Open Journal of Physical Chemistry, 2011, 1, 61-69
doi:10.4236/ojpc.2011.13009 Published Online November 2011 (http://www.SciRP.org/journal/ojpc)
Copyright © 2011 SciRes. OJPC
61
Data Consistency Tests through the Use of Neural
Networks and Virial Equation. Application of the Proposed
Methodology to Critical Study of Dens i ty Data
Serge Laugier1*, Hakim Madani2, Abdeslam Hassen Meniai3, Dominique Richon4
1I2m, Umr Cnrs 5295 Enscbp, Pessac, France
2Laboratoire détudes des Systèmes Energétiques Industriels, Université de Batna, Batna, Algeria
3Laboratoire de lIngénieri e de s P roc édés dEnvironnement, Université Mentouri Constantine, Constantine, Algeria
4MINES ParisTech, Centre Énergétique et Procédés, 35 Rue Saint Honoré, Fontainebleau, France
*E-mail: laugier@enscbp.fr
Received May 30, 2011; revised August 9, 2011; ac c e pt ed September 11, 2011
Abstract
This paper focuses on a very important point which consists in evaluating experimental data prior to their use
for chemical process designs. Hexafluoropropylene P, ρ, T data measured at 11 temperatures from 263 to
362 K and at pressures up to 10 MPa have been examined through a consistency test presented herein and
based on the use of a methodology implying both neural networks and Virial equation. Such a methodology
appears as very powerful to identify erroneous data and could be conveniently handled for quick checks of
databases previously to modeling through classical thermodynamic models and equations of state. As an ap-
plication to liquid and vapor phase densities of hexafluoropropylene, a more reliable database is provided
after removing out layer data.
Keywords: Consistency Tests, Hexafluoropropylene, Neural Networks, Vibrating Tube Densimeter, Virial Equation
1. Introduction
Nowadays a great concern towards the environment pro-
tection is shown by the refrigeration industry which is
urged to find new fluids as refrigerants substitutes. In
Montreal Protocol (1987) it was decided to phase out and
replace ozone-depleting refrigerants like chlorofluo r oc a-
rbons (CFC’s) and hydrochlorofluorocarbons (HCFC’s).
Consequently CFCs were prohibited in 1996 by the
signing of Montreal Protocol countries and the fixed
deadline for a total banishment of HCFC’s which have
low ozone de pletion potential was set , i.e. 2030 [1].
Alternative compounds must be found and hexafluor-
opropene (HFP, R1216), CAS Number 116-15-4 is good
candidate with its 0.86 GWP value [2], eventually mixed
to other components. Numerous data concerning volu-
metric properties of this compound obtained using the
vibr ating-tube densitometer technique are available in
Coquelet et al. paper [3]. They correspond to P,
ρ
, T
triplets belonging to 11 isotherms reported partly in
tables 2 and 3. It is highly recommended prior to using
data for process design to have serious estimation about
their reliability and their accuracy, mainly in ord er not to
take too big error margins. The best recommendation that
could be done to experimental laboratories would be the
use of several experimental techniques to check for re-
producibility of data [4] and consequently provide guar-
anteed data. Young researchers must be encouraged to
handle experimental works that are so useful to industry
and theoricists [5]. Collaborations between laboratories
with complementary skills in either experimental, mod-
eling and simulation are a real advantage for presenting
worth and reliable data. Starting a French thermodynam-
ic research federation under the auspices of CNRS is in
project along with the setup of an international network.
The following sentence [6]: “All of the just mentioned
points need to be addressed in the frame of a thermody-
namicists’ network. This is one of the urgent goals to be
achieved in the near future” is found just after a list of
comments and remarks done during round table discus-
sions.
For low pressure PVT data, a very simple test consists
in verifying the data do follow the virial equation and
agree with the perfect gas law when pressure tends to
S. LAUGIER ET AL.
Copyright © 2011 SciRes. OJPC
62
zero (density tends to zero while pressure tends to zero).
For high pressure PVT data, the use of a flexible mod-
el, as a neural network based model, allows testing in-
ternal constancy of d ata.
These approaches are enlightened in this paper. They
are complementary tests to over, for example see refer-
ences [7-9], but are the most convenient here, in the con-
sidered pre ssure range.
2. Thermodynamic Consistency of Data in
Vapor Phase through the Virial Equation
At low pressures the use of the virial equation is quite
convenient to check for data consistency with respect to
vapor phase. Indeed, it is well known that the virial equ-
ation, truncated after the 3rd term allows acceptable re-
presentation of PVT data of pure compounds at low and
medium pressures (up to about 2 MPa).
This equation is written as:
P.v/R.T = 1 + B/v + C/v2 (1)
with B and C being the 2n d a nd 3rd virial c oefficient s .
These two parameters can be adjusted from isothermal
experimental data by rewriting Equation (1) as:
(P.v/R.T-1 ). v = B + C/v (2)
By tracing the evolution of the ((P.v/R.T)-1). v term
versus 1/v, it is possible to check for the linearity of Eq-
uation (2). Applied to data at 348.1 K, the results are
shown in Figure 1. Perfect linear ity is observed for 1 / v
above 0.4 kmol·m–3 i.e. above 1.0 MPa. But, at lower
pressures, the experimental points depart significantly
from linearity. Experimental u ncertainties cannot explain
these discrepancies while the virial equation is all the
more justified for the lowest pre ssures.
