Abstract
Most reliability models assume components and systems experience one failure mode. Several systems such as hardware, however, are prone to more than one mode of failure. Past two failure mode research derives equations to maximize reliability or minimize cost by identifying the optimal number of components. However, these equations are derived from models that make the simplifying assumption that components fail in a statistically independent manner. This paper develops models to assess the impact of correlation on two failure mode system reliability and derives corresponding expressions for optimal reliability designs. Our illustrations demonstrate that, despite correlation, the approach identifies reliability optimal designs.