Thermodynamics / Migration from Mathcad 15 - Non linear equation solving - Сообщения
#21 Опубликовано: 12.01.2023 13:59:26
Red line looks good.
It is in line with the expected response of the system.
Red line plot may be integrated providing a model fit is given.
Attach your data set, I will attempt a model.
Cheers ... Jean.
#22 Опубликовано: 12.01.2023 23:51:13
Thanks for the help.
I am swamped with another project right now. But I will follow up this week end.
Pierre
I am swamped with another project right now. But I will follow up this week end.
Pierre
#23 Опубликовано: 13.01.2023 02:51:17
Hello,
I need to double check the data from the plot as I picked them up one by one.
And I want to give you sound data.
I will take care of it this week end.
I kept debugging the program. I found a few silly errors.
Hovering the mouse on the program shows the matrix.
The number of loop is reduced to 9 for troubleshooting.
But it's neither displayed outside of the program nor saved in a file.
Is there a way to fix that ?
Thanks for the help,
Pierre
VSD06 G14 12 Test.sm
I need to double check the data from the plot as I picked them up one by one.
And I want to give you sound data.
I will take care of it this week end.
I kept debugging the program. I found a few silly errors.
Hovering the mouse on the program shows the matrix.
The number of loop is reduced to 9 for troubleshooting.
But it's neither displayed outside of the program nor saved in a file.
Is there a way to fix that ?
Thanks for the help,
Pierre
VSD06 G14 12 Test.sm
#24 Опубликовано: 13.01.2023 07:53:50
Hi Peter. You still haven't solved the problem that f doesn't find any root for the initial conditions you propose, so the 3rd and 4th rows are zero. In addition, the way in which you build the range for the values of pz(1) and pz(2) does not seem adequate, since its limits depend on N, and it should not be like that (only the intermediate points should depend on N, not both ends of the interval). To save the data to a file, "execute" the command outside of the loop, ie, set exportData(...) = ... at the worksheet level, and not in the loop. Finally, the use of try ... on error is not particularly useful for numerical computation problems, and is actually a convenient and quick way of avoiding actual error handling, as well as being a lousy recommendation for this particular case.
Other problems would also be that you use some subscripts that are units (in blue) or that to graph you must use a double index on the MTX matrix, that is, something like augment(transpose(row(MTX,4)),transpose(row( MTX,5))), but the fundamental thing is that you establish the domain on which you want to calculate and then what are your seed or guess values. Since I do not know the thermodynamics that you are applying in this new version, I am not in a position to recommend any.
Best regards.
Alvaro.
Other problems would also be that you use some subscripts that are units (in blue) or that to graph you must use a double index on the MTX matrix, that is, something like augment(transpose(row(MTX,4)),transpose(row( MTX,5))), but the fundamental thing is that you establish the domain on which you want to calculate and then what are your seed or guess values. Since I do not know the thermodynamics that you are applying in this new version, I am not in a position to recommend any.
Best regards.
Alvaro.
#25 Опубликовано: 13.01.2023 09:09:38
Salut Pierre.
As you can imagine, I'm trying to see something from
these indigestible too big too small values.
Maybe it goes nowhere or hits the wall.
Cheers ... Jean.
VSD06 G14 12 Binary (2).sm
As you can imagine, I'm trying to see something from
these indigestible too big too small values.
Maybe it goes nowhere or hits the wall.
Cheers ... Jean.
VSD06 G14 12 Binary (2).sm
#26 Опубликовано: 13.01.2023 10:38:23
Thank you for your contributions Alvaro and Jean.
I understand better the problem.
I had issues with Mathcad and I wanted to tried another SW as a sanity check.
The approach is occasionally successful on Mathcad. But it's certainly the same challenge.
I need to clean up my programs a bit and see if I can get a good dataset again.
Also, I got the data export working.
Give me a bit a time to digest the new feedbacks.
Alvaro, I am not sure I got your comment right, but the solutions for the corner values are known.
It's 0 and 1 (pure elements).
FYI, there is an alternative algorithm to get to the solution.
It seems more popular, maybe for a good reason.
Let me look into it and attempt to code it.
You will know how it goes.
Thanks again,
Pierre
I understand better the problem.
I had issues with Mathcad and I wanted to tried another SW as a sanity check.
The approach is occasionally successful on Mathcad. But it's certainly the same challenge.
I need to clean up my programs a bit and see if I can get a good dataset again.
Also, I got the data export working.
Give me a bit a time to digest the new feedbacks.
Alvaro, I am not sure I got your comment right, but the solutions for the corner values are known.
It's 0 and 1 (pure elements).
FYI, there is an alternative algorithm to get to the solution.
It seems more popular, maybe for a good reason.
Let me look into it and attempt to code it.
You will know how it goes.
Thanks again,
Pierre
#27 Опубликовано: 13.01.2023 10:58:17
Alvaro, I am not sure I got your comment right, but the solutions for the corner values are known.
It's 0 and 1 (pure elements).
Hi Pierre. I talk about pz(1), pz(2) and N as in the notation you use in your worksheet. Usually N determines the number of intervals into which a range is divided, but here it is not the case since it also affects the values of the edges of the interval. I mean these corners
Best regards.
