Different results

Hi,
I just ran a case of a bidisperse flow using versions 2013 and 2021 of MFiX-TFM. The case is the same as the legacy tutorial except for the gas inlet velocity and initial solid volume fractions in the bed. However, the results are completely different. The settings are the same and one can easily check that by opening the mfix.dat of the 2013 version using MFiX 2021 GUI and only changes the value for the following parameters as the keywords have changed:

Solid density(1)=2524 kg/m^3
Solid density(2)=2524 kg/m^3
and
IC_ep_s(1)=0.1475 (Which is calculated by dividing the ROP_s(1) by Ro_s(1) in 2013 version)
IC_ep_s(1)=0.4425 (Which is calculated by dividing the ROP_s(2) by Ro_s(2) in 2013 version)

You also need to uncheck and check the detect stall in GUI to correct the keyword value in the file. Finally, after saving the file in GUI, you may need to delete the unknown keywords section from the mfix file.

I have also compared the files line by line and I could not find any differences that might lead to different results.

I have attached the 2013 and 2021 versions of the case file along with the average bed
heights of the different phases vs time plots. It should be mentioned that the results using the 2013 version are in better agreement with the related experimental data.

I appreciate it if you can comment on any issue that may cause the difference.

Thanks, Mohsen

SGE_Case1_2021.mfx (7.8 KB)
mfix.dat (6.5 KB)

Hi Mohsen.

Can you clear something up for me? Are you comparing the results of running the two input files in the current MFiX, or are you running the code from 2013? There are 4 combinations of (code, input file), I’m not sure which 2 you are comparing.

Loading the old mfix.dat into current MFiX will automatically apply a number of changes. Note that the console window shows all the changes that are applied, as well as reporting 7 warnings:


CGS units detected! Converted to SI. Original file saved as mfix.dat.cgs

Warning: mfix.dat: line 54: unknown keyword: ro_s
Warning: mfix.dat: CGS units detected! Converted to SI. Original file saved as mfix.dat.cgs
Warning: mfix.dat: rdf_type 'None' invalid for polydisperse cases, setting to LEBOWITZ
Warning: mfix.dat: no species for phase 1
Warning: mfix.dat: no species for phase 2
Warning: mfix.dat: ic_rop_s(2,1) and ic_ep_s(2,1) both set
Warning: mfix.dat: ic_rop_s(2,2) and ic_ep_s(2,2) both set
Warning: Project has been converted to SI.
Please check result of conversion.
Loaded mfix.dat with 7 warnings

I applied a little normalization script to the files to make them easier to compare. It removes all comments and keywords which are at default values. Then sorted the lines.

Here’s the complete set of diffs:


