Create variables from instantaneous averages?

Moderators: davetaflin, salter, steve, jprenaud, straton, Chris Idso

akiss20
Experienced user
Posts: 10
Joined: Mon Nov 25, 2013 7:34 am
Location: Cambridge, MA

Create variables from instantaneous averages?

Postby akiss20 » Tue Jun 10, 2014 7:54 am

Hi All,

I am diving into the world of unsteady CFD and have some questions about unsteady post-processing in Tecplot. To preface, I am loading a transient 3D CGNS file into Tecplot 360 2011/2012 (depending on the computer) R1 for analysis. I am simulating the stall of a rotor-only calculation of a low speed compressor.

One thing I want to be able to do is to create a variable based on the instantaneous area/mass average at that given time. For example, I would like to define a total-to-static pressure rise coefficient which is defined as

(P_static - massAverage(P_total@inlet))/(0.5*rho*U_tip^2)

I want the mass average of total pressure at the inlet to be a function of time rather than a time average. Is this possible to do in Tecplot? I also want this to be a variable, not just a single value, so I can see how this coefficient is distributed spatially and temporally.

On a more nitty gritty note, I also am struggling with data organization in Tecplot. It seems when I load an unsteady CGNS file it creates a separate zone for each timestep and spatial zone. So say if I have 10 spatial zones (inlet,outlet,blade Omesh etc etc) and 20 time steps, I now have 200 zones. Tecplot will lump them together via time strands in the zone style dialogue box, but if I go to Dataset Info it shows all 200 zones. This would make re-naming the spatial zones, or taking sub-zones very difficult. Is there any way around this?

Thanks!

akiss20
Experienced user
Posts: 10
Joined: Mon Nov 25, 2013 7:34 am
Location: Cambridge, MA

Re: Create variables from instantaneous averages?

Postby akiss20 » Mon Jun 16, 2014 11:21 am

For anyone facing the same challenge I have figured out a way to do it, but it isn't very elegant.

The way my CGNS loads (and I assume all standard 3D unsteady CGNS files load) is that each spatial zone and each timestep is a different zone in tecplot. Thus for M spatial zones and N timesteps there are M*N zones in tecplot. I first create a variable in zones called, for example, "Psi" and set it to a dummy value. I then loop through each spatial zone and time-step and calculate my Psi value based on the inlet conditions at that time using the integration calls and storing the results as a dummy variable. This takes quite a while but it gets the job done. The same thing can be done to rename each individual zone (something like "Inlet Time X").

This requires that you know a-priori the number of spatial zones in your CGNS file, but that shouldn't be too bad.


Return to “Analyzer”

Who is online

Users browsing this forum: No registered users and 1 guest

cron