Hi Daniel,
on running a sensitivity study I encountered a problem:
I have two streams going to a mixer and after that to a heater which is heating to 1% vapor fraction.
The study increases the feed stream from 0 to 200 Kg/h, the other one is always at 100 Kg/h.
After finishing with all runs of study the original parameters are supposed to be restored. This is not happening correctely! Actually the second to last one (=180 Kg/h) is stored back.
As this mixture behaves somehow dificult to calculate i used Inside-Out flash calculation. But if I use standard method Nested-Loops I do get some wrong results in recorded data and temperature is changed even if not beeing manipulated!
After finishing the study and closing the window you go back and open the window again. You still see the results from last run. Only the last study is deselected. As soon as you select that study again values of independent variable in result table are changed but values of recorded variable are unchanged.
In my view, last study should be kept selected on opening the window again. The result table should be emptied in case of selecting an other study or after changing definition of actual study.
What's about my former suggestions to record several variables in parallel in one study and to get a button to export result data to clipboard (to Excel)?
regards
Gregor
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
1 - The code that retrieves the initial values of the independent variables was inside the loop that runs the simulation from the lower to the upper bounds, so, my bad... hahah
2 - That's weird, can you attach the simulation here?
3 - Fixed.
To export the data, just select the table, then Ctrl+C, Ctrl+V on Excel...
I'm finishing enhancing the utility to support more than one dependent variable. Will send you a link to the installer so you can take a look at it.
Regards,
Daniel
Last edit: Daniel Medeiros 2013-01-04
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
The sensitivity analysis utility was enhanced as you asked. Also, try to save your simulation as a DWXML file and take a look at its contents using any text editor... it should be intuitive enough so you can change any parameter that you want and then reload the simulation.
Have fun!
Daniel
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Daniel,
i prepared a sample file to demonstrate the described problem.
Please run sensitivity study with "nested loops" first and have a look at temperature and massflow of stream ethanol.
For comparison do the same with "inside out" method after restoring the original values first.
regards
Gregor
p.s.: DWSim is always asking to store simulation on closing the program even if it was stored immediately bevore. Storing is always done in "save as" mode even if a name for that simulation is already assigned.
The temperature changes you observed in the ethanol stream occurred because when there is only one compound in the stream, DWSIM tries to calculate a PH-Flash because there can be a partial vaporization... and in that case it couldn't find a solution for it, going all the way down with the temperature... I'll see what happened there. Probably an instability in the temperature convergence block.
That restoring bug was fixed in the latest dwsim 3 build that I've uploaded yesterday, which unfortunately you won't be able to test with your simulation... some of the previous simulations can be opened normally in dwsim 3 while others don't, and the exception message that visual studio throws doesn't tell me too much... Can you recreate you simulation in DWSIM 3 and see if it works?
That closing/saving confusion exists since the beginning and I always forget to fix it.. I'm glad you noticed it, will work on it asap.
Regards,
Daniel
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Got it! In your study, the first mass flow is zero, which gives an error in the calculation of the stream and DWSIM is not able to store the stream enthalpy. In the second run, it tries to read pressure and enthalpy (that doesn't exist) and then you see the temperature change.
I'll fix it now.
Daniel
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Daniel,
i tested the new update patch :) - it's working perfectly now! Even with troublesome "nested loops" method there is no problem anymore.
But one more new small bug remains: when renaming a study and saving it, the new name is not displayed in list of available studies. Now you select a different study and select the renamed study again. The renamed one will not be displayed anymore. The window needs to be reopend again to display the list with renamed study. Swapping of studys is working now again.
regards
Gregor
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi Daniel,
on running a sensitivity study I encountered a problem:
I have two streams going to a mixer and after that to a heater which is heating to 1% vapor fraction.
The study increases the feed stream from 0 to 200 Kg/h, the other one is always at 100 Kg/h.
After finishing with all runs of study the original parameters are supposed to be restored. This is not happening correctely! Actually the second to last one (=180 Kg/h) is stored back.
