Design and validation of a versatile thermomechanical fatigue test system retrofit kit

Abstract

This paper presents the design and experimental validation of a low-cost thermomechanical fatigue (TMF) test system that can be retrofitted to a dynamic load universal test machine with minimal modifications. The motivation for this study is to fabricate a cost-effective and universally compatible thermomechanical fatigue testing system that has the ability to evaluate ferromagnetic aerospace materials. Typically, TMF test systems are available from an original equipment manufacturer (OEM) for a complete system or as a retrofit kit, costing as much as $500,000 or $300,000, respectively. Also, OEM equipment is not interchangeable with other systems and requires manufacturer specific components to operate. An analytical study was performed to design and size components to meet American Society for Testing and Materials test standard E2368 with minimal cost. The resulting design consists of an Arduino micro controller that reads instrumentation signals and a package of components that do not rely on interfacing with external software or hardware to perform a thermomechanical fatigue test. The specific test system hardware components consist of a 10 kW induction heater, industrial chiller for induction coil operation, cooling flow nozzles coupled with an air compressor, thermocouples mounted outside the gauge section, an infrared pyrometer for temperature monitoring within the specimen gauge section, and an axial extensometer to synchronize the thermal and mechanical loading cycles. Validation of the system is completed by evaluating the synchronization of thermal and mechanical loading cycles over a test matrix with varying cycle times and loading parameters. The system operation follows a set mechanical load and alters the thermal profile in response to match the profiles. Initial experiments were conducted to tune heating and cooling process parameters to achieve minimal load variance between mechanical and thermal loading profiles. Subsequent experiments characterized response time variance at different loading cycle frequencies and amplitudes. From experimental testing, it was determined the system control scheme is correctly designed with existing components to maintain synchronization within a certain frequency band and amplitude band for an indefinite number of cycles. This study enables development of fatigue model validation for predicting needed preventative maintenance, and improving supply chain management. Further recommendations for this system may include an improved induction heater that uses digital active control to regulate power output magnitude and an improved air cooling system to regulate air flow rate in real time.