[Haiqin]

Hi,
I work in 2D to simulate a fluidized bed, after finishing the simulation, I would like to obtain time-averaged profiles of parameters, for example, time-averaged solid volume fraction at specific bed height, time-averaged gas velocity at specific bed height etc. But I don't know how, do anybody have any idea?
Thanks in advance!

Haiqin

[fluid23]

That's sort of a loaded question. I am assuming this was an unsteady analysis, no? What kind of analysis; RANS, LES, DNS?

[Haiqin]

Originally Posted by MBdonCFD

That's sort of a loaded question. I am assuming this was an unsteady analysis, no? What kind of analysis; RANS, LES, DNS?

Yes, it is a transient problem. It is a gas-solid fluidized bed, and simulated with multiphase flow model.

[fluid23]

So RANS, LES or DNS? If you aren't sure, just tell me what turbulence model you are using.

[Haiqin]

Originally Posted by MBdonCFD

So RANS, LES or DNS? If you aren't sure, just tell me what turbulence model you are using.

well, in my case, both the gas phase and solid phase were modeled with laminar model.

[fluid23]

Ok... so transient laminar problem? Do you have time-variant boundary conditions? It seems odd that you would setup a model in this way, but I am not familiar with fluidized bed analysis.

Lets approach this from a different angle. Can you share a screen shot of your physics continua selections? I just want to make sure I give you the best answer I can.

[Haiqin]

well, perhaps I should describe my question in another way:
for a simulation of transient problem, one parameter (the solid volume fraction, for example) has been recorded in a history file at each time step during the simulation. So when the simulation finished, there should be a set of solid volume fraction values recorded in the history file that correspond to different time step.
And what I want to do is to extract the time series solid volume fraction values.

[fluid23]

That is a much better question. The way it was phrased before I was going to tell you that RANS analysis results (which I assume you have) are already time averaged across each time step. If your flow is laminar and you aren't looking at varying boundary conditions or how the bed fills from an empty state then your results are not going to change from time step to time step therefore all of your results will be time-averaged if you look at a single time step.

Based on your rephrased question, I would ask whether you read the help documentation > user guide > analyzing > recording and viewing solutions? Its not clear if your volume fraction is a global value or shown for each cell (a scalar plot). Solution histories are maybe not the right tool to do this if it is a global value. They are for saving scalar, vector and xy plots at each time step so they can be recalled. If its a global value (a single value volume fraction of your domain at a given time step) then you need to setup a report and monitor to capture this in a report plot.

[Haiqin]

The parameter of solid volume fraction in my case is not a global value. Now I can extract the value for specific position at specific time step from the history file (this has been done by following the user guide), but I hope to extract values at all time steps with one operation. I can fulfill the objective by repeating the operation of extracting value at specific time step, but obviously it is time-consuming and kind of stupid.

[fluid23]

So if you need to get volume fraction for a specific X,Y,Z at each time step the easiest thing to have set it up before you ran. You can probably create a macro to step through the solution histories based extract the data. Or if you have time I would setup a derived part probe > point then create a report and monitor for volume fraction. Make sure the monitor is set to update at each time step, not each iteration then create a report plot from the monitor. Also, make sure the x-axis is set to physical time not iteration. You will end up with volume fraction vs time at X,Y,Z at the end of each time step.

[me3840]

What you want to look in to is a field mean monitor. There is data on it in the user guide. You have to be very careful with it though for large grids if you use the windowing function, since you can easily eat up your system memory.

[Haiqin]

Originally Posted by MBdonCFD

So if you need to get volume fraction for a specific X,Y,Z at each time step the easiest thing to have set it up before you ran. You can probably create a macro to step through the solution histories based extract the data. Or if you have time I would setup a derived part probe > point then create a report and monitor for volume fraction. Make sure the monitor is set to update at each time step, not each iteration then create a report plot from the monitor. Also, make sure the x-axis is set to physical time not iteration. You will end up with volume fraction vs time at X,Y,Z at the end of each time step.

Actually, that is what I am doing now. But it is still kind of time-consuming in my case, because I need values in lots of points. I think I need to find out an easier way.
But thanks very much for your suggestions and time.

[Haiqin]

Originally Posted by me3840

What you want to look in to is a field mean monitor. There is data on it in the user guide. You have to be very careful with it though for large grids if you use the windowing function, since you can easily eat up your system memory.

That may be a good way, I should try. Thanks for the suggestion.

[fluid23]

You said you needed one point. If you need multiple points, do what me3840 suggested.