1: Boltzmann-Uehling-Uhlenbeck (BUU) transport models are based on the Boltzmann equation, which was modified by Nordheim, Uehling and Uhlenbeck to incorporate quantum statistics. A brief description of the formalism is given below. \new{For a detailed discussion concerning the physical input for pion-induced reactions we refer the reader to~\cite{Buss:2006vh} and the references therein. A more general description of the whole GiBUU model will be given in a forthcoming paper~\cite{GiBUUPaper}.}
2:
3: \subsection{The BUU equation}\label{buuEQ}
4: The BUU equation actually consists of a series of coupled differential equations, which describe the time evolution of the \newla{single}-particle \newla{phase-space densities $f_a(\vec{r},\vec{p},t)$. The index $a=\pi,\omega,N,\Delta,\ldots$ denotes the different particle species in our model.} A large number of mesonic and baryonic states is actually included, but at the energies of interest for this study, the relevant ones are $\pi,N$ and the $\Delta(1232)$ resonance.
5:
6: For a particle of species X, its time evolution is given by
7: \begin{eqnarray}
8: \frac{d f_X(\vec{r},\vec{p},t)}{d t}&=&\frac{\partial f_X(\vec{r},\vec{p},t)}{\partial t} +\frac{\partial H_X}{\partial \vec{p}}\frac{\partial f_X(\vec{r},\vec{p},t)}{\partial \vec{r}}
9: -\frac{\partial H_X}{\partial \vec{r}}\frac{\partial f_{X}(\vec{r},\vec{p},t)}{\partial \vec{p}} \nonumber \\ &=&I_{coll}\left(f_X,f_a,f_b,\ldots\right)
10: \label{BUUEquation}
11: \end{eqnarray}
12: with the one-body Hamilton function
13: \begin{eqnarray}
14: %H_X(\vec{r},\vec{p})&=&\sqrt{\left(\vec{p}+\vec{A}_X(\vec{r},\vec{p})\right)^{2}+m_X^{2}+U_X(\vec{r},\vec{p})}\nonumber\\ && +A_X^{0}(\vec{r},\vec{p})
15: H_X(\vec{r},\vec{p})&=&\sqrt{\left(\vec{p}+\vec{A}_X(\vec{r},\vec{p})\right)^{2}+m_X^{2}+U_X(\vec{r},\vec{p})}\nonumber +A_X^{0}(\vec{r},\vec{p}) \; .
16: \label{HamiltonFunc}
17: \end{eqnarray}
18: The scalar potential $U_X$ and the vector potential $A^\mu_X$ of species X may in principle depend upon the phase space densities of all other species. Hence, the differential equations are already coupled through the mean fields. In the limit of $I_{coll}=0$, eq.~(\ref{BUUEquation}) becomes the well-known Vlasov equation. The collision term $I_{coll}$ on the right-hand side of eq.~(\ref{BUUEquation}) incorporates explicitly all scattering processes among the particles. The reaction probabilities used in this collision term are chosen to match the elementary collisions among the particles in vacuum.
19: Within the BUU framework \newla{the $\pi N$ reaction cross section is given by an incoherent sum of resonance contributions and a direct, i.e. point-like, contribution.} Interference effects are therefore neglected.
20:
21: \subsection{Elementary processes}
22: \begin{figure}[b]
23: \begin{center}
24: %\includegraphics[angle=0,width=9cm]{./BUUData/potentials/deltaPotential.eps}
25: \includegraphics[]{./data/elementary/PlotEle.eps}
26: \end{center}
27: \caption{The angular distributions for the charge exchange process $\pi^- p \to \pi^0 n$ in the CM-frame of pion and proton. The plots are labeled by the kinetic energies of the pions in the laboratory frame. The data are taken from ref. \cite{Sadler:2004yq}.}
28: \label{eleX}
29: \end{figure}
30:
31: DCX emerges from the interplay of two elementary mechanisms: pion-nucleon quasielastic scattering, with or without charge exchange, and pion absorption inside the nucleus. In our model, the cross section for quasi-elastic scattering is given by an incoherent sum of background $\sigma^{\BG}$ and resonance contributions
32: \begin{eqnarray}
33: \sigma_{\pi N \rightarrow \pi N}&=&\sigma_{\pi N \rightarrow R \rightarrow \pi N}+\sigma^{\BG}_{\pi N \rightarrow \pi N} \; . \nonumber \label{backDef}
34: \end{eqnarray}
35: The resonance cross sections are obtained from the partial wave analysis of ref.~\cite{ManleySaleski}. The background cross sections denoted by $\sigma^{\BG}$ are chosen in such a manner, that the elementary cross section data in the vacuum are reproduced. Background contributions are instantaneous in space-time, whereas the resonances propagate along their classical trajectories until they decay or interact with one or two nucleons in the medium.
