Intuitive

Hi All, I'm trying to model the behaviour of pile-raft (250mm thk. slab and 300mm dia. piles) supporting a 1-storey building with the piles modelled as spring support. The software I'm using is ROBOT structural analysis. After running the analysis, as you would expect, the support at the centre where the loads are at its highest gives a larger reaction than the ones on the edge. Once I reduce the spring constant value, the reactions still exhibits a similar behaviour, however I realized that now the loads redistributes itself in a way that the support receiving higher loads at the centre are now giving a lower reaction (as compared to the case with larger spring constant value), and the edge supports are giving a higher reaction (as compared to the case with larger spring constant value). I have also created various other models as experiments, and it seems to be giving the same trend. It's almost like the loads try to ''even out'' itself everytime I reduced the spring constant value. So I'm just wondering, is this the sort of behaviour we should be expecting from a geotechnical point of view? From what I understand, the load distribution should be the same if the spring constant value for all supports are the same as the stiffness are all equal. And regardless of what value of spring constant you use, the sum of reactions should be the same. Replies will be much appreciated, many thanks in advance.