Author |
Message |
Issue Ho (ih1209)
New member Username: ih1209
Post Number: 1 Registered: 12-2004
| Posted on Thursday, December 02, 2004 - 06:20 pm: | |
In the example to resize domain in z- direction, the new differential equation is defined as dx[k*dx(Tp)]/Zscale + dy[k*dy(Tp)]/ZScale + dz[k*zscale*dz(Tp)] + Q/zscale= 0 What is the different if I rearrange the "zscale" factor and change the equation to dx[k*dx(Tp)] + dy[k*dy(Tp)] + dz[k*zscale*dz(Tp)]*zscale + Q = 0
|
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 268 Registered: 06-2003
| Posted on Thursday, December 02, 2004 - 07:09 pm: | |
As explained in the Technical Note "Coordinate Scaling", the scale factor is divided out of the equation so that flux continuity will be maintained in passing between layers with different scale. As long as "zcale" is the same in all layers, the equation you propose will work. If you have different scales in different layers, then you must use the form with the scale divided out. |
|