The Aspen converged process became non-convergent after adding sensitivity analysis

There could be various reasons for this issue; here are some possible explanations and solutions: 1. Unstable network connection: If you are using Aspen over the Internet, an unstable network connection may prevent Aspen from converging or may cause errors. Please ensure that your network connection is stable, or try using Aspen without an internet connection. 2. Improper model parameter settings: The setting of certain model parameters may prevent Aspen from converging or cause errors. Please carefully check whether the model parameter settings are correct and try to make adjustments. 3. Data entry errors: There may be mistakes or inconsistencies in the entered data, which can cause errors in Aspen. Please check whether your input data is correct and ensure it matches the main process simulation and sensitivity analysis. 4. Model complexity issue: If your model is too complex, it may prevent Aspen from converging or cause errors. Try to simplify the model or use other software for simulation. If the above methods do not resolve the issue, please try contacting Aspen technical support or other professionals for further assistance. .

Also, I encountered a situation where adding different sensitivity analyses caused a sensitivity result that was previously convergent to become non-convergent; the convergence was restored after removing the newly added sensitivities. What could be the reason for this? Everyone is welcome to discuss.

This may be because, after adding different sensitivity analyses, Aspen recalculates parameters such as the model’s gradient and Jacobi matrix. These parameters may differ from those that led to convergence previously, thereby preventing the model from converging again. Solutions may include: 1. Trying to use a more advanced numerical solver or attempting to adjust the convergence properties of the solver to make it easier for the model to converge. 2. Check whether your sensitivity analysis settings are correct. Certain settings, such as those that require a large number of iterations or the introduction of perturbations, may cause the model to fail to converge. 3. If you have added new input variables to the model, try reducing the number of input variables, or use simpler variables to replace complex ones. 4. If you have added new parameters that span multiple cells (such as process parameters), try separating these parameters from the cell parameters, as this can improve the convergence of the model. 5. If none of the above methods work, it is recommended to redesign the model or use other software for simulation analysis. .

Is anyone else here? Need help – the main workflow doesn’t converge, but the sensitivity is fine. Can this data still be used?

Hello, if the simulation of the main process does not converge, but the data from the sensitivity analysis shows normal values, it is necessary to carefully check whether the parameters of the main process model and the material flow are correct, and try to make the model converge by adjusting input conditions, calculation methods, etc. If reasonable results cannot be obtained even after multiple attempts, it is recommended that you redesign the model or use other software for simulation. Whether the sensitivity analysis results are usable depends on the metrics and application scenarios you are concerned about. If the results of the sensitivity analysis still provide valuable information, it can be considered valid data; otherwise, a new sensitivity analysis may be required. .