Extended Antoine Coefficients.
rain.on.moon
I want to know if there is an equation or correlation by which we can
convert the 'Conventional Antoine Coefficients (A,B and C)' of any
compound to the newer 'Extended Antoine Coefficients (A,B,C,D,E and
F)'? Waiting for replies.
Regards,
Atta.
charliew2
"rain.on.moon" <rain.o...@gmail.com> wrote in message
news:1122755389.3...@g44g2000cwa.googlegroups.com...
I doubt that there is an equation or correlation that can do this. However,
there should be a program or method which can do this. Take the original
Antoine equation and generate several data points in the range of
applicability (i.e., in a range of interest, where you know that the data
are good). Then, re-fit that data for the extended Antoine coefficients.
By the way, what do you hope to gain by going through this procedure?
rain.on.moon
have ChemSep Lite. I want to add the coefficients for a few compounds
of my interest to its library. I only know the traditional
coefficients.I just want to modify its library according to my needs. I
hope u can understand now.
Regards,
Atta.
John
news:11eo0ro...@corp.supernews.com...
It is best to get real data to fit the Extended Antoine, but I have seen
people use the Antoine to generate data as you said. Be sure to include the
critical temperature and pressure of the component in the regression of it's
Extended Antoine.
--
John
charliew2
No, I haven't heard of ChemSep. Send me an example of the functional form
of the extended Antoine equation via email, and I'll take a look.
J DUDLEY
equation should degenerate to the Antoine form is D = E = F = 0. If you are
starting with data from a numerical Antoine model, not data, then D = E = F
= 0, by definition, and you will get no benefits.
Give us the functional form, and we can provide more information.
"John" <moor...@chartermi.removethisandonedot.net> wrote in message
news:tZ_Ge.1538$4z....@fe02.lga...
John
range of applicability of the original Antoine equation.
The original poster will have to supply the Extended Antoine equation.
--
John
"J DUDLEY" <sbdp...@blueyonder.co.uk> wrote in message
news:4SbHe.45536$dN6....@fe1.news.blueyonder.co.uk...
rain.on.moon
is
ln P' = A - [B/(C+T)]
Where P' is Vapor pressure in MPa and T is in K
The extended form is given as below
ln P' = A + [B/(C+T)] + (D*T) + (E*lnT) + [F*(TexpG)]
Where P' and T have same units as above.
Now consider here an example of water
Conventional :
--------------------
A = 23.1963
B = 3816.44
C = -46.13
Extended :
-------------------
A = 68.555
B = -7050
C = 0
D = 3.589/1000
E = -6.6689
F = -8.505/10000000
G = 2.0
These will give pressure in Pa and T is in K.
These extended soefficients mentioned above are from ChemSep library.
Cany one suggest a correlation which can help me to jump from
conventional form to extended one?
Waiting anxiously.
Regards,
Atta.
charliew2
> The general form of conventional Antoine equation is, as you all know,
> is
>
> ln P' = A - [B/(C+T)]
>
> Where P' is Vapor pressure in MPa and T is in K
P' and T can be in any units that you find convenient.
Mathematically, you need a minimum of 3 data points to find the coefficients
of the Antoine equation. Also mathematically, you need a minimum of 7 data
points to find the coefficients of the extended Antoine equation. Due to
this, there is a strong implication that no correlation exists which can
directly translate 3 coefficients into 7 coefficients. If such a
correlation existed, I suspect that enough information would be contained in
the normal Antoine equation to make the extended Antoine equation
irrelevant.
Due to the above reasoning, I suggest you get data and fit the extended
Antoine equation functional form to it. If you don't know how to do this, I
can make some suggestions which would be helpful. Let me know.
A.H.
rain.on.moon wrote:
> The general form of conventional Antoine equation is, as you all know,
> is
>
> ln P' = A - [B/(C+T)]
>
> Where P' is Vapor pressure in MPa and T is in K
>
> The extended form is given as below
>
> ln P' = A + [B/(C+T)] + (D*T) + (E*lnT) + [F*(TexpG)]
>
> Where P' and T have same units as above.
Given the definitions above, it is trivial to "convert" Antoine to
extended Antoine.
Set D, E and F to zero, change the sign of B, and you are done, as the
resulting equation will exactly reproduce your "old" Antoine equation
which is the best you can ask for.
