Time Step and CEL expressions for Mesh Deformation

rbarrett · July 1, 2011, 05:10

Hi all,

I am relatively new to CFX and I was just wondering if anyone could help me with a few problems that I have been having with a mesh deformation problem. I am modelling the compression of a piston within a cylinder. The actual geometry consists of two pistons within one cylinder, so there is a line of symmetry at mid chamber location. My model is also axisymmetric.

The problem that I am having is that the simulation deformation is over predicting the applied displacement and I was wondering if there is an issue with (i) my CEL expressions or (ii) my time step.

(i) The CEL expressions that I am using are ('WeightFunction' is a linear weighting function and 'Li' is the total length of the compression chamber):

Disp = sDisp*WeightFunction
Li = 0.178 [m]
WeightFunction = 1-((z-sDisp)/(Li-sDisp))
sDisp = velCV*t
velCV = 8 [m/s]

(ii) The time step that I am using is 5.2e-4 s for a total compression time of 16.6 ms.

Any help with this issue is greatly appreciated.

ghorrocks · July 2, 2011, 07:50

You seem to be describing a strange motion. Doesn't your function describe a displacement proportional to z height? That means the piston will pivot, not slide.

The best way to debug mesh motion stuff is to do a test run with all the solvers turned off using the expert parameters, and saving a result file (including the mesh!) every time step. Then you can quickly run the mesh motion and check you are getting the motion you intend.

stumpy · July 4, 2011, 10:16

There might be a problem with your weighting function here. Instead of using sDisp in WeightFunction use a callback to get the piston position, e.g areaAve(zGlobal)@movingend.
The callback is lagged to the previous timestep, but sDisp is not, which may be causing the unexpected behaviour.

rbarrett · July 6, 2011, 11:13

The issue was related to sDisp in the weighting function. I have now replaced this with a function based on t-tStep and the model is deforming as desired. Thanks a million for your help.

July 1, 2011, 05:10	Time Step and CEL expressions for Mesh Deformation	#1
rbarrett New Member Richard Barrett Join Date: Jun 2011 Posts: 14 Rep Power: 15	Hi all, I am relatively new to CFX and I was just wondering if anyone could help me with a few problems that I have been having with a mesh deformation problem. I am modelling the compression of a piston within a cylinder. The actual geometry consists of two pistons within one cylinder, so there is a line of symmetry at mid chamber location. My model is also axisymmetric. The problem that I am having is that the simulation deformation is over predicting the applied displacement and I was wondering if there is an issue with (i) my CEL expressions or (ii) my time step. (i) The CEL expressions that I am using are ('WeightFunction' is a linear weighting function and 'Li' is the total length of the compression chamber): Disp = sDispWeightFunction Li = 0.178 [m] WeightFunction = 1-((z-sDisp)/(Li-sDisp)) sDisp = velCVt velCV = 8 [m/s] (ii) The time step that I am using is 5.2e-4 s for a total compression time of 16.6 ms. Any help with this issue is greatly appreciated.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
air bubble is disappear increasing time using vof	xujjun	CFX	9	June 9, 2009 08:59
Cel expressions for trasient problem	Jervds	CFX	0	March 4, 2008 11:03
error using combination of step function	xujjun	CFX	1	January 15, 2008 17:46
CEL using different time step for different domain	new@cfx	CFX	3	September 30, 2005 09:03
How does STEP function work in CEL setup in CFX5.5	joseph	CFX	1	April 6, 2002 00:06

July 2, 2011, 07:50		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	You seem to be describing a strange motion. Doesn't your function describe a displacement proportional to z height? That means the piston will pivot, not slide. The best way to debug mesh motion stuff is to do a test run with all the solvers turned off using the expert parameters, and saving a result file (including the mesh!) every time step. Then you can quickly run the mesh motion and check you are getting the motion you intend.

July 4, 2011, 10:16		#3
stumpy Senior Member Join Date: Apr 2009 Posts: 531 Rep Power: 21	There might be a problem with your weighting function here. Instead of using sDisp in WeightFunction use a callback to get the piston position, e.g areaAve(zGlobal)@movingend. The callback is lagged to the previous timestep, but sDisp is not, which may be causing the unexpected behaviour.

July 6, 2011, 11:13		#4
rbarrett New Member Richard Barrett Join Date: Jun 2011 Posts: 14 Rep Power: 15	The issue was related to sDisp in the weighting function. I have now replaced this with a function based on t-tStep and the model is deforming as desired. Thanks a million for your help.