DETECTION OF MICRO-CRACKS ON METAL SURFACES USING NEAR-FIELD MICROWAVE RESONATORS J. Kerouedan 1,2 , P. Quéffélec 1 , P. Talbot 1 , C. Quendo 1 , S. De Blasi 1 , and A. Le Brun 2 1 Lab-STICC (UMR CNRS 3192), Université de Bretagne Occidentale, CS 93836, 6 avenue Le Gorgeu, 29238 Brest Cedex 3, France. 2 EDF R&D / STEP department, 6 quai Watier, BP 49, 78401 Chatou Cedex, France. ABSTRACT. This paper demonstrates how micro-cracks at the surface of metals can be detected and imaged by using near-field microwave resonators. It deals with two novel sensors: a first-order dual-behavior resonator (DBR) band-pass filter probe and a first-order DBR filter with an opening in the ground plane. Measurements results carried out on a stainless steel mock-up with several 200 μm wide EDM rectangular surface notches are presented showing the high sensitivity of the DBR probes and their ability to differentiate between notches of different depths. Keywords: near-field microwave resonators, microstrip dual-behavior resonator (DBR) filters, surface defects on metals, sensitivity. PACS: 07.57.-c, 81.70.Ex INTRODUCTION The fatigue and ageing of metal materials under operation conditions are major concerns in energy production plants. An early and nondestructive diagnostic of surface defects would allow one to carry out relevant preventive maintenance operations without dismantling or prematurely changing healthy components. Today, most of the automated nondestructive evaluation (NDE) solutions available to detect the surface-breaking defects are based on ultrasound or eddy current techniques. Despite their high sensitivity and resolution, they are unable to meet all the requirements of every real situation. Eddy currents are sensitive to different external parameters what sometimes makes signal analysis difficult and ultrasounds are not suitable for the detection of small depth (with a depth below 3 mm) defects located near the inspection surface. Consequently, it sounded us relevant to evaluate the potential of microwave-based techniques to detect the surface defects with a depth shallower than 3 mm. Over the last years, far- and near-field approaches were investigated to detect surface defects. With a far-field characterization [1], the spatial resolution is of the order of a half wavelength (λ/2). So, to detect micro-cracks it is necessary to work at very high frequencies, which causes high measurement equipment costs and a reduction in the sensor size. In addition, at very high frequencies, signal-to-noise ratio (SNR) issues can appear.