Author |
Message |
Andreas Ferber (madprofessor)
Junior Member Username: madprofessor
Post Number: 3 Registered: 04-2009
| Posted on Wednesday, April 15, 2009 - 07:30 am: | |
Hello there Im trying to get some information about the transfer behavior of my system. Therefor I must evaluate the Flexpde script thousand times but for different Peclet-numbers Pe. I tried to make it easy by using the array function for the Peclet-number but unfortunately it didnīt work. Another way is to do it manual but this is very time-consuming. Particularly because I must alter other Constants as well. My question is: Is there any way to run a FelxPDE script several times automaticaly and thereby changing the value of an Constant in the defenition section(in my case Pe). Or can I implement an arry of Pecletnumbers so that I get the information in one run? My script is down below. Thanks for your help Andreas TITLE 'Dynamisches Kompressionsverhalten Kartesich' COORDINATES cartesian1 VARIABLES Phii Phir Thetai Thetar DEFINITIONS gamma = 1.4 h0=5e-3 h=.001e-3 p0=10e5 T0=273.15+23 Rg=287 rho0=p0/Rg/T0 lambda=0.025 cp=gamma/(gamma-1)*Rg cv=cp/gamma a_s=sqrt(T0*Rg*gamma) D=lambda/rho0/cp !kappa = a_s*h0/D kappa=10^5 !hplus = h/h0 hplus=0.1 !Pe = h0^2*omega/D Pe= pi/2*kappa ! 10^(-1) 10^(0.625) 10^4 pi*kappa Pr = 0.71 omega=Pe*D/h0^2 EQUATIONS Thetar: Thetar + hplus*(gamma - 1)*dx(Phir) + (gamma/Pe)*( - dxx(Thetai)) = 0 Thetai: Thetai + hplus*(gamma - 1)*dx(Phii) + (gamma/Pe)*( dxx(Thetar) ) = 0 Phir: gamma*(Pe^2)*Phir - ((kappa^2)/hplus)*dx(Thetar) + (kappa^2)*(dxx(Phir)) + (4/3)*gamma*Pr*Pe*(-dxx(Phii)) = 0 Phii: gamma*(Pe^2)*Phii - ((kappa^2)/hplus)*dx(Thetai) + (kappa^2)*(dxx(Phii)) + (4/3)*gamma*Pr*Pe*( dxx(Phir)) = 0 BOUNDARIES { The domain definition } REGION 1 { For each material region } START(0) Point Value(Phir) = 0 Point Value(Phii) = 0 Point natural(Thetar) = 0 Point natural(Thetai) = 0 LINE TO (1) Point Value(Phir) = 1 Point Value(Phii) = 0 Point value(Thetar) = 0 Point value(Thetai) = 0 ! TIME 0 TO 1 { if time dependent } MONITORS { show progress } PLOTS { save result displays } elevation( ABS(sqrt(Phir^2+Phii^2)))from(0)to(1)RANGE( 0,1 ) elevation( ABS(sqrt(Thetar^2+Thetai^2)/hplus))from(0)to(1) RANGE( 0,1 ) elevation( ABS(sqrt( dx(Phir)^2 + dx(Phii)^2 )) ) from(0)to(1)RANGE( 0,2 ) elevation( ABS(sqrt(Thetar^2+Thetai^2)/hplus +{-} sqrt( dx(Phir)^2+ dx(Phii)^2 ) ))from(0)to(1)RANGE( 0,2 ) END |
Jared Barber (jared_barber)
Member Username: jared_barber
Post Number: 42 Registered: 01-2007
| Posted on Wednesday, April 15, 2009 - 12:57 pm: | |