As this mixture behaves somehow dificult to calculate i used Inside-Out flash calculation. But if I use standard method Nested-Loops I do get some wrong results in recorded data and temperature is changed even if not beeing manipulated!
After finishing the study and closing the window you go back and open the window again. You still see the results from last run. Only the last study is deselected. As soon as you select that study again values of independent variable in result table are changed but values of recorded variable are unchanged.
In my view, last study should be kept selected on opening the window again. The result table should be emptied in case of selecting an other study or after changing definition of actual study.
What's about my former suggestions to record several variables in parallel in one study and to get a button to export result data to clipboard (to Excel)?
regards
Gregor
Ok, let's go... :-)
1 - The code that retrieves the initial values of the independent variables was inside the loop that runs the simulation from the lower to the upper bounds, so, my bad... hahah
2 - That's weird, can you attach the simulation here?
3 - Fixed.
To export the data, just select the table, then Ctrl+C, Ctrl+V on Excel...
I'm finishing enhancing the utility to support more than one dependent variable. Will send you a link to the installer so you can take a look at it.
Regards,
Daniel
Last edit: Daniel Medeiros 2013-01-04
Hi Gregor, try this => http://sourceforge.net/projects/dwsim/files/DWSIM/DWSIM%203.0/DWSIM_bin_v30_preview3_setup.exe/download
The sensitivity analysis utility was enhanced as you asked. Also, try to save your simulation as a DWXML file and take a look at its contents using any text editor... it should be intuitive enough so you can change any parameter that you want and then reload the simulation.
Have fun!
Daniel
Hi Daniel,
i prepared a sample file to demonstrate the described problem.
Please run sensitivity study with "nested loops" first and have a look at temperature and massflow of stream ethanol.
For comparison do the same with "inside out" method after restoring the original values first.
regards
Gregor
p.s.: DWSim is always asking to store simulation on closing the program even if it was stored immediately bevore. Storing is always done in "save as" mode even if a name for that simulation is already assigned.
Hi Daniel,
i installed the version 3 now and tried to load simulation files of Version 2.1. This gives an error an is not possible.
Gregor
Gregor,
The temperature changes you observed in the ethanol stream occurred because when there is only one compound in the stream, DWSIM tries to calculate a PH-Flash because there can be a partial vaporization... and in that case it couldn't find a solution for it, going all the way down with the temperature... I'll see what happened there. Probably an instability in the temperature convergence block.
That restoring bug was fixed in the latest dwsim 3 build that I've uploaded yesterday, which unfortunately you won't be able to test with your simulation... some of the previous simulations can be opened normally in dwsim 3 while others don't, and the exception message that visual studio throws doesn't tell me too much... Can you recreate you simulation in DWSIM 3 and see if it works?
That closing/saving confusion exists since the beginning and I always forget to fix it.. I'm glad you noticed it, will work on it asap.
Regards,
Daniel
Got it! In your study, the first mass flow is zero, which gives an error in the calculation of the stream and DWSIM is not able to store the stream enthalpy. In the second run, it tries to read pressure and enthalpy (that doesn't exist) and then you see the temperature change.
I'll fix it now.
Daniel
Gregor, try this patch for DWSIM 2.1, so you can load your simulation on it and check if it is working ok now (except for the nested loops instability which is on a different level of difficulty...) = > http://sourceforge.net/projects/dwsim/files/DWSIM/DWSIM%202.1/DWSIM_bin_v21_patch_b4503_to_4753.zip/download
Regards,
Daniel
Hi Daniel,
i tested the new update patch :) - it's working perfectly now! Even with troublesome "nested loops" method there is no problem anymore.
But one more new small bug remains: when renaming a study and saving it, the new name is not displayed in list of available studies. Now you select a different study and select the renamed study again. The renamed one will not be displayed anymore. The window needs to be reopend again to display the list with renamed study. Swapping of studys is working now again.
regards
Gregor