Such departures, at low pressures, displayed at 348.1
K are observed for all of the other temperatures. This
observation proves that although the densimeter vibrat-
ing tube [3] is ideal for measuring medium or high den-
sity values it is no longer the ideal tool for very low den-
sities.
Although a low pressure “pressure transducer” was
used, when working at the lowest pressures, this was not
the solution for ensuring quality density values. It is r ea-
sonable to believe that densimeter designed for working
at pressures up to 40 MPa , is not convenient at low pres-
sures because of its thick-walled (not sensitive enough)
vibrating tube.
For low pressures, the values of data P.V.T can be
extrapolated preferably and conveniently from the data
obtained at higher pressures, by using the virial equa-
tion. In Table 1, are g iven the values of the 2nd and 3rd
virial coefficients calculated from the reliable part of
the data.
-0.4
-0.35
-0.3
-0.25
-0.2
-0.15
-0.1 00.2 0.4 0.6 0.811.2 1.4 1.6 1.8
((Pv/RT) -1) . v
(m3.kmol-1)
1/v(kmo l . m
-3
)
Figure 1. Data departure from virial equation at 348.1 K.
Table 1. 2nd and 3rd virial coefficients as a function of
temperature.
T B C
K m3·kmol–1 m6·kmol–2
283.24
303.28
323.21
343.26
348.12
353.12
355.27
357.06
358.16
362.90
0.5400
–0.4880
–0.4283
–0.3720
–0.3556
–0.3432
–0.3384
–0.3344
–0.3325
–0.3201
0.17000
–0.01124
0.04116
0.04303
0.04015
0.03953
0.03920
0.03912
0.03903
0.03743
3. Thermodynamic Consistency of Data in
Liquid Phase through a Multiparameter
Model
A multiparameter model allows the representation of ex-
perimental data within their experimental uncertainties,
provided the uses of a sufficient number of parameters
and the availability of great enou gh number of da t a .
When a reduced number of data are not consistent with
respect to main part of database, the model will lead to a
representation presenting deviations much higher than
estimated values of experimental uncertainties. Conse-
quently, doubtful data will be easily identified through
deviati on plots.
Multiparameter models must have certain characteris-
tics: they must be very flexible and able to admit a great
number of parameters. A neural network based model has
all these characteristics. It will be used in this work. De-
tailed description of neural network models is given
elsewhere [10,11]. Figure 2 shows the topology of a
neural network with only one hidden layer. The activation
function, used herein, is the sigmoid one. Through opti-
mization procedure the adequate number of neurons in
hidden layer was found to be 7, allowing both under and
over-fittings. The two neurons in the input layer represent
the following independent variables: temperature T and
density ρ, while the output variable is the pressure P. All
variables are normalized between 0.1 and 0.9 to deal with
S. LAUGIER ET AL.
Copyright © 2011 SciRes. OJPC
63
T
1
2
7
.
.
.
.
ρ
P
Bias
Bias
Input Hidden Output
Laye r Layer Lay er
Figure 2. Neural network topology.
large variable ranges.
We have selected 4300 triplets in a random way among
not the published data 3 but the raw data as they were got
by the authors through their data acquisition unit (about
48,000 (P,
ρ
, T) triplets), for the training the network
(adjustment of its 29 parameters). This permits observing
the deviations between experimental and calculated pres-
sures for these 48,000 triplets. It is worth to use raw data
instead of published data as they are not rounded and
consequently have more digits av ailable for mathematical
treatments. Special behavior is pointed out at 348.2 K, see
Figure 3. In fact, between 8.5 and 10 MPa, several points
evolve in a way different from most of the others. To en-
light clearly thi s problem, we hav e plotted (see Figure 4)
the density as a function of pressure in the (8-10) MPa
range. One observes two distinct series of data. The series
corresponding to the lowest density values but also to the
highest Pexp-Pcal deviations must be eliminated from the
corresponding isothermal set data (30 experimental po-
ints). A new neural network treatment on the 348 K iso-
thermal data after removing doubtful data was carried out
adjusting a ga i n 29 param et e rs . Eve n a ft e r discarding thes e
30 experimental points too high deviations are still ob-
served; they concern 450 data that must be also discarded.
Four successive treatments ha ve been necessary to obtain
a clean coherent dataset (deviations in agreement with
estimated experimental uncertainties).
Behaviors similar to those displayed in Figures 3 and 4
are found at all temperatures. Corresponding comple-
mentary figures are available upon request to the corres-
ponding author of this paper. All isothermal data at 355.2
K must be rejected, the reason of rejection being the high
standard deviation that is observed between experimental
and calculated pressures; It is, in fact, 5 times bigger for
this isotherm than for all other isotherms. Finally, 42,240
triplets are considered as trustable over the 48,000. Pres-
sure deviations corresponding to these 42,240 triplets are
plotted in Figure 5. They are contained in +/0.04 MPa
pressure range, although authors [3] claimed for pressure
uncertai nt ies within 0.0003 or 0.000 6 M P a , depending on
-0.05
0
0.05
0.1
0.15
0.2
0.25
2 4 6 810 12
P
exp
-P
cal
(MPa)
P
exp
(MPa)
Figure 3. Deviations between experimental and calculated
pressures at 348.2 K.
1155
1160
1165
1170
1175
1180
1185
8.5 99.5 10 10.5
ρ
exp (kg . m-3)
Pexp (MPa)
Figure 4. Density as a function of pressure at 348.2 K.