Alvaro.
#28 Опубликовано: 13.01.2023 13:29:13
I understand better the problem.
I had issues with Mathcad and I wanted to tried another SW as a sanity check.
The approach is occasionally successful on Mathcad. But it's certainly the same challenge.
Salut Pierre.
Outside of Mathcad and else, do you have the maths/formula of that system
from a chemical book c/w graph. No sweat to digitize graph.
See you soon... Jean.
#29 Опубликовано: 13.01.2023 13:41:31
Yes I do. I actually published it as well.
I post it this week end.
Pierre
I post it this week end.
Pierre
#30 Опубликовано: 14.01.2023 10:32:29
Yes I do. I actually published it as well.
I post it this week end.
Pierre
Refreshed version. Original MCD Thiele.
Final is the export for plugging in other document.
Next: will test if a seeded scanner can improve timing.
Cheers Pierre, take care ... Jean.
VSD06 G14 12 Binary (2) Copy.sm
#31 Опубликовано: 14.01.2023 23:37:40
Salut Jean,
here is a reference paper.
This work is the reference in the field.
https://www.sciencedirect.com/science/article/pii/0022024883903007
A slightly older paper but with similar maths
https://iopscience.iop.org/article/10.1149/1.2401687/meta
JCG62_83_225.pdf
JES121_74_1357.pdf
here is a reference paper.
This work is the reference in the field.
https://www.sciencedirect.com/science/article/pii/0022024883903007
A slightly older paper but with similar maths
https://iopscience.iop.org/article/10.1149/1.2401687/meta
JCG62_83_225.pdf
JES121_74_1357.pdf
#32 Опубликовано: 14.01.2023 23:43:20
Salut Jean,
Here is a set of data apparently legit (plot) with the source worksheet.
VSD AB Plot.xlsx
VSD06 AB Single.sm
Here is a set of data apparently legit (plot) with the source worksheet.
VSD AB Plot.xlsx
VSD06 AB Single.sm
#33 Опубликовано: 14.01.2023 23:46:22
Here is a mathcad worksheet with the results.
I attached an Smath worksheet with the parameters.
VSD Si-Ge 23-01-14.xmcd
MCD VSD Si-Ge 23-01-14.pdf
MCD VSD Si-Ge 23-01-14.xlsx
VSD06 G14 12 SiGe.sm
I attached an Smath worksheet with the parameters.
VSD Si-Ge 23-01-14.xmcd
MCD VSD Si-Ge 23-01-14.pdf
MCD VSD Si-Ge 23-01-14.xlsx
VSD06 G14 12 SiGe.sm
#34 Опубликовано: 14.01.2023 23:55:27
Finally a sketch of the chemical vapor deposition (CVD) reactor and notation.
It's a continuous flow reactor.
But with a 0D approach, the "instantaneous" partial pressures are only considered.
p° (noted pz) are virtual partial pressure based on the flows.
peq is the partial pressure at equilibrium
(pz-peq) give the amount of solid deposited. It's the supersaturation.
It's the same physics for salt and sugar.
If amounts below solubility are introduced, no solid is grown.
The worksheet attempts to calculate the "solubility" of components A and B in order to estimate the composition of the solid.
In this system, the elements are coupled (interaction parameter omega).
It's a continuous flow reactor.
But with a 0D approach, the "instantaneous" partial pressures are only considered.
p° (noted pz) are virtual partial pressure based on the flows.
peq is the partial pressure at equilibrium
(pz-peq) give the amount of solid deposited. It's the supersaturation.
It's the same physics for salt and sugar.
If amounts below solubility are introduced, no solid is grown.
The worksheet attempts to calculate the "solubility" of components A and B in order to estimate the composition of the solid.
In this system, the elements are coupled (interaction parameter omega).
#35 Опубликовано: 15.01.2023 11:35:59
Salut Jean,
Here is a set of data apparently legit (plot) with the source worksheet.
Sorry Pierre, NONE of your Smath document work on my system.
The Mathcad table does not plot much.
Cheers ... Jean.
Page6 Pierre.sm
#36 Опубликовано: 17.01.2023 12:03:22
#37 Опубликовано: 17.01.2023 12:53:02
Thank you Jean.
I need a little time to digest the information.
Pierre
I need a little time to digest the information.
Pierre
#38 Опубликовано: 17.01.2023 16:53:22
Thank you Jean.
I need a little time to digest the information.
Pierre
I played my ass a bit in there ... Jean
Page2 Freak.sm
#39 Опубликовано: 17.01.2023 17:48:13
In your attached PDF, a function has this;
Same function in your smath files is this;
Causing this;
If this was not on purpose, here is your file fixed;
VSD06 G14 12 SiGe.sm
Same function in your smath files is this;
Causing this;
If this was not on purpose, here is your file fixed;
VSD06 G14 12 SiGe.sm
#40 Опубликовано: 17.01.2023 22:40:54
Thanks you for spotting and fixing the typo.
Pierre
Pierre
-
Новые сообщения
-
Нет новых сообщений