CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > FLUENT

C_UDMI Question in DPM

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Display Modes
Old   February 27, 2009, 18:25
Default C_UDMI Question in DPM
  #1
Prashanth
Guest
 
Posts: n/a
Hi all

I am using the DPM model to do particle tracking and I'm using a UDF to determine the mass and no. of particles that stick to the wall based on the particle normal and critical velocities. I am using a C_UDMI command to store the mass & no. of particles. My question is this: Can I just use the Define->User Defined->Memory option in the GUI to define 2 UDM locations, then use the Display->Contours option to display the UDM's so they get initialized, then compile my UDF and run my case? Is the above method right? Or do I have to use DEFINE_ON_DEMAND or some other macro to define the UDM's and set them to zero inside the UDF itself?

I have given my UDF below:

#include "udf.h" #include "dpm.h"

DEFINE_DPM_BC(deposition_bc,p,t,f,f_normal,dim) { /*#if !RP_HOST*/ Domain *d; cell_t c; d=Get_Domain(4); real Tavg,Ep,vcr; /*real norm_coeff=1.; real tang_coeff=1.;*/ real vn=0.; real R=287.; real normal[3]; int i,idim=dim; real x[ND_ND];

for (i=0.; i<idim; i++)

normal[i] = f_normal[i];

if(p->type==DPM_TYPE_INERT)

{

/*computing normal velocity*/

for(i=0.;i<idim;i++)

vn += p->state.V[i]*normal[i];

/*computing critical velocity*/

Tavg = (F_T(f,t)+P_T(p))/2.;

Ep = (3.*(pow(10.,20.)))*exp(-0.02365*Tavg);

vcr = pow(((2.*Ep)/P_DIAM(p)),(10./7.));

/*particle continues its path - no sticking*/

if (vn > vcr)

{

return PATH_ACTIVE;

}

/*particle sticks*/

else

{

thread_loop_c(t,d)

{

begin_c_loop(c,t)

{

C_UDMI(c,t,0) += P_MASS(p); /*mass of particles deposited*/

}

end_c_loop(c,t)

}

return PATH_ABORT;

}

}

/*#endif*/ }

  Reply With Quote

Old   February 28, 2009, 19:55
Default Re: C_UDMI Question in DPM
  #2
giovanni franchello
Guest
 
Posts: n/a
Hi Prashanth

You have to 1) compile your macro and load 2) run the case. At the beginning the UDMI values should be zero. However, to be sure, you can include a macro DEFINE_INIT (or DEFINE_ON_DEMAND) where you initialize 3) In your macro is not necessary to do the loop over all the cells: in this way the value of one particles is used to update all the domain, obtaining the same value in all the cells ! If you include the following: cell_t c = RP_CELL(&p->cCell); Thread *tcell = RP_THREAD(&p->cCell); you obtain the pointers to the cell where the particle is. Note that t in your DEFINE_DPM_BC is a face thread, not a cell Thread. Your UDMI should be updated as follows C_UDMI(c,tcell,0) += P_MASS(p); /*mass of particles */

If you whant the mass of the entire parcel, include C_UDMI(c,tcell,0) += P_MASS(p)*P_N(p); /*mass of particles */ Ciao Gian

  Reply With Quote

Old   March 1, 2009, 04:32
Default Re: C_UDMI Question in DPM
  #3
Prashanth
Guest
 
Posts: n/a
Hi Gian

Thanks a bunch for your reply. I have to get this UDF working within the next 4 days for a Thesis meeting with sponsors and your reply has really raised my hopes of not turning up at the meeting empty-handed. I am now able to run my UDF with a group injection but the UDM's still show zero after the particle tracking is done. I do think I'm closer to getting this working. It would be very helpful if you can take a look at the changes I have made and see if I'm missing something.

I'm particularly concerned about the domain ID in my DPM_BC and DEFINE_ON_DEMAND macros. My fluid zone ID is 1 and the ID for the wall on which the DPM_BC is applied is 4. When I initialize my UDM's using the DEFINE_ON_DEMAND, am I doing it right?

I have given my code below.

