IMPROVEMENTS ON FEA WITH A TWO-STEP SIMULATION OF EXPERIMENTAL PROCEDURES IN TURBINE BLADE CRACK DETECTION IN SONIC IR NDE

We showed our work on modeling turbine blade crack detection in Sonic Infrared (IR) Imaging with a method of creating flat crack surface in finite element analysis (FEA) in last year's QNDE paper. This modeling has been carried out continuously as part of model-assisted study on crack detection in aircraft engine turbine blades. We have presented that Sonic IR Imaging NDE is a viable method to detect defects in various structures. It combines ultrasound excitation for frictional heating in defects and infrared imaging to sense this heating, and thus to identify the defects. It is a fast wide-area imaging technology. It only takes a second to image a large area of a target sample. When an aircraft is in flight, the turbine engine blades operate under high temperature and high cyclic stresses. Thus, fatigue cracks can form after many hours of operation. Sonic IR Imaging can be used to detect such cracks. However, we still need to better understand contributions of parameters/factors in the crack detection process with Sonic IR Imaging. FEA modeling can help us to reveal certain aspects through the data it produces where experimental work cannot achieve. Upon the model we presented last year, a two-step simulation process was designed to simulate the important aspects in our experiments. These include a newly designed model for the ultrasound transducer which delivers mechanical energy to the sample and the implementation of static force while engaging the transducer to the sample. In this paper, we present the ideas and the results from the new model.