Benchmarking with HFODD

To demonstrate that the isospin-invariant formalism has been properly implemented, we provide a detailed comparison between the HFODD [27] and HFBTHO frameworks. This benchmarking is meaningful as the two pnEDF codes were developed independently and have fairly different structures. In particular, the HF equations in HFODD [28,29] are solved in three-dimensional Cartesian basis while HFBTHO employs the two-dimensional cylindrical basis.

Calculations were performed for the $ A=40$, $ T\simeq8$ deformed IASs with SkM* EDF parametrization using the s.p. basis space of $ N_{\rm sh} = 10$. The oscillator length was assumed to be $ b$ = 1.697626fm. The mass constant in Eq. (5) was fixed at $ \hbar^2/2m = 20.73$MeVfm$ ^2$. As far as integration is concerned, we used $ N_{\rm GH} = 26$ Gauss-Hermite nodes for each Cartesian coordinate in HFODD, whereas in HFBTHO, the numbers of Gauss-Hermite ($ \rho$-direction) and Gauss-Laguerre ($ z$-direction) nodes were assumed to be equal: $ N_{\rm GH} = N_{\rm GL}=40$. In addition, in HFBTHO, the number of Gauss-Legendre nodes used in the integration of the direct Coulomb field was set to 80, and the Coulomb length scale was taken to be $ L=50$fm. This set of parameters was recommended as a default value in the latest version of the HFBTHO, as it provides a sufficient precision on the direct Coulomb energy [24]. Without Coulomb, the isocranking frequency was set to $ \lambda ' = 27.092394$MeV and $ \lambda_{\text{off}} = 0$MeV. With Coulomb, we took the values $ \lambda ' =28.613615$MeV and $ \lambda_{\text{off}} = -6.010741$MeV.


Table: Benchmarking of HFODD with HFBTHO for the deformed $ A=40, T\simeq 8$ IASs using SkM* EDF ( $ N_{\rm sh} = 10$), and $ \theta '=0^\circ $ ($ ^{40}$Mg). Other parameters are: $ \lambda ' = 27.092394$MeV, $ \lambda _{\rm off} = 0$MeV (without Coulomb), and $ \lambda ' =28.613615$MeV, $ \lambda _{\rm off} = -6.010741$MeV (with Coulomb). Shown are various contributions to the total binding energy $ E_{\rm tot}$ (in MeV), r.m.s. radii (in fm), expectation values of $ T^2, T_z$, and $ T_x$, and the quadrupole deformation $ \beta _2$. The digits which do not coincide in HFODD and HFBTHO are marked in bold.
  HFODD HFBTHO HFODD HFBTHO
  Without Coulomb With Coulomb
         
$ E_{\rm tot}$ -303.42519 -303.42520 -276.47641 -276.47643
$ E_{\rm kin}^{({\rm n})}$ 498.448466 498.448464 495.53929 495.53930
$ E_{\rm kin}^{({\rm p})}$ 175.371764 175.371762 171.30205 171.30206
$ E_{\rm pot}$ -977.24542 -977.24543 -970.09923 -970.09926
$ E_{\rm SO}$ -34.357903 -34.357905 -33.184812 -33.184816
$ E_{\rm Cou}^{({\rm dir})}$     30.920704 30.920697
$ E_{\rm Cou}^{({\rm exc})}$     -4.139228 -4.139228
$ r_{\rm rms}^{({\rm n})}$ 3.697718 3.697718 3.709975 3.709975
$ r_{\rm rms}^{({\rm p})}$ 3.176356 3.176356 3.217587 3.217587
$ T^2$ 72.022743 72.022743 72.023123 72.023123
$ T_z$ 8.000000 8.000000 8.000000 8.000000
$ T_x$ 0.000000 0.000000 0.000000 0.000000
$ \beta _2$ 0.304201 0.304201 0.311518 0.311518


Table 2: Similar to Table 1 but for $ \theta '=90^\circ $ ($ ^{40}$Ca).
  HFODD HFBTHO HFODD HFBTHO
  Without Coulomb With Coulomb
         
