CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   HELP URGENT: model turbulent LAUNDER SHARMA (http://www.cfd-online.com/Forums/fluent/27818-help-urgent-model-turbulent-launder-sharma.html)

dory August 5, 2000 13:55

HELP URGENT: model turbulent LAUNDER SHARMA
 
I NEED INFORMATION OVER THE NEAR WALL TREATMENT FOR FULL LOW-RE MODEL , Y PLUS MUST BE =1? SORRY FOR MY ENGLISH, I AM CHILEAN, IF YOU CAN READ SPANISH PLEASE READ MY PROBLEM:

ESTOY MODELANDO CONVECCIÓN NATURAL SOBRE UN TUBO ALETADO, MAS ESPECIFICAMENTE UNA SOLA ALETA, Y DEBO UTILIZAR UN MODELO DE BAJO REYNOLDS, POR PANTALLA ENTRE A: DEFINE/MODELS/VISCOUS/LOW RE KE

PERO EN EL MANUAL DE FLUENT NO APARECE LAS CARACTERISTICAS DEL Y+ PARA EL TRATAMIENTO CERCANO A LA PARED, Y ESTOY AGRANDANDO LA CAPA LÍMITE PARA QUE ME Y+=1, PERO YA ES MUY GRANDE(2MM) Y AUN NO OBTENGO BUENOS RESULTADOS, YA QUE LA TEMPERATURA DE LA ALETA, QUE ES DE ALUMINIO, VARÍA 0.18k DESDE LA BASE AL EXTREMO DE LA ALETA, LO QUE ES POCO REAL, YA QUE LA CONDUCTIVIDAD NO ES COMO PARA QUE EL RENDIMIENTO DE LA ALETA SEA DE UN 100% ?¿

John C. Chien August 5, 2000 14:39

Re: HELP URGENT: model turbulent LAUNDER SHARMA
 
(1). The reason why one need to use Y+=1 (or something like that) for the first grid point away from the wall (or the first cell) is that based on the experimental data, for equilibrium boundary layer, there is a near wall viscous sublayer which covers the region Y+ less than 5. (2). So, if your mesh size is such that Y+=5, then you don't have any mesh points inside this region, thus you are not going to see it in the result. If Y+=1, at least you have roughly 5 points or so in this region, thus you can compute the near wall region. (3). Since Y+ is not the same as y measured from the wall, in many cases, the real y at the first mesh point away from the wall will be very small, say 0.001 or 0.0001 inch. (for Reynolds number of order 100,000 or 1,000,000) So, the actual y value will depend on the conditions you are trying to compute. (4). Y+=1 is just one suggested value based on the simple test data, in real problem, you need to refine it until the solution is mesh independent. And Y+=1 condition is good for any low Reynolds number model which covers the near wall region. (5). You really don't have to use Spanish at all, because we can guess at the question and answer, even if you use very simple English. For example, you can post the message as:" Y+=1, why?, low Reynolds number model, please explain."

dory August 5, 2000 15:38

Re: HELP URGENT: model turbulent LAUNDER SHARMA
 
THEN, IF MY Y+=0.02 APPROXIMATELY, DOES NOT MEAN THAT THE LAYER LIMITS THIS BAD ONE AND MUST ENLARGE IT?

John C. Chien August 5, 2000 19:18

Re: HELP URGENT: model turbulent LAUNDER SHARMA
 
(1). If your Y+=0.02, then it is fine and you should not change it. (2). The actual condition on Y+ at the first mesh point (or cell) should read: Y+ equal or less than one. Since it is 0.02, it is less than one. (3). You are free to change it to Y+=1.0 , but you should check the results to see the effect. If the result does not change, then it does not matter. (4). So, the condition is Y+ must be smaller than one. And you are all right to use Y+=0.02.


All times are GMT -4. The time now is 20:10.