The correct value of init noise and iin for turbulent pipe flow simulation

[Madhurjya814]

Hi, may I know one thing. The value of init noise for channel flow case is 0.125 and the same init noise for pipe flow simulation is 3.0. Why is this much of difference and which one is the correct value for pipe flow init noise. Because, a simulation performed with init noise of 4.0 and iin=1 and 33,000,000 grid resolution, the flow became laminar at Re=5300. Also, is iin=1correct or if not then what should be the iin (iin=1,2,...) value for turbulent pipe flow?
Thanks...

[mathrack]

Hello,

The CFD code Incompact3d is open-source : you can open the source code and try to understand.

What is the impact of iin ? Check for the variable iin in the code. The impact of this variable is mostly case-specific. The only exception is iin = 3 (some files must be provided).

For the Pipe flow, the impact of iin in the file Case-Pipe-flow.f90 is relatively clear. Please let us know what is the difficulty. Please notice the factor um=exp(-fifteen*r*r) applied to the noise to modulate the amplitude.

I am afraid you will have to understand the code and the case you simulate to understand what is correct.

[Madhurjya814]

The problem I am facing is that with iin=1 and init noise of 3.0, 4.0 or 5.0 the flow is becoming laminar. This should not happen because the Re is 5300. Is it because of iin=1. Should I try with iin=2?

[mathrack]

Random numbers are bad precursors for turbulent simulations because they are quickly damped by diffusion.

[Madhurjya814]

That means I have to try SEM?