[Bisht]

Hi,

I am using a linearly increasing velocity profile with time at the inlet. Here is UDf.

#include "udf.h"
DEFINE_PROFILE(velocity_magnitude, t, i)
{
real default_early_velocity_mag,default_late_velocity_m ag,velocity_mag;

real start_velocity,end_velocity;
real ramp_start,ramp_end;
real line_slope,the_current_time;

face_t f;

default_early_velocity_mag=0.01;
default_late_velocity_mag=0.1;
start_velocity=0.01;
end_velocity=0.1;
ramp_start=0.0;
ramp_end=45.0;


velocity_mag=default_early_velocity_mag;
the_current_time = CURRENT_TIME;

line_slope=(end_velocity-start_velocity)/(ramp_end-ramp_start);

if ((the_current_time>=ramp_start) && (the_current_time<=ramp_end))
{
velocity_mag=line_slope*the_current_time;
}

if ((the_current_time>ramp_end))
{
velocity_mag=default_late_velocity_mag;
}
begin_f_loop(f,t)
{
F_PROFILE(f,t,i) = velocity_mag;
}
end_f_loop(f,t)
}

The simulation is running okay for the first 35 seconds after that I got the error message.

Unable to parse: [free
Error: undefined read macro
Error Object: ()
Error: > (greater-than): invalid argument [2]: wrong type [not a number]
Error Object: #f
Halting due to end of file on input.
Ending computation


Any idea why is it happening?

[pakk]

No.

However, I suspect that you forgot one term:

velocity_mag = start_velocity + line_slope*the_current_time;

This causes a discontinuity in your boundary condition (it jumps at 45 s from 0.09 m/s to 0.10 m/s), but I don't expect that this causes the error that you see. Still good to fix it, though...

[Bisht]

Thanks for the highlighting the improvement.
Yeah it shouldn't create a problem at 35 seconds of physical time. I am using a journal file to run my simulation on cluster.
Could it be due to it?

[obscureed]

As pakk says, there isn't much wrong with the UDF (apart from the discontinuity -- well spotted!), so I would go looking at the journal. (Does the journal contain the word "free", as mentioned in the error message?)

There is one error in the UDF as listed: a space in the fourth line "default_late_velocity_m ag". But I would be surprised if the libudf compiled with that space there, so perhaps it just crept in during transfer to this forum.

[Bisht]

I included the start velocity in my UDF as suggested by pakk and it worked without interuppting the simulation.