-0.05
-0.04
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
0.04
0.05
0510
P
exp
-P
cal
(MPa)
P
exp
(MPa)
Figure 5. Deviations between experimental and calculated
pressures for remaining liquid densities after consistency
test treatment.
pressure sensor used. The standard deviation on the
42,240 points is 0.005 MPa, this is about ten times larger
than their estimated pressure uncertainties. We can con-
clude uncertainties on pressures were evidently underes-
timated by Coquelet et al. [3].
Table 2 contains only the trustable data from Tables 2
S. LAUGIER ET AL.
Copyright © 2011 SciRes. OJPC
64
Table 1. Densities of HFP.
Vapor densit i es
T P
ρ
T P
ρ
T P
ρ
T P
ρ
K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3
263.41 0.1302 9.2809 263.40 0.2274 16.933 283.24 0.4445 32.350 303.28 0.7729 56.178
263.41 0.1330 9.5029 263.40 0.2294 17.053 283.24 0.4483 32.640 303.28 0.7821 57.024
263.40 0.1358 9.7287 263.41 0.2314 17.292 283.24 0.4521 32.992 303.28 0.7912 57.852
263.40 0.1386 9.9512 283.25 0.4559 33.232 303.28 0.8003 58.744
263.41 0.1415 10.136 283.23 0.2895 19.903 283.25 0.4605 33.768 303.28 0.8105 59.713
263.41 0.1442 10.379 283.24 0.2953 20.300 303.26 0.8187 60.522
263.41 0.1471 10.581 283.24 0.3035 20.988 303.28 0.4206 27.450 303.27 0.8270 61.310
263.40 0.1498 10.801 283.24 0.3084 21.365 303.27 0.4316 28.275 303.28 0.8340 62.058
263.40 0.1525 10.973 283.23 0.3133 21.707 303.28 0.4439 29.151 303.28 0.8420 62.873
263.40 0.1553 11.212 283.24 0.3174 22.036 303.28 0.4548 29.939
263.40 0.1579 11.414 283.24 0.3222 22.368 303.28 0.4657 30.696 323.21 0.5008 30.574
263.40 0.1606 11.647 283.24 0.3270 22.770 303.28 0.4763 31.544 323.20 0.5407 33.232
263.40 0.1632 11.833 283.24 0.3312 23.117 303.28 0.4884 32.458 323.21 0.5686 35.175
263.41 0.1659 12.035 283.24 0.3352 23.372 303.28 0.5004 33.397 323.20 0.6033 37.627
263.41 0.1686 12.218 283.24 0.3400 23.845 303.27 0.5136 34.318 323.20 0.6395 40.168
263.41 0.1712 12.456 283.24 0.3442 24.108 303.28 0.5241 35.196 323.21 0.6750 42.784
263.41 0.1738 12.654 283.24 0.3482 24.411 303.28 0.5345 35.971 323.20 0.7098 45.300
263.41 0.1764 12.818 283.25 0.3531 24.764 303.26 0.5450 36.847 323.21 0.7373 47.337
263.40 0.1789 13.078 283.24 0.3571 25.072 303.27 0.5555 37.603 323.20 0.7846 50.994
263.40 0.1815 13.237 283.25 0.3611 25.445 303.28 0.5658 38.412 323.21 0.8254 54.173
263.41 0.1839 13.463 283.24 0.3659 25.853 303.28 0.5760 39.225 323.21 0.8558 56.561
263.41 0.1864 13.619 283.24 0.3700 26.146 303.28 0.5875 40.136 323.21 0.8825 58.745
263.41 0.1890 13.817 283.24 0.3747 26.552 303.28 0.5977 41.017 323.21 0.9164 61.512
263.41 0.1913 14.012 283.24 0.3787 26.899 303.28 0.6079 41.804 323.20 0.9493 64.319
263.40 0.1938 14.200 283.24 0.3828 27.236 303.28 0.6206 42.906 323.21 0.9832 67.213
263.40 0.1961 14.396 283.24 0.3867 27.563 303.27 0.6306 43.671 323.20 1.0027 69.010
263.40 0.1985 14.629 283.25 0.3907 27.782 303.26 0.6431 44.725 323.21 1.0261 71.062
263.41 0.2009 14.793 283.24 0.3946 28.179 303.26 0.6532 45.583 323.21 1.0477 72.912
263.40 0.2032 14.983 283.24 0.3987 28.482 303.27 0.6630 46.436 323.21 1.0673 74.738
263.40 0.2055 15.135 283.24 0.4034 28.816 303.28 0.6730 47.215 323.20 1.0834 76.287
263.40 0.2078 15.384 283.24 0.4074 29.204 303.28 0.6827 48.129 323.21 1.1048 78.288
263.40 0.2099 15.521 283.24 0.4114 29.532 303.28 0.6936 49.011 323.20 1.1256 80.357
263.40 0.2123 15.707 283.25 0.4162 29.928 303.27 0.7033 49.884 323.21 1.1401 81.784
263.40 0.2146 15.892 283.24 0.4209 30.373 303.27 0.7129 50.724 323.20 1.1552 83.245
263.41 0.2167 16.078 283.24 0.4248 30.649 303.28 0.7237 51.658 323.21 1.1675 84.433
263.40 0.2188 16.238 283.24 0.4287 31.045 303.28 0.7344 52.626 323.21 1.1932 87.022
263.40 0.2211 16.431 283.25 0.4327 31.286 303.28 0.7450 53.664 323.20 1.2070 88.441
263.40 0.2232 16.590 283.24 0.4366 31.702 303.27 0.7544 54.477 323.20 1.2204 89.889