Of course if you can go back to the original data, or if you can get a
bunch of good-quality numbers (such as for water from the steam
tables), then you would want to fit the extended Antoine to the data.
But if all you have is the regular Antoine coefficients and no other
information, then the right thing to do is to exactly reproduce the
regular Antoine expression, as I described above.
Dr. Allan H. Harvey, Boulder, CO
"Insert usual disclaimers here."
rain.on.moon
Suppose I set D, E and F to zero. Now I change the sign of B as u
proposed. Now insert the given values of A, B and C for extended
version; do I get the same answer as for conventional coefficients. I
guess not. How do u interpret the interconversion of values?
Dear charliew2,
As u said if I do not know how to do it, then I should ask you. So here
I am. Guide me how to get the unknown Extended coefficients from known
Conventional coefficients.
Regards.
Quaoar
Just so this discussion is based on the correct extended Antoine
equations, NIST http://www.nist.gov/srd/webguide/nist85/85_1.htm
uses the following:
Antoine
Log10(P) = A - B/(C+T)
Extended Antoine
Log10(P) = A - B/(C+T) +
D((T-T0)/Tc)**n + E((T-T0)/Tc)**8 +
F((T-T0)/Tc)**12
P in bar, t in C
Note the sign of the term B/(C+T) is the same in both (presuming NIST
has published the referenced page correctly!)
What this shows is that the extended equation is attempting to determine
an error term e((T-T0)/Tc)) implicit in the traditional Antione
prediction. This error term can be determined only by comparing the
predicted vapor pressure from the traditional Antoine equation against
experimental data.
The fact that the constants A, B, C derived from water data for the
traditional and extended Antione equations are not the same is that
there are cross-correlations in the extended form that muddy the results
when the P-T data as a whole are used for regressing the extended
constants. The constants would have been identical if only the error in
the traditional prediction had been regressed against the indicated form
of the error term.
In any event, there is no way on God's Green Earth to determine the
extended constants knowing only the traditional constants.
Q
A.H.
rain.on.moon wrote:
> Dear Dr. Allan H. Harvey,
> Suppose I set D, E and F to zero. Now I change the sign of B as u
> proposed. Now insert the given values of A, B and C for extended
> version; do I get the same answer as for conventional coefficients. I
> guess not. How do u interpret the interconversion of values?
Here is my understanding of what you said your problem was:
1) You have some fluids for which you have traditional Antoine
coefficients.
2) Your software requires "extended" Antoine coefficients.
So, you want to use the "extended" equation and duplicate the vapor
pressures you already have. If you set D, E, and F to zero, the
"extended" equation you posted reduces to:
ln(P) = A + B/(T+C)
This is *exactly* the same form (except for the sign on the B term) as
the conventional Antoine equation. Therefore, if you plug in the
coefficients from your conventional equation (changing the sign of B if
needed), you must get *exactly* the same vapor pressure.
I see now in your earlier message that you had P in MPa in your
conventional equation and Pa in your extended equation. If that is the
case, then of course you must take an additional step to fix the units,
which will involve adding or subtracting ln(1000000) to the constant
(A) term.
Finally a more general note -- there is no such thing as THE extended
Antoine equation. Actually, one has to be careful in working with even
the original Antoine equation, because some people put different signs
in front of B and/or C and some use ln(P) while others use log-10(P).
Then various people have proposed various "extended" Antoine equations
to introduce more fitting parameters to try to describe data where the
3 traditional parameters of the Antoine equation are not considered
good enough. So anybody wishing to talk about an "extended Antoine
equation" needs to be clear about WHICH extended form they are talking
about.
Dr. Allan H. Harvey, Boulder, CO
Usual disclaimers here
WinSim Support
Replies by Q & DrAH's should suffice here...
All expressions which have constants (extended or otherwise) are relevant
ONLY when they can fit the collected expt. vapor pressure data in a
reasonable range. One set of constants have never led to another set of
constants for the same property by any expression in use of "first
principles". The final arbiter is real data. Please refer to the vapor
pressure section of "The Properties of Gases & Liquids " (has many editions
with different authors) for a near comprehensive treatise of the subject.
Thanks
"A.H." <stea...@aol.com> wrote in message
news:1122930549....@f14g2000cwb.googlegroups.com...