|
[Sponsors] |
May 31, 2010, 06:48 |
monitor y+
|
#1 |
Senior Member
Join Date: Feb 2010
Posts: 213
Rep Power: 17 |
I use simpleFoam and I use yPlusRAS to check the quality of the simulation.I'd like to know how to monitor y+ while running. In the forum I find only threads about y+ post-processing. Can anyone help me, please?
|
|
June 1, 2010, 14:37 |
|
#2 |
Senior Member
Marco A. Turcios
Join Date: Mar 2009
Location: Vancouver, BC, Canada
Posts: 740
Rep Power: 28 |
I was just looking around for this, so check this thread out. The direction you want to take is:
1) Add the initial creation of the yPlus field to createFields.H 2) Add the calculation of yPlus somewhere in the solver code (I think the thread suggested at the end of the PISO loop, which makes sense to me) Code for both tasks can be easily lifted from the original yPlusRAS.C source. While I'm here, I'll ask anyone else who is reading what modifications need to be made for using the utility with compressible flows. My post is in the thread I linked to above, so please reply there. |
|
June 3, 2010, 09:28 |
|
#3 |
Senior Member
Join Date: Feb 2010
Posts: 213
Rep Power: 17 |
Does it mean I have to edit simpleFoam code and re-compile it? I don't want to play with solver codes, I'm not such an 'expert' and I'm afraid of making something wrong. I hope there is a simpler way, such as dictionary editing or utilities like yPlusRAS
|
|
June 3, 2010, 13:57 |
Not that bad actually
|
#4 |
Senior Member
Marco A. Turcios
Join Date: Mar 2009
Location: Vancouver, BC, Canada
Posts: 740
Rep Power: 28 |
I thought it would be bad at first to do things like this as well, but after some practice it isn't that bad. Especially since yPlus is just a part of the turbulence model classes anyhow. A great tutorial for information on how to modify solvers can be found on the OFWiki here.
If you are unfamiliar with C/C++ and object oriented programming, a great place to get up to speed is CPlusPlus.com Your situation is slightly different, as you only need to output a quantity that already exists somewhere in the solver already. Have a look at $FOAM_APP/utilities/postProcessing/wall/yPlusRAS/yPlusRAS.C to see that the actual output of y+ doesn't take too many lines. And for anyone else reading this, I'm looking to modify yPlusRAS to work with compressible flow. I'm thinking I just need to change the creation of the turbulence model to what we see in the createFields.H file of reactingFoam. Sound reasonable? |
|
June 3, 2010, 21:22 |
|
#5 |
Senior Member
Sandy Lee
Join Date: Mar 2009
Posts: 213
Rep Power: 18 |
Yes, I am finding the tutorial about it.
|
|
June 4, 2010, 03:54 |
|
#6 |
Senior Member
Join Date: Feb 2010
Posts: 213
Rep Power: 17 |
Thank you for your suggestions, Marco. I hope someone else has a no-programming solution for me. In the meantime I'll study this stuff.
|
|
March 14, 2011, 05:58 |
|
#7 |
Senior Member
Join Date: Feb 2010
Posts: 213
Rep Power: 17 |
Some time ago, I asked for a method to monitor y+ while the computation is running. I was answered to include some yPlusRAS code in the solver file. I'm not a programmer or a C++ expert and I'm afraid to get in trouble, so I dropped.
Now I ask for that again, I really need it. I work with 36 hours computations on a cluster and the person in charge for it and coworkers could kill me if I try to mess around with OF and solvers. Maybe in the meantime someone has implemented a function to obtain y+ on the fly. Anyway, I can't understand why nobody made it before. I think it would be very useful to avoid very time consumption simulations and find out a bed y+ only at end of running. Maybe the matter is not clear to me and I could estimate a good y+ value before launching the solver, but I'm involved with marine propellers and I think the grid spacing calculators are not appropriate to me. Can someone help me or explain why I'm wrong? Maybe I miss something important about the question. I'll be gratefull for your corrections. |
|
May 30, 2011, 12:32 |
|
#8 |
Member
The True
Join Date: Dec 2010
Posts: 80
Rep Power: 15 |
just using : yPlusRAS calculates the yPlus values. But I don't know how it is calculated. Can i trust the values coming from yPlusRAS ? For some case I had value above 3000
i see that your question was different, sorry. And if the yplus value is oke in the begin (0 file) than it should remain oke , isn't it ? |
|
May 31, 2011, 22:39 |
|
#9 |
Senior Member
Sandy Lee
Join Date: Mar 2009
Posts: 213
Rep Power: 18 |
No, it will vary with the different friction velocity, namely wall viscosity-stress.
|
|
Thread Tools | Search this Thread |
Display Modes | |
|
|