It seems that for some table formulas the simulation results are not correct even with RestartSolver=true for all table points
TableDissolution.zip
E.g. in the attached example we have a table formulation with higher slope:
Comparing the simulation results with RestartSolver=false vs. RestartSolver=true for all formulation data points shows almost no difference:
However restricting the maximal solver step to HMax=0.001 (minutes) - there is a significant difference
To prevent this, integration should be restricted by the next restart point via
CVodeSetStopTime
(s. also https://computing.llnl.gov/projects/sundials/usage-notes
It seems that for some table formulas the simulation results are not correct even with
RestartSolver=truefor all table pointsTableDissolution.zip
E.g. in the attached example we have a table formulation with higher slope:
Comparing the simulation results with
RestartSolver=falsevs.RestartSolver=truefor all formulation data points shows almost no difference:However restricting the maximal solver step to HMax=0.001 (minutes) - there is a significant difference
To prevent this, integration should be restricted by the next restart point via
CVodeSetStopTime
(s. also https://computing.llnl.gov/projects/sundials/usage-notes