A nonlinearly constrained minimization problem can be solved by the exact penalty approach involving nondifferentiable functions $\sum_{i} |c_i(x)|$ and $\sum_{i}\max(0,c_i(x))$. In this paper, a trust region approach based on a 2-norm subproblem is proposed for solving a nonlinear $l_1$ problem. The (quadratic) approximation and the trust region subproblem are defined using affine scaling techniques. Explicit sufficient decrease conditions based on the approximations are suggested for obtaining a limit point satisfying complementarity, Kuhn-Tucker conditions, and second order necessary conditions. The global convergence analysis of the method is presented in \cite{Li94b}.