Data driven prognosis and diagnosis are highly dependent on the quantity and quality of system operation data, including fault data. In order to obtain gas-path fault data of a turboshaft engine, a static nonlinear mathematic model of the engine is established by adopting the component models method. The change tendency of component performance caused by the gas path fault is studied. The component performance map is modified by degradation factors simulating different component faults. Then, methods for simulating the combined fault and various degrees of fault severity are investigated. Based on the mathematic model and fault simulation methods, the impact of various gas-path faults, degrees of fault severity, and combined fault on the total pressure, total temperature and output power are simulated, respectively. Analysis of the simulation reveals some features of the engine gas-path fault. The feasibility of this fault-simulation method is confirmed, and a method to obtain the faults data for data-driven engine gas-path fault diagnosis and prognosis is provided.