#include "udf.h"

#include "dpm.h"

#define NUM_UDM 2

static int udm_offset = UDM_UNRESERVED;

DEFINE_EXECUTE_ON_LOADING(on_loading,libname) { if (udm_offset == UDM_UNRESERVED)

udm_offset = Reserve_User_Memory_Vars(NUM_UDM);

if (udm_offset == UDM_UNRESERVED)

Message("\nNeed to define upto %d extra UDM's in GUI and reload %s\n",NUM_UDM,libname);

else

{

Message("%d UDM's have been reserved %s\n",NUM_UDM,libname);

Set_User_Memory_Name(udm_offset,"UDMmass");

Set_User_Memory_Name(udm_offset+1,"UDMnum");

}

Message("\nUDMoffset for current lib = %d",udm_offset); }

DEFINE_ON_DEMAND(set_udms) { Domain *d; Thread *ct; cell_t c; int i; d=Get_Domain(1); /*ID for fluid*/

if (udm_offset != UDM_UNRESERVED)

{

Message("Setting UDMs\n");

for (i=0;i<NUM_UDM;i++)

{

thread_loop_c(ct,d)

{

begin_c_loop(c,ct)

{

C_UDMI(c,ct,udm_offset+i)=0.0;

}

end_c_loop(c,ct)

}

}

}

else

Message("UDM's not reserved\n"); }

DEFINE_DPM_BC(deposition_bc,p,t,f,f_normal,dim) {

Domain *d; cell_t c=RP_CELL(&p->cCell); Thread *tcell=RP_THREAD(&p->cCell); real Tavg,Ep,vcr; d=Get_Domain(4); real vn=0.; real R=287.; real normal[3]; int i,idim=dim; real x[ND_ND];

for (i=0.; i<idim; i++)

normal[i] = f_normal[i];

if(p->type==DPM_TYPE_INERT)

{

/*computing normal velocity*/

for(i=0.;i<idim;i++)

vn += p->state.V[i]*normal[i];

/*computing critical velocity*/

Tavg = (F_T(f,t)+P_T(p))/2.;

Ep = (3.*(pow(10.,20.)))*exp(-0.02365*Tavg);

vcr = pow(((2.*Ep)/P_DIAM(p)),(10./7.));

/*particle continues its path - no sticking*/

if (vn > vcr)

{

return PATH_ACTIVE;

}

/*particle sticks*/

else

{

C_UDMI(c,tcell,0) += P_MASS(p); /* mass of particles deposited*/

C_UDMI(c,tcell,1) += 1.0;

Message("No. of particles deposited is %d",C_UDMI(c,tcell,1));

return PATH_ABORT;

}

}

}
  Reply With Quote

Old   March 1, 2009, 08:02
Default Re: C_UDMI Question in DPM
  #4
giovanni franchello
Guest
 
Posts: n/a
Now I have not time to debug your UDF. You should control the DEFINE_DPM_BC, loops like for(i=0.;ivn += p->state.V[i]*normal[i]; that are not clear. As I see in your macro the pointer domain is not used. Include messages to control if what is calculated correspond to what you expect. Include additional udmi to count no. of particles not sticking. Note that you don't count particles, but parcels. I suppose 4 day are not enougth. Sei ancora in alto mare Ciao Gian
  Reply With Quote

Old   March 1, 2009, 08:22
Default Re: C_UDMI Question in DPM
  #5
Prashanth
Guest
 
Posts: n/a
Hi Gian

Thanks a lot for your help. You have pointed out the direction to look in and thank you for that. I'll go back to working on it.

Thanks

Prashanth
  Reply With Quote

Old   August 20, 2009, 04:35
Default define reflect boundary conditions
  #6
New Member
 
