Learning Forum

[eannakennedy]

I'm trying to simulate the flow of cells in a microfluidic channel and for my results I will need to know the X and Y positions of the particles as they pass through the outlet plane but I can't figure out how to do this. Any help would be appreciated, thanks.

[DrAmine]

Check particle sampling at your outlet as it outputs what you are looking for in a separate file.

[eannakennedy]

Great I'll have a look now, thanks!

 

[eannakennedy]

I can't seem to find anything like this, I'm using Ansys 19.2. I saw mentions online of sample reports for older versions of Fluent but can't find anything for the new version. I'm very new to Fluent and I'm still figuring out where everything is. Do I have to create a user defined function for this?

 

[DrAmine]

Look after Sampling of trajectories in User's manual chapter 23.8.8.

[eannakennedy]

Where can I find the user manual for version 19? The latest version i can find online is version 17. Thanks again.

[Rob]

Click on "Help" on whichever panel you have open in Fluent and then work through the index. We can't give links as you need to log into one of our systems, but using the software bypasses that step (don't ask!). 

[eannakennedy]

From the Help panel i'm guessing i need to view the .dpm file but how can I view the alphanumeric data from this file?

[eannakennedy]

I'm getting quite confused looking through the help files now. I'll explain my project a bit more.

I want to design a microchannel which focuses a flow of particles into single file so that they can be easily detected. I want to record the positions of the particles at the outlet and on several other planes. Then I want to use this to compute the standard deviation of particles from the centre of the channel and their coefficient of variance. 

I am use Steady-State uncoupled solution with no other physics enabled at the moment. I am injecting 30 particles from the inlet surface at 1m/s (the same as the fluid velocity). For some reason only 28 particles are exiting while 2 remain incomplete. How do I run the simulation until all particles exit?

I tried using the step-by-step particle reporting to write a file with the particle positions, but this creates a .his file and I don't know how to work with this.

Once I can get the the model to work properly and view the results I need I then want to use either the adjoint solver or parameter based optimization to minimise the standard deviation of the particles about the channel centre. Will i need to use UDF's for this?

[Rob]

Create a number of planes (or ideally, build in a number of interior faces when building the model). Set up a DPM Summary on each face, this will create a .dpm file for each surface. Release the particles and read the data using a text editor or Excel. 

If you're using a surface release it's possible that two parcels are getting trapped in corners somewhere: plot the particle tracks and see where they finish up. Also, you may want/need to increase the maximum number of steps: default is 500 or 1000 (can't remember), 5000 may be more suitable. Don't just use a "high" number for maximum number of steps as it'll slow the calculation. 

[eannakennedy]

Thanks for the help again. I have the numbers I need in Excel now. When optimizing this channel will I need to create a UDF to optimize for Standard Deviation?

 

As for the two missing particles they are not trapped, they are incomplete. And I have used up to max 50,000 steps for the DPM.

[Rob]

Incomplete means they're stuck (trapped - the terms interchange in English for the non-native speakers) somewhere in the domain, usually in low/no flow regions. If you add enough steps they may eventually leave the domain but the extra cpu & clock time may not be worth it. 

If you're trying to optimise the shape based on the particle positions you may need an external tool such as RBF Morpher, I don't know how that hooks in so will leave that for someone else to explain.  

 

[eannakennedy]

I've also noticed that varying the mass flow rate of the injection doesn't affect the number of particles being injected, it's stuck at 30 and only 28 completing. I'll look into RBF Morpher, but if I just perform parametric optimization, do you think I will be able to do that in Fluent without add-ons?

[Rob]

You're tracking parcels rather than particles: each parcel has an amount of mass based on the injection mass and number of parcels. Each parcel is tracked based on it's diameter & material properties. When it hits something/evaporates etc the parcel mass is used to determine the effect.  It's a simple concept that always causes confusion as we generally talk about particles when we should say parcels..... 

If you can get the value of interest out of the solver and back into your optimiser code then it should be possible, however I don't know any details. 

 

[eannakennedy]

Thanks again for the help.

 

I've been looking at creating an output parameter by writing a UDF, is the macro DEFINE_OUTPUT_PARAMETER still available or has this been replaced by something else?

I've come across it in older help files but its not in the help files for version 19.2 which I am using.

 

[Rob]

There were some changes which improved the reporting tools and their promotion to parameters which made some of the UDFs obsolete. Check 20.3 in the Beta feature manual for an alternative.  

[eannakennedy]

I checked the manual and I think I see now how to go about this now, but there is still one thing i don't understand.

I want to record the data of each particle as it passes through a plane for later use. I have read the source code for the dpm sample reports that fluent uses to do this, and it seems that the function is called each time a particle passes through the sampling plane and writes it a file.

But i can't find a way to read data from the .dpm file to compute the output parameter i need in another UDF.

So I think I either need some way of adding a counter to the DEFINE_DPM_OUTPUT macro  so i can save the data i need to an array i can actually use. Or I need a way to read the .dpm file.

[sachvan]

Hi sir, 

Can you tell how to create the DPM summary for each plane? 

I don't seem to find that option anywhere on Fluent.