67 replies to this topic

[DanW]

Hello my friends,

An update for DWSIM 2.1 for Windows, Linux and OS X was released. The following is the list of changes since Build 4503:

Version 2.1 Build 4606

- [CHG] COSMO-SAC database loading is now done only on-demand instead of during startup (Windows only)
- [CHG] Petalas-Aziz pressure drop model now uses a native library by the authors (Windows only)

Version 2.1 Build 4605

- [NEW] Simultaneous Adjust Solver
- [NEW] Added the option to edit pure compound properties through the Pure Compound Properties utility
- [NEW] Added automatic calculation of PR and SRK Peneloux volume translation (shift) coefficients for pseudocomponents
- [FIX] Several fixes to the Pipe Segment (and pressure drop) model, includes Joule-Thomson cooling option
- [FIX] Fixed drawing of Adjust's line connectors

Version 2.1 Build 4602

- [NEW] Added Peneloux volume translation support for PR and SRK Property Packages (Configure Property Package > General Options > Use Peneloux Volume translation => set to 1)
- [FIX] Fixed material stream composition editing in commmand line mode
- [FIX] Fixed ambient temperature not being set on Thermal Profile Editor (Pipe Segment)

Version 2.1 Build 4589

- [CHG] Enabled calculation of material streams when there is no mass/mole flow
- [FIX] Fixed mixer calculation when some of the inlet streams have no flow
- [FIX] Fixed a bug with the mass balance in the gas-liquid separator

Version 2.1 Build 4569

- [NEW] Save/Restore simulation states
- [NEW] Capture flowsheet snapshot and send to clipboard
- [FIX] Fixed command line run mode
- [FIX] Fixed heat capacity coefficients generated by the Compound Creator
- [FIX] Minor fixes to petroleum charact. utilities

Version 2.1 Build 4534

- [FIX] More fixes and enhancements to the flash algorithms
- [FIX] Fixed a bug in the excel interface
- [NEW] Added "reset settings" functionality by pressing Shift during startup

Version 2.1 Build 4526

- [NEW] Added Lua scripting support
- [FIX] Further fixes and improvements to the flash algorithms
- [FIX] Fixed ChemSep column operation with pseudocomponents/hypotheticals
- [FIX] Fixed Zc/Vc calculation for pseudocomponents
- [FIX] Fixed some bugs in the CAPE-OPEN Thermo interfaces

Version 2.1 Build 4513

- [FIX] Enhanced stability and reliability for Pressure-Enthalpy flash calculations

For more details and links to installer, docs and tutorials, go to http://dwsim.inforside.com.br

Thanks,
Daniel

Edited by DanW, 11 August 2012 - 04:32 PM.

[struchem]

Good luck to you.








Stru Chem Co.,LTD If you have any inquiry,pl feel free to contact with us.(kathy@struchem.com marketing@struchem.com )
CAS No:24214-72-0 cyclopentyl hydrazine hydrochloride
CAS No:27398-39-6 3-Chloro-2-Pyrazine-Carboxylic Acid
CAS No:887590-25-2 3,4-dihydro-2h-quinoxaline-1-carboxylic acid tert-butyl ester
CAS No:22353-40-8 2,3-dichloro-5-nitropyridine
CAS No:887592-68-9 2-(2-cyano-acetyl)-piperidine-1-carboxylic acid tert-butyl ester
CAS No:27611-63-8 2-cyano-benzoyl chloride
CAS No:24922-00-7 3-cyclopentyl-3-oxo-propionic acid ethyl ester
CAS No:15971-92-3 3-cyclohexyl-3-oxo-propionic acid ethyl ester
CAS No:129488-10-4 N-(1)-Boc-5-Amino-Indazole
CAS No:887589-58-4 (4-piperidin-4-yl-phenyl)-carbamic acid tert-butyl ester
CAS No:3468-17-5 indole-6-methylamine
CAS No:19747-79-6 1h-indole-3-carboximidic acid ethyl ester
CAS No:69975-65-1 6-amino-indan-1-one
CAS No:41959-45-9 5-nitro-1,2,3,4-tetrahydroisoquinoline
CAS No:397864-14-1 5-nitro-3,3-dihydro-1H-isoquinoline-2-carboxylic acid tert-butyl ester
CAS No:105945-24-2 3-(4-chloro-phenoxy)-phenylamine
CAS No:263266-23-5 3-(3-(trifluoromethoxy)phenoxy)benzenamine
CAS No:436151-35-8 3-(4-trifluoromethoxy-phenoxy)-phenylamine
CAS No:55586-26-0 4-amino-3-hydroxy-benzonitrile
CAS No:60523-82-2 3-(aminomethyl)-5-methoxyindole
CAS No:17617-05-9 pyridin-3-yl-methanethiol
CAS No:25672-97-3 2-Aminomethylbenzoic acid
CAS No:30953-58-3 (2-chlorophenethyl)hydrazine
CAS No:42166-50-7 Pyridin-3-yl-hydrazine
CAS No:118128-53-3 Fluoren-2-yl-hydrazine
CAS No:538342-98-2 4-Propylbenzylamine
CAS No:168899-61-4 3-Amino-2-bromo-benzoic acid

