Nuclear Models#

Coherent#

In coherent scattering the whole nucleus participates and the nucleus remains in its ground state. The hadronic current is proportional to the sum of initial and final nuclear four-momenta weighted by the nuclear form factor:

\[J^\mu_{\rm coh} = \bigl(p_A^\mu + p_A^{\prime\,\mu}\bigr)\,F_{\rm coh}(Q^2)\]

where \(F_{\rm coh}(Q^2)\) is the coherent (nuclear charge) form factor. Only the coherent entry in the form factor file is used. The initial-state weight is unity (the nucleus is always in its ground state).

QESpectral#

Quasi-elastic scattering is treated in the impulse approximation: a single off-shell nucleon with three-momentum \(\boldsymbol{p}\) and removal energy \(E_r = m_N - E\) is struck by the virtual boson and ejected from the nucleus.

Initial-state sampling. The joint probability for finding a nucleon with momentum \(|\boldsymbol{p}|\) and removal energy \(E_r\) is given by the nucleon spectral function:

\[w_{\rm IS} = N_{p/n}\,P(|\boldsymbol{p}|,\,E_r)\]

where \(N_{p/n}\) is the number of protons or neutrons and \(P\) is read from SpectralP / SpectralN. The spectral function is stored as a tabulated 2D array in momentum and removal energy and is interpolated at run time.

Hadronic current. The current uses the standard nucleon vertex with Dirac (\(F_1\)), Pauli (\(F_2\)), and axial (\(F_A\), \(F_{AP}\)) form factors:

\[J^\mu = \bar{u}(p^\prime)\Bigl[ F_1\,\gamma^\mu + i\,\frac{F_2}{2m_N}\,\sigma^{\mu\nu}q_\nu + F_A\,\gamma^\mu\gamma^5 + \frac{F_{AP}}{m_N}\,\gamma^5 q^\mu \Bigr]u(p)\]

where \(u(p)\) and \(\bar{u}(p^\prime)\) are free Dirac spinors for the initial and final nucleons. The initial nucleon energy is set to the free on-shell value \(E_{\rm free} = \sqrt{|\boldsymbol{p}|^2 + m_N^2}\), while the energy transfer is shifted by the off-shell correction \(\omega \to \omega + E - E_{\rm free}\). Restoration of the Ward identity is handled by the choice of scheme.

All four spin-state combinations \((s, s^\prime) \in \{-1,+1\}^2\) are summed incoherently, giving 4 entries in the current array.

HyperonSpectral#

Hyperon production follows the same spectral-function formalism as QESpectral for the initial state. The hadronic current uses separate Dirac, Pauli, and axial form factors for each hyperon channel (\(\Lambda^0\), \(\Sigma^0\), \(\Sigma^-\)), taken from the hyperon block of the form factor file. The outgoing baryon mass in the Pauli term is replaced by the hyperon mass.

FortranModel#

FortranModel acts as a bridge between the C++ framework and a user-supplied Fortran nuclear model. The C++ class calls the Fortran routines described in Fortran Interface to obtain the hadronic current and initial-state weight. The nuclear mode and phase-space generator are also determined by the Fortran model at runtime, so a FortranModel can implement any of the modes listed in Nuclear Modes.

Three Fortran models ship with Achilles and are registered automatically at startup. They are summarised below. For all three the run-card Model key must be set to FortranModel and the Name key must match the model’s registered name exactly.

Built-in Fortran models at a glance#

Registered name (Name)

Source file

Mode

Frame

InspireHEP

QE_Spectral_Func

qe_spectral_model.f90

Quasielastic (2)

Lab (0)

Rocco:2018mwt

RES_Spectral_Func

res_spectral_model.f90

Resonance (4)

QZ (1)

Nakamura:2015rta, Rocco:2019gfb

Intf_Spectral_Func

intf_spectral_model.f90

Interference QE–MEC (7)

Lab (0)

Lovato:2023khk

QE_Spectral_Func — Quasielastic spectral function#

This is the Fortran implementation of the quasi-elastic spectral-function model, equivalent in physics to the C++ QESpectral model. It samples a nucleon from the nuclear spectral function and computes the one-body hadronic current with Dirac, Pauli, and axial form factors.

Physics. The hadronic current is:

\[J^\mu = \bar{u}(p^\prime)\Bigl[ F_1\,\gamma^\mu + i\,\frac{F_2}{2m_N}\,\sigma^{\mu\nu}q_\nu + F_A\,\gamma^\mu\gamma^5 + \frac{F_{AP}}{m_N}\,\gamma^5 q^\mu \Bigr]u(p)\]

with the off-shell energy correction \(\omega \to \omega + E - E_{\rm free}\) applied to the energy transfer. Four spin combinations are summed (nspin = 4). The initial-state weight is \(w = N_{p/n}\,P(|\boldsymbol{p}|, E_r)\).

Form factors used: F1, F2, FA, FAP (from the vector and axial categories of the form factor file).

