Contents lists available at ScienceDirect NDT and E International journal homepage: www.elsevier.com/locate/ndteint Characterization of microstructural anisotropy in pearlitic steel with mode- converted ultrasonic scattering Hualong Du a , Joseph A. Turner b , Ping Hu c,* a Applied Research Associates, Inc., Littleton, CO, 80127, USA b Mechanical and Materials Engineering, University of Nebraska-Lincoln, W342 Nebraska Hall, Lincoln, NE, 68588, USA c School of Power and Mechanical Engineering, Wuhan University, Wuhan, China ARTICLE INFO Keywords: Ultrasonic scattering Pearlitic steel Microstructural anisotropy Mode-converted Duplex microstructure ABSTRACT A mode-converted (longitudinal-to-transverse, L-T) ultrasonic scattering technique was applied to evaluate the variation of microstructural anisotropy in a railroad wheel sample. The anisotropy was examined using a pitch- catch longitudinal-to-transverse measurement configuration in two perpendicular directions. Then the whole scan area was divided into 14 sections to calculate the variance of scattered signals for each section. The variance curves were fit with the theoretical models to determine amplitudes for each section. The experimental results show the variance amplitudes of the L-T response are very similar for two perpendicular directions near the tread surface which indicates microstructural isotropy. For measurement areas near the center of the sample from the tread surface, the variance amplitudes in two perpendicular directions split and the difference increases with depth. The result indicates growing microstructural anisotropy in regions associated with the coarse duplex pearlite microstructure. The fine and coarse duplex microstructures have already been examined by optical analysis, but the microstructural anisotropy cannot be characterized from metallographic images. The mode- converted ultrasonic scattering approach presented in this paper provides a nondestructive evaluation (NDE) method for characterization of microstructural anisotropy in complex metallic materials. 1. Introduction Ultrasonic scattering resulting from internal inhomogeneities in structural materials and biological tissues has a significant negative effect on ultrasonic imaging in medical and nondestructive evaluation applications. However, the scattered signals carry very important in- formation about the sample microstructure and can be used to quantify the length scales associated with grains or intergranular microstructure if properly modeled [1,3–5]. Because most polycrystalline materials have randomly oriented grains, acquired backscatter signals have a large spatial dependence. Thus, statistical models are usually employed to process the measured signals. Several models have already been proposed to study ultrasonic scattering in polycrystalline media [1–4]. Rose [5,6] developed the first backscatter model for polycrystals with equiaxed grains. Han and Thompson [7] expanded the backscatter model for titanium alloys with duplex microstructure. Lobkis, et al. [8,9] and Yang, et al. [10] presented an ultrasonic backscatter model in polycrystals with elongated single phase and duplex microstructures. Ghoshal, et al. [11,12] developed a time-dependent spatial variance model of ultrasonic backscatter with respect to an assumption of a singly scattered response (SSR) to microstructural properties. Based on the time-dependent spatial variance model developed by Ghoshal, et al. [11,12] and the ultrasonic backscatter model for materials with duplex microstructure presented by Lobkis, et al. [8,9], Du, et al. [13,14] de- veloped an ultrasonic backscatter model for pearlitic steel with lamellar duplex microstructure. All these ultrasonic scattering models were de- veloped for backscatter experiments in the pulse-echo configuration. The majority of research to date has been devoted to the longitudinal- to-longitudinal (L-L) approach. However, the diffuse scattering field also includes other scattering modes besides the L-L mode. Weaver [2] first developed the necessary inner products between a covariance tensor containing elastic modulus fluctuations and scattered wave vectors to describe arbitrary scattering modes. Recently, Hu, et al. [15] developed a time-dependent mode-converted (longitudinal-to-trans- verse, L-T) ultrasonic scattering model in a pitch-catch configuration to characterize grain size in polycrystalline media. The results show that the grain size of the steel with equiaxed grains measured by the L-T measurement is equal to that measured from the L-L measurement. One advantage of the L-T measurement over the traditional L-L measure- ment is that the ultrasonic transverse scattering in a specific direction https://doi.org/10.1016/j.ndteint.2018.11.016 Received 8 June 2018; Received in revised form 15 October 2018; Accepted 22 November 2018 * Corresponding author. E-mail addresses: hualongdu@gmail.com (H. Du), jaturner@unl.edu (J.A. Turner), ping.hu@whu.edu.cn (P. Hu). NDT and E International 102 (2019) 189–193 Available online 23 November 2018 0963-8695/ © 2018 Elsevier Ltd. All rights reserved. T