[kkala]

It has been a glad surprize for me to read the whole thread. Presently using a mobile connection, I have just downloaded the mobile version of DWSIM for my education (my simulation knowledge is limited to some use of WinSim's Design II). Hopefully I could offer some comments after a long time, if possible.
Such patient effort deserves encouragement and help to the extent possible. Not only because this is purpose of the Forum; the commercial simulators had better "feel" some competition from Open Source software. They may not feel Daniel's "shadow" now, but should feel the "shadow" of Open Source as a whole.

[amitray]

Dear Sir Daniel,

   I am very much thankful to you for utlising your spare time in such a fruitful way to come out with such an excellent software, that too open source. I have to do a project on Tributylphosphate, but right now I am in a bit of trouble as I am unable to find PO4 or PO bond details in joback method to get parameters for "Tributyl phosphate(TBP)". I have to add that compound to the list but unfortunately I am unable to do so. Kindly help me. The CAS number for that element is 126-73-8.

Regards

Amit

[Zubair Exclaim]

hey ... great work bro ... i am definitely gonna put my hands on this ..,. lets see if i can come up with a complete unit simuation on it.....

 

i have experience with HYSYS , Aspen pus etc etc ... let me know if you want additional help

 

[chemproc]

Good luck man!

[DanW]

Hi guys,

 

DWSIM 3.0 Build 5031 was released for Windows, Linux and OS X. This is a major release with a number of new features and enhancements.

 

Changes since 2.1 Build 4819:

 

- [NEW] Added Aqueous Electrolyte systems simulation support (beta) with two new Property Packages: LIQUAC* and Extended UNIQUAC

- [NEW] New Unit Operations for Solid-handling simulations: Solids Separator and Continuous Cake Filter
- [NEW] Unified code base and single executable for .NET/Mono, compiled for CLR v4.0
- [NEW] New XML simulation file format for full compatibility between platforms (Windows/Linux/OS X)
- [NEW] New Parallel Calculations engine with support for multicore CPUs and CUDA/OpenCL-capable GPUs
- [NEW] 'Copy Data to Clipboard' function added to flowsheet objects
- [CHG] Stability enhancements to Rigorous Column model and Boston-Britt Inside-Out solver
- [CHG] Report tool now fully supports Excel Spreadsheet file format
- [CHG] Added more material stream composition input options: Mass/Mole Flows, Standard Liquid Volumetric Fractions, Molarity and Molality (for electrolyte simulations)
- [CHG] Mixer and Separator Vessel models now supports up to six inlet streams
- [CHG] NRTL/UNIQUAC Interaction Parameters are now temperature-dependent
- [CHG] All Activity coefficient models now include an option to model the Vapor Phase as ideal. The Data Regression utility also supports this option in order to regress data to find the interaction parameters accordingly
- [CHG] The Compound Creator was enhanced to include more Joback/UNIFAC groups, Element information and Solid Phase properties estimation, including temperature-dependent ones (Cp, Density). The regression graphs now show calculated properties for every temperature dependent property
- [CHG] Phase Envelope utilities now have cancellation support
- [CHG] The Binary Envelope Utility was enhanced to show every kind of equilibrium line supported by DWSIM (VLE, LLE, SLE and Critical)
- [CHG] The Data Regression utility now supports fixing some interaction parameters, enhanced LLE /SLE regression
- [CHG] The Pure Compound Property viewer now includes molecular properties, solid properties and property tables

 

Visit the Wiki for the latest info and go to the Downloads page to get the installers.