panyadi
Join Date: Aug 2009
Location: nanjing
Posts: 2
Rep Power: 0
panyd is on a distinguished road
hi Prashanth
resently, i study the particle deposition on tube surface. but during the interation with my defined reflect boundary,fluent display "Error:divergence detected in AMG solver: x-momentum" .my Define_dpm_bc is as follows:
#include "udf.h"
#include "dpm.h"
#include "mem.h"
#include <stdlib.h>
#include <math.h>
#include <iostream.h>
#include <time.h>
#define p_Ela_mol 3.6e11
#define p_sur_vos 0.15
#define p_den 1550
#define angle_cri 1.27
#define p_po 0.23
#define vn_y 7.5
DEFINE_DPM_BC(bc_reflect,p,t,f,f_normal,dim)
{
real alpha,alpha_a; /* angle of particle path with face normal */
real vn=0.;
real vt=0.;
real nor_coeff,tan_coeff;
real vn_cri,vnm_cri;
real fs_coeff=0.4;
real normal[3];
real c_num,c_xsh,poro;
int i, idim = dim;
int num_acc;
real NV_VEC(x);
real Eeff;
Thread *t0;
cell_t c0;

Eeff=p_Ela_mol/(1-p_po*p_po);
num_acc = F_UDMI(f,t,1);
c_num=num_acc*F_UDMI(f,t,2)/(0.0508*angle_cri/2);
c_xsh=pow((c_num/20),-0.282);
poro=pow(c_xsh,3)/(pow(c_xsh,3)+0.852);
if (poro<0.55)
vn_cri=0.0445+0.0074*exp(5.8828*poro);
else
vn_cri=0.2692+2.0303*pow(10,-7)*exp(20.8835*poro);
srand( (unsigned)time( NULL ) );
for (i=0; i<idim; i++)
normal[i] = f_normal[i];
if(p->type==DPM_TYPE_INERT)
{
alpha = M_PI/2. - acos(MAX(-1.,MIN(1.,NV_DOT(normal,p->state.V)/
MAX(NV_MAG(p->state.V),DPM_SMALL)))); alpha_a=alpha+5*M_PI*gaussrand()/180;
for(i=0; i<idim; i++)
vn +=p->state.V[i]*normal[i];
if ((NNULLP(t)) && (THREAD_TYPE(t) == THREAD_F_WALL))
F_CENTROID(x,f,t);

if (alpha>angle_cri )
{
nor_coeff=1;
tan_coeff=1-fs_coeff*(1+nor_coeff)/tan(alpha_a);
}
else if (vn<=vn_cri)
{
nor_coeff=0;
tan_coeff=0;
}
else if (vn>vn_cri && vn<=vn_y)
{
nor_coeff=sqrt(1-pow(vn_cri/vn,2.0));
tan_coeff=1-fs_coeff*(1+nor_coeff)/tan(alpha_a);
}
else
{
nor_coeff=sqrt(1.2*sqrt(3)*(1-pow(vn_y/vn,2.0)/6)*sqrt((vn_y/vn)/((vn_y/vn)+2*sqrt(1.2-0.2*pow(vn_y/vn,2.0))))-pow(vn_cri/vn,2.0));
tan_coeff=1-fs_coeff*(1+nor_coeff)/tan(alpha_a);
}

for(i=0; i<idim; i++)
p->state.V[i] -=vn*normal[i];
for(i=0; i<idim; i++)
p->state.V[i] *=tan_coeff;
for(i=0; i<idim; i++)
p->state.V[i] -=nor_coeff*vn*normal[i];
for(i=0; i<idim; i++)
p->state0.V[i]=p->state.V[i];

if (nor_coeff==0)
{
return PATH_ABORT;
}
else
{
return PATH_ACTIVE;
}
}
return PATH_ABORT;
}

in the programme of DEFINE_DPM_EROSION(), i set F_UDMI(f, t, 1) to store the number of deposited particle and F_UDMI(f, t, 2) to store the mean diameter of deposited particle.

would you please help me to check my programme and point out the mistakes
panyd is offline   Reply With Quote

Reply

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
C_UDMI(c,t) bohis FLUENT 0 April 1, 2008 10:44
C_UDMI and F_UDMI Lucy FLUENT 0 April 28, 2006 04:16
gradient of C_UDMI(c,t,i) joyce FLUENT 2 April 13, 2005 09:51
Problems with C_UDMI nicolas FLUENT 0 February 16, 2005 09:41
Storing vectors with C_UDMI nicolas FLUENT 4 February 16, 2005 06:31


All times are GMT -4. The time now is 02:16.