[hmasenger]

Hello my friends
I need to rotate a butterfly valve's disk from 45 degree to 90 degree installed in a pipe.
first of all I activated the mesh deformation and set its parameters as can be seen in pic no1.next , I defined a coordinate system in the middle of the disk an then set an expression
“rotation=0.08[m/s]*t”
to rotate the disk clockwise by 5 degree(0.08radian) per every time step Although i am using steady state results for initial condition .I think something is wrong with this expression because solver doesn’t run and crashes out.
.
Please help me with the expression.
5 degree is equal to 0.08 radians in the expression so I multiplied it to [m/s] to have a [m] dimension after multiplying by time .i don’t know why cfx doesn’t have degree unit for theta parameter in cylindrical displacement (pic no 3)and the expression’s dimension must be in meter or millimeter !
Is it right to use “t” parameter to define time step in expression?

[ghorrocks]

CFX will take lengths entered with any length dimension (so feet, yards, mm, kilometres are all fine) and they will eb internally converted to metres.... assuming you have not changed the solver units from SI units.

Likewise for angles. You can enter degrees or radians and it will be internally converted to radians. You just need to specify what the unit is and it will happen like magic.

Yes, you can use time (t) in expressions.

Also, have you considered using a rotating frame of reference to model this? If the motion can be reduced to a simple rotation then that will simplify things greatly and movign mesh is quite a bit more complex.

[hmasenger]

thanks for your answer ghorrocks .your answers always helps me through CFX
I think I am going to forget about rotating the disk in one transient run !maybe it is better to solve the problem in several different runs in various valve openings.i mean setting up the problem for all target openings and separately solve it.(lots of errors about initial condition happens while I am trying to solve a transient run.i am a little confused )

[ghorrocks]

Good to see you are thinking about what is actually important. A series of steady state runs is much easier than a transient run, if the steady state assumption is good.