 

Thanks,

Daniel

[chris turner]

Daniel, I think you need to up the caveats on the splash screen 

 

I didn't get the impression from your first post or the Disclaimer that its a one man effort.

 

But a very impressive one man effort!   

[StealthProg]

Although it is a tremedous effor for one person to accomplish it should be used with some caution.  A good example of this is the lee kesler enthalpy departure functions, i've just taken these out DWSIM and placed them in VBA code in excel and did quick comparison against the published compressibility factors from the original paper.  Whilst most of the time DWSIM code does give the correct answer (bearing in mind that the published values have to be interpreted very carefully around saturation points and beyond).  It does not allways do so.  So it is therefore possible that in some situations the calculated enthalpy will not be correct.

 

The basic reason for this is Lee kesler equation has multiple roots for some combinations of Tr and Pr and most convergence schemes can not gaurantee converging on the correct root.  Various roots may correspond to either liquid or vapour compressibility or be completely meaningless.

 

Note that errors are not limited to amateur software, all of the major process simulators have had bugs in them at various times that would make your eyes water if only you knew they were there, which most people don't. (because they don't check the results!).

[DanW]

Well, it seems that this behavior is not limited to DWSIM as you wrote yourself. Some sacrifice has to be made to speed up things. Of course there's a safe way of checking every possible root, select the one with the smallest Gibbs energy and identify it as a liquid or vapor. It can be done.

 

You're experienced and check the results of your simulations, as every chemical engineer should do - I always say that to my colleagues. I have lost count of the number of errors I've found in HYSYS or UNISIM for instance, but I understand and accept their limitations. As a programmer of thermo models, I know (and you probably know too) how difficult it is to get some things working correctly as you want. Sometimes it just seems impossible...

 

I'm proud of what I've done and still am doing in DWSIM, but it is not a one man operation anymore. For about a year or two, or even more, I'm having a huge help from Gregor Reichert, a german engineer which works on DWSIM almost every day as I do. He's already found lots of bugs, corrects the code by himself and already understands the code enough to do his own implementations. He's the one who did the Excel Unit Operation, the LLE diagram, upgraded the Compound Creator and recently revamped the VLLE flash algorithm, greatly enhancing its reliability. It has been really great working with him so far.

 

Gustavo León, a Venezuelan engineer, also worked on the rigorous column solvers, but it has been some time since his last contribution.

 

Well, to sum things up, anyone can contribute and any help is much appreciated. If you want to help, just sign up to GitHub, clone the repository and push some bits to the server: https://github.com/DanWBR/dwsim3

 

By the way, DWSIM is already on version 3.3, with Build 5517 just released (http://dwsim.inforsi...title=Main_Page), and my next task is making the solver work on Azure Virtual Machines, so if your computer isn't powerful enough to solve a big simulation in a decent time, you can use a virtual machine on the cloud to get it solved for you while you wait for the results in a few seconds (hopefully). 

 

Although it is a tremedous effor for one person to accomplish it should be used with some caution.  A good example of this is the lee kesler enthalpy departure functions, i've just taken these out DWSIM and placed them in VBA code in excel and did quick comparison against the published compressibility factors from the original paper.  Whilst most of the time DWSIM code does give the correct answer (bearing in mind that the published values have to be interpreted very carefully around saturation points and beyond).  It does not allways do so.  So it is therefore possible that in some situations the calculated enthalpy will not be correct.

 

The basic reason for this is Lee kesler equation has multiple roots for some combinations of Tr and Pr and most convergence schemes can not gaurantee converging on the correct root.  Various roots may correspond to either liquid or vapour compressibility or be completely meaningless.

 

Note that errors are not limited to amateur software, all of the major process simulators have had bugs in them at various times that would make your eyes water if only you knew they were there, which most people don't. (because they don't check the results!).

