|
[Sponsors] |
April 10, 2017, 11:26 |
DPM and MHD
|
#1 |
New Member
James Kim
Join Date: Mar 2017
Posts: 13
Rep Power: 9 |
Hi, I'm trying to simulate the trajectories of magnetic particles in fluid channel with external magnetic field applied, using discrete phase modelling (DPM) with magnetohydrodynamic (MHD) addon module. I can get the magnetic field data (B0) from Magnetostatic module, and apply it into Fluent and after running calculation, it would show me that the magnetic field (B0) applied in Fluent is equivalent to what was obtained from the Magnetostatic. However, even if I crank up the DPM material property (magnetic permeability) to see how the applied magnetic field would effect the trajectory of the magnetic particle, I wouldn't see any change in the trajectory at all. (So I would apply -1 T in the y direction throughout the channel (which should pull down the particles) but I would only see straight line in the particle trajectory) From the Forum, I've learned that I'm supposed to apply the magnetic field after the initialization so I've been doing that, but it wouldn't affect the particle trajectory at all so I was wondering if there is other specific steps that I would have to take. I also set up the UDF in the DPM so that the body force is equivalent to mhd_dpm_force::mhd and source as mhd_dpm_source::mhd. Does anyone have experience with this to help me out or give me any hints?
Thank you, James Kim |
|
June 8, 2017, 21:05 |
|
#2 |
New Member
Daisy
Join Date: Jun 2017
Location: Indiana
Posts: 9
Rep Power: 9 |
Hi, James
I wonder how you apply the magnetostatic data to fluent? Can you give me some hints? Thanks so much. |
|
June 9, 2017, 10:53 |
|
#3 |
New Member
James Kim
Join Date: Mar 2017
Posts: 13
Rep Power: 9 |
Hi Daisy,
in the ANSYS manual, if you look up mag data, in "appendix C. External magnetic Field data format", there is an explanation there. Also, to apply the mag data to fluent, you need to look at the MHD Module section in the manual. It should have an explanation of how MHD module works. Hope this helps! |
|
June 9, 2017, 11:57 |
|
#4 |
New Member
Daisy
Join Date: Jun 2017
Location: Indiana
Posts: 9
Rep Power: 9 |
Hi, James.
I already checked all the explanation in the help documents. I try to follow the steps and apply the datas to fluent, it just doesn't work. Can you give me more details about how you get the fluent running well? |
|
June 9, 2017, 12:25 |
|
#5 |
New Member
James Kim
Join Date: Mar 2017
Posts: 13
Rep Power: 9 |
Hey Daisy,
So this is how I ran my MHD module. You first load the MHD module unto Fluent, then set up your flow condition (BC and everything), initialize the solution, then apply the mag data from MHD tab (models), initialize MHD solution, apply the magnetic field (it is important that you click the apply button right before you go for calculation), and then click run. Then after the calculation is done, you should be able to see the mag field applied through user defined memory! Hope this helps, james |
|
June 9, 2017, 12:35 |
|
#6 |
New Member
Daisy
Join Date: Jun 2017
Location: Indiana
Posts: 9
Rep Power: 9 |
OK, I just did exactly what you said, but the problem is I can't apply the magnetic field. It always come out F1 error in fluent. Have you met this in your simulation?
|
|
June 9, 2017, 12:43 |
|
#7 | |
New Member
James Kim
Join Date: Mar 2017
Posts: 13
Rep Power: 9 |
No not really.... maybe the magdata format is wrong?
This is the mag data I've used when I was doing trouble shooting. Quote:
|
||
June 9, 2017, 12:55 |
|
#8 |
New Member
Daisy
Join Date: Jun 2017
Location: Indiana
Posts: 9
Rep Power: 9 |
I have same format datas as yours. But why the numbers in your second line are not same? like 100 100 100.
|
|
June 9, 2017, 12:57 |
|
#9 |
New Member
James Kim
Join Date: Mar 2017
Posts: 13
Rep Power: 9 |
It means I want 100 division in x axis, 10 division in y axis, 1 division in z axis. so it will look like a thin, long, rectangular channel.
|
|
June 9, 2017, 13:02 |
|
#10 |
New Member
Daisy
Join Date: Jun 2017
Location: Indiana
Posts: 9
Rep Power: 9 |
But the help document: "The second line defines the number of data points in the X , Y and Z directions."
So I think these three number should be same. What do you think? |
|
June 9, 2017, 14:07 |
|
#11 |
New Member
James Kim
Join Date: Mar 2017
Posts: 13
Rep Power: 9 |
For my case, I had different number of mesh division (100 X, 10 Y, 1 Z) so the numbers on the second lines depend on your geometry and the magnetic field data.
|
|
June 9, 2017, 16:10 |
|
#12 |
New Member
Daisy
Join Date: Jun 2017
Location: Indiana
Posts: 9
Rep Power: 9 |
OK,I finally figure out my problem. The datas I exported are random, maybe I cannot apply them directly into fluent and I need to interpolate them to structural datas. I decide to use matlab to interpolate datas. How about you? Did you use any software to process the datas exported from ansys magnetostatic?
|
|
June 9, 2017, 16:28 |
|
#13 |
New Member
James Kim
Join Date: Mar 2017
Posts: 13
Rep Power: 9 |
Yup, I played around with the mesh size and number in magnetostatic to get the mag data into the right format!
|
|
June 14, 2017, 14:12 |
|
#14 |
New Member
Daisy
Join Date: Jun 2017
Location: Indiana
Posts: 9
Rep Power: 9 |
Hi, James
I wonder how do you get "magnetic field (B0) applied in Fluent is equivalent to what was obtained from the Magnetostatic" like compare the contours of B0 in fluent and magnetostatic? |
|
June 14, 2017, 15:55 |
|
#15 |
New Member
James Kim
Join Date: Mar 2017
Posts: 13
Rep Power: 9 |
Hi Daisy,
After you calculate your result, in the result/graphics section in fluent, you can view contour plot of user defined memory(which will be B0 in this case). I checked my magnetic field that way. |
|
June 14, 2017, 20:03 |
|
#16 |
New Member
Daisy
Join Date: Jun 2017
Location: Indiana
Posts: 9
Rep Power: 9 |
Hi, James
Thanks so much. I have checked the results like you said, but the question I have now is that the distribution of values (like B0) on each direction cannot match the results in magnetostatic, but the ranges (max & min) of values on each direction are equivalent. I still try to figure that out. |
|
June 16, 2017, 17:52 |
|
#17 |
New Member
Daisy
Join Date: Jun 2017
Location: Indiana
Posts: 9
Rep Power: 9 |
Hi, James.
I have figured out all the problems of exporting data into fluent.Thanks so much. I wonder did you validate the module of magetostatic? |
|
November 24, 2018, 11:54 |
|
#18 |
New Member
Melisa Cardona
Join Date: Apr 2018
Location: Colombia
Posts: 5
Rep Power: 8 |
Hi James
Did you figure it out how to affect the particles with the magnetic field? I am struggling with the same issue |
|
Tags |
dpm, mhd |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Lorentz force in MHD module | amirbahador | FLUENT | 13 | March 26, 2019 04:34 |