[Juun]

Hi everyone, I am doing a simulation about absorption in a pipe.

some background
Mixture gas (A B C) flow in a pipe, gas A is absorbed on the inside wall.
The inlet conditions are constant (u, t, gas mixture fraction ect…) from 0s-500s, after that temperature and mixture will change. Since gas A is absorbed into the wall, the outlet gas A should remain at 0 for a period at the beginning, then raise to a level that same as inlet (as wall surface cannot take anymore).

Problem
I use transient to simulate this, the result is odd. The trend of gas A level at outlet is correct, but it doesn’t go according to time, but the number of iteration.

For example, gas A at outlet reach 300ppm when the solver run about 2000 iterations, no matter what time step size is used

That means

if time step size is 0.1s, 20 itr/time_step, outlet gas A get to 300ppm in 10s;
if it is 0.5s, 20 itr/time_step, it reaches 300ppm about 50s
if use 0.05s, 20 itr/time_step, it only takes 5s to reach 300ppm

And in fact, if I used 1s on time step size, the result is close to the experiment (yes, by luck), reach 300ppm in 100 second.

It seems like it has nothing to do with time here, but only the total number of the iterations.

Anyone has idea what goes wrong here?