Formal robustness analysis of failure detection and accommodation systems could play an important role in their validation and ultimate certification. However, formulation of linear fractional transformation (LFT) models can be very difficult for complex parameter-dependent systems. The paper presents a matrix-based computational approach that can be implemented in software for obtaining LFT models of complex systems involving parametric uncertainties, faults, and failures. Several examples are presented, including, a modeling framework for performing robustness analyses of faulty uncertain systems and failure detection and accommodation systems.