The field of materials science and engineering (MSE) is based on the fundamental principle that microstructure controls properties. Traditionally, the study of material structure has been limited by sectioning and post-mortem observations. This approach is often inaccurate or inadequate for solving many fundamental problems. It is also often laborious and time-consuming. Advances in experimental methods, analytical techniques, and computational approaches, have now enabled the development of in situ techniques that allow us to probe the behavior of materials in real-time, i.e., 4D materials science. The study of microstructures under an external stimulus (e.g., stress, temperature, environment) as a function of time is particularly exciting.

X-ray micro and nano-tomography provides a wonderful means of characterization damage in materials non-destructively. In this talk, I will describe experiments and simulations that address the critical link between microstructure and deformation behavior of metallic materials, by using a three-dimensional (3D) virtual microstructure obtained by x-ray synchrotron tomography. The approach involves capturing the microstructure by novel and sophisticated in situ testing in the lab or using a synchrotron facility, followed by x-ray tomography and image analysis, and 3D reconstruction of the microstructure.

Image processing and segmentation of 4D tomographic data is extremely time intensive. Moreover, factors such as phase transformation or defect propagation during a timeevolved tomography experiment limits the scan time and/or number of scan iterations. Thus, there is also a need to establish robust algorithms that can render time-dependent xray datasets accurately and efficiently. I will describe the application of Deep Convolutional Neural Network (DCNN) and modified Generative Adversarial Network (GAN) algorithms for X-ray image quality enhancement and segmentation. Our results point to the ability to drastically reduce x-ray data acquisition times, thereby opening a window for efficient 4D experiments.