[robrt2]

 

Well, it seems that this behavior is not limited to DWSIM as you wrote yourself. Some sacrifice has to be made to speed up things. Of course there's a safe way of checking every possible root, select the one with the smallest Gibbs energy and identify it as a liquid or vapor. It can be done.

 

You're experienced and check the results of your simulations, as every chemical engineer should do - I always say that to my colleagues. I have lost count of the number of errors I've found in HYSYS or UNISIM for instance, but I understand and accept their limitations. As a programmer of thermo models, I know (and you probably know too) how difficult it is to get some things working correctly as you want. Sometimes it just seems impossible...

 

I'm proud of what I've done and still am doing in DWSIM, but it is not a one man operation anymore. For about a year or two, or even more, I'm having a huge help from Gregor Reichert, a german engineer which works on DWSIM almost every day as I do. He's already found lots of bugs, corrects the code by himself and already understands the code enough to do his own implementations. He's the one who did the Excel Unit Operation, the LLE diagram, upgraded the Compound Creator and recently revamped the VLLE flash algorithm, greatly enhancing its reliability. It has been really great working with him so far.

 

Gustavo León, a Venezuelan engineer, also worked on the rigorous column solvers, but it has been some time since his last contribution.

 

Well, to sum things up, anyone can contribute and any help is much appreciated. If you want to help, just sign up to GitHub, clone the repository and push some bits to the server: https://github.com/DanWBR/dwsim3

 

By the way, DWSIM is already on version 3.3, with Build 5517 just released (http://dwsim.inforsi...title=Main_Page), and my next task is making the solver work on Azure Virtual Machines, so if your computer isn't powerful enough to solve a big simulation in a decent time, you can use a virtual machine on the cloud to get it solved for you while you wait for the results in a few seconds (hopefully). 

 

Although it is a tremedous effor for one person to accomplish it should be used with some caution.  A good example of this is the lee kesler enthalpy departure functions, i've just taken these out DWSIM and placed them in VBA code in excel and did quick comparison against the published compressibility factors from the original paper.  Whilst most of the time DWSIM code does give the correct answer (bearing in mind that the published values have to be interpreted very carefully around saturation points and beyond).  It does not allways do so.  So it is therefore possible that in some situations the calculated enthalpy will not be correct.

 

The basic reason for this is Lee kesler equation has multiple roots for some combinations of Tr and Pr and most convergence schemes can not gaurantee converging on the correct root.  Various roots may correspond to either liquid or vapour compressibility or be completely meaningless.

 

Note that errors are not limited to amateur software, all of the major process simulators have had bugs in them at various times that would make your eyes water if only you knew they were there, which most people don't. (because they don't check the results!).

 

 

Hello daniel, I have been following the development of DWSIM since a few months ago. I am a Procees Engineer working at the Venezuelan main Oil Company, and I remember I was told of your Open Source process simulation software when I was visiting an Extra-Heavy Crude Oil Upgrader. I'm actually working on the Basic Engineering design and FEED of some major projects for new upcoming Upgraders, and we are using Aspen HYSYS for calculations. I have been involved in Crude Distillation Unit, Vacuum Distillation Unit, Utilities, but mainly in Sour Water Stripping Unit, where Hysys hasn't been 100% accurate on some components interactions. I will definitely will start to try out the designs we have for SWS at DWSIM as soon as I find some spare time. Whether results come out positvely or come with some bugs to be fixed, I am sure it'll be helpful.

One more time, great job you're doing with that software, great effort has its rewards. Hope the best for you and DWSIM development.

 

Best regards,

Roberto

[StealthProg]

 

Well, it seems that this behavior is not limited to DWSIM as you wrote yourself. Some sacrifice has to be made to speed up things. Of course there's a safe way of checking every possible root, select the one with the smallest Gibbs energy and identify it as a liquid or vapor. It can be done.

 

You're experienced and check the results of your simulations, as every chemical engineer should do - I always say that to my colleagues. I have lost count of the number of errors I've found in HYSYS or UNISIM for instance, but I understand and accept their limitations. As a programmer of thermo models, I know (and you probably know too) how difficult it is to get some things working correctly as you want. Sometimes it just seems impossible...