mfix.dat                                                      SGE_Case1_2021.mfx

                                                              > bc_ep_s(1,1) = 0.0
                                                              > bc_ep_s(1,2) = 0.0
                                                              > bc_t_g(1) = 293.15
                                                              > bc_t_s(1,1) = 293.15
                                                              > bc_t_s(1,2) = 293.15
bc_v_g(1) = 120.0                                             | bc_u_s(1,1) = 0.0
bc_x_e(1) = 15.0                                              | bc_u_s(1,2) = 0.0
bc_x_e(2) = 15.0                                              | bc_v_g(1) = 1.2
                                                              > bc_v_s(1,1) = 0.0
                                                              > bc_v_s(1,2) = 0.0
                                                              > bc_w_g(1) = 0.0
                                                              > bc_w_s(1,1) = 0.0
                                                              > bc_w_s(1,2) = 0.0
                                                              > bc_x_e(1) = 0.15
                                                              > bc_x_e(2) = 0.15
bc_y_n(2) = 50.0                                              | bc_y_n(2) = 0.5
bc_y_s(2) = 50.0                                              | bc_y_s(2) = 0.5
close_packed = .TRUE.  .TRUE.                                 <
d_p0 = 0.15D0 0.25D0                                          | d_p0(1) = 0.0015
                                                              > d_p0(2) = 0.0025
detect_stall = .FALSE                                         | detect_stall = .FALSE.
discretize = 9*2                                              | discretize(1) = 2.0
                                                              > discretize(2) = 2.0
                                                              > discretize(3) = 2.0
                                                              > discretize(4) = 2.0
                                                              > discretize(5) = 2.0
                                                              > discretize(6) = 2.0
                                                              > discretize(7) = 2.0
                                                              > discretize(8) = 2.0
                                                              > discretize(9) = 2.0
                                                              > gravity_x = 0.0
                                                              > gravity_y = -9.80665
                                                              > gravity_z = 0.0
ic_rop_s(1,1) = 0.37229                                       | ic_ep_s(1,1) = 0.1475
ic_rop_s(1,2) = 1.11687                                       | ic_ep_s(1,2) = 0.4425
ic_rop_s(2,1) = 0.0                                           | ic_ep_s(2,1) = 0.0
ic_rop_s(2,2) = 0.0                                           | ic_ep_s(2,2) = 0.0
                                                              > ic_p_star(1) = 0.0
                                                              > ic_p_star(2) = 0.0
                                                              > ic_t_g(1) = 293.15
                                                              > ic_t_g(2) = 293.15
                                                              > ic_t_s(1,1) = 293.15
                                                              > ic_t_s(1,2) = 293.15
                                                              > ic_t_s(2,1) = 293.15
                                                              > ic_t_s(2,2) = 293.15
ic_x_e(1) = 15.0                                              | ic_w_g(1) = 0.0
ic_x_e(2) = 15.0                                              | ic_w_g(2) = 0.0
                                                              > ic_w_s(1,1) = 0.0
                                                              > ic_w_s(1,2) = 0.0
                                                              > ic_w_s(2,1) = 0.0
                                                              > ic_w_s(2,2) = 0.0
                                                              > ic_x_e(1) = 0.15
                                                              > ic_x_e(2) = 0.15
ic_y_n(1) = 15.0                                              | ic_y_n(1) = 0.15
ic_y_n(2) = 50.0                                              | ic_y_n(2) = 0.5
ic_y_s(2) = 15.0                                              | ic_y_s(2) = 0.15
                                                              > k_s0(1) = 1.0
                                                              > k_s0(2) = 1.0
                                                              > ks_model(1) = 'BAUER'
                                                              > ks_model(2) = 'BAUER'
mu_g0 = 0.00018                                               | mu_g0 = 1.8e-05
                                                              > nmax_s(1) = 0.0
                                                              > nmax_s(2) = 0.0
nodesi = 1   NODESJ = 1   NODESK = 1                          | nodesi = 1.0
                                                              > nodesj = 2.0
                                                              > nodesk = 1.0
                                                              > project_version = '5'
                                                              > rdf_type = 'LEBOWITZ'
resid_string = 'P0' 'P1' 'U0' 'R2' 'V0' 'V1' 'V2'             | resid_string(1) = 'P0'
ro_g0 = 0.0012                                                | resid_string(2) = 'P1'
ro_s = 2.524D0 2.524D0                                        | resid_string(3) = 'U0'
run_name = 'SGE'                                              | resid_string(4) = 'R2'
                                                              > resid_string(5) = 'V0'
                                                              > resid_string(6) = 'V1'
                                                              > resid_string(7) = 'V2'
                                                              > resid_string(8) = 'U1'
                                                              > ro_g0 = 1.2
                                                              > ro_s0(1) = 2524.0
                                                              > ro_s0(2) = 2524.0
                                                              > run_name = 'SGE_CASE1_2021'
species_eq = .FALSE.    .FALSE.    .FALSE.                    | species_eq(0) = .FALSE.
spx_dt = 0.01 0.01      0.01  0.01  0.01      100. 100.   0.0 | species_eq(1) = .FALSE.
                                                              > species_eq(2) = .FALSE.
                                                              > spx_dt(1) = 0.01
                                                              > spx_dt(10) = 1.0
                                                              > spx_dt(11) = 1.0
                                                              > spx_dt(2) = 0.01
                                                              > spx_dt(3) = 0.01
                                                              > spx_dt(4) = 0.01
                                                              > spx_dt(5) = 0.01
                                                              > spx_dt(6) = 100.0
                                                              > spx_dt(7) = 100.0
                                                              > spx_dt(8) = 0.01
                                                              > spx_dt(9) = 100.0
tstop = 60.0                                                  | tstop = 30.0
units = 'CGS'                                                 | units = 'SI'
xlength = 15.0                                                | x_max = 0.15
ylength = 50.0                                                | x_min = 0.0
                                                              > y_max = 0.5
                                                              > y_min = 0.0


	

There’s quite a bit here and I won’t comment on every single change. But the noteworthy items, I think, are:

  1. The file has been converted from cgs to SI. This is probably OK but checking the cgs-SI conversion is never a bad idea (as the warning messages suggest)

  2. Some of these diffs, e.g. discretize, are spurious, due to the older mfix.dat using formats like ‘9*2’ which are now deprecated. Same for d_p0, species_eq, close_packed, spx_dt and a few others. These values have not changed.

  3. A lot of BC and IC properties are defined in the 2021 file that are not in the 2013 version. In the 2013 version ic_rop_s is specified, in the 2021 version the ICs are set up by specifying volume fraction ic_ep_s. This conversion should be checked.
    Also, default temperatures for BCs and ICs have been auto-applied.

  4. The 2021 file specifies KS_MODEL = 'BAUER' and RDF_TYPE='LEBOWITZ', these are both non-default values which are not specified in the 2013 file

– Charles