material utilization - mesh resolution + model view

Hello,

I am working on a model of concrete building containing of slabs (shells) + columns and beams (beams). I have two questions regarding material utilization in Karamba. I went through this topic in manual but didn't find any answers.

First - I noticed that mesh resolution has impact on utilization results (from Utilization component). It is reasonable that finer mesh would give more precise results but differences are huge. For example, going from resolution of 1 to resolution of 2 results in almost 2x smaller utilization (0.6->0.3) of shells (slabs) and also influences utilization of beam elements (0.9 -> 0.7). Same rule applies to calculated deflections. What suprised me is that it was intuitive for me to expect bigger utilization and displacements when density value is smaller - as bigger mesh faces would result in bigger forces applied to vertices. In the meantime it is the other way round.

What is the reason for that? Is it recommended to ALWAYS go with mesh resolution of 1, as there is such big difference in results? Can i assume that with resolution of 1 results are closer to truth?

The second thing i wanted to ask about is the difference between legend tags (Shell View and Beam View components) and Utilization component. When i get max. utilization value of 0.6 from Utilization component i would expect extreme value to be 60%/-60% in Beam/Shell View components. In the meantime I get around +/-150% as max. utilization in legends in that case. What is the reason for that?