(431f) Comparison of Blend Time Measurements: An Experimental and Computational Study

The understanding and prediction of blend times in agitated vessels is industrially important for many applications (e.g. chemical reactions, thermal homogeneity, additive blending). The measurement of blend time can be done in different ways, depending on the application. Understanding the measurement techniques and being able to interpret the information from them is important. Depending on the application, one or more of these techniques can be used to successfully measure the blend time.

Most work in estimating blend times is done with “clean” fluids that minimize noise in the measurement signal. Techniques for measuring blend times include decolourization via chemical reaction (e.g. acid/base with an indicator), colourization (e.g. adding a dye), and temperature or concentration monitoring via probes at particular points in a vessel. Comparing the values of these different techniques can be challenging because of the stochastic nature of turbulent blending and various definitions of “blended”. Examining different measurement techniques experimentally simultaneously is time consuming and challenging from an experimental standpoint. With recent increase in computing power, high fidelity CFD simulations that were strongly in the academic domain a decade ago are now available as engineering tools. Using these types of tools to examine blending can help understand the relationships between various types of blending measurements.

The work to be presented will include both experimental and computational elements. Data from experimentation and published literature will be compared with CFD results to form a basis for validation of the CFD. Once validation, the CFD can be used to examine the several different blending criteria simultaneously. This includes acid/base reactions with an indicator, REAX by FMP, and probe-based responses. Because the simulation allows the concentration to be known at all locations within the vessel (with limits at the simulation resolution), comparisons to the uniformity of the entire “population” of the vessel positions will also be done