| Author | Message | |||
     mefisto faust (mefistofaust) New member Username: mefistofaust Post Number: 1 Registered: 08-2007 |
Hi I have problems to write these equations in FlexPDE. Can you help me please? Itīs here:  | |||
| Robert G. Nelson (rgnelson) Moderator Username: rgnelson Post Number: 927 Registered: 06-2003 |
This equation is the same as the one used as an example in our tutorial www.pdesolutions.com click Product Info -> Tutorial. Also see the example problems in the folder Samples|Time_Dependent|Heatflow. Incidentally, the way you have written the equation is correct only if k has a single, constant value throughout the domain. If there is any variation in k, or differing material values, then you must write the k inside the divergence: dt(u) = div(K*grad(u))+q | |||
| mefisto faust (mefistofaust) Junior Member Username: mefistofaust Post Number: 3 Registered: 08-2007 |
Hi, thank you much for help. I tried to write the script. But I have something wrong in the part of boundaries. Hope everything else is OK. Can you look at it,please?
I wrote a commentary to script, so hope you will understand. Thank much for your help. =* | |||
| Robert G. Nelson (rgnelson) Moderator Username: rgnelson Post Number: 930 Registered: 06-2003 |
You've simply drawn a diagonal line. This is not a 2D figure, as it has no area. Re-read the Help section "User Guide->Basic Usage->Mapping the Domain", where we show you how to draw a box. | |||
| mefisto faust (mefistofaust) Member Username: mefistofaust Post Number: 4 Registered: 08-2007 |
Thanks for advice, I think I generated the boundaries now correctly, but I am not sure, if the PLOT and MONITORS is right writen. Can you look at it please, and write me if itīs ok and if not, whatīs right? I have there one dialog and is empty, so I have something wrong, probably. Thank you very much for your time 
| |||
| Robert G. Nelson (rgnelson) Moderator Username: rgnelson Post Number: 931 Registered: 06-2003 |
"ON x=0" and "ON y=0" are 3D cut-plane specifiers. They are meaningless in 2D problems, and confuse the plot generator. Remove them. Also, you have no boundary conditions, so the default is a fully insulated body with an internal source. The temperature is therefore uniform throughout the domain, and growing steadily and without bound in time. Provide some kind of boundary condition. | |||
| mefisto faust (mefistofaust) Member Username: mefistofaust Post Number: 6 Registered: 08-2007 |
Thanks for advice, I wrote my source as a circle in the middle of the plate. The code is added.
It looks now good,  but why I saw the numbers in negative? Not +number but -number? Please can you write me, if I have done it good or something missing me there? Thanks many Robert | |||
| mefisto faust (mefistofaust) Member Username: mefistofaust Post Number: 7 Registered: 08-2007 |
Soory I work out more on it. I was wrong before. Now I have source Q the constant number and define the NATURAL(u) and VALUE(u) in BOUNDARIES. (The numbers shown in PLOT pictures is a temperature, I hope so.Thanks for answer.) Hope now itīs ok. Can you check it, please? Thank much.
| |||
| Robert G. Nelson (rgnelson) Moderator Username: rgnelson Post Number: 935 Registered: 06-2003 |
The sign is wrong on your Natural BC. The Natural defines the outward surface normal component of the argument of the divergence when the divergence is on the left side of the equation. In your case, this means that the Natural BC defines the outward surface normal component of -k*grad(u), or the outward heat flux. Natural(u)=70 means that 70 units of heat are drawn out of the system in each time unit, regardless of the value of U. | |||
| mefisto faust (mefistofaust) Member Username: mefistofaust Post Number: 11 Registered: 08-2007 |
Dear Robert I think I got what you said (NATURAL(u)=normal(k*grad(u))). Thanks for your excellent comment and your help on this script. I hope the second problem will be done better from me Thanks Jelena
|
