smeagol
Differences
This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision | ||
|
smeagol [2010/11/01 13:41] prokop |
smeagol [2012/12/04 15:59] (current) |
||
|---|---|---|---|
| Line 1: | Line 1: | ||
| + | ====== Equilibriom Kpoins ====== | ||
| + | |||
| + | melo by to fungovat, skousel jsem 1D H chain, H111 bulk, a Si100 bulk, vsude to fungovalo pro vice k-bodu. | ||
| + | |||
| + | Chyba vznikala pri jednom k-bodu, protoze se v tom pripade pouzivala na inverovani GF lapack subrutina pro symetricke matice | ||
| + | Predelal sem to aby se vzdy pouzivala subrutina pro obecne matice | ||
| + | |||
| + | Voltage se aplykuje v diag_k_NET_eq.f90 na diagonalu hamiltonianu. | ||
| + | |||
| + | == DEV NOTES == | ||
| + | |||
| + | v Smeagolu neni zadne jine misto (if statement) ktere by zalezelo na poctu k-bodu. Nepouziva se tam ani nikde jinde symetricky solver (krome invert.f ktera se ale podle vseho nepouziva vubec). | ||
| + | |||
| + | ====== Nonequlibrim Khon-Sham ====== | ||
| + | |||
| + | Aktualne to konverguje (pri vysoke teplote tempfe=1000) a dava "nejakou" vodivost. Ale vysledny potencial (fftpot.xsf) nevypada vubec smysluplne - je symetricky. | ||
| + | |||
| + | === To Do === | ||
| + | |||
| + | === Problemy === | ||
| + | * Hlavni problem ze potencial pro nenulovy VBias nekonverguje k smysluplnemu reseni (asymetricky prohnute rampe), ale k necemu kde jsou LEADy +/- symetricky posunute oproti scattenring regionu. :-( | ||
| + | * Smeagol z nejakeho duvodu nezintegruje spravne hustotu stavu leadu. To se pokousi resit workarround v kspace_NET_KS.f90 v sekci //CORRECTION TO EDGE//, ale nejsem si jisty jetli by to tak melo byt, asi ne. | ||
| + | * zda se mi ze to nenacita/neuklada sprave denmat.dat. protoze kdyz spustim smeagol znovu, jen s trochu pozmenenym napetim, tak je dost velika sigma=1E-1. => __Podivat se jak se nacita a uklada s denmat.__ | ||
| + | * FIXME denmat se ve smeaglu nacita/uklada spatne, :-( kdyz spustim 2x beh programu (1x uz zkonvergovany) je stejne v prvnim kromu sigma=1E-1 | ||
| + | |||
| + | |||
| + | === Kosmeticke vady === | ||
| + | * Jednotkova konvecne je chaoticka, nektere veci se konvertuji do Hratree untis uvnitr smeagol (HL0,HR0 unvitr negf() ) jine uvnitr fireball (V_bias) | ||
| + | * celkove je kod chaoticky - rozstrkany po ruznych subrutinach fireballu a podadresarich (GRID, LOOPS, MODULES, SOLVESH_DIAG, UTILS, SM_INTERFACE, SM_NETRANSPORT ) | ||
| + | |||
| + | === FAQ === | ||
| + | * denmat.dat se nacita. Vypise to "Read denmat.dat". deje se to v initbasics.f90 -> initdenmat.f90. ale funguje to nejak spatne asi, nechapu zatim proc, ale nevypada to ze by uz jednou zkonvergovany vypocet konvergoval vyrazne rychleji. | ||
| + | * "tempfe" ovlivnuje "Tf" ktera se pouziva v negfk.F a selfenergy.F/CHARGEBIAS_EQ, prenasobi se kB=6.33363D-6 [Ry/K] | ||
| + | * Delta resp Sigma (proste imaginarni cast, rozsireni peaku stavu) se nastavuje podle teploty - zavisi na tempfe | ||
| + | * Je lepsi pouzivat fermi level ze smeagol-SCF nez z LEADu. S Ef z LEADu to nekonverguje vubec. Mozna by si mel smeagol __urcit fermi-level sam pomoci integrace__????? | ||
| + | |||
| + | |||
| + | ====== Zmene soubory od puvodniho fireballu ====== | ||
smeagol.1288615277.txt.gz ยท Last modified: 2011/02/18 13:14 (external edit)
Page Tools
Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution-Noncommercial-Share Alike 4.0 International
