[Geehrte]

The post starter wanted to use TUI command to switch boundary condition type when the outlet pressure reach the critical point(10000Pa). The UDF and TUI commands are being used in this case.
Here is my idea in handling this. I defined a variable "outlet_pressure" in scheme file and then it's called in UDF by "RP_GET_Real" command. In UDF I computed the average pressure on outlet surface and then update this variable. Finally, by using "rpgetvar", I can access the value of outlet pressure and decide whether to change the outlet boundary condition or not.
However, the post starter found that after a few calculation steps, the TUI can't access the updated UDF "outlet_pressure" value. In Fluent console panel I tapped (rpgetvar 'outlet_pressure) command and it returned the initially set value 1.
I used EXECUTE_AT_END macro in UDF, here is UDF code:

#include "udf.h"

DEFINE_EXECUTE_AT_END(pres_check)
	/* Average pressure on outlet surface */
{
	int zone_ID_outlet = 10;
	real A[ND_ND];
        Thread *thread;
	face_t f;
	real o_pressure, outlet_pressure;
	real P_ave = 1.0, P_sum = 0.0, A_sum = 0.0;
	Domain *d;
	d = Get_Domain(1);
	thread = Lookup_Thread(d, zone_ID_outlet);
	o_pressure = RP_Get_Real("outlet_pressure");
	begin_f_loop(f, thread)
	{
		F_AREA(A,f,thread);
		A_sum += NV_MAG(A);
		P_sum += NV_MAG(A)*F_P(f,thread);
	}
	end_f_loop(f, thread)
	P_ave = P_sum/A_sum;
	if (P_ave > o_pressure)
		o_pressure = P_ave;
	RP_Set_Real("outlet_pressure",o_pressure);
	printf("Outlet Pressure = %f \n",o_pressure);
}

The scheme file is executed every time step. Here is scheme code:

(rp-var-define 'outlet_pressure 1.0 'real #f)
(cond
((<= (%rpgetvar 'outlet_pressure) 10000) 
(ti-menu-load-string "define/boundary-conditions/zone-type 10 wall"))
(else (ti-menu-load-string "define/boundary-conditions/zone-type 10 pressure-outlet"))
)

So why I can't get a calculated value? The "o_pressure" value obviously updated every time step! Can you help me figure it out? Thanks for your kindness help!

[obscureed]

One possibility is that the rp-var-define command is constantly overwriting the value. It is fairly common to make sure that rp-var-define is called only once with something like

(if (not (rp-var-object 'rpvarname))(rp-var-define 'rpvarname 0.0 'real #f))

or you could just do it once per case and take it out of the repeated Scheme.

Really I think that you need to debug the sequence yourself:
-- Set outlet_pressure to a large negative value; run pres_check; confirm that outlet_pressure has been updated to a sensible value.
-- Check that outlet_pressure still has a sensible value when the Scheme starts and ends.
-- Check that the ti-menu-load-string commands do what you intend.
-- Check that the Scheme cond command does what you intend.
-- etc
If every step works, the whole thing will work. If not, and if you can't fix the broken step, you'll have a more specific question.

[Geehrte]

Hi, Obscureed! I noticed that you are a veteran in this forum and I have learnt a lot from your former posts. Thanks for your advice.
Now I have replaced my rp-var-define command into this to make sure that the rp-var-define macro execute only once.

(define (make-new-rpvar 'outlet_pressure -999.0 'real)
(if (not (rp-var-object 'outlet_pressure))
(rp-var-define 'outlet_pressure -999.0 'real #f)))
(make-new-rpvar 'outlet_pressure -999.0 'real)

After I hooked the EXECUTE_AT_END macro and made sure my scheme file works every time step, you can see the running results as following(from 65-75 iterations)

Updating solution at time level N...
 done.
  iter  continuity  x-velocity  y-velocity      energy           k     epsilon     time/iter
    65  9.1066e-03  8.5559e-04  2.0053e-03  9.3002e-08  2.6643e-04  2.2503e-03  0:00:00    5
    66  9.5180e-03  1.8930e-03  1.4070e-03  1.3492e-07  2.7044e-04  2.2077e-03  0:00:00    4
    67  9.3741e-03  1.3079e-03  3.3517e-03  7.6416e-08  2.7824e-04  2.2561e-03  0:00:00    3
    68  9.3179e-03  1.2106e-03  1.6374e-03  7.9531e-08  2.8102e-04  2.1450e-03  0:00:00    2
    69  9.2098e-03  1.5034e-03  4.6282e-03  8.4639e-08  2.8510e-04  2.0802e-03  0:00:00    1
    70  9.1362e-03  1.0286e-03  1.6592e-03  7.8024e-08  2.9011e-04  2.0498e-03  0:00:00    0
Outlet Pressure = 10261.125977 
/file/read-macro "scheme3.scm"Loading ".\.\scheme3.scm"
define/boundary-conditions/zone-type 10 wall
 Zone not slit.Done.

