Hello,
I hope you are doing great.
I have been working on MFiX-GSP model since recently.
I want to vary the densities of each individual GSP particle, which I am doing in the particle_input.dat file formatting, however, as I run the case and check the particle_output.dat file, the Density column changes to F, I am not sure which of the density value is being used in the MFiX now, can you please let me know?
Also to check how do we read the total mass of the particles generated after running the simulation of mfix?
I am attaching the zip file of my case.
I am looking forward to hearing from you soon.
Sincerely,
Sumanth Penmetcha
GSP_528_08_19.zip (640.3 KB)
I have used the Monitor region to know the sum of total particle mass.
For sanity check I wanted to compare the printed result with manually calculating (Volume)x(Density), but unfortunately it is not matching and the difference of margin being huge.
I am attaching both the particle_input.dat file, particle_output.dat and the csv file from the Monitor Region, can anyone please let me know where am I making an error in calculation or modeling?
Any help would be greatly appreciated.
If you could at least tell me how the bounding diameter is used to calculate the volume, that would be immensely helpful!
Thanks!
particle_input.dat (34.3 KB)
particle_output.dat (600.6 KB)
SUM_OF_MASS_FOR_WHOLE_REGION.csv (544 Bytes)
gsp_config.csv (1.9 KB)
gsp_config.dat (2.7 KB)
gsp_properties.dat (468 Bytes)
In current version of MFiX for GSP, bounding diameters of glued particles are given in particle_input.dat file. In partilce_output.dat file, the diameters and densities of each COMPONENT spheres are saved.
I calculated the total mass of particles based on your particle_output.dat file. It is the same as what is given in the csv file from the monitor.
Thank you so much!
That really solved one of the major problems I was facing, for some reason, the output.dat file was not showing me the density column, but I had to re-install the software and then it started showing up.
I do have one question though, can you please tell me what is the relation between the bounding diameter that I insert in the particle_input.dat file to the diameters that print out in the particle_output.dat file.
Like for example one of the bounding diameter that I am inserting is b_d=0.011349022 m and the corresponding component spheres I am getting are:
0.19666834E-02
0.19916058E-02
0.19934330E-02
0.19945452E-02
0.19949424E-02
0.19946247E-02
0.19935806E-02
0.19918215E-02
0.19669671E-02
How are the component spheres(in this case 9) are formed, is there a specific scaling ratio that is applied?
I would be grateful for your clarification.
Once again thank you so much.
Sincerely,
Sumanth Penmetcha
The bounding diameter in the particle_input.dat is the diameter of the bonding sphere around your non-spherical (glued) particles.
The diameters in the particle_output.dat are the diameters of each component sphere in your glued particles. This depends on how you discretize the glued particle.
Component spheres can be generated in Solids->Materials->Shape->Glued-sphere designer. You can specify how many spheres you want in each direction.
1 Like
Thanks again for your quick response, I really appreciate it!!
I understand that the diameters in the particle_output.dat file are diameters in each component.
I am discretizing in such a way that in the direction of x=y=1 and in direction of z=9 component spheres.
So, it looks like a cylinder; I am attaching an image for your reference.
So when I am just calculating the ratio between bounding diameter to one of respective component sphere, I am getting it as 5.7706, I wonder where did this ratio come from? How is the solver choosing that?
For your reference I am also uploading the GSP properties data files
Once again thank you so much for your prompt responses.
gsp_config.dat (2.7 KB)
gsp_properties.dat (468 Bytes)
Hi Sumanth,
Currently, MFIX does not support write particle output function for GSP model, we are working on that and probably release it at the end of September. So at this point, don’t use particle_output.dat to initialize a new simulation.
For bounding diameter, you can think a big sphere who’s center is in the glued sphere particle center (in your case, in the cylinder center). This bounding diameter ensure this big sphere can include all the component spheres of a glued sphere particle.
This bounding diameter is only used to control the size of the glued sphere, if you look at gsp_config.dat, values inside are actually the real value divided by bounding diameter. For example, you found a value of 0.173291 in gsp_config.dat as relative diameter, 0.173291*bounding diameter provides the real diameter of the component sphere. In this way, when you change bounding diameter in particle_input.dat, you can change the size of gsp particles. you will get like big cylinders and small cylinders, but they still have same aspect ratio.
1 Like
Hello Renjie,
Thank you so much; this clears up a lot of things!
I think I have a better understanding of the GSP!
Until next time,
Sumanth Penmetcha
Hello again,
I was wondering if I were have to calculate the length of the z-axis of each individual GSP particle(because there are overlaps between component spheres), like the one I am attaching for reference, do we have a scaling ratio for that as well?
If so, can you please let me know?
Is it similar to the one in this thread - How are the heights of the SQP particles created, are determined? - #21 by cgw
Thank you so much, again!
GSP particles can be considered an approximation of the original particle shape, so you might see component spheres intersecting with the original geometry. It’s not possible to achieve a 100% match, even if you increase the number of component spheres. Currently, we haven’t added a scaling ratio for adjustment, but if you wish, you can directly modify the gsp_config.dat file. The XYZ values in the file represent the position of each component relative to the particle center, so you can uniformly multiply the Z values by a coefficient, for example, 0.95.
Thank you for your response.
I do not actually want to change any z-axis. The particles are already generated, I do have the GSP Particle’s bounding diameter and also the diameter of each component sphere.
I would like to know the z-axis of each particle that is (already) generated, so that I can calculate the maximum projected area of each GSP Particle.
if it helps, I am also attaching the GSP_config.dat. and properties.dat
gsp_config.dat (2.7 KB)
gsp_properties.dat (468 Bytes)
Thank you for your time, Renjie.
you have locations of each component sphere, you also have diameter of each component sphere. Like I said before, they are all in gsp_config.dat as relative values. You just need to multiply them with bounding diameter and do some calculations to extract the z-axis value. That should not be difficult to get any projection area you want.
Hello Renjie,
Thank you so much for your response!
I think I have understood the solution, I am putting down a sample solution here, if you’ve got time please let me know if it’s correct to further put confidence of my understanding.
From gsp_config.dat I see the
Minimum Relative Z-coordinate: Zmin, rel=−0.413354
Maximum Relative Z-coordinate: Zmax, rel=0.413280
For example lets take a bounding diameter(b_d)= 0.011349022m
then Z-axis length would be absolute value sum of (Zmin x b_d) + (Zmax x b_d)= |−0.00469116| + |0.00469032| = 0.00938149m
Once again, I highly appreciate the responses and answers from the MFiX Forum; this has been immensely helpful!
Sincerely,
Sumanth Penmetcha
Yeah, that make sense to me. Those are coordinates of component spheres’ center, so in your case, maybe you want to also add the radius?
You can easily retrieve the diameter values of the specific component spheres associated with the Z_min and Z_max.
Thanks for the affirmation.
How do we retrieve the diameter values associated with Z_min and Z_max? Do you mean particle_output.dat file?
its in the gsp_config.dat, next column to the coordinate, ofc they are relative values, you want to multiply them with b_d as well. And they are diameters of component spheres BTW.