S. LAUGIER ET AL.
Copyright © 2011 SciRes. OJPC
65
263.40 0.2254 16.763 283.24 0.4405 31.968 303.27 0.7637 55.342 323.22 1.2326 91.079
T P
ρ
T P
ρ
T P
ρ
T P
ρ
K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3
323.21 1.2442 92.409 343.23 2.1308 184.26 353.13 2.7613 297.64 355.25 2.8628 314.75
323.20 1.2549 93.544 343.22 2.1433 186.60 353.13 2.7695 302.00 355.25 2.8754 322.19
323.20 1.2762 95.862 343.24 2.1553 188.96 353.13 2.7781 307.16 355.25 2.8897 331.26
323.21 1.2944 97.960 343.25 2.1668 191.26 353.13 2.7816 309.46 355.25 2.9013 340.48
323.20 1.3116 99.800 343.25 2.1791 193.71
323.20 1.3280 101.85 343.21 2.1901 196.25 355.25 1.5818 101.98 357.02 1.6048 102.77
323.20 1.3436 103.60 343.25 2.2014 198.59 355.25 1.6143 104.74 357.02 1.6737 108.67
323.21 1.3567 105.23 343.21 2.2117 200.97 355.25 1.6455 107.51 357.01 1.7418 114.76
355.26 1.6750 110.13 357.02 1.8085 120.92
343.24 1.5744 110.45 353.12 1.9032 134.30 355.25 1.7016 112.53 357.02 1.8630 126.06
343.24 1.5997 113.00 353.13 1.9573 140.03 355.25 1.7255 114.77 357.02 1.9167 131.37
343.25 1.6236 115.44 353.12 2.0110 146.07 355.25 1.7498 116.98 357.03 1.9716 137.03
343.23 1.6467 117.95 353.13 2.0621 152.10 355.25 1.7730 119.17 357.04 2.0258 142.64
343.24 1.6692 120.33 353.13 2.1096 157.94 355.25 1.7961 121.37 357.04 2.0760 148.19
343.25 1.6910 122.68 353.13 2.1541 163.58 355.26 1.8170 123.34 357.04 2.1232 153.54
343.25 1.7127 125.08 353.13 2.1957 169.10 355.25 1.8369 125.28 357.04 2.1646 158.38
343.25 1.7340 127.50 353.13 2.2439 175.75 355.25 1.8567 127.27 357.04 2.2085 163.72
343.25 1.7549 129.82 353.13 2.2824 181.43 355.25 1.9085 132.51 357.04 2.2521 169.16
343.25 1.7746 132.14 353.13 2.3162 186.58 355.26 1.9678 138.66 357.04 2.2920 174.38
343.23 1.7948 134.67 353.14 2.3510 192.05 355.26 2.0205 144.41 357.03 2.3295 179.35
343.24 1.8160 137.14 353.14 2.3839 197.48 355.25 2.0652 149.45 357.03 2.3675 184.68
343.24 1.8338 139.44 353.14 2.4158 202.92 355.25 2.1148 155.31 357.02 2.4042 190.03
343.25 1.8513 141.49 353.14 2.4435 208.12 355.25 2.1600 160.70 357.02 2.4384 195.20
343.20 1.8694 143.90 353.14 2.4703 213.14 355.24 2.2060 166.59 357.03 2.4724 200.51
343.23 1.8871 146.21 353.13 2.4969 218.39 355.24 2.2495 172.29 357.02 2.5056 206.01
343.25 1.9054 148.67 353.13 2.5216 223.60 355.24 2.2886 177.69 357.03 2.5368 211.33
343.24 1.9225 151.01 353.13 2.5455 228.69 355.24 2.3246 182.73 357.03 2.5673 216.75
343.22 1.9391 153.27 353.14 2.5669 233.65 355.24 2.3632 188.54 357.03 2.5958 221.98
343.21 1.9556 155.71 353.13 2.5871 238.54 355.24 2.4181 197.16 357.04 2.6233 227.25
343.18 1.9720 158.13 353.13 2.6065 243.48 355.24 2.4750 206.70 357.04 2.6502 232.68
343.22 1.9832 159.75 353.13 2.6241 248.17 355.24 2.5285 216.42 357.03 2.6749 237.96
343.24 1.9963 161.60 353.13 2.6411 252.95 355.24 2.5783 226.29 357.04 2.6927 241.91
343.23 2.0108 163.82 353.13 2.6582 257.83 355.24 2.6212 235.50 357.04 2.7212 248.45
343.25 2.0246 165.97 353.13 2.6734 262.52 355.24 2.6635 245.20 357.04 2.7567 257.14
343.24 2.0379 168.09 353.13 2.6872 267.15 355.24 2.6996 254.54 357.04 2.7878 265.41
343.23 2.0520 170.29 353.13 2.7007 271.80 355.25 2.7333 263.91 357.04 2.8115 272.21
343.25 2.0658 172.70 353.13 2.7134 276.52 355.24 2.7628 273.22 357.04 2.8375 280.22
343.25 2.0793 174.97 353.13 2.7249 281.02 355.24 2.7902 282.60 357.04 2.8621 288.49
343.25 2.0927 177.24 353.13 2.7342 285.02 355.24 2.8116 290.74 357.03 2.8830 296.29
343.25 2.1060 179.66 353.13 2.7444 289.45 355.24 2.8311 299.13 357.03 2.8951 301.32