Updating solution at time level N...
 done.
  iter  continuity  x-velocity  y-velocity      energy           k     epsilon     time/iter
    70  9.1362e-03  1.0286e-03  1.6592e-03  7.8024e-08  2.9011e-04  2.0498e-03  0:00:00    5
    71  9.5955e-03  1.0906e-03  3.6865e-03  1.6078e-07  2.9754e-04  2.1202e-03  0:00:00    4
    72  9.5147e-03  3.4935e-03  6.3283e-03  7.0936e-08  3.0514e-04  2.1359e-03  0:00:00    3
    73  9.4124e-03  1.8094e-03  5.3499e-03  8.1904e-08  3.0944e-04  2.0748e-03  0:00:00    2
    74  9.3880e-03  1.5320e-03  4.6221e-03  7.7788e-08  3.1366e-04  1.9670e-03  0:00:00    1
    75  9.2679e-03  2.0402e-03  1.1904e-03  7.5247e-08  3.1794e-04  1.9048e-03  0:00:00    0
Outlet Pressure = 10719.134766 
/file/read-macro "scheme3.scm"Loading ".\.\scheme3.scm"
define/boundary-conditions/zone-type 10 wall
 Zone not slit.Done.

So I made sure that the pressure calculation works well in UDF, the value of "o_pressure" updates every time step. But, when I stopped running calculation, and check the value of "outlet_pressure". It was still the original defined value.

(rpgetvar 'outlet_pressure)
-999

So I'm confused that whether it's RP_Set_Real command in UDF don't work or simply the %rpgetvar command in scheme don't work.

[Geehrte]

I have tried a lot of times these days, but I can't get it through. Is there anyone could help me? Thanks a lot!

[AlexanderZ]

if you are using fluent version v19 and earlier, code should be parallelized

so for your case command RP_Set_Real("outlet_pressure",o_pressure);
should be executed on host
compile following code

#include "udf.h"

DEFINE_EXECUTE_AT_END(pres_check)
	/* Average pressure on outlet surface */
{
	int zone_ID_outlet = 10;
	real A[ND_ND];
        Thread *thread;
	face_t f;
	real o_pressure, outlet_pressure;
	real P_ave = 1.0, P_sum = 0.0, A_sum = 0.0;
	Domain *d;
	d = Get_Domain(1);
	thread = Lookup_Thread(d, zone_ID_outlet);
	#if !RP_NODE
		o_pressure = RP_Get_Real("outlet_pressure");
	#endif
	host_to_node_real_1(o_pressure);
	Message0("o_pressure on node from scheme = %f \n",o_pressure);	
	begin_f_loop(f, thread)
	{
		F_AREA(A,f,thread);
		A_sum += NV_MAG(A);
		P_sum += NV_MAG(A)*F_P(f,thread);
	}
	end_f_loop(f, thread)
	P_ave = P_sum/A_sum;
	if (P_ave > o_pressure)
		o_pressure = P_ave;
	Message0("Outlet Pressure on node = %f \n",o_pressure);	
	node_to_host_real_1(o_pressure);
	#if !RP_NODE
		RP_Set_Real("outlet_pressure",o_pressure);
		Message("Outlet Pressure on host = %f \n",o_pressure);
	#endif	
}

scheme

(define (make-new-rpvar name default type)(if (not (rp-var-object name))(rp-var-define name default type #f)))
(make-new-rpvar 'outlet_pressure 1.0  'real)
(cond
((<= (%rpgetvar 'outlet_pressure) 10000) 
(ti-menu-load-string "define/boundary-conditions/zone-type 10 wall"))
(else (ti-menu-load-string "define/boundary-conditions/zone-type 10 pressure-outlet"))
)

scheme should be executed on each timestep

[Geehrte]

Thank you so much, Master AlexanderZ. Since I tried your modified codes in parallel, an connection between UDF and TUI was established!