way their hysteresis works, but also the way that the strength deteriorates. From your description, and since you do not face numerical difficulties, it seems that you are introducing significant deterioration with the parameters of the model. Please check all these parameters, carrying also sensitivity studies before.

In the case of masonry element, simplified hysteretic curves are introduced and default values are employed for the shear curve. Note however that the response is also determined by the behavior in bending of the internal fiber-based element.

Seismosoft Support