Hello: I may be misinterpreting but there appear to be a few inconsistencies in the PHOENICS docs about the use of CHEMKIN:

1. (/phoenics/D_POLIS/D_LECS/Chemkin.htm, #11) states that input is in CGS units but the example mechanism (#10) file is in JOULES/MOLE (not CGS)

2. Given question #1 above, what are INLET() pressure units? (CGS/Atm/MKS?)

3. Also in section #11 of these docs, the inlet composition may be "Mass, Mole, or concentration fractions" - no examples of how to select mole fraction (preferred unit) are given.

4. CHEMKIN/PREMIX normalizes input mole fractions to "1.000.." - how does PHOENICS handle this?

5. Are mixture-averaged diffusivities still activated only by: DO K=16 to KS PRNDTL(KS)=-GRND9 ENDDO ...and should these be set?

6. Should TEMP0 be set? How?

7. Has the Curtiss-Hirschfelder model been fixed yet?

Thank you very much for your time PattiMichelle

Hi Patti, here are quick answers to your questions. It would take me too much time to investigate each one in depth to give a definate answer, but I hope this is better than nothing. I agree that the CHEMKIN documentation is unclear and needs improvement. You may be interested to know that we talking to the vendors of CHEMKIN III with a view to upgrading the PHOENICS-CHEMKIN interface so as to permit problem set-up in VR. This will make it much easier to use and the documentation will be modified accordingly.

1. (/phoenics/D_POLIS/D_LECS/Chemkin.htm, #11) states that input is in CGS units but the example mechanism (#10) file is in JOULES/MOLE (not CGS)

The PHOENICS-CHEMKIN calculation is always carried out in CGS units because the CHEMKIN software works in those units. The PHOENICS input must be CGS units, but there are the exceptions to accommodate the CHEMKIN software. Thus, the input for the mechanism file is exactly that documented on p36 of the CHEMKIN II SAND89-8009 report, and so the CHEMKIN practice of allowing alternative units in the mechansim file has been retained in the PHOENICS-CHEMKIN interface.

2. Given question #1 above, what are INLET() pressure units? (CGS/Atm/MKS?)

As I recall the operating pressure is specified in atmospheres and this is converted to CGS units by the interface. The computed pressures are in CGS units. I am not sure what you mean by INLET() pressure, unless you mean the COVAL for P1 for a mass inflow boundary. In this case, the units would be CGS for the mass inflow rate per unit area for a PATCH with area TYPE.

3. Also in section #11 of these docs, the inlet composition may be "Mass, Mole, or concentration fractions" - no examples of how to select mole fraction (preferred unit) are given.

Apart from mass fractions, I know only of the option for specifying mol fractions. The procedure is the same as for mass fractions, but the 8th character of the PATCH name must be X.

4. CHEMKIN/PREMIX normalizes input mole fractions to "1.000.." - how does PHOENICS handle this?

The onus is on the user to ensure his input is consistent and meaningful. The inlet value need not be specified for the last species, i.e. mass fraction that is not solved.

5. Are mixture-averaged diffusivities still activated only by: DO K=16 to KS PRNDTL(KS)=-GRND9 ENDDO ...and should these be set?

I think the answer is yes to the first part. However, I am uncertain about thesecond part because there was some internal inconsistency problem with one the diffusion models, but I think it was the Curtiss-Hirschfelder model as you note in your next question.

6. Should TEMP0 be set? How?

As I remember, if the temperature TEM1 is in K, then TEMP0 is set to zero. However, if TEM1 is in deg C, then TEMP0 is set to 273K.

7. Has the Curtiss-Hirschfelder model been fixed yet?

No.

Thanks Dr. Malin!!

> "The procedure is the same as for mass fractions, but the 8th character of the PATCH name must be X."

By this you mean naming INLET's as NOCPCK1X, NOCPCK2X, etc. causes the associated VALUE statement inputs to be treated as mole fractions, correct?