$ E_{\rm tot}$ -303.42519 -303.42520 -234.429 -234.419
$ E_{\rm kin}^{({\rm n})}$ 336.910115 336.910113 333.68 333.71
$ E_{\rm kin}^{({\rm p})}$ 336.910115 336.910113 318.704 318.713
$ E_{\rm pot}$ -977.245420 -977.24543 -954.79 -954.83
$ E_{\rm SO}$ -34.357903 -34.357905 -31.49 -31.51
$ E_{\rm Cou}^{({\rm dir})}$     75.105 75.106
$ E_{\rm Cou}^{({\rm exc})}$     -7.12522 -7.12525
$ r_{\rm rms}^{({\rm n})}$ 3.549360 3.549360 3.5930 3.5928
$ r_{\rm rms}^{({\rm p})}$ 3.549360 3.549360 3.63519 3.63512
$ T^2$ 72.022743 72.022743 72.149 72.155
$ T_z$ 0.000000 0.000000 0.15649 0.15652
$ T_x$ 8.000000 8.000000 8.0051 8.0054
$ \beta _2$ 0.304201 0.304201 0.3184 0.3182

The benchmarking results for the deformed case are shown in Tables 1 and 2 for $ \theta '=0^\circ $ and $ 90^\circ$, respectively. In the absence of the Coulomb term, the difference in the total energy $ E_{\rm tot}$ is less than 20eV, and the total isospin $ \langle\,\hat{T}^2 \,\rangle $ agrees up to the sixth decimal place. With the inclusion of the Coulomb term, the agreement is slightly reduced but is still excellent. A comparison between HFBTHO and HFODD was also performed for the spherical $ A=48$, $ T\simeq4$ IASs and the results are presented in Table 3 for $ \theta '=90^\circ $ where the differences between the two codes are largest. In this case, it is found that deviation in the total energy is about 20eV.

It is to be noted that both HFODD and HFBTHO use the same number of basis harmonic oscillator states. Moreover, as it has been demonstrated previously [24], using a sufficient number of quadrature points in HFBTHO and HFODD, the results of both solvers agree with a high accuracy of several eV. The differences between the two codes with the Coulomb potential turned on, can be traced back to different techniques used to compute the direct Coulomb field: the solver HFODD, uses a more accurate Green's function approach. The benchmark examples discussed in this section demonstrate that the p-n mixing has been implemented correctly in both codes. Presently, we are in process of implementing the cylindrical Green's function treatment of the Coulomb potential into HFBTHO and it is expected that the agreement between the two codes will further improve.


Table: Similar to in Table 1 but for the spherical $ A=48, T\simeq 4$ IASs and $ \theta '=90^\circ $ ($ ^{48}$Cr). Other parameters are: $ \lambda ' = 11.0$MeV, $ \lambda _{\rm off} = 0$MeV (without Coulomb), and $ \lambda ' =12.0$MeV, $ \lambda _{\rm off} = -8.0$MeV (with Coulomb).
  HFODD HFBTHO HFODD HFBTHO
  Without Coulomb With Coulomb
         
$ E_{\rm tot}$ -491.243706 -491.243724 -389.8454 -389.8438
$ E_{\rm kin}^{({\rm n})}$ 422.93152 422.93158 415.69 415.71
$ E_{\rm kin}^{({\rm p})}$ 422.93152 422.93143 404.673 404.679
$ E_{\rm pot}$ -1337.10675 -1337.10673 -1310.44 -1310.46
$ E_{\rm SO}$ -36.736418 -36.736417 -34.1216 -34.1240
$ E_{\rm Cou}^{({\rm dir})}$     109.3774 109.3781
$ E_{\rm Cou}^{({\rm exc})}$     -9.14576 -9.14581
$ r_{\rm rms}^{({\rm n})}$ 3.497939 3.497940 3.52633 3.52628
$ r_{\rm rms}^{({\rm p})}$ 3.497939 3.497939 3.57787 3.57784
$ T^2$ 20.037818 20.037818 20.0756 20.0771
$ T_z$ 0.000000 0.000003 -0.012676 -0.012663
$ T_x$ 4.000000 4.000000 4.00264 4.00278

Jacek Dobaczewski 2014-12-07