Francesco Casella
Francesco Casella
OK with the new nightly the lexer issue is fixed. The syntax highlighting issue is still there. BTW, I ran the model and I also go ``` Process crashed Simulation...
Adding @AndreaBartolini, @deepak19015 to the loop.
@mwetter I guess you will also port the fix to 10.0.x and 9.1.x, right?
I think @phannebohm's analysis is correct. In fact, the problem arises after the fault has cleared, when unfortunately the nonlinear solver latches on a bogus solution with extremely low voltage...
I checked further the `SpaceCooling` model and I think I have a good trail. The model has three Boolean variables` mulStaDX.watVapEva.off`, `sinSpeDX.watVapEva.off`, and `varSpeDx.watVapEva.off`, which are switching a large number...
Hmm, I added `max = 0.05` to the declaration of `XEvaWetBulOut` but I still get divisions by zero. I turned on -lv=LOG_NLS_V, it turns out the first divisions by zero...
> `xloc` is the local name for the array of iteration variables in an algebraic loop inside our generated C code. We do a NaN check at runtime, currently for...
@bilderbuchi see #12193.
I guess some simple heuristics could already do wonders. If you have this pattern ```modelica if then v1 = ; v2 = ; v3 = ; else v3 = ;...
See #3915 and #5346