CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)

 nhlstd June 18, 2017 06:42

defining a blade with mesh motion in cfx

Hi all,
I'm working on a blade of a turbine and I want to change the angle of the blade during a transient simulation. it's mesh should move and the total rotation of the blade is about 3 degrees in about 2.5 seconds.
I chose dynamic mesh for my domain and I chose previous mesh as displacement relativity. As my blade is rotating relative to a point different from (0,0,0), I defiened a new coordination system at the rotation point (0.289073,0.184303,0).
Then I set my blade as a boundary and it's coordination frame as coord 1 (the new coordination system that I mentioned above). in boundary details tab, for mesh motion I chose "specified displacement" and for displacement object I chose cylindrical component. Since my rotation axis is in z direction, I set axial and radial components to 0.
For theta component I used some expressions as below:
expression>insert>r1> ((x-0.289073[m])^2+(y-0.184303[m])^2)^0.5 (radius relative to my new coordination system)
expression>insert>omega>0.008666666666666666666666 66666667[rev s^-1]
expression>insert>rw>r1*omega
expression>insert>dx>rw*dtstep
So I wrote dx as an expression for displacement theta component. No slip wall and mesh motion were chosen.

When it solves mesh, I recognized that it just moves in the first time step and it doesn't move even a bit in the next time steps.
I tried reducing my time steps but it didn't change.
I suppose that it's because of r1's definition. I think It just chooses the initial x and y locations of the blade and it's not updating after the first time step so it's just repeating the first motion!!!
Does anyone has any idea about my case? Is my hypothesis true? What is your recommendation?

 nhlstd August 6, 2017 06:48

hi again!
I solved the problem myself! you should use time in your simulation parameters in order to change the blade angle.
So for solving the problem above, you should define dx as rw*t. It seems that CFX understands the position of x and y only in the first time step. so for applying their new position you should change something else, such as time that is a variable and I used here!

 gratify3 December 4, 2019 01:17

Quote:
 Originally Posted by nhlstd (Post 659651) hi again! I solved the problem myself! you should use time in your simulation parameters in order to change the blade angle. So for solving the problem above, you should define dx as rw*t. It seems that CFX understands the position of x and y only in the first time step. so for applying their new position you should change something else, such as time that is a variable and I used here!
Hi,
I have tried with the same option. But for my case it is not working with specified displacement. Can you help me how to define the rotating mesh motion about 60 deg with time.

 All times are GMT -4. The time now is 16:50.