S. LAUGIER ET AL.
Copyright © 2011 SciRes. OJPC
66
343.21 2.1185 181.99 353.12 2.7537 293.77 355.24 2.8486 307.39 357.03 2.9003 303.89
T P
ρ
T P
ρ
T P
ρ
T P
ρ
K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3
357.03 2.9107 308.24 358.11 2.3484 180.01 358.14 3.0297 350.44 362.88 2.5683 199.28
357.03 2.9234 313.91 358.12 2.3729 183.30 358.16 3.0785 391.31 362.88 2.6231 207.21
357.03 2.9363 320.08 358.12 2.3945 186.40 362.87 2.6736 215.10
357.03 2.9582 331.37 358.12 2.4147 189.24 362.89 1.6026 98.83 362.88 2.7182 222.25
357.03 2.9777 343.50 358.12 2.4316 191.77 362.90 1.6666 104.27 362.88 2.7545 228.42
357.05 3.0070 365.19 358.13 2.4492 194.38 362.90 1.7208 108.65 362.87 2.7843 233.81
358.12 2.4637 196.64 362.90 1.7711 112.78 362.87 2.8064 237.98
358.10 1.6004 101.77 358.12 2.4802 199.08 362.89 1.8181 116.83 362.87 2.8272 241.92
358.10 1.6902 109.47 358.13 2.4961 201.53 362.89 1.8613 120.81 362.86 2.8487 246.13
358.10 1.7717 116.73 358.12 2.5087 203.68 362.89 1.8920 123.41 362.87 2.8840 253.16
358.10 1.8445 123.53 358.12 2.5211 205.69 362.89 1.9269 126.55 362.87 2.9014 256.91
358.11 1.9099 129.77 358.13 2.5339 207.80 362.89 1.9594 129.59 362.87 2.9269 262.44
358.10 1.9529 133.98 358.13 2.5466 209.81 362.89 1.9911 132.66 362.87 2.9783 274.33
358.11 2.0055 139.40 358.12 2.5553 211.41 362.88 2.0213 135.65 362.88 3.0746 299.37
358.11 2.0530 144.47 358.13 2.6325 225.14 362.88 2.0506 138.29 362.90 3.1534 325.03
358.11 2.0992 149.52 358.13 2.7116 241.11 362.88 2.0762 141.04 362.89 3.2150 350.25
358.10 2.1400 154.11 358.13 2.7772 256.23 362.88 2.1156 144.99 362.89 3.2634 375.24
358.10 2.1760 158.32 358.13 2.8259 269.03 362.88 2.1828 152.15 362.89 3.3028 401.16
358.10 2.2109 162.48 358.13 2.8632 279.95 362.88 2.2493 159.13 362.89 3.3321 426.48
358.11 2.2430 166.31 358.13 2.8932 289.65 362.88 2.3136 166.43 362.90 3.3554 452.56
358.11 2.2741 170.22 358.14 2.9456 308.34 362.88 2.3824 174.89 362.90 3.3735 478.83
358.11 2.2994 173.49 358.13 2.9749 320.98 362.87 2.4454 182.79 362.90 3.3855 500.09
358.11 2.3229 176.62 358.13 2.9952 330.70 362.88 2.5108 191.35
Liquid and supercritical densities
T P
ρ
T P
ρ
T P
ρ
T P
ρ
K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3
263.49 0.2344 1462.60 263.49 3.3443 1477.99 263.50 6.8018 1493.30 283.17 0.5712 1382.89
263.49 0.4005 1463.48 263.50 3.4480 1478.47 263.51 7.0484 1494.28 283.17 0.6658 1383.60
263.49 0.5098 1464.06 263.50 3.5897 1479.11 263.50 7.2675 1495.21 283.17 0.7214 1383.94
263.50 0.6902 1464.99 263.49 3.7777 1480.02 263.51 7.5076 1496.20 283.16 0.8365 1384.86
263.50 0.7793 1465.45 263.50 3.9414 1480.72 263.49 7.7506 1497.24 283.16 1.0396 1386.34
263.49 0.8440 1465.81 263.50 4.1241 1481.55 263.52 8.0413 1498.34 283.17 1.6324 1390.59
263.50 1.1187 1467.16 263.51 4.3378 1482.55 263.51 8.3648 1499.70 283.16 2.2501 1394.87
263.49 1.4771 1469.01 263.50 4.6165 1483.82 263.50 8.5930 1500.66 283.16 2.6648 1397.57
263.49 1.8313 1470.78 263.50 4.9541 1485.33 263.50 8.8186 1501.54 283.15 2.7118 1397.93
263.49 2.1807 1472.49 263.50 5.2620 1486.69 263.50 9.0842 1502.59 283.14 2.8054 1398.54
263.50 2.5772 1474.41 263.50 5.6047 1488.17 263.50 9.3944 1503.80 283.16 2.9493 1399.47
263.49 2.8130 1475.50 263.50 6.0257 1490.01 263.50 9.7162 1505.07 283.16 2.9927 1399.77
263.49 2.9795 1476.28 263.49 6.2416 1490.94 263.51 9.99 1506.07 283.17 2.9971 1399.77
