custom heat capacity problem

[Mikele]

Hello everybody!
I'm trying to put custom properties correlations for a liquid component that is not in Aspen databanks (i.e. viscosity and heat capacity); while for viscosity I had no problems, I didn't manage to put the heat capacity correlation in the simulator. Better, I put a polinomial correlation in the Properties->Parameter->Pure components section, but it seems that Aspen refuses to use it!
I don't know how to solve this, please help me!

[juncruz]

You can try to do regression in "Data" in properties (thermodynamic, then CPL or CPL-M for liquid heat capacity). A simpler way is to just do "estimation" then estimate all missing properties. Hope this helps

[Mikele]

I tried what you suggest, but it didn't work T.T
unfortunately the estimation tool is not accurate (more than 20% error) so I can't use it.
The only "way" I found to change the CPL is to modify the CPIG relation that Aspen estimates, but I don't know if it is useful

[juncruz]

CPIG is for gas, so I don't know if that will relate to CPL which is liquid. What I do is regress data as follows:
1. Put in the component in the components Specifications box
2. Go to “Molecular Structure” and draw the molecule
3. Go to “Data” and “New”. Enter a name for the ID (example CPLM)
4. Select type: “Pure comp”
5. Property: CPL- (liquid heat mass capacity)
6. Component: point to the component
7. You can then enter a constant parameter, either pressure or temperature. Normally I put a constant pressure of 1 atm for CPL
8. Go to Data tab; Below “STD-DEV”, Choose “DATA” and enter the values for the temperature, and corresponding CP value (you can change the unit). Put many data (different temperatures). The more the better. I normally put 5 – 8 data.
9. Then I go back to “Components” “Molecular Structure”, choose the component then “Structure”, then “Calculate Bonds”
10. Then I click on “TDE NIST”. Choose the component, then click “Evaluate now”
11. The I go to “Regression”, “New”. On Data set, choose the ID that you made on the data, (example CPLM)
12. Go to “Parameters”, choose “Type” Parameter. Element “1”. On the “Name”, there are many options for CP. I normally use CPLIKC but it depends on the type of component. Then the component. You have to make 1 -7 elements.
13. Then run the estimation.
Sometimes you have to determine what regression set works. Aside from CPLIKC, there is CPLPO, CPLTDECS, CPLSTMLPO, CPLXP1, CPLXP2. Especially if you have many components and properties to regress, it can be very very tedious.
Good luck!

[Mikele]

CPIG is for gas, so I don't know if that will relate to CPL which is liquid. What I do is regress data as follows:
1. Put in the component in the components Specifications box
2. Go to “Molecular Structure” and draw the molecule
3. Go to “Data” and “New”. Enter a name for the ID (example CPLM)
4. Select type: “Pure comp”
5. Property: CPL- (liquid heat mass capacity)
6. Component: point to the component
7. You can then enter a constant parameter, either pressure or temperature. Normally I put a constant pressure of 1 atm for CPL
8. Go to Data tab; Below “STD-DEV”, Choose “DATA” and enter the values for the temperature, and corresponding CP value (you can change the unit). Put many data (different temperatures). The more the better. I normally put 5 – 8 data.
9. Then I go back to “Components” “Molecular Structure”, choose the component then “Structure”, then “Calculate Bonds”
10. Then I click on “TDE NIST”. Choose the component, then click “Evaluate now”
11. The I go to “Regression”, “New”. On Data set, choose the ID that you made on the data, (example CPLM)
12. Go to “Parameters”, choose “Type” Parameter. Element “1”. On the “Name”, there are many options for CP. I normally use CPLIKC but it depends on the type of component. Then the component. You have to make 1 -7 elements.
13. Then run the estimation.
Sometimes you have to determine what regression set works. Aside from CPLIKC, there is CPLPO, CPLTDECS, CPLSTMLPO, CPLXP1, CPLXP2. Especially if you have many components and properties to regress, it can be very very tedious.
Good luck!

First of all I want to thank you very much for your interes. I did everything you said but I always got the same problems: Aspen continues ignoring my datas.
I post you some images of the input I put and the results I got, maybe they could be useful

[link Point to another website Only the registered members can access]

[link Point to another website Only the registered members can access]

[link Point to another website Only the registered members can access]

[Mikele]

PS: I tried all the possible CP options but it doesn't change very much, the standard deviation is always huge...

[juncruz]

can you give me the component molecular structure and CP data that you are trying to regress? I will try it
Regards

[Mikele]

I've just sent you a pm with all you requested. Thanks

[juncruz]

Yes I got the data but I need the mol structure drawing. I noticed you only have name/element "CPLIKC 1" you need to repeat this "CPLIKC 2,3,4,5,6,7"

[Mikele]

Problem solved! thanks to juncruz!!