Insights

When a clean-looking powder line still produces contaminated product, compare sample location with when the contaminant appears, then investigate suspected reservoirs with recovery-tested surface or rinse sampling.

When a rotary airlock valve leaks, wears, or jams, the cause usually lies in the powder rather than the equipment. This insight maps cohesion, abrasiveness, bulk density, and particle size distribution to distinct failure patterns and explains what each pattern tells you about the powder state.

September 2026's moringa leaf powder recall, set against a year of Salmonella events touching powdered milk and supergreens blends, shows why bulk powder sampling plans can miss low-prevalence contamination and why low water activity slows but does not eliminate pathogen survival.

Hausner ratio, shear cell, and dynamic flow energy results on the same PBF powder lot rarely agree exactly. That disagreement is a decision input, not proof that a method or lab got it wrong.

DEM input parameters are fitted to match a bulk observable, not derived from first principles. A parameter set calibrated against an angle of repose test constrains one point in a multi-dimensional parameter space. When the hopper geometry changes, the stress state changes, and unconstrained parameters become the dominant source of prediction error.

CFD-DEM simulations show triboelectrically charged particles can reverse direction in low-drag duct corner zones during pneumatic conveying, with the delay between feed pulses driving the effect more than pulse length.

LED-based particle image velocimetry (PIV) images the velocity field inside a moving powder bed during blending, adding spatial resolution that static angle of repose, Hausner ratio, and shear cell tests cannot provide. Here is what that resolution can and cannot support for a process decision.

In powder-fed laser cladding, a stable feeder setpoint and an in-spec particle size distribution do not guarantee a stable deposit. Catchment efficiency and stream divergence sit between the feeder and the melt pool, and both can drift independently of the powder's PSD.

The power or torque trace in high-shear wet granulation rises and plateaus as granules consolidate, but the signal is a supporting process indicator, not a direct particle size measurement, and its interpretation is bounded by batch size, impeller geometry, and formulation.

A Pharmaceutical Research study trained Random Forest and XGBoost classifiers on particle size, shape, surface energy, and silica dry-coating data to predict powder blend flowability category. The accuracy holds up as a screening tool and breaks down in specific, predictable places.

Pairing laser diffraction with dynamic image analysis adds per-particle shape descriptors, aspect ratio, circularity, convexity, that help explain an anomalous D50 or D90 and show where the combined result still cannot replace a dedicated particle morphology study.

Gas pycnometry and BET surface area each report a single number. X-ray micro-CT builds a three-dimensional map of the pore network, showing connectivity, tortuosity, and closed porosity that single-value tests cannot resolve, within real resolution and sample-size limits.