(423d) CFD Validation in Gas-Liquid Systems Using Flow Flowing Sensors

There is a constant need to validate computational fluid dynamics models against real data to test and improve methods. Many means of validating CFD against experiments have been proposed and utilized, but few allow testing in industrial scale operations. By using flow-following sensors, neutrally buoyant probes approximately 30 mm in diameter, we have been able to measure position, velocity, and scalar fields at industrial scale in gas/liquid systems, with sampling periods of multiple days, on actual systems, during normal operations. For sparged tanks, with and without agitation, we measured the position, velocity, and scalar fields with a few probes for a few days of operation. We have also simulated these systems using a Lattice Boltzmann Method, with hundreds of probes, collecting data for many probes over shorter periods of time, to produce datasets of similar size. Appealing to the principle of ergodicity, we compare the distributions of probe positions, velocities, and probe-circulation times. We find that the simulated and experimental position, velocity, and circulation time distributions match with an R-squared of 0.90 or greater. Analyzing the datasets as distributions also provides a simple means of comparing different operating conditions, such as operation with and without agitation. It similarly provides means of comparing differences from model formalism, such as different choices of breakup and coalescence models in sparged systems.