A CFD Digital Twin to Understand Miscible Fluid Blending.

Area of Science:

• Fluid dynamics
• Biopharmaceutical manufacturing
• Computational modeling

Background:

• Mixing stratified miscible fluids with differing properties is crucial in biopharmaceutical manufacturing.
• Density and viscosity differences significantly increase blend times compared to single-fluid systems.
• Mixing performance in two-fluid systems is highly sensitive to the Richardson number.

Purpose of the Study:

• To develop accelerated digital twins for simulating physical mixing tanks.
• To predict real-time fluid mechanics with high fidelity and lower cost than physical testing.
• To explore the physics of fluid blending in stratified two-fluid systems and provide practical guidance.

Main Methods:

• Utilizing lattice Boltzmann transport algorithms combined with graphics processing unit (GPU) computing.
• Building and validating accelerated digital twins against experimental mixing data.
• Employing digital twins to investigate the influence of fluid properties and Richardson number on mixing.

Main Results:

• Digital twins accurately simulated single- and multi-fluid mixing processes.
• The models provided insights into the physics governing stratified two-fluid blending.
• Validated digital twins offer a cost-effective alternative to physical experimentation for process optimization.

Conclusions:

• Accelerated digital twins are powerful tools for understanding and optimizing biopharmaceutical mixing.
• The developed models can guide stratified two-fluid mixing processes and inform the creation of similar models for other systems.
• This approach significantly reduces the cost and time associated with physical testing and process development.