263.49 3.1959 1477.28 263.50 6.4174 1491.65 283.17 2.9998 1399.80
S. LAUGIER ET AL.
Copyright © 2011 SciRes. OJPC
67
263.50 3.2614 1477.57 263.50 6.5911 1492.39 283.17 0.4698 1382.04 283.17 3.0382 1400.05
T P
ρ
T P
ρ
T P
ρ
T P
ρ
K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3
283.16 3.1615 1400.87 303.37 2.7849 1309.16 323.36 3.2630 1209.83 343.18 3.2770 1072.74
283.16 3.2296 1401.33 303.36 2.9185 1310.51 323.37 3.5588 1214.48 343.14 3.4333 1078.74
283.16 3.2728 1401.61 303.36 2.9307 1310.67 323.36 3.6720 1216.27 343.15 3.5963 1084.50
283.16 3.3637 1402.20 303.38 2.9933 1311.27 323.36 3.7893 1218.14 343.10 3.8060 1091.78
283.17 3.5116 1403.13 303.37 3.0866 1312.17 323.36 3.9397 1220.38 343.10 3.9148 1095.09
283.17 3.6663 1404.11 303.38 3.1857 1313.13 323.36 4.0934 1222.65 343.10 4.0523 1099.41
283.15 3.8026 1405.02 303.37 3.2693 1313.88 323.36 4.2825 1225.41 343.15 4.1954 1103.09
283.15 3.9872 1406.17 303.37 3.3126 1314.39 323.37 4.4794 1228.20 343.15 4.3449 1107.25
283.15 4.1863 1407.41 303.37 3.4926 1316.15 323.36 4.6979 1231.27 343.08 4.4776 1111.27
283.15 4.4758 1409.19 303.37 3.7235 1318.30 323.36 4.9172 1234.22 343.13 4.6043 1114.20
283.16 4.8504 1411.46 303.37 3.9594 1320.48 323.36 5.1432 1237.21 343.14 4.7474 1117.85
283.16 5.1823 1413.45 303.36 4.2842 1323.48 323.37 5.3859 1240.33 343.13 4.8853 1121.19
283.16 5.4802 1415.20 303.37 4.7074 1327.25 323.36 5.6525 1243.74 343.14 5.0337 1124.78
283.17 5.6908 1416.42 303.37 5.1888 1331.39 323.36 5.8633 1246.31 343.16 5.1610 1127.67
283.17 5.9249 1417.79 303.37 5.5877 1334.67 323.36 6.0183 1248.17 343.11 5.2809 1130.51
283.16 6.1474 1419.06 303.36 5.7419 1335.96 323.36 6.2000 1250.33 343.15 5.4296 1133.72
283.17 6.4221 1420.61 303.38 5.9178 1337.36 323.37 6.3922 1252.57 343.13 5.5803 1136.99
283.16 6.6614 1421.98 303.37 6.0895 1338.75 323.36 6.5607 1254.59 343.13 5.7458 1140.49
283.16 6.9272 1423.44 303.36 6.2521 1340.07 323.37 6.7801 1257.08 343.09 5.9144 1144.10
283.16 7.1758 1424.82 303.37 6.4774 1341.86 323.36 7.0081 1259.57 343.13 6.1494 1148.49
283.16 7.4145 1426.14 303.37 6.7016 1343.59 323.37 7.2360 1262.04 343.15 6.3754 1152.64
283.16 7.8586 1428.50 303.36 6.9381 1345.42 323.36 7.4970 1264.90 343.12 6.6420 1157.63
283.16 8.3282 1431.01 303.37 7.2007 1347.40 323.37 7.7482 1267.53 343.12 6.9374 1162.77
283.16 8.6936 1432.92 303.37 7.4843 1349.56 323.36 8.0400 1270.52 343.16 7.1456 1166.09
283.16 9.1257 1435.16 303.37 7.7694 1351.63 323.37 8.3358 1273.49 343.13 7.3143 1169.11
283.17 9.6012 1437.53 303.37 8.0990 1354.02 323.36 8.6540 1276.62 343.15 7.5136 1172.21
283.16 10.00 1439.52 303.38 8.4005 1356.17 323.36 8.8450 1278.47 343.15 7.7097 1175.33
303.38 8.7837 1358.84 323.36 9.0517 1280.45 343.13 7.9291 1178.74
303.36 0.85 1287.97 303.38 9.1799 1361.62 323.36 9.2694 1282.50 343.13 8.1778 1182.49
303.37 0.9378 1288.97 303.37 9.7063 1365.14 323.36 9.4714 1284.41 343.13 8.4071 1185.80
303.37 0.9511 1289.14 303.37 9.99 1367.03 323.36 9.7097 1286.55 343.10 8.6707 1189.51
303.36 1.0541 1290.38 323.37 1.42 1174.31 323.36 9.9458 1288.68 343.12 9.0310 1194.37
303.37 1.1573 1291.56 323.36 1.5746 1177.86 323.36 9.99 1289.15 343.17 9.4439 1199.72
303.36 1.2988 1293.20 323.36 1.7619 1181.93 343.07 9.8832 1205.55
303.37 1.4773 1295.32 323.36 1.9856 1186.54 343.11 2.38 1030.41 343.13 9.9823 1206.55
303.36 1.6090 1296.75 323.37 2.1895 1190.59 343.18 2.5875 1041.95
303.36 1.8082 1298.96 323.36 2.4451 1195.49 343.18 2.7999 1052.25 348.15 2.5063 959.57
303.36 1.9951 1301.01 323.36 2.7330 1200.72 343.17 2.8793 1056.54 348.15 2.5807 967.62
303.37 2.2154 1303.37 323.36 2.8503 1202.76 343.17 2.9726 1060.25 348.15 2.6650 975.74
303.36 2.4610 1305.98 323.36 2.9981 1205.34 343.14 3.0434 1063.62 348.15 2.7670 984.43
