|
[Sponsors] |
May 11, 2011, 23:20 |
Princeton Ocean Model POM
|
#1 |
New Member
JCS
Join Date: May 2011
Posts: 2
Rep Power: 0 |
Hi, i am new here. i had run a couple of searches for Princeton Ocean Model but i could not find a dedicated thread. i am posting this for us to share our knowledge or encounters with this CFD.
Right now, i am in need of assistance to find out a problem i've been trying to solve for weeks. My model which has been converted from original daily to hourly run, with the time steps adjusted and calculated well. Example: Days is changed to hours Original: IEND=DAYS*24*3600/DTI+2 After modification: IEND=HOURS*3600/DTI+2 so let's say 3 days = 72 hours...the time steps would be equal. The symptom of the problem is my model only can run in homogeneous state (Salinity and Temperature are constants). i am quite sure i had applied the correct boundary conditions as listed below: 1. Elevation uses Sommerfeld radiation BC is applied. C EAST DO 120 J=1,JM GA=DTE*sqrt((H(imm1-1,j)+H(imm1,J))*.5*GRAV) & /((DX(imm1-1,j)+DX(imm1,J))*.5) ELF(imm1,J)= & GA*(.25*el(imm1-1,j-1)+ & .5*el(imm1-1,j)+ .25*el(imm1-1,j+1))+ & (1-GA)*(.25*el(imm1,j-1)+ & .5*el(imm1,j)+ .25*el(imm1,j+1)) ELF(IM,J)=ELF(IMM1,J) 2. Barotropic velocity uses radiation along with specified velocity (ms-1) and elevation (m) to calculate the external waves . Example: C EAST DEPTH=(H(imm1,j)+H(im,j)/2.) DXX=(DX(im,j)+DX(imm1,j))*0.5 GA=sqrt(GRAV/H(imm1,j))*DEPTH*DXX uaf(im,j)=(uabound(imm1,j)* RAMP + GA* & (el(imm1,j)-el1bound(imm1,j)*ramp)) & /(DEPTH*DXX) UAF(im,j)=UAF(imm1,j)*ramp VAF(IM,J)=0.0 i.el1bound= predefined elevation (tidal calculations) ii.uabound= Averaged velocity from CORA data. 3. Baroclinic velocity uses radiation bcond similar to the elevation B.C. 4. The salinity and temperature boundary condition is also a radiating B.C. example: C EAST U1=2.*U(IM,J,K)*DTI/(DX(IM,J)+DX(IMM1,J)) IF(U1.LE.0.) THEN UF(IM,J,K)=T(IM,J,K)-U1*(TBound(im,J,K) -T(IM,J,K)) VF(IM,J,K)=S(IM,J,K)-U1*(sBound(im,J,K) -S(IM,J,K)) ELSE UF(IM,J,K)=T(IM,J,K)-U1*(T(IM,J,K)-T(IMM1,J,K)) VF(IM,J,K)=S(IM,J,K)-U1*(S(IM,J,K)-S(IMM1,J,K)) IF(K.NE.1.AND.K.NE.KBM1) THEN WM=.5*(W(IMM1,J,K)+W(IMM1,J,K+1)) 1 *DTI/((ZZ(K-1)-ZZ(K+1))*DT(IMM1,J)) UF(IM,J,K)=UF(IM,J,K)-WM*(T(IMM1,J,K-1)-T(IMM1,J,K+1)) VF(IM,J,K)=VF(IM,J,K)-WM*(S(IMM1,J,K-1)-S(IMM1,J,K+1)) ENDIF ENDIF 5. The vertical and the rest are predefined boundary conditions from the original script. i hope someone could help me point out where the problem could lie. Thanks. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
LES and combustion model | Margherita Cadorin | CFX | 0 | October 29, 2008 05:24 |
model | han | Main CFD Forum | 0 | September 27, 2005 08:35 |
regarding eddy break up model | dj | CFX | 0 | September 30, 2003 02:05 |
Advanced Turbulence Modeling in Fluent, Realizable k-epsilon Model | Jonas Larsson | FLUENT | 5 | March 13, 2000 03:27 |
Ocean Wave Model with FLOW3D | Haryo Armono | Main CFD Forum | 0 | January 28, 1999 13:11 |