36:
37: As an improvement in the earlier treatment \cite{Buss:2006vh,Effenberger:1996rc,Engel:1993jh}, we now included a more realistic angular distribution for the elastic scattering of the pions. Due to the $P$-wave nature of the $\Delta(1232)$ resonance, we assume for $\pi N \to \Delta \to \pi N$ in the \new{resonance} rest frame a distribution of the pion scattering angle $\theta$ according to
38: \[ f^\Delta(s,\theta)=\left(1+3\cos^2(\theta) \right) g(s,\theta) \]
39: which is peaked in forward and backward scattering angles. The function $g(s,\theta)$, depending on Mandelstam $s$, parameterizes the energy dependence of the $\pi N$ angular distribution. In a coherent calculation the angular distribution is generated by interference effects, which can not be accomplished by our transport model. In our ansatz we need to split the cross section in an incoherent way to preserve our semi classical resonance picture. Therefore we take
40: \[
41: g(s,\theta)=\left(\alpha-\cos(\theta) \right)^{\beta \left(m_\Delta-\sqrt{s}\right)/m_\Delta}
42: \]
43: with the $\Delta$ pole mass $m_\Delta=1.232 \GeV$. For the background events we assume
44: \[ f^\BG(s,\theta)=g(s,\theta) \; . \]
45: The constants $\alpha=1.9$ and $\beta=26.5$ are fitted to the angular distributions measured by Crystal Ball~\cite{Sadler:2004yq}; a comparison of our parameterization to this data is shown in \fig{\ref{eleX}}.
46:
47: Concerning pion absorption, the most important mechanisms are a two step processes in which $\pi N \to \Delta$ is followed by $\Delta N \to NN$ \new{or $\Delta N N \to NNN$}, and a one step background process $\pi N N \to N N$.
48: As medium modifications, we include Pauli-blocking and Fermi-motion of the nucleons, Coulomb forces and hadronic potentials for baryons; we account as well for the collisional broadening in the $\Delta$ resonance width. \new{In \cite{Buss:2006vh} we discussed the influence of the real part of the pion self energy on absorption processes in nuclei. Since DCX is mostly sensitive to the surface (see the discussion in sec. \ref{resu}), the influence of this real part has turned out to be negligible due to the low effective density and has not been included in the calculations presented here.}
49: For details see ref. \cite{Buss:2006vh} and the references therein.
50:
51: \subsection{Nuclear densities} \label{densSec}
52: The nucleons are initialized in a local density approximation.
53: For $^{16}_{8}O$, $^{40}_{20}Ca$, $^{208}_{82}Pb$ we have implemented density profiles $\rho(r)$ according to the parameterizations collected in ref.~\cite{osetPionicAtoms}, which are of Woods-Saxon type for heavier nuclei ($Ca$, $Pb$) and of harmonic-oscillator type for lighter ($O$) ones. The proton densities are based on the compilation of ref.~\cite{DeJager:1974dg} from electron \newla{scattering. The neutron densities are provided by Hartree-Fock calculations.
54: For $^{103}_{45}Rh$, we use a Woods-Saxon density distribution
55: \begin{eqnarray}
56: \rho_{n}(r)&=& \frac{\rho_n^0}{1+exp((r-R_n)/a_n)} \nonumber \\
57: \rho_{p}(r)&=& \frac{\rho_p^0}{1+exp((r-R_p)/a_p)}\; ,
58: \label{densWS}
59: \end{eqnarray}
60: with the parameters given in table I~\cite{lenske}. }
61: \begin{table}
62: \begin{center}
63: \begin{tabular}{cc|cc|cc}
64: $\rho_p^0 [fm^{-3}]$& $\rho_n^0 [fm^{-3}]$ & $R_p [fm]$ & $R_n [fm]$ & $a_p [fm]$ &$a_n [fm]$\\
65: \hline
66: 0.0708 & 0.0835 & 5.194 & 5.358 & 0.4743 & 0.4780
67: \label{WS_Rh_table}
68: \end{tabular}
69: \caption{Parameters of the Woods-Saxon parametrizations for $_{45}^{103} Rh$.}
70: \end{center}
71: \end{table}
72: The larger neutron radii of heavy nuclei play a relevant role in DCX as we show in Sec.~\ref{resu}.
73:
74: