Granular mixing has a significant influence on the resulting products in a variety of industries, such as pharmaceutical production, food processing, and energy generation. Most commonly, granular mixing processes seek a high degree of homogeneity, and in some particular instances, the purpose is to influence simultaneous processing, such as chemical reactions and heat and/or mass transfer rates. Granular mixing is a complex process because it exhibits multiple rheological differences, many times simultaneously within the same granular bed. A fundamental problem commonly encountered during the mixing of granular materials is the tendency for mixtures to segregate due to differences in particle size, shape, and/or density. While significant effort has been made to improve the granular mixing process, the vast majority of granular mixing research efforts have focused on relatively simple mixer geometries and idealized granular materials compared to what is actually used in industrial practice. For example, the renewable energy industry, or more specifically the biomass thermochemical conversion sector, relies on double screw pyrolyzers to convert cellulosic biomass into bio-oil. Currently, there is a large amount of research that relates the feedstocks being used to the bio-oil yields; however, minimal research efforts have been directed toward studying the granular mixing dynamics inside the double screw pyrolyzer due to the opaque nature of the reactor and the granular flow itself.

The overall goal of this project is to thoroughly understand, characterize, and optimize the granular mixing of low density red oak chips and high density glass beads within a double screw mixer, which geometrically replicates screw pyrolyzers, under various operating conditions. Moreover, it aims to provide an improved understanding of the biomass and heat carrier media granular mixing dynamics inside the screw pyrolyzer. The increased mixing effectiveness and insight gained through this project will lead to increased heat transfer rates and could enhance product yield and/or quality; thereby improving the economic viability of a screw pyrolyzer.

Throughout this project, a number of different visualization and quantification techniques have been developed and used to evaluate the mixing effectiveness of red oak chips and glass beads inside a screw mixer that was designed and constructed specifically for granular mixing studies. Four parameters were initially investigated: (i) screw rotation speed, (ii) dimensionless screw pitch, (iii) screw rotation orientation, and (iv) material injection configuration. After identifying the influence that these parameters had on the mixing effectiveness, the effect of the dimensionless mixing length was also investigated.

Advanced optical visualization methods were first developed by capturing, and then spatially aligning and temporally syncing four independent projections of the screw mixer. Next, an improved granular sampling procedure which satisfies the two "golden rules of sampling" was developed. Composition analysis of the samples was then coupled to analysis of variance (ANOVA) statistical methods to qualitatively assess the spatial heterogeneity of the granular mixing process. Together, the qualitative optical visualization and quantitative composition and statistical analysis lead to the optimization of the screw mixer's operating conditions. For the range of parameters that were studied, these conditions were determined to be a screw rotation speed of ω = 60 rpm, a dimensionless screw pitch of p/D = 1.75, a counter-rotating down-pumping screw rotation orientation (CtrR DP), and a material injection configuration with the red oak chips and glass beads injected into port one and two (RO 1, GB 2), respectively. Studies which investigated the mixing effectiveness of the screw mixer as a function of the dimensionless mixing length also indicated that this operating condition lead to the most homogeneous mixture at each of the tested dimensionless mixing lengths.

An improved cone-beam compensated back-projection algorithm enabling X-ray particle tracking velocimetry (XPTV) of the screw mixer was then developed to visualize and quantify the 3D granular flow structures inside the screw mixer under various operating condition using "modified" red oak chip tracer particles. It was shown that these improved XPTV methods provided a significant reduction in error associated with the tracer particle's position. The influence of the aforementioned four factors on the granular flow structures inside the screw mixer were investigated using XPTV, and similar results in terms of the desired operating conditions to those from the previous studies were noted. Moreover, the counter-rotating down-pumping screw rotation orientation, which was found to be the most influential factor using the composition and statistical analysis methods, was also shown to increase the material residence time, resulting in a longer mixing time.

A number of different methods were developed and used to characterize the granular mixing process inside the screw mixer. While each of the methods provided a different view on mixing behavior, the studies converged on a single double screw mixer operating condition, leading to a conclusive mixing effectiveness optimization.