Process Optimization via Numerical Experiment
CFD modeling provides time-accurate, three-dimensional insights into fluid motion, reaction chemistry, particle trajectories, heat transfer, and free surface dynamics. The development of new fluid modeling algorithms, combined with modern computational architectures, now enable engineers to simulate the performance of fluid handling equipment with a fidelity that rivals what can be measured experimentally. PROCESS utilizes a software package called M-Star CFD, which packages these algorithms into Windows-based pre- and post-processing user interfaces. This now allows for complex systems commonly found in the process industries to be efficiently, accurately, and cost effectively analyzed in such a way as to significantly increase the confidence factor that processing systems will perform as expected. Just a few examples of where we can use CFD:
- Multi-Phase Newtonian and Non-Newtonian Fluids (Liquid/Liquid, Liquid/Solid, Liquid/Vapor, Solid/Liquid/Vapor, etc.)
- Mixed tanks
- Agitator Performance
- Velocity Profiles, Shear Rate, etc.
- Vortex Formation
- Surface Dynamics
- Simulated Dye Injection
- Dead Spots
- Blending Time/Mixing Uniformity
- Solid Dispersion
- Gas Sparging
- Reactant or Impurity Concentrations
- Impact of Internals (Baffles, Internal Coils, etc.)
- Agitator Performance
- Flow profile through piping, orifice, mixing baffles
- Flashing Liquids
- Heat transfer / mass transfer systems
- Chemical reactions
- Much, much more.
Simplistically speaking, PROCESS first creates a geometric model (3D CAD model) of the system to be studied (mixed tank, piping system, sparged reactor, etc.). Properties information for the fluids/solids/gases in the system are entered, as are dynamic aspects such as flow rates, impeller speeds, etc. View some of our CFD Projects here.
The M-Star software then uses direct numerical simulations (DNS) for laminar and transitional systems, and large eddy simulations (LES) for turbulent systems to run a solution for the system as previously defined. The robustness of these time-accurate approaches eliminates the ambiguity associated with traditional turbulence models. This software modeling captures transient flow patterns that cannot be resolved using time-averaged CFD, and correctly predicts the effects of scale-up on flow dynamics.
Upon completion of a case run, the software generates open source output files (ASCII output data and VTK output) that are easily translated into videos and pictures for accurate visualization of the results and/or numerical evaluation of data points. This allows the results of the simulation to be brought to life with real-time animations and output. Output data may consist of information such as automated fluid mechanics output (velocity, strain, vorticity, energy dissipation, rates (EDR), time-average velocity, pressure, kinetic energy, temperature), solid mechanics output (transient force distribution on all immersed objects, transient torque on all immersed objects, power number for rotating equipment, pumping number for rotating equipment), and other system outputs (mean age, residence time distributions (RTD), blend times).