Hello, So I am running simulations of a simple agitator in a bucket (see agitator pic). I am interested in calculating the torque of a spinning agitator but my results seem odd when I use tangential history on the gran/wall and pair_style fix. When I plot the f_agitator_post[6] (from everything I read, it seems that this would be the torque acting on the agitator body in the z-direction), the values jump between positive and negative 2 N-m (see tangent_history_chart). This result seems rather odd because the blade is only spinning in one direction. When I re-run this simulation with tangential history removed from the pair style and the gran/wall, the results look more realistic (see no_history_chart).
Here is the portion of the code where I place the geometry and define the torque; my plots are of f_agitator_post[6]. When I re-ran the code I just took out the words tangential history. My first impression would be that I am not measuring the right axis, but I do not think this is the case because I recreated the stl file several times and the blade clearly runs centered. I even tried looking at the stl file to figure out if it was off centered. From my understanding, tangential history is the friction component that applies load parallel to the face. Unless I am missing something conceptually, I do not understand how the tangential component is producing this result.
#*****************************************************************
#New pair style
pair_style gran model hertz tangential history
pair_coeff * *
timestep ${dt}
# Setup integration
fix 1 all nve/sphere
# Setup gravity
fix 2 all gravity 9.81 vector 0.0 0.0 -1.0
#*********************** Walls ***************************
#import triangular mesh
fix agitator_post all mesh/surface/stress file CAD/agitator_post.stl type 2 scale 0.001 curvature 1E-6
fix hopper_post all mesh/surface file CAD/hopper_post.stl type 3 scale 0.001 curvature 1E-6
fix base_post all mesh/surface file CAD/base_post.stl type 4 scale 0.001 curvature 1E-6
variable ForceX equal f_agitator_post[1]
variable ForceY equal f_agitator_post[2]
variable ForceZ equal f_agitator_post[3]
variable CoupleX equal f_agitator_post[4]
variable CoupleY equal f_agitator_post[5]
variable CoupleZ equal f_agitator_post[6]
#define rotation
fix move all move/mesh mesh agitator_post rotate/variable origin 0. 0. 0. axis 0. 0. 1. omega v_angVelo
#use the imported mesh as granular wall
fix granwalls_post all wall/gran model hertz tangential history mesh mesh n_meshes 3 meshes agitator_post hopper_post base_post
#**********************************************************
Thank you,
Aj.Spin
Attachment | Size |
---|---|
![]() | 29.08 KB |
![]() | 42.54 KB |
![]() | 24.52 KB |
ckloss | Thu, 09/25/2014 - 08:44
Hi Aj.Spin,
Hi Aj.Spin,
1) which version are you using?
2) is the average value the same?
2) remote diagnosis is difficult, if not impossible. If you're using the latest version of LIGGGHTS, we can have a look if you post a very simple example that runs within say a minute or 2.
Cheers
Christoph
AJ.Spin | Mon, 10/06/2014 - 14:46
The model was 3.0.2 but i
The model was 3.0.2 but i have since upgrades to 3.0.3. There simliar amount of torque with the tangent history. I attempted to make a test case to send by decreasing the diameter of my tube and increasing my particle size. The negative values went away with the larger particles so I cannot show it with a simple example. However, seeing that the particle size influenced the result, gave me new insight. Basically your questions, allowed it to make sense so I think I am good now.
Thank you,
Anthony
saeedga74 | Sun, 09/27/2020 - 20:23
I have the same Issue and I cant understand why
I have the same Issue and when I use tangential history the graph is so odd and I'm using liggghts 3.0.7 what should I do to get more realistic results