[Other] How to change zone type without floating point exception

ringel_f1 · February 1, 2021, 04:28

Hello,
I am trying to model a very simple piston pump. So we have one intake channel and we have one outlet channel. There is the compression chamber where we have a moveable piston (rigid body). The intake and outlet channels are connected to the compression chamber and an the walls between these parts I have defined as components. My idea is, by making it easier to modell the intake and outtake valve, I just have to change the zone type of these connection walls. When there is the intake stroke I change the wall at the outtake channel from interiour to wall. During the compression stroke both walls form the intake and outtake channel are changed from interiour to wall. And for the exhaust stroke the wall at the outtake channel is changed from wall to interiour again. And this is where the problem occurs. So my boundary conditions are pressure inlet and outlet both the 1 bar absolute pressure. The whole simulation works until the exhaust stroke starts, where I get the message "floating point exception". This one directly comes up, when the change of the zone type was done from wall to interior... Does anybody know why this happens? Do you have an alternative way to modell this problem?

Thanks and looking forward to hearing from you!

aero_head · February 1, 2021, 11:56

Hello,

Since this is an unsteady simulation, you may need to check and modify your time-step size, what are you using currently and what numerical scheme are you using? As well, are you running your simulation in double precision?

Otherwise, there could be an issue with the mesh; what kind of mesh have you used, structured or unstructured? An example issue may be with cells that have very low aspect ratio.

Try using some of the mesh diagnostic tools, as well as looking at your volume mesh with some of the cell quality scalars.

February 1, 2021, 04:28	How to change zone type without floating point exception	#1
ringel_f1 New Member Maximilian Ringel Join Date: Jul 2020 Posts: 6 Rep Power: 6	Hello, I am trying to model a very simple piston pump. So we have one intake channel and we have one outlet channel. There is the compression chamber where we have a moveable piston (rigid body). The intake and outlet channels are connected to the compression chamber and an the walls between these parts I have defined as components. My idea is, by making it easier to modell the intake and outtake valve, I just have to change the zone type of these connection walls. When there is the intake stroke I change the wall at the outtake channel from interiour to wall. During the compression stroke both walls form the intake and outtake channel are changed from interiour to wall. And for the exhaust stroke the wall at the outtake channel is changed from wall to interiour again. And this is where the problem occurs. So my boundary conditions are pressure inlet and outlet both the 1 bar absolute pressure. The whole simulation works until the exhaust stroke starts, where I get the message "floating point exception". This one directly comes up, when the change of the zone type was done from wall to interior... Does anybody know why this happens? Do you have an alternative way to modell this problem? Thanks and looking forward to hearing from you!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
activeBaffleVelocity boundary condition ?	om3ro	OpenFOAM Programming & Development	10	November 17, 2020 00:26
Film cooling problem	arun1994	OpenFOAM	0	October 26, 2020 10:08
Foam::error::printStack(Foam::Ostream&) at ??:?	mdhfiz	OpenFOAM Running, Solving & CFD	1	September 19, 2019 10:46
inlet pressure is higher than outlet pressure for fan	sivakumar	OpenFOAM Pre-Processing	16	December 30, 2017 15:16
Error during initialization of "rhoSimpleFoam"	kornickel	OpenFOAM Running, Solving & CFD	8	September 17, 2013 06:37

February 1, 2021, 11:56		#2
aero_head Senior Member Kira Join Date: Nov 2020 Location: Canada Posts: 435 Rep Power: 9	Hello, Since this is an unsteady simulation, you may need to check and modify your time-step size, what are you using currently and what numerical scheme are you using? As well, are you running your simulation in double precision? Otherwise, there could be an issue with the mesh; what kind of mesh have you used, structured or unstructured? An example issue may be with cells that have very low aspect ratio. Try using some of the mesh diagnostic tools, as well as looking at your volume mesh with some of the cell quality scalars.