Innovative Developments of Infrared Thermography for Non Destructive Evaluation

点击翻译按钮获取中文摘要

Infrared Thermography can be applied to the Non Destructive Evaluation (NDE) of materials and structures: it is called Infrared Thermography for NDE (or TNDE for short). Basically TNDE can be either passive or active. In the active mode, an external thermal stimulation is brought to the sample and response to this stimulus, recorded with an infrared camera, allows to determine the subsurface structure of the inspected component such as the presence of potential subsurface defects (delaminations, etc.). In the passive mode, isotherms are collected as is without perturbation of the thermal state of the object. The proposal concerns the active mode. Despite being increasingly popular, TNDE still presents challenges since many problems arise either at the stimulation stage or at the interpretation stage. The objectives of the proposed research program are oriented towards solution of these problems and they are in continuity with the long term goal of my research activities. The present proposed research program offers good potential for novel and significant research activities as follows: // - STIMULATION/OBSERVATION STAGES //: Objectives are to study in depth, through thermal modeling and experiments, the effects of various inspection stimulation schemes in order to enhance response from subsurface artifacts of interest. Proposal includes development of the following: • Eddy current thermography: exploit the eddy current injected for heating to get more details from a given flaw. The flaw is thus observed with two NDE schemes: thermography and «traditional» eddy current. • Cold Stimulation Thermography using a «cold» instead of a «warm» stimulation for particular applications (such as in restricted areas for which a warm approach is not possible due to over-heating). • Robotic deployment with adapted heating strategies. Robotic deployment is becoming the new TNDE frontier. // - INTERPRETATION STAGE //: Objectives consist to pursue the development of advanced processing methodologies to extract the relevant features of the acquired dataset to compute quantitative information (depth, size, thermal properties) and enable reliable interpretation of data. Signal processing is a key procedure in the analysis of thermal images. Proposal includes development of the following: • Partial Least Squares regression (PLS). PLS enables to decompose and subsequently reconstruct a dataset with the benefit to omit certain unwanted features (related to non-uniform heating/cooling). • Pulsed phase thermography (PPT). PPT explores the dataset in the frequency domain but looses the time information needed to access quantitative information (defect depth). Extension of PPT with Wavelet Transform (WT) is planned to circumvent this limitation. Next, PLS reconstruction will be combined with WT PPT. • Data fusion of signals from Eddy Current (EC) and Eddy Current Thermography to calculate the probability of an event (defect detection) thus improving the NDE diagnosis. EC is sensitive to surface cracks but has difficulties to detect subsurface defects for which thermography is good since thermal diffusion extend beyond the penetration of eddy currents. Work presented here is promising since it will bring fresh ideas to the TNDE community and industrial users while solving critical problems. Moreover, several HQPs will be involved at all stages of the proposed research activities. The significance of the proposed activities can be seen on two sides. The basis of the work has a strong theoretical foundation, while results of the proposed program are well applied, for instance in Canadian industries where contacts have already been established during previous research activities, in order to ensure valuable spins-off.