Introduction & Context

In comminution circuits, the Rosin-Rammler (RR) distribution is the industry-standard model for describing the size distribution of milled particles. The two-parameter RR model is uniquely suited to on-line diagnostics because it collapses an entire sieve curve into a single pair of numbers: the characteristic size \(D'\) and the uniformity index \(n\). By tracking how far the live RR parameters deviate from the design baseline, plant engineers can detect excursions in ultrafines (< 150 µm) long before they propagate to downstream classification, dewatering, or tailings-handling bottlenecks. This diagnostic is therefore embedded in supervisory control loops for ball mills, vertical roller mills, and high-pressure grinding rolls across cement, coal, and iron-ore operations.

Methodology & Formulas

Particle-size distribution model
The RR equation gives the mass-percent retained on a sieve of aperture \(d\): \[ R(d)=100\;\exp\!\left[-\left(\frac{d}{D'}\right)^{\!n}\right] \] where
- \(D'\) is the characteristic size (µm) at which 36.8 % of the mass is retained,
- \(n\) is the uniformity index (dimensionless); higher \(n\) ⇒ steeper slope ⇒ narrower distribution.
Ultrafines calculation
The mass-percent passing the target sieve \(d_{\text{sieve}}\) is \[ P(d_{\text{sieve}})=100-R(d_{\text{sieve}}). \] Excess fines are declared whenever \(P(d_{\text{sieve}})\) exceeds the set-point \(P_{\text{target}}\).

Operating envelope

Parameter	Lower limit	Upper limit	Remarks
Uniformity index \(n\)	0.8	1.5	RR correlation validated only within this band
Moisture content	12 %	15 %	Brittleness assumptions hold; prevents coating
Characteristic size \(D'\)	> 0 µm	—	Negative values are physically impossible

Pressure-drop rise across downstream filters or baghouses faster than historical baseline.
Increased dust emissions at stack or fugitive leak points.
Product bulk-density drop >5 % versus spec or visible segregation in bins.
Mill motor amps trending downward while classifier amps rise, indicating over-grinding.
Yield loss reported by packaging line due to off-spec particle-size distribution.

Isolate a 1–2 kg slice sample from the entire cross-section of a transfer chute using a stopped-belt cutter or Vezin sampler.
Combine six equally spaced increments taken during a 10-minute production window to cancel out short-term segregation.
Split the composite to 200 g using a rotary riffler, then run dry-sieve or laser-diffraction analysis; report the percent passing 45 µm as the diagnostic fines index.

Record baseline classifier RPM and corresponding fines %.
Reduce RPM by 5 % while keeping feed rate constant; collect product sample after 5 residence times.
If fines % decreases by more than 1 %-point, classifier speed is a key lever; if change is <0.3 %, investigate feed size or mill internals instead.

Fluted roll surface edges becoming rounded, reducing nip angle and increasing recycle load.
Scraper blades worn to <3 mm thickness, allowing material slip and repeated compaction passes.
Edge seal plates with >1 mm gap, causing bypass of semi-compacted cake that later breaks into fines.

Worked Example – Diagnosing Excessive Fines at the 150 µm Sieve

A shift team at a 1 t h⁻^-1 detergent granulation line notices that the on-line laser sizer is reporting more than 30 % of product below 150 µm, well above the 5 % target. They suspect the classifier rotor tip speed has drifted low and want to confirm whether the current cut-point (d′) and sharpness index (n) can explain the observed fines level.

Knowns

Target fines below 150 µm: 5 %
Current cut-point d′: 380 µm
Current sharpness index n: 0.95
Sieve aperture: 150 µm
Moisture: 12 % (within 12–15 % spec)
Temperature: 25 °C
Empirical n range: 0.8–1.5

Step-by-step calculation

Compute the dimensionless ratio \( \frac{d}{d'} \) for the 150 µm sieve: \[ \frac{150}{380} = 0.395 \]
Evaluate the cumulative oversize fraction \( R_{150} \) using the Rosin–Rammler equation: \[ R_{150} = e^{ -\left( \frac{150}{380} \right)^{0.95} } \] \[ R_{150} = e^{ -0.395^{0.95} } = e^{ -0.411 } = 0.662 \]
Convert to percentage oversize: \[ R_{150} = 0.662 \times 100 = 66.132\% \]
Calculate the fines fraction below 150 µm: \[ \text{Below 150 µm} = 100 - 66.132 = 33.868\% \]

Final Answer

The model predicts 33.9 % of product below 150 µm, confirming that the low rotor tip speed (reflected in d′ = 380 µm and n = 0.95) is the root cause of the excessive fines. Restoring d′ to the baseline 450 µm and n to 1.2 is expected to bring the fines level back to the 5 % target.

"Un projet n'est jamais trop grand s'il est bien conçu." — André Citroën

"La difficulté attire l'homme de caractère, car c'est en l'étreignant qu'il se réalise." — Charles de Gaulle