CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Need to measure transient pressure at one moving point (

fated-oracle October 17, 2011 11:43

Need to measure transient pressure at one moving point

I'm currently stuck with my simulation so I would be really grateful if anyone could help me. I'm simulating the dropping of a wedge onto the free water surface. I've been using the rigid body solution for the wedge (it is in fact only a beta feature in Ansys v12.1, which is the only available at uni) and the interphase between the two fluids is modelled as a free surface. The simulation is obviously transient. In fact, the tools I used and my set up seem to work all right, as I can see from the keyframes of the pressure and water volume fraction distributions. However, what I need in fact in order to validate my results is the pressure for all time steps at a specific point that lies on the wedge surface. However, the problem is that this surface moves along (in the negative y direction) with the rigid body (wedge). Hence, I cannot specify a signle point by inputting the coordinate files. However, what I could do is to model the mesh so that I have a node lying on the surface exactly in the location I need the pressure measurements to be taken at. Nevertheless, I don't know what to call this point in Ansys CFX-Pre.
If anyone knows how to deal with this or has a different idea I'd be really grateful. thank you in advance!

gpwakeham November 15, 2011 05:01

does it have to be at one specific point? Because instead you could monitor an area average of the pressure at the bottom surface, for example


where bottom is the bottom face of the wedge

ghorrocks November 15, 2011 05:22

This is probably easiest done in post processing where you can extract the pressure at a series of points which represents the motion of your sensing point as the body moves. It means you will have to produce lots of results files but if you only include the variables you are interested in (and the mesh as it is moving) they should not be too big.

D.B November 15, 2011 06:31

To add to ghorrocks' point, what you could do is, simply define a point on the body, by either node number or variable max/min ( use X,Y,or Z as the variable) and then just export the data for each time step.


stumpy November 15, 2011 10:51

I think you can just use a standard monitor point. Even though you enter x,y,z coordinates for monitor points, they get associated to the closest node (see the .out file). The values reported are for that node, so it moves with the mesh. If you perform re-meshing then you'd need a new monitor point.

ghorrocks November 15, 2011 20:26

Run-time monitor points snap to the nearest node, but post-processing points interpolate to the point location.

D.B November 16, 2011 00:00

The best way to do is to figure out the co-ordinates of the point at the initial point of time (initial location), then create a point there. When you do that, you can see the nearest node number below that. Now Just select the Node number option in point deifinition and mention that specific node number you noted earlier, and then start lodaing the timesteps and exporting the data. I don't think you can use this method in pre, otherwise you could have created a monitor point at that node location. The post method is tedious but doable.

All times are GMT -4. The time now is 03:39.