Implementing JJ BC for solid phase and NSW for gas phase

[mohsenclick]

Hi,
I wonder if there is any difference between the following formats for implementing the JJ BC for the solid phase and NS BC for the gas phase? As I read in the forum, they should be the same, but the 2nd format that is created by the GUI does not exactly satisfy the NS for the gas phase. Please see the values of the V_g of two wall adjacent cells, 1st interior cell and ghost cell:

Y = 0.225000E-0001
Z = 0.125000E-0002
Time = 0.100001E+0002
X V_g
-.125000E-0002 -.1090446949E+0001
0.125000E-0002 0.1090439200E+0001

However, the 1st format exactly satisfies the NS bc for the gas phase on the wall

(I am running an old case that has been run using the 2015 or 2016 version of the code and when I open it in GUI the 1st format will be saved for the BC but when I try to use the GUI to set the same boundary condition from scratch, I get the 2nd format.)

1st format---------------------------------------------
bc_type(3) = ‘NSW’
bc_x_e(3) = 0.3
bc_x_w(3) = 0.3
bc_y_s(3) = 0.0
bc_y_n(3) = 0.7
bc_z_b(3) = 0.0
bc_z_t(3) = 0.015
bc_c_t_g(3) = 0.0
bc_hw_t_g(3) = 0.0
bc_jj_ps(3) = 1
bc_uw_g(3) = 0.0
bc_vw_g(3) = 0.0
bc_ww_g(3) = 0.0
# Solid 1
bc_c_t_s(3,1) = 0.0
bc_hw_t_s(3,1) = 0.0
bc_thetaw_m(3,1) = 0.0
bc_uw_s(3,1) = 0.0
bc_vw_s(3,1) = 0.0
bc_ww_s(3,1) = 0.0
2nd format-------------------------------------------------------------------------
bc_type(4) = ‘PSW’
bc_x_e(4) = 0.3
bc_x_w(4) = 0.3
bc_y_s(4) = 0.0
bc_y_n(4) = 0.7
bc_z_b(4) = 0.0
bc_z_t(4) = 0.015
bc_jj_ps(4) = 1
bc_uw_g(4) = 0.0
bc_vw_g(4) = 0.0
bc_ww_g(4) = 0.0
# Solid 1
bc_c_theta_m(4,1) = 0.0
bc_hw_s(4,1) = 0.0
bc_hw_theta_m(4,1) = 0.0
bc_uw_s(4,1) = 0
bc_vw_s(4,1) = 0
bc_ww_s(4,1) = 0

[cgw]

Hi Mohsen

I don’t think these two forms of BC are equivalent (as your reported numbers show).

Here’s a side-by-side summary of the differences - produced by using ‘sdiff’ after some slight adjustments (changing 4 to 3 in the second listing)

bc_c_t_g(3) = 0.0                <
bc_c_t_s(3,1) = 0.0              <
bc_hw_t_g(3) = 0.0               <
bc_hw_t_s(3,1) = 0.0             <
bc_thetaw_m(3,1) = 0.0           <

                                 >  bc_c_theta_m(3,1) = 0.0
                                 >  bc_hw_s(3,1) = 0.0

bc_jj_ps(3) = 1                     bc_jj_ps(3) = 1
bc_type(3) = ‘NSW’               |  bc_type(3) = ‘PSW’
bc_uw_g(3) = 0.0                    bc_uw_g(3) = 0.0
bc_uw_s(3,1) = 0.0                  bc_uw_s(3,1) = 0.0
bc_vw_g(3) = 0.0                    bc_vw_g(3) = 0.0
bc_vw_s(3,1) = 0.0                  bc_vw_s(3,1) = 0.0
bc_ww_g(3) = 0.0                    bc_ww_g(3) = 0.0
bc_ww_s(3,1) = 0.0                  bc_ww_s(3,1) = 0.0
bc_x_e(3) = 0.3                     bc_x_e(3) = 0.3
bc_x_w(3) = 0.3                     bc_x_w(3) = 0.3
bc_y_n(3) = 0.7                     bc_y_n(3) = 0.7
bc_y_s(3) = 0.0                     bc_y_s(3) = 0.0
bc_z_b(3) = 0.0                     bc_z_b(3) = 0.0
bc_z_t(3) = 0.015                   bc_z_t(3) = 0.015
# Solid 1                           # Solid 1

I think that what you need to do is create a “Mixed wall” and set “Wall type” to NSW for the fluid phase and “PSW” for the solids. Is that what you tried to do?

Looking more closely at the docs, I think it’s possible that the current handling of JJ boundary conditions in the GUI is not quite right. Can you share your project file?

– Charles

[mohsenclick]

@cgw HI Charles,
YOU are right. I am trying to implement the NSW for the fluid phase and JJ BC for the solid phase. I used the Mixed wall type in GUI as you mentioned. However, the *.mfix file is like the 2nd format. According to the:

Wall BC for TFM simulation

I did not find any issue. Please see the definition of wall boundary condition (bold sentence), that is why you see all wall boundary conditions are set to PSW no matter what boundary condition you chose. |FREE_SLIP_WALL| |Velocity gradients at the wall vanish. If BC_JJ_PS is equal to 1, the Johnson-Jackson boundary condition is used for solids. A FSW is equivalent to using a PSW with hw=0.| |NO_SLIP_WALL| |All components of the velocity vanish at the wall. If BC_JJ_PS is equal to 1, …

, it should be correct, but I think the BC does not implement correctly. What do you think?
Also, when we use the JJ BC for the solid phase, and the energy equation is not solved, the following terms should not be used in the solution:
bc_c_t_g(3) = 0.0 <
bc_c_t_s(3,1) = 0.0 <
bc_hw_t_g(3) = 0.0 <
bc_hw_t_s(3,1) = 0.0 <
bc_thetaw_m(3,1) = 0.0 <

                             >  bc_c_theta_m(3,1) = 0.0
                             >  bc_hw_s(3,1) = 0.0

6.mfx (18.9 KB)

Thanks,
Mohsen