Fiber-optic-based distributed acoustic sensors (DAS) are a new technology that can be deployed in a well and are continuously interrogated during operations. These sensors measure the strain (or strain rate) at all points along the fiber and have been used extensively to monitor hydraulic stimulations. The data from these sensors indicate that they are sensitive to high-frequency signals associated with microseismicity and low-frequency signals associated with fracture growth. We have developed a set of idealized models to simulate these signals and to identify interpretation methods that may be used to estimate fracture location, geometry, and extent. We use a multiphysics code that includes rock physics, fluid flow, and elastic-wave propagation to generate synthetic DAS measurements from a set of simple models that mimic hydraulic fracturing. We then relate the signals observed in the synthetic DAS to specific features in the model such as fracture height, width, and aperture. Our results demonstrate that the synthetic DAS measurements may be used to interpret field DAS measurements and to optimize the design of future sensor deployments for sensitivity to fracture attributes.