[srv537]

i simulated turbulent flow in a channel, i created a custom field functions and after 4 flow through time i averaged and plot turbulent stress u'u', v'v' and u'v,

-u'u' =(2 / density) * viscosity-turb * (dx-velocity-dx) - (2 / 3) * turb-kinetic-energy

-v'v' = (2 / density) * viscosity-turb * (dy-velocity-dy) - (2 / 3) * turb-kinetic-energy

-u'v' = (1 / density) * viscosity-turb * ((dx-velocity-dy) + (dy-velocity-dx))

i used kw-sst model

but i am getting large error as compare to experiment result , please suggest me where is mistake.

thank you

[Shamoon Jamshed]

What variable or parameter are you comparing with experiment

[Shamoon Jamshed]

I am not satisfied with the velocity stress, since k-w k-epsilon, or SA or kw std or keps (standard) all are designed to model Reynolds stress, which otherwise are computed using RSM model. I think you will not get something right with intentionally computing u'v' yourself. and also if your are using custom field function , first you will need to compute kinetic energy as well.

[srv537]

Originally Posted by Shamoon Jamshed

I am not satisfied with the velocity stress, since k-w k-epsilon, or SA or kw std or keps (standard) all are designed to model Reynolds stress, which otherwise are computed using RSM model. I think you will not get something right with intentionally computing u'v' yourself. and also if your are using custom field function , first you will need to compute kinetic energy as well.

how can i compute u'v', is there any other method to calculate u'v', u'u' and v'v'.
please give me some detail

[srv537]

Originally Posted by Shamoon Jamshed

What variable or parameter are you comparing with experiment

i want to compare with experiment result of average turbulent stress u'u', v'v' and u'v'

[Shamoon Jamshed]

Try to use RSM model, if your computational power allows, otherwise try to use the custom FF of Kinetic energy then compute u'v', but I do not think this will give you some answers with k-w or k-epsilon , just try to do it. But first try to use RSM

[srv537]

Originally Posted by Shamoon Jamshed

Try to use RSM model, if your computational power allows, otherwise try to use the custom FF of Kinetic energy then compute u'v', but I do not think this will give you some answers with k-w or k-epsilon , just try to do it. But first try to use RSM

actually in fluent there is kinetic energy as a variable is available can i use that in place of custom FF for kinetic energy? and if not then can you please tell me how can i create kinetic energy custom field function.

thank you

[Shamoon Jamshed]

I think you should be able to use TKE=0.5*(mean(ux'ux')+mean(uy'uy')+mean(uz'uz'))