Grain Drying Device: 2-way DPM Model

Grain Drying Device CFD Simulation Using Two-Way DPM Model, ANSYS Fluent Training

Description

This project studies a grain drying device using the two-way Discrete Phase Model (DPM) combined with the Species model in ANSYS Fluent. Hot air enters the drying device, and 120,000 rice grains carrying 10% moisture are injected randomly over a 6-second period, with evaporation continuing for an additional 9 seconds after injection completes. The device's hot surfaces are maintained through contact with hot exhaust smoke from an engine, providing the elevated temperature conditions needed for efficient drying.

Freshly harvested rice typically carries 20-30% moisture — a level that can corrupt the grains quickly if left untreated. Drying the grain before storage and milling is therefore essential, and this rice drying device provides a mechanical means of exposing grains to ambient hot air to accelerate moisture evaporation.

The 3D geometry was built in Design Modeler, representing a 3 m × 1 m channel box containing four triangular passages, each 20 cm long. The domain was meshed in ANSYS Meshing using an unstructured grid totaling 556,145 elements.

Methodology

Several assumptions were applied to this simulation: it was run as transient (unsteady) to capture the time-dependent behavior of both the fluid and the injected particles, a pressure-based solver was used given the working fluid's incompressibility, the two-way DPM tracked the injected rice grain particles under the desired conditions, and gravitational acceleration was included at -9.81 m/s² in the y-direction.

Key simulation settings included:

• Models: Energy equation enabled; Realizable k-epsilon viscous model with standard wall functions; Species Transport model using a mixture-template; DPM with continuous-phase interaction and unsteady particle tracking both enabled

• Injection: Surface-type injection at the inlet, 10,000 streams, modeled as evaporating water-liquid droplets (10% volatile component fraction) with a uniform diameter distribution of 0.005 m, injected at 363.15 K with a total flow rate of 0.5 kg/s over a 0–6 second injection window

• Boundary conditions: Velocity inlet at 1 m/s, 5% turbulent intensity, 363.15 K, with DPM set to escape; outlet wall with DPM set to reflect; hot walls held at 773.15 K with DPM set to reflect

• Solution methods: SIMPLE pressure-velocity coupling, least-squares cell-based gradient scheme, second-order discretization for pressure, momentum, H₂O, and energy, and first-order upwind for turbulent kinetic energy and dissipation rate

• Run settings: Time step size of 0.05 s, 300 total time steps, maximum 40 iterations per time step

Conclusion

The simulation tracks 10,000 injected rice grains over the 6-second injection period, with evaporation beginning almost immediately (though simplified/ignored during this initial injection phase for modeling purposes). The resulting H₂O mass fraction graph shows moisture peaking at approximately 0.015 at the end of injection, then declining steadily to zero over the following 9 seconds as evaporation removes moisture from the grains — confirming that the device successfully dries the injected rice grains within the modeled timeframe.