S. LAUGIER ET AL.
Copyright © 2011 SciRes. OJPC
68
303.37 2.6970 1308.28 323.36 3.0089 1205.47 343.18 3.1646 1068.25 348.14 2.8350 989.71
T P
ρ
T P
ρ
T P
ρ
T P
ρ
K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3 K MPa kg·m–3
348.15 2.8998 994.57 353.10 2.9255 912.33 357.02 3.1111 833.74 362.88 3.5137 727.62
348.15 2.9771 999.82 353.12 2.9622 917.80 357.02 3.1712 852.76 362.90 3.5720 762.04
348.15 2.9924 1000.80 353.12 2.9893 921.43 357.01 3.2529 871.60 362.92 3.6667 797.39
348.13 3.0241 1003.27 353.13 2.9998 922.83 357.01 3.3072 882.00 362.92 3.7773 826.35
348.15 3.1686 1011.94 353.13 3.0507 929.65 357.02 3.3780 893.55 362.93 3.8981 849.28
348.14 3.3388 1021.49 353.10 3.1188 938.02 357.03 3.4703 906.34 362.92 3.9491 857.62
348.13 3.5586 1032.40 353.13 3.2195 948.23 357.03 3.5628 917.41 362.92 4.0139 867.25
348.14 3.6615 1037.05 353.14 3.3175 957.33 357.01 3.6121 923.16 362.92 4.0824 876.48
348.14 3.7808 1042.19 353.13 3.4302 966.82 357.00 3.6702 929.31 362.92 4.1528 885.28
348.15 3.8951 1046.96 353.13 3.5523 976.05 357.01 3.7558 937.42 362.91 4.2311 894.18
348.14 4.0246 1052.08 353.13 3.6916 985.56 357.01 3.8510 945.66 362.91 4.3262 903.88
348.13 4.1628 1057.39 353.12 3.8631 996.09 357.01 3.9570 954.08 362.91 4.4244 913.15
348.13 4.3298 1063.41 353.13 4.0712 1007.44 357.02 4.0890 963.54 362.90 4.5292 922.20
348.14 4.5222 1069.87 353.12 4.3084 1019.03 357.02 4.2311 972.86 362.89 4.6511 931.84
348.14 4.7012 1075.55 353.12 4.5492 1029.60 357.02 4.4027 983.12 362.90 4.8021 942.47
348.14 4.8352 1079.64 353.12 4.7878 1039.11 357.02 4.5908 993.37 362.92 4.9814 953.96
348.13 4.9330 1082.55 353.13 5.0040 1047.06 357.02 4.8261 1004.87 362.92 5.1603 964.26
348.13 5.0774 1086.71 353.13 5.2259 1054.73 357.02 5.0853 1016.35 362.92 5.2834 970.93
348.14 5.2077 1090.33 353.13 5.4583 1062.16 357.02 5.3577 1027.14 362.93 5.4501 979.39
348.14 5.3649 1094.65 353.13 5.6592 1068.32 357.03 5.5121 1032.88 362.92 5.5832 985.74
348.14 5.5644 1099.78 353.13 5.8803 1074.71 357.02 5.6709 1038.54 362.93 5.7812 994.50
348.13 5.7912 1105.44 353.13 6.0511 1079.42 357.02 5.8365 1044.18 362.93 5.9198 1000.30
348.14 6.0365 1111.21 353.14 6.2400 1084.40 357.02 6.0264 1050.30 362.92 6.0494 1005.54
348.13 6.3062 1117.33 353.13 6.4190 1089.01 357.02 6.1898 1055.34 362.92 6.2176 1012.03
348.14 6.6047 1123.73 353.15 6.6542 1094.70 357.00 6.3067 1058.93 362.92 6.3845 1018.10
348.13 6.8303 1128.36 353.14 6.8350 1099.00 357.02 6.4757 1063.71 362.92 6.5749 1024.73
348.12 7.0392 1132.55 353.14 7.0278 1103.39 357.01 6.6403 1068.34 362.91 6.7692 1031.15
348.12 7.2462 1136.50 353.15 7.2508 1108.26 357.02 6.8284 1073.32 362.91 6.9763 1037.63
348.12 7.4780 1140.85 353.14 7.4607 1112.69 357.02 7.0299 1078.51 362.91 7.2208 1044.90
348.11 7.7121 1145.16 353.14 7.6879 1117.39 357.01 7.2375 1083.67 362.90 7.4626 1051.74
348.12 7.9409 1149.12 353.13 7.9294 1122.16 357.02 7.4729 1089.23 362.89 7.7376 1059.06
348.12 8.1996 1153.54 353.12 8.1999 1127.34 357.01 7.7275 1095.06 362.89 8.0418 1066.74
348.12 8.4235 1157.17 353.09 8.4395 1131.92 357.01 7.9973 1100.94 362.89 8.3779 1074.76
348.12 8.6747 1161.26 353.06 8.6677 1136.08 357.01 8.2949 1107.17 362.89 8.7464 1083.00
348.13 8.9460 1165.39 353.06 9.0076 1141.95 357.02 8.6373 1113.95 362.89 8.9576 1087.48
348.14 9.2995 1170.72 353.07 9.3594 1147.74 357.02 8.9868 1120.56 362.88 9.1065 1090.63
348.16 9.95 1179.71 353.07 9.7590 1154.06 357.01 9.3930 1127.88 362.88 9.2777 1094.12
353.13 2.79 883.82 353.06 10.00 1157.73 357.02 9.8823 1136.21 362.88 9.4571 1097.66
353.12 2.8200 892.36 357.02 9.9954 1138.05 362.88 9.6459 1101.34
353.10 2.8439 897.65 357.00 3.02 777.77 362.88 9.8470 1105.11
S. LAUGIER ET AL.
Copyright © 2011 SciRes. OJPC
69
353.11 2.8887 906.09 357.02 3.0532 807.50 362.88 3.47 691.30 362.88 9.9964 1107.95
and 3 of original paper [3]. Other trustable data (about
10500 over the 48000 original raw data) are available
from corre s ponding author: laugier@enscbp.fr.