> "if TEM1 is in deg C, then TEMP0 is set to 273K"

I feel pretty dumb asking this one, but after looking in the docs, I relize I don't know how to insure the units to K - are you saying that including the PIL command TEMP0=0.0 does this?

> "the CHEMKIN practice of allowing alternative units in the mechansim file has been retained in the PHOENICS-CHEMKIN interface."

I just wanted to make sure that if I input a CHEMKIN mechanism in JOULES/MOLE then PHOENICS would interpret the rate constant data correctly.

Thank You Very Much, PattiMichelle

> By this you mean naming INLET's as NOCPCK1X,
: NOCPCK2X, etc. causes the associated VALUE statement > inputs to be treated as mole fractions, correct?

Yes that's right.

> I relize I don't know how to insure the units to K - > are you saying that including the PIL command
: TEMP0=0.0 does this?

I think that TEMP0 = 0.0 by default, and so the default is that PHOENICS presumes that any temperatures defined by the user are in K.

> I just wanted to make sure that if I input a CHEMKIN > mechanism in JOULES/MOLE then PHOENICS would
: interpret the rate constant data correctly.

Yes, it should be OK.

> 4. CHEMKIN/PREMIX normalizes input mole fractions to "1.000.." - how does PHOENICS handle this?

> The onus is on the user to ensure his input is consistent and meaningful. The inlet value need not be specified for the last species, i.e. mass fraction that is not solved.

My concern was that there are only 6 decimal places available to manually make sure the sum of the input species fractions add to "1". If I add trace species, say, at the 10-8 level, how is it possible to specify these relative to the dominant species inputs, for which only 6 decimal places of specification are available?

My understanding is that 6 decimal places will be accepted after the decimal point in the 1.E-8 format.

I think I wasn't clear - if I input species A=1.999999E+00 and then want to add species B=1.000000E-08, then there is 9.900000E-07 that is unspecified. Actually, I suppose this is a fairly minor point.

I also tried using your suggestion of substituting NOCPCK1X & NOCPCK2X for NOCPCK1 & NOCPCK2, in a q1 file that I knew ran well, to allow me to input Mole Fractions. After the substitution, EARTH stops with an error. At the end of the result file is:

GROUND from FN1 & grex3

but no setting made. Please check. PATCH NAME = NOCPCK1 INDVAR = 148 f index = 9972 f value = HEATBL

> 4. CHEMKIN/PREMIX normalizes input mole fractions to "1.000.." - how does PHOENICS handle this?

> The onus is on the user to ensure his input is consistent and meaningful. The inlet value need not be specified for the last species, i.e. mass fraction that is not solved.

I was having trouble getting a file to run, Earth replied with an error telling me to check my data. I isolated the problem to an INIT PATCH, in which species fractions appear to need to sum to less than unity. The following PATCH will not run if the 'H' INIT is changed to 1E-3. I am wondering if NOCPCK INLET's have the same constraint.

PATCH (FLINIT ,INIVAL,2,0,0,0,0,0,1,1)

INIT(FLINIT ,TEM1, 0.000000E+00, 1.218320E+03)

INIT(FLINIT ,H2* , 0.000000E+00, 1.000000E-10)

INIT(FLINIT ,H , 0.000000E+00, 1.000000E-05)

INIT(FLINIT ,O , 0.000000E+00, 1.000000E-02)

INIT(FLINIT ,O2 , 0.000000E+00, 1.000000E-10)

INIT(FLINIT ,OH , 0.000000E+00, 1.000000E-10)

INIT(FLINIT ,H2O , 0.000000E+00, 1.000000E-10)

INIT(FLINIT ,SO , 0.000000E+00, 1.000000E-10)

INIT(FLINIT ,SO2 , 0.000000E+00, 1.000000E-05)

INIT(FLINIT ,HSO2, 0.000000E+00, 1.000000E-10)

INIT(FLINIT ,HSO , 0.000000E+00, 1.000000E-10)

INIT(FLINIT ,AR , 0.000000E+00, 5.000000E-01)

INIT(FLINIT ,N2 , 0.000000E+00, 5.000000E-01)