[foolfishzju]

Hi everyone,

I would like to get the gradient of species mass concentration to compute the mass transfer coefficient based on Fick's law. The code runs in parallel mode.

I am stuck with the issue of getting zero gradients of UDS. But when I check the value of UDS, there are non-zero values.

My UDF code is as below,

----------------------------------------------------------------------------------

#include "udf.h"

DEFINE_ADJUST(spec_interf_concen,mixture_domain)
{

#if !RP_HOST

int liq_phase_id = 0;
int gas_phase_id = 1;
int co2_species_id = 0;
int yi = 0;

......

Domain *pDomain = DOMAIN_SUB_DOMAIN(mixture_domain,gas_phase_id);
{

Alloc_Storage_Vars(pDomain,SV_VOF_RG,SV_VOF_G,SV_N ULL);
Scalar_Reconstruction(pDomain, SV_VOF,-1,SV_VOF_RG,NULL);
Scalar_Derivatives(pDomain,SV_VOF,-1,SV_VOF_G,SV_VOF_RG,Vof_Deriv_Accumulate);

}

mp_thread_loop_c(mix_thread, mixture_domain, pt)
{

if (FLUID_THREAD_P(mix_thread))
{

begin_c_loop(c,mix_thread)
{

liq_thread = THREAD_SUB_THREAD(mix_thread,liq_phase_id);
gas_thread = THREAD_SUB_THREAD(mix_thread,gas_phase_id);

......

C_UDSI(c,pt[liq_phase_id],yi) = C_YI(c,pt[liq_phase_id],co2_species_id);

if((vof_l_c > 1e-6)&&(vof_l_c < 0.999999))
{
ca = press/KH_px*rho_l/Ml*vof_l_c;
C_UDSI(c,pt[liq_phase_id],yi) = ca*Mg/(ca*Mg+rho_l*vof_l_c);
}

......

}
end_c_loop_int(c,mix_thread)

}

}

/* Free_Storage_Vars(pDomain,SV_VOF_RG,SV_VOF_G,SV_NU LL); */

#endif

}


DEFINE_LINEARIZED_MASS_TRANSFER(surface_renewal, c, mixture_thread, from_phase_index, from_species_index, to_phase_index, to_species_index, lin_from, lin_to)
{

Thread *liq;
Thread **pt;

int liquid_species_id = 0;
int gas_phase_ID = 1;
int co2_species_id = 0;
int yi = 0;

......

real rho_l, rho_g, r_rho_lg;
real cg_G[3];

liq = THREAD_SUB_THREAD(mixture_thread, to_phase_index);
pt = THREAD_SUB_THREADS(mixture_thread);

rho_l = C_R(c,liq);
rho_g = C_R(c,pt[gas_phase_ID]);

vof_l = C_VOF(c,liq);
vof_g = C_VOF(c,pt[gas_phase_ID]);

if (NULL != THREAD_STORAGE(liq,SV_UDSI_G(yi)))
{

/* gradient of species mass concentration */
NV_V(cg_G,=,C_UDSI_G(c,liq,yi));
Message("Magnitude of species mass concentration: %f \n",NV_MAG(cg_G));

}
else
{

NV_S(cg_G,=,0.0);

}

......

return m_source;

}

----------------------------------------------------------------------------------

Moreover, I did the following setup to keep the memory from freeing:
/solve/set/expert

but the TUI shows the option is incompatible with the adaption in parallel. I wonder whether this option will affect the value of UDS gradient.

[AlexanderZ]

manual

the solver continually removes data from memory that it does not need. In order to retain the
gradient data (when you want to set up user-defined scalar transport equations, for example), you can
prevent the solver from freeing up memory by issuing the text command solve/set/expert and
then answering yes to the question, “Keep temporary solver memory from being freed?”
Note that when you do this, all of the gradient data is retained, but the calculation requires more
memory to run.

so looks like you can't access gradient data without this setting

[foolfishzju]

Originally Posted by AlexanderZ

manual


so looks like you can't access gradient data without this setting

Hi Alexander,

Do you know how to fix this issue?

I found a similar topic in Ansys forum:
https://forum.ansys.com/discussion/1...on-not-only-in

A coordinator mentioned that there is an unofficial method retrieving the gradient values, but he did not give the solution.

[AlexanderZ]

unofficial way is to use reconstruction gradient macros, which you are using already

 Alloc_Storage_Vars(...)
 Scalar_Reconstruction(...)
 Scalar_Derivatives(...)
 Free_Storage_Vars(...)

you may check if memory for temperature gradient is allocated or not using following code

thread_loop_c(t, d)
 {
 if (!SV_ALLOCATED_P(t, SV_T)) {
 Message("Energy Equation not active in zone %d on compute
 node %d\n", THREAD_ID(t), myid);
 } else {
 Message("Energy Equation active in zone %d on compute node
 %d\n", THREAD_ID(t), myid);
 }