4. Conclusions
Through the present study we have pointed out the im-
portance of reliable and accurate data and the usefulness
of simple data consistency tests. These simple tests jus-
tify the development of various performing techniques
such as tho s e based on neur a l ne twork.
Density of gases and gas mixtures must tend to zero
when pressure tends to zero and additionally the virial
law must be followed.
Neural network based approach presently used to as-
sess the quality of experimental hexafluoropropylene
densities gives encouraging results.
For supercritical HFP only one isotherm has been
conserved from previous work [3]. Due to high thermal
effects in the vicinity of critical po int, it would have been
certainly necessary to be more careful and more patient
recording the data (very small pressure changes are re-
quested as a function of time).
Finally, we conclude neural network modeling is
worth in the assessment of data consistency.
As far as high pressure vibrating tube densimeters are
used, vapor densities at very low pressures are better
determined through extrapolation of higher pressure val-
ues using virial equation of state.
Only trustable densities for the three vapor, liquid and
supercritical physical states have been reported herein.
Supporti ng Informa tion Availab l e:
Density values of HFP at various temperatures and in
its various states: vapor, liquid and supercritical. This
material is available upon demand to corresponding au-
thor: laugier@enscbp.fr. It has to be used preferably to
the material presented in Coquelet et al.’s paper [3] that
unfortunately contains inaccurate and erroneous data.
5. References
[1] H. Madani, A. Valtz, C. Coquelet, A. H. Meniai and D.
Richon, “(Vapor plus Liquid) Equilibrium Data for (Car-
bon Dioxide + 1,1-Difluoroethane) System at Tempera-
tures from 258 to 343 K and Pressures up to about 8
MPa,” The Journal of Chemical Thermodynamics, Vol.
40, No. 10, 2008, pp. 1490-1494.
doi:10.1016/j.jct.2008.06.002
[2] G. Acerboni, J. A. Buekes, N. R. Jensen, J. Hjorth, G.
Myhre, C. J. Nielsen and J. K. Sundet, “Atmospheric De-
gradation and Global Warming Potentials of Three
Perfluoroalkenes,” Atmospheric Environment, Vol. 35,
No. 24, 2001, pp. 4113-4123.
doi:10.1016/S1352-2310(01)00209-6
[3] C. Coquelet, D. Ramjugernath, H. Madani, A. Valtz, P.
Naidoo and A. H. Meniai, “Experimental Measurements
of Vapor Pressures and Densities of Pure Hexafluoro-
propylene,” Journal of Chemical & Engineering Data,
Vol. 55, No. 6, 2010, pp. 2093-2099.
doi:10.1021/je900596d
[4] D. Richon, “Experimental Techniques for the Determina-
tion of Thermophysical Properties to Enhance Chemical
Processes,” Pure and Applied Chemistry, Vol. 81, No. 10,
2009, pp. 1769-1782. doi:10.1351/PAC-CON-08-09-06
[5] C. Coquelet, L. A. Galicia-Luna, A. H. Mohammadi and
D. Richon, “The Essential Importance of Experimental
Research and the Use of Experimental Thermodynamics
to the Benefit of Industry,” Fluid Phase Equilibria, Vol.
296, No. 1, 2010, pp. 2-3.
doi:10.1016/j.fluid.2010.04.004
[6] D. Richon, 2009.
http://www.wix.com/drichon/thermoadvices
[7] A. Eslamimanesh, A. H. Mohammadi and D. Richon,
“Thermodynamic Consistency Test for Experimental Da-
ta of Water Content of Methane,AIChE Journal, Vol.
57, No. 9, 2011, pp. 2566-2573. doi:10.1002/aic.12462
[8] A. Eslamimanesh, M. Yazdizadeh, A. H. Mohammadi
and D. Richon, “Experimental Data Assessment Test for
Diamondoids Solubility in Gaseous System,Journal of
Chemical & Engineering Data, Vol. 56, No. 5, 2011, pp.
2655-2659. doi:10.1021/je200193n
[9] A. Eslamimanesh, A. H. Mohammadi and D. Richon,
“Thermodynamic Consistency Test for Experimental Da-
ta of Sulfur Content of Hydrogen Sulfide,” Journal of
Chemical & Engineering Data, Vol. 50, No. 6, 2011, pp.
3555-3563. doi:10.1021/ie1017332
[10] A. Chouai, S. Laugier and D. Richon, “Modeling of
Thermodynamic Properties Using Neural Networks
Application to Refrigerants,” Fluid Phase Equilibria, Vol.
199, No. 1-2, 2002, pp. 53-62.
doi:10.1016/S0378-3812(01)00801-9
[11] S. Laugier, F. Rivolet and D. Richon, “New Volume
Translation for Cubic Equations of State,” Fluid Phase
Equilibria, Vol. 259, No. 1, 2007, pp. 99-104.
doi:10.1016/j.fluid.2007.04.032