16
∑k(2¾Γ1 with 6k = ħ~k21(2m). Note that we have ignored the Hartree shift in the BdG
equation. In the mean-field level, it is reasonable to thedrop the Hartree term at unitarity
limit [22].
3.2 Self-consistent numerical solutions of the BDG equa-
tion
Our aim is to search for the ground state by solving the BdG equation above self-consistently.
Such self-consistent calculation requires huge computational resources and time, therefore
an effective numerical scheme and parallelization of the program will be necessary. We
adopt a hybrid strategy as detailed in Ref. [22], where it has been used to successfully
solve BdG equation in an effective ID case. In short, the basic idea for this strategy is
that a cutoff energy Ec is introduced. Quasiparticle modes with ∣E7I < Ec are calculated
by solving Eqs. (3.2); while the high-energy modes with ∣E7∙∣ > Ec are calculated using
the semi-classical method in the spirit of the LDA. In our practice, we have found this
technique to be very efficient in the sense that it does not require a very large Ec (typical
values of Ec we used is a few times Fermi energy) and the results are essentially cutoff-
independent. For each run, we choose T, as and λ. We then either fix the values for Nσ, or
that for μσ, which is referred as canonical or grand-canonical mode respectively and solve
Eqs. (3.2) Self-Consistently together with the above-cutoff modes.
More intriguing information
1. EDUCATIONAL ACTIVITIES IN TENNESSEE ON WATER USE AND CONTROL - AGRICULTURAL PHASES2. The name is absent
3. Campanile Orchestra
4. Party Groups and Policy Positions in the European Parliament
5. Female Empowerment: Impact of a Commitment Savings Product in the Philippines
6. The name is absent
7. Strategic monetary policy in a monetary union with non-atomistic wage setters
8. Cyber-pharmacies and emerging concerns on marketing drugs Online
9. Transport system as an element of sustainable economic growth in the tourist region
10. TOMOGRAPHIC IMAGE RECONSTRUCTION OF FAN-BEAM PROJECTIONS WITH EQUIDISTANT DETECTORS USING PARTIALLY CONNECTED NEURAL NETWORKS