It seems that when using particle_input.dat the values for particle parameters like diameter, density, etc which are set in the MFiX GUI are ignored completely. (Potentially confusing)
We need to use a Version 3 particle input file. Although the Version 3 format is not well documented yet, by looking at the functionWRITE_PART_OUT_V3P0 in read_particle_input.f you can see what the header and instructions look like. Now the data starts on line 39 instead of 35 and there are additional fields for SuperDEM parameters.
Here’s your file in version 3 format. I went ahead and removed the constant colums.
Version 3.0
========================================================================
Instructions:
Dimension: Enter "2" for 2D, "3" for 3D (Integer)
Particles: Number of particles to read from this file (Integer)
For each variable, enter whether it will be read from the file
("T" to be read (True), "F" to not be read (False)). When not read, enter the
default values assign to all particles.
Coordinates are always read (X,Y in 2D, X,Y,Z in 3D)
Phase_ID, Diameter, Density, Temperature are scalars
Velocity requires U,V in 2D, U,V,W in 3D
Temperature is only read/set if the energy equation is solved
Species are only read/set if the species equations are solved, and requires
all species to be set (if phase 1 has 2 species and phase 2 has 4 species,
all particles must have 4 entries, even phase 1 particles (pad list with zeros)
User scalars need DES_USR_VAR_SIZE values
Data starts at line 39. Each column correspond to a variable.
SQP_semi-axes, SQP_roundness, and SQP_quaternion are only needed when using SuperDEM.
========================================================================
Dimension: 3
Particles: 4500
========================================================================
Variable Read(T/F) Default value (when Read=F)
========================================================================
Coordinates: T Must always be T
Phase_ID: F 1
Diameter: F 6.42583409E-03
Density: F 2.65000000E+03
Velocity: F 0 0 0
Temperature: F (Ignored if energy eq. is not solved)
Species: F (Ignored if species eq. are not solved)
User_Scalar: F (Ignored if no user scalars are defined)
SuperDEM_semiaxis F 0.00635 0.00635 0.00175
SuperDEM_roundness F 2 8
SuperDEM_quaternion F 1 0 0 0
===========================================
X (m) Y (m) Z (m)
===========================================
1.34940000E-02 1.10180000E-02 1.55820000E-02
2.69890000E-02 1.10180000E-02 1.55820000E-02
4.04830000E-02 1.10180000E-02 1.55820000E-02
...
particle_input.dat (199.7 KB)
Using this input file, the case runs! And here’s a picture to prove it:
Look at those pretty non-spherical particles! 
But alas, after a while there is a new error:
Backtrace for this error:
#0 wrap_pow
at xpow.c:31
#1 sq_equivalent_radius_mod_MOD_sq_gradient
at des/sq_equivalent_radius.f:53
#2 sq_contact_newton_dpmethod_mod_MOD_func_dp_a
at des/sq_contact_detection_newton.f:949
#3 sq_contact_newton_dpmethod_mod_MOD_sq_contact_newton_dp_a
at des/sq_contact_detection_newton.f:267
#4 sq_calc_force_mod_MOD_calc_force_superdem
at des/sq_calc_force_superdem.f:341
#5 des_time_march_MOD_des_time_step
at des/des_time_march.f:192
#6 run_dem
at mfix.f:211
#7 run_mfix
at mfix.f:146
#8 main_MOD_run_mfix0
at main.f:79
we are trying to compute x*x where x=6.6205546496795669e+154 and it’s an overflow. It’s also Friday, I will return to this topic next week
– Charles