Question
When solving the same problem twice, I notice that the contact forces in some areas change slightly, why might this be?
Solution
The contact forces will vary due to:
- the static indeterminacy of your structure, and
- the way in which modern microprocessors perform complex numerical calculations (basically numerical truncations mean that the solver will not necessarily reach exactly the same solution if there are many which are equally valid).
In LimitState:RING the solver is trying to find the critical distribution of forces within the structure. In regions of movement the critical distribution of forces is generally uniquely defined, so contact forces there are also uniquely defined (and hence won't change between solves). However, away from regions of movement there are many possible distributions of forces, and the software will choose, and display, just one of them.
However, because there are many equally valid solutions, and because of modern microprocessor design, precisely the same distribution won't necessarily be shown in successive runs.
If you want to increase the likelihood of forces being uniquely determined then you need to reduce the degree of static indeterminacy.