 

I'm proud of what I've done and still am doing in DWSIM, but it is not a one man operation anymore. For about a year or two, or even more, I'm having a huge help from Gregor Reichert, a german engineer which works on DWSIM almost every day as I do. He's already found lots of bugs, corrects the code by himself and already understands the code enough to do his own implementations. He's the one who did the Excel Unit Operation, the LLE diagram, upgraded the Compound Creator and recently revamped the VLLE flash algorithm, greatly enhancing its reliability. It has been really great working with him so far.

 

Gustavo León, a Venezuelan engineer, also worked on the rigorous column solvers, but it has been some time since his last contribution.

 

Well, to sum things up, anyone can contribute and any help is much appreciated. If you want to help, just sign up to GitHub, clone the repository and push some bits to the server: https://github.com/DanWBR/dwsim3

 

By the way, DWSIM is already on version 3.3, with Build 5517 just released (http://dwsim.inforsi...title=Main_Page), and my next task is making the solver work on Azure Virtual Machines, so if your computer isn't powerful enough to solve a big simulation in a decent time, you can use a virtual machine on the cloud to get it solved for you while you wait for the results in a few seconds (hopefully). 

 

Although it is a tremedous effor for one person to accomplish it should be used with some caution.  A good example of this is the lee kesler enthalpy departure functions, i've just taken these out DWSIM and placed them in VBA code in excel and did quick comparison against the published compressibility factors from the original paper.  Whilst most of the time DWSIM code does give the correct answer (bearing in mind that the published values have to be interpreted very carefully around saturation points and beyond).  It does not allways do so.  So it is therefore possible that in some situations the calculated enthalpy will not be correct.

 

The basic reason for this is Lee kesler equation has multiple roots for some combinations of Tr and Pr and most convergence schemes can not gaurantee converging on the correct root.  Various roots may correspond to either liquid or vapour compressibility or be completely meaningless.

 

Note that errors are not limited to amateur software, all of the major process simulators have had bugs in them at various times that would make your eyes water if only you knew they were there, which most people don't. (because they don't check the results!).

 

 

Determining whether it is liquid or vapour doesnt help in this case because the problem is finding the correct root for a known phase, Lee Kelser actually can have several roots in some areas.

 

Anyway the error is probably actually rather small.

 

Like I said the comment is a general comment on simulations and not meant to be picking solely on DWSIM.

 

 

As to contributing to the code, personally i would never find the time to do that and their would be legal implications anyway.

[DanW]

StealthProg, what I was trying to say is that the correct root is the one with the smallest Gibbs energy. After finding it, we can do some extra calculations to check if the phase behaves as a liquid or as a vapor, but they are computationally expensive. DWSIM has options to do them, but not in LK Property Package,

 

Do you think it is worthy to implement an algorithm to get the correct root in every single case? Remember that there's the speed tradeoff, but this calculation can be made faster with a parallel implementation, i.e. we can calculate the gibbs energy for every root in parallel, but the root finding routine will have to be sequential as it is now.

[DanW]

For those interested in learning a bit more about phase equilibrium thermodynamics and the associated (complex) math work, check this very interesting book by M. Michelsen: http://www.tie-tech....id=32_0_1_0_M13

[Poe]

DanW, great job and good luck, gretings from Santa Cruz Bolivia... sigue adelante !

Edited by Poe, 19 February 2016 - 07:37 AM.

[Devanand Telgote]

Hi Daniel,

I am unable to find all the components present in the Naphtha fraction as I am trying to do Simulations. I would like to mention the components 2,2,3-trimethyl hexane, 2,2,3,3-tetramethyl hexane, 2,2,5,5-tetramethyl hexane, 3 methylundecane, Isopropyl cyclohexane, isobytylcyclohexane, N-hexylcyclohexane, n-heptylcyclohexane, n-ocylcyclohexane, N-nonylcyclohexane,n-decylcyclohexane, and similar cyclopentanes. Can you please tell me how can I add then? or at least provide me link if you have already answered it somewhere. Thank you.

[DanW]

Hi Devanand,

 

If they're not in DWSIM or ChemSep databases, you can use the Compound Creator to create these compounds: http://dwsim.inforsi...Creator_Utility

[rezaee415]

How much simulator software's results can be trustworthy (for industrial design)?

can anyone introduce any book (or article) about this topic?

 tnx.