Hello, You may be looking for the "staged" function. You should look it up. Your code line starting Pe: Pe= pi/2*kappa ! 10^(-1) 10^(0.625) 10^4 pi*kappa would become Pe= staged(value1,value2,value3). You'll notice it will start saving files (if you are exporting files) like: filename_#.p01 where # is the stage number corresponding to the Pe value being used during the solve used to create that file. Important "select" solution controls to look at are "autostage" and "reinitialize". For your code then, e.g.: TITLE 'Dynamisches Kompressionsverhalten Kartesich' SELECT autostage = "off" {pauses in between each "stage" for you to look at stuff if you want...not necessary, it is "on" by default and doesn't pause between each stage} reinitialize = "on" {doesn't use any initial guesses or initial meshes from the previous time step} COORDINATES cartesian1 VARIABLES Phii Phir Thetai Thetar DEFINITIONS gamma = 1.4 h0=5e-3 h=.001e-3 p0=10e5 T0=273.15+23 Rg=287 rho0=p0/Rg/T0 lambda=0.025 cp=gamma/(gamma-1)*Rg cv=cp/gamma a_s=sqrt(T0*Rg*gamma) D=lambda/rho0/cp !kappa = a_s*h0/D kappa=10^5 !hplus = h/h0 hplus=0.1 !Pe = h0^2*omega/D Pe= pi/2*kappa ! 10^(-1) 10^(0.625) 10^4 pi*kappa Pr = 0.71 omega=Pe*D/h0^2 EQUATIONS Thetar: Thetar + hplus*(gamma - 1)*dx(Phir) + (gamma/Pe)*( - dxx(Thetai)) = 0 Thetai: Thetai + hplus*(gamma - 1)*dx(Phii) + (gamma/Pe)*( dxx(Thetar) ) = 0 Phir: gamma*(Pe^2)*Phir - ((kappa^2)/hplus)*dx(Thetar) + (kappa^2)*(dxx(Phir)) + (4/3)*gamma*Pr*Pe*(-dxx(Phii)) = 0 Phii: gamma*(Pe^2)*Phii - ((kappa^2)/hplus)*dx(Thetai) + (kappa^2)*(dxx(Phii)) + (4/3)*gamma*Pr*Pe*( dxx(Phir)) = 0 BOUNDARIES { The domain definition } REGION 1 { For each material region } START(0) Point Value(Phir) = 0 Point Value(Phii) = 0 Point natural(Thetar) = 0 Point natural(Thetai) = 0 LINE TO (1) Point Value(Phir) = 1 Point Value(Phii) = 0 Point value(Thetar) = 0 Point value(Thetai) = 0 ! TIME 0 TO 1 { if time dependent } MONITORS { show progress } PLOTS { save result displays } elevation( ABS(sqrt(Phir^2+Phii^2)))from(0)to(1)RANGE( 0,1 ) elevation( ABS(sqrt(Thetar^2+Thetai^2)/hplus))from(0)to(1) RANGE( 0,1 ) elevation( ABS(sqrt( dx(Phir)^2 + dx(Phii)^2 )) ) from(0)to(1)RANGE( 0,2 ) elevation( ABS(sqrt(Thetar^2+Thetai^2)/hplus +{-} sqrt( dx(Phir)^2+ dx(Phii)^2 ) ))from(0)to(1)RANGE( 0,2 ) END I should say to make sure to check the staged answers with unstaged answers as "staged" in the past has given me a couple of problems in that the answers it gives me for, say, value3 doesn't always correspond to the answer I got if I just plugged in Pe = value3 with no staging. I'll leave it at that for now and good luck. Jared |
Robert G. Nelson (rgnelson)
Moderator Username: rgnelson
Post Number: 1245 Registered: 06-2003
| Posted on Wednesday, April 15, 2009 - 01:15 pm: | |
See Help->User Guide->Some Common Variations->Parameter Studies Using STAGES See also "STAGED" in the Help Index.
|
Andreas Ferber (madprofessor)
Member Username: madprofessor
Post Number: 4 Registered: 04-2009
| Posted on Wednesday, April 15, 2009 - 03:38 pm: | |
Thanks a lot! owing to your suport I think now I have finished my first work with FlexPDE. Its a very nice Programm. kind regards
|
|