Industrial research is continuously increasing efforts in designing new-tailored light-weight materials in order to meet the high demands regarding efficiency, environment, safety as well as comfort. Especially in the aeronautics industry a high demand for advanced composite materials is observable. The new generations of aircrafts are made of more than 50 % of these novel composite materials. Carbon fiber reinforced polymers (CFRPs) are currently considered as the most promising candidate since this material is outperforming the majority of conventional materials. As a result of the manufacturing process this material tends to have pores inside. Pores in the material are typically inclusions of air. As they have an impact on the mechanical proper- ties of the component, their determination and evaluation is an important task in quality control and a particular challenge for non-destructive testing (NDT) practitioners. Besides the characterization of individual pores, their spatial distribution in the tested component is a relevant factor. For example, a high concentration of pores in certain regions leads to different material characteristics as compared to a homogenous distribution of the pores. This work is based on 3D X-ray Computed Tomography (XCT) to gain new insight into CFRP components. Based on domain experts-questions, specific tasks were derived. Besides the quantitative porosity determination, the main visualization tasks are: giving a fast porosity overview, exploring the individual pores, and tracking features over time based on XCT time-series. In this thesis, three novel visual analysis tools are presented to solve these tasks. To enhance the evaluation workflow for non-destructive testing (NDT) practitioners, a visualization pipeline for the interactive exploration and visual analysis of CFRP specimens is developed. After the calculation of local pore properties, i.e., volume, surface, extents and shape factors, a drill-down approach is employed to explore pores in a CFRP specimen. Therefore Porosity Maps (PM) are presented to allow for a fast porosity overview and selecting a region of interest. Pores in this region may be filtered and visualized with a parallel-coordinates selection. Furthermore a novel visualization technique which allows for a fast porosity overview and exploration of pores by focusing more on their shapes is proposed. In this method, all objects (pores) are clustered into a Mean Object (MObject). To explore this MObject, the visualization of mean object sets (MObject Sets) in a radial and a parallel alignment is introduced. By selecting a specific property such as the volume or shape factor and the desired number of classes, a MObject is split up into sub-classes. With this approach, intended classifications and visualizations of MObjects may be explored by the user. These representative MObjects may be exported as volumetric datasets to serve as input for successive calculations and simulations. For an overview of the pore properties in the dataset local MObjects are calculated in a grid and combined with a color-coded homogeneity visualization. Both approaches were evaluated with real-world CFRP specimens. To go one step further, time as a fourth dimension is added to analyze a process over time, e.g., how the features evolve and formate over time. Therefore features in a series of XCT scans are tracked with the Fuzzy Feature Tracking approach and are then visualized together with the extracted events in multiple linked-views, each emphasizing individual aspects of the 4D time-series data. Spatial feature information, global temporal overview, and global temporal evolution of how the features are tracked and connected over the whole time-series are covered with the visual-analysis system. The results and advantages of the Fuzzy Feature Tracking tool are demonstrated using various real-world applications, such as AlSiC alloys under thermal load or wood shrinkage analyses.