Configuration. The ConfigFile is a plain-text file with two lines giving the paths to the proton and neutron spectral function data files:

data/Spectral_Functions/pke12p_tot.data
data/Spectral_Functions/pke12n_tot.data

Pre-supplied config files: data/info_C12_pke.data, data/info_O16_pke.data, data/info_Ar40_pke.data.

Run-card example:

NuclearModels:
  - NuclearModel:
      Model: FortranModel
      Name: QE_Spectral_Func
      ConfigFile: data/info_C12_pke.data
      FormFactorFile: FormFactors.yml
      Ward: None

RES_Spectral_Func — Resonance spectral function#

Single-pion production through baryon resonances (primarily the \(\Delta(1232)\)), combined with a spectral-function description of the nuclear initial state. All pion–nucleon channels consistent with charge conservation are included automatically.

Physics. The hadronic state has one incoming nucleon (with spectral-function weight) and one outgoing nucleon plus one pion. The current is computed in the frame where \(\boldsymbol{q}\) is aligned along the \(z\)-axis (frame = QZ), which simplifies the angular integration. The vector coupling is set by the FResV form factor; the axial coupling is set by FResA. The has_axial flag is true whenever FResA is non-zero, enabling axial contributions for weak interactions while switching them off for purely electromagnetic processes.

Form factors used: FResV, FResA (from the resonancevector and resonanceaxial categories).

Configuration. The ConfigFile uses the same two-line format as QE_Spectral_Func, supplying proton and neutron spectral function paths:

data/Spectral_Functions/pke12p_tot.data
data/Spectral_Functions/pke12n_tot.data

The same pre-supplied config files (data/info_C12_pke.data, etc.) can be re-used. No ModelParamsFile is needed.

Run-card example:

NuclearModels:
  - NuclearModel:
      Model: FortranModel
      Name: RES_Spectral_Func
      ConfigFile: data/info_C12_pke.data
      FormFactorFile: FormFactors.yml
      Ward: None

Intf_Spectral_Func — QE–MEC interference spectral function#

This model computes the interference between the one-body quasi-elastic amplitude and the two-body meson-exchange current (MEC) amplitude. It corresponds to nuclear mode Interference_QE_MEC (7) and requires its own phase-space generator (IntfSpectral), which samples both an active nucleon and a spectator nucleon from the nucleus.

Physics. The nuclear current is the cross-term between the one-body current \(J^\mu_{1b}\) (computed as in QE_Spectral_Func) and the two-body MEC current \(J^\mu_{2b}\). The two-body current receives contributions from both pion-exchange and \(\Delta\)-exchange diagrams, in direct and exchange topologies:

\[J^\mu_{2b} = \frac{1}{2 E_{p_2}}\Bigl( J^\mu_{\pi,{\rm dir}} + J^\mu_{\Delta,{\rm dir}} + J^\mu_{\pi,{\rm exc}} + J^\mu_{\Delta,{\rm exc}} \Bigr)\]

where \(E_{p_2}\) is the energy of the spectator nucleon.

Because the QE and MEC amplitudes share the same hadronic phase-space point, the model alternates between returning the one-body current and the two-body current on successive calls (controlled by an internal flag compute_1body). This paired-call structure is handled transparently by the Achilles framework.

Initial-state weight. The weight is the product of two spectral-function factors — one for the active (hole) nucleon and one for the spectator:

\[w = \mathcal{N}\,P(|\boldsymbol{p}_1|, E_{r,1}) \;\times\; \mathcal{N}\,P(|\boldsymbol{p}_2|)\]

where \(\mathcal{N}\) is the spectral function normalisation and the spectator factor is obtained by integrating over removal energy.

Form factors used: F1, F2, FA, FAP (one-body part); FMecV3, FMecV4, FMecV5, FMecA5, FPiEM (two-body MEC part).

Configuration. The ConfigFile lists the mean-field proton and neutron spectral function files (distinct from the total spectral functions used in the QE model):

data/Spectral_Functions/pke12p_MF.data
data/Spectral_Functions/pke12n_MF.data

Pre-supplied config files: data/info_intf_C12.data, data/info_intf_Ar40.data.

A ModelParamsFile is required for this model. It supplies the hadronic coupling constants and momentum cut-offs for the MEC vertex functions:

Parameters:
  fpind:  0.54      # f_{pi N Delta} coupling
  fstar:  2.13      # f* Delta coupling
  fpinn2: 1.01789   # f_{pi N N}^2 coupling
  ga:     1.26      # axial coupling g_A
  lpi:    1300.0    # Lambda_pi cut-off (MeV)
  lpind:  1150.0    # Lambda_{pi N Delta} cut-off (MeV)

The default parameter file data/intf_params.yml provides the values above, which were used in [LRS25].

Run-card example:

NuclearModels:
  - NuclearModel:
      Model: FortranModel
      Name: Intf_Spectral_Func
      ConfigFile: data/info_intf_C12.data
      FormFactorFile: FormFactors.yml
      ModelParamsFile: data/intf_params.yml
      Ward: None