How Is the Particle Size Distribution of Pulse Flour Measured and Controlled During Production

Consistent particle‑size distribution (PSD) is the core quality metric for pulse flour destined for dry protein‑starch fractionation. Even with well‑configured grinding‑classification equipment, raw‑material variation, feeder drift and component wear can shift PSD throughout continuous production. Reliable measurement workflows paired with closed‑loop process control ensure flour stays within the target operating window, typically D90 10‑65 μm for dry‑fractionation feed flour. This article covers standard measurement techniques, sampling protocols, key control variables and common troubleshooting for industrial pulse‑milling lines.

Standard Methods for Measuring Pulse Flour PSD

1. Laser Diffraction (Primary industrial reference method)

Laser diffraction measures volume‑weighted particle‑size distribution, outputting D10, D50, D90, D97 and PSD span, the key parameters for pulse‑flour quality evaluation.

• Dry‑dispersion mode (preferred for pulse flour): Powder is dispersed by compressed air without liquid solvent. It avoids starch granule swelling and particle aggregation caused by water contact, delivering results matching actual milling‑process conditions. Instruments such as Malvern Mastersizer with Scirocco dry‑dispersion module are widely used.
• Wet‑dispersion mode: Suspends powder inside liquid solvent. Not recommended as the primary test for pulse flour; starch and protein particles swell in aqueous environments, shifting PSD readings to larger values and creating misleading data.
• Critical measurement parameters: Set appropriate dispersion air pressure to break soft agglomerates without physically fracturing intact starch granules. Over‑high dispersion pressure will artificially produce finer test results.

2. Sieve analysis (Secondary screening method)

Vibratory sieve testing separates powder by physical mesh size. It is low‑cost and easy‑to‑operate, but limited for superfine pulse flour below 75 μm. Fine particles easily agglomerate during sieving. Sieve analysis is mainly used to detect oversized coarse outliers rather than acquiring full continuous PSD curves. It cannot deliver accurate D50 / D90 values for the 10‑65 μm target range.

3. In‑line real‑time particle‑size monitoring

Installed directly on the production line at the mill outlet, in‑line laser sensors continuously capture PSD data without manual sampling. Data feeds to the plant control system, enabling trend tracking and early‑stage alarm triggering for PSD drift. In‑line instruments require regular calibration against offline laboratory laser‑diffraction reference measurements to offset noise from dust and moisture.

Representative Sampling Principles for Production Testing

Measurement data is only reliable with correct sampling:

1. Extract representative samples from the finished flour discharge point, not from static hopper surfaces where particle segregation occurs.
2. Implement time‑interval sampling for continuous production; grab‑spot sampling easily misses short‑term PSD fluctuations caused by feeder surges.
3. Seal samples immediately to prevent moisture absorption, which induces particle agglomeration before laboratory analysis.

Process Parameters to Control PSD in Integrated Mill‑Classifier Systems

For mechanical‑impact mills fitted with dynamic air classifiers, PSD is governed by four interconnected core variables. No single parameter can be adjusted in isolation.

1. Dynamic classifier rotor speed

The primary fine‑tuning lever for PSD.

• Increase classifier wheel speed → smaller cut‑point; finer D90, narrower PSD span under stable conditions.
• Decrease classifier wheel speed → larger cut‑point; coarser finished flour.
• Too high rotor speed raises system recirculation load and increases heat generation risk.

2. Feed rate

Stable loss‑in‑weight feeding is essential for consistent PSD.

• Higher feed rate: particle loading inside grinding chamber rises; D90 shifts coarser, PSD span broadens, incomplete cell‑liberation occurs.
• Lower feed rate: particles receive excessive impact energy; D90 becomes finer, risk of starch over‑grinding and ultrafine‑dust build‑up.
• Feed‑rate oscillation causes cyclic PSD drift which classifier tuning alone cannot compensate.

3. Process air volume

Air flow transports powder through the grinding‑classification loop and provides cooling.

• Higher air volume: carries more fine particles out of the grinding chamber; shifts product finer.
• Lower air volume: extends material residence time inside grinding zone, increases re‑grinding intensity.
• Air volume must match feed rate and classifier speed; mismatched air flow creates distorted PSD and thermal overheating.

4. Mill rotor / impact speed

Rotor speed determines impact energy for cell‑wall rupture.

• Higher rotor speed enhances particle fracture, producing finer flour, yet elevates frictional‑heat generation.
• Lower rotor speed delivers gentler impact, reduces heat risk, but may fail to fully liberate protein‑starch composite particles.

Manual and Closed‑Loop PSD Control Workflow in Production

Off‑line laboratory‑based control (conventional factory workflow)

1. Collect periodic samples from production line.
2. Test PSD via dry‑dispersion laser diffraction.
3. Compare D50, D90 and span against target specification.
4. Manually adjust classifier speed, feed‑rate or air volume according to deviation.
5. Run production for a short stabilisation period then re‑sample and verify.

Automated closed‑loop control (advanced industrial configuration)

1. In‑line PSD sensor continuously outputs real‑time D90 and span data to PLC control system.
2. When measured PSD drifts outside set tolerance windows, the system automatically modulates classifier rotor speed as the primary adjustment.
3. If deviation persists after classifier adjustment, the system makes secondary fine‑tuning to feed‑rate set‑points within safe operating limits.
4. Trigger production alarms when parameters reach safety boundaries to avoid producing large volumes of off‑spec flour.

Common PSD Deviations and Root‑Cause Troubleshooting

1. D90 becomes unexpectedly coarse, span widens
   Possible causes: feed‑rate spike; classifier wheel wear; insufficient air volume; incoming raw material with higher seed hardness or moisture.
2. Excess ultrafine fraction, D90 drops too low
   Possible causes: feed‑rate drops below design window; classifier speed set excessively high; over‑high rotor impact speed.
3. Periodic cyclic PSD fluctuation
   Possible causes: feeder bridging and surging; unstable raw‑material supply; inconsistent dehulling quality introducing variable fibre content.
4. D90 meets target value, yet span remains too broad
   Possible causes: worn classifier rotor blades; improper dispersion‑pressure during lab measurement; large coarse fibre outliers from poor dehulling.

Key Practical Notes

• PSD control cannot ignore raw‑material properties. For different pulse varieties, seed hardness, oil‑content and moisture require adjusting the whole parameter set even when targeting the same D90 window.
• PSD must always be evaluated alongside secondary quality indicators: milling temperature, starch‑damage level and protein PDI. Reaching target D90 is meaningless if over‑grinding induces protein denaturation.
• Stand‑alone hammer‑mills without dynamic classification cannot achieve tight‑span PSD; measurement will show D90 within range but excessively broad span, leading to poor dry‑fractionation performance.

Pulse‑flour particle‑size distribution is primarily measured by dry‑dispersion laser diffraction, complemented by sieve screening and in‑line real‑time monitoring. In industrial production, PSD is controlled mainly by coordinating classifier rotor speed, stable feed‑rate, process‑air volume and mill impact speed. Periodic offline laboratory verification or automated closed‑loop control prevents PSD drift caused by raw‑material changes and equipment wear. Achieving target D90 value is insufficient; manufacturers must maintain narrow PSD span while controlling starch‑damage and thermal exposure to guarantee high‑efficiency dry protein‑starch separation.