|
[Sponsors] |
August 18, 2005, 05:54 |
SP in USRSRC-CFX4
|
#1 |
Guest
Posts: n/a
|
Hi,
In the Source representation: S= SU+(PHI*SP) When putting variables into SP, does one need to differentiate the term with respect to the variable? For example: IF the source(=TERM): TERM=(CNST1*H/CNST2+H), CNST1 & CNST2=constants, H=enthalpy If this is a negative source and needs to be placed in SP, is it simply: SP(INODE,IPHASE)=SP(INODE,IPHASE)-(CNST1/CNST2+H) or does one need to dTERM/dh, and then put the solution to this into SP. If it is the latter, could you also explain why is this so? And if it applies to SP, why does it not apply to SU? Hope someone can shed some light on this issue... Many thanks, Robin |
|
August 19, 2005, 20:04 |
Re: SP in USRSRC-CFX4
|
#2 |
Guest
Posts: n/a
|
No, you do not need the gradient terms in the sources.
The thing to remember is that the code multiplies your entire source in SP by the variable PHI (in your case H). Your Source C1*H/C2 + H Can be factored into H * (C1/C2 + 1) Thus SP should be set to SP(INODE,IPHASE)= SP(INODE,IPHASE) + (C1/C2 + 1) You put the source terms in SP if they are negative (i.e. C1 or C2 in this expression is negative, but not both). You don't actually "force" them to be negative by subtracting from SP. If (C1/C2+ 1) is positive, just multiply the whole thing by the local H(INODE,IPHASE) and add it to SU. Hope this helps. Jeff |
|
August 20, 2005, 05:41 |
Re: SP in USRSRC-CFX4
|
#3 |
Guest
Posts: n/a
|
Hi Jeff,
Many thanks Jeff for this and your previous replies to my requests. Robin |
|
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
CFX4 dump file format | Glenn Horrocks | CFX | 2 | October 12, 2003 19:16 |
remote user of CFX4 | John | CFX | 5 | September 7, 2003 18:20 |
Grid generation for conical nose shpe in CFX4 | zahid | CFX | 8 | June 2, 2003 21:51 |
Unigraphics -> CFX4 Build | Vincent | CFX | 1 | February 11, 2003 16:04 |
CFX4 Validation | Glenn Horrocks | CFX | 0 | January 30, 2001 22:56 |