Two coupled product properties
A fine wet screen can be judged by the size composition of its undersize. A dewatering deck can be judged by the wet-basis moisture of its oversize and the amount of solids carried away with the drainage. Both operations separate solids and water. Their natural fitting targets differ. If the oversize moisture is measured but the solids lost with drainage are unknown, a water balance can still be reconciled with several different combinations of product solids and water. The size distribution and total concentration in the underpan are particularly useful for distinguishing them.
A bulk water split can be specified for wet classification. The bypass of fine solids often has a separate fitted value. In a dewatering model, the moisture of retained solids may instead determine the water split. In either case estimate both outlet solids and outlet water on the same timestamp and sampling basis. A product moisture reported as water per dry solids must be converted before comparing with a wet-basis measurement.
Fine wet-screen kinetics
The Mwale form for oversize recovery includes a loading term that depends on dry solids flow S, feed water W and total open area A. Let C = S/(S+W) be feed solids mass fraction, K a fitted rate in t/(h·m²), and h the aperture. With x and h expressed in the same units,
The first term is zero at x = 0 in the limiting sense. The second term has a fine recovery floor bꜰ that declines with x. It is a fitted fine-solids bypass term, separate from water recovery to oversize. The corrected midpoint for the first term alone is
The pressure or feed distribution conditions that achieved a fitted K and δ are part of that calibration. Since bꜰ grows with both solids flow and solids concentration, extrapolation to unusually concentrated slurry can produce raw recovery above one. A bounded curve still closes mass but can contain extensive clipping. Inspect the raw prediction or fit residual, especially near the fine end. In the source work, Mwale investigated fine screening over a size range around 45–150 µm. Application to a new feed, aperture family or material requires performance evidence.
Worked case with fines bypass
Let S = 12 t/h, W = 18 t/h, A = 0.80 m², K = 3.432 t/(h·m²), α = 2.518, δ = 0.010 h/t and h = 106 µm. Then C = 0.40, L = (12 × 0.40)/(0.80 × 3.432) = 1.74825 and bꜰ = 0.010 × 12/0.60 = 0.20. The corrected midpoint is 98.21 µm. The table gives recoveries for three representative sizes.
| Representative size, µm | Feed solids, t/h | Actual oversize partition | Oversize solids, t/h |
|---|---|---|---|
| 200 | 3.0 | 0.89252 | 2.67756 |
| 100 | 5.0 | 0.53421 | 2.67105 |
| 40 | 4.0 | 0.07862 | 0.31449 |
| Total | 12.0 | 5.66310 |
The oversize receives 47.19% of all feed solids. If a measured 25% of the water reports there, it receives 4.50 t/h water, giving 44.28% wet-basis moisture. The wet undersize contains the complementary 6.337 t/h solids and 13.50 t/h water. The fitted asymptotic fines bypass is 20%. The 40 µm class falls below that level because the second term already declined, while the first term contributes very little at that size. A constant-floor cyclone correction would give the wrong shape here.
Holding S and A fixed and increasing water to 28 t/h gives C = 0.30, L = 1.3112 and bꜰ = 0.17143. The idealized corrected cut grows by (1.74825/1.3112)^(1/2.518) ≈ 1.121. This is the response of the calibrated equation. In a plant, higher water will also change feed distribution and operating conditions to an extent that this calculation does not separately resolve.
Vibration, drainage and residence time
In the Ng representation, deck length Lᴅ and modeled bed velocity v establish a residence time t = Lᴅ/v. Convert speed f from revolutions per minute to angular frequency ω = 2πf/60. For deck inclination θ, vibration direction ϕ and displacement amplitude a in metres, define the ratio
Use radians when evaluating trigonometric functions and the empirical slope term in the equation. The exponent q can be fitted against aperture w in mm as q = q₀+q₁(w−1). With wet-basis saturation moisture M∞ and fitted drainage coefficient k, the retention moisture M is
For positive k and q, longer residence approaches M∞. The maximum moisture is limited to the feed value in this model because the deck has no independent addition of water in this closure. For an oversize solids rate Sₒ and feed water W, the corresponding oversize water is
In the coupled default, eligible solids loss E is multiplied by the fraction rᵥ of feed water reporting to undersize. Consequently Sₒ = S−E rᵥ and Wₒ = W(1−rᵥ). Setting r = M/(1−M) and substituting gives
Handle dry feed and empty solids separately. Panel alternatives determine the solids split from aperture geometry and solve for an effective exposure against a recovery target, or use directly fitted exposure. Then the same moisture relation assigns water to the predicted oversize solids. A fitted passage trial density and a fitted drainage exponent answer different questions and can be constrained from different measurements.
Worked coupled balance
Take screen length 2.0 m, vibration frequency 900 rpm, amplitude 1.5 mm, horizontal deck and 45° vibration direction. Choose kᵥ = 0.50, m₀ = 1.0 and take q = 0.40 at the selected aperture. Then ω = 94.248 s⁻¹, gᵧ = 1.04088, v = 0.017651 m/s and t = 113.31 s. With M∞ = 0.080 and k = 0.150 on a seconds basis, calculated retained moisture is 0.080 + 0.150 × 113.31⁻⁰·⁴ = 0.10261. For 100 t/h feed solids, 30 t/h feed water and 25 t/h eligible solids, r = 0.11435 and rᵥ ≈ 0.6840. Thus Sₒ ≈ 82.90 t/h, Wₒ ≈ 9.48 t/h, and the underpan receives approximately 17.10 t/h solids and 20.52 t/h water. Reconstruct Wₒ/(Sₒ+Wₒ) to confirm about 10.26% moisture.
A separate capacity estimate can be derived from feed F₈₀ in mm and deck width B, using capacity per metre [20 + 2.5F₈₀] for F₈₀ < 12 mm or [50 + 0.3F₈₀] otherwise, multiplied by 1−exp(−4w). For example, F₈₀ = 8 mm, w = 0.50 mm and B = 1.5 m give 40 × (1−exp(−2)) × 1.5 = 51.88 t/h estimated capacity. A model can solve a moisture balance at 100 t/h even though this empirical capacity diagnostic suggests a need for different sizing or material-specific tests. The drainage prediction alone cannot certify mechanical capability.
Exercises and solutions
1. Change the residence time
Retain the same drainage coefficients as in the worked example. If the residence time doubles, estimate moisture.
Solution. The excess moisture above M∞ is multiplied by 2⁻⁰·⁴ ≈ 0.7579. The moisture becomes 0.080 + (0.10261−0.080) × 0.7579 ≈ 0.09714, or 9.71%.
2. Open area or kinetics
A fine screen has a fitted throughput parameter AK = 2.4 t/h. Identify A and K from the partition curve at one measured operating point.
Solution. The equation depends on the product AK. An aperture measurement cannot break the degeneracy by itself. Use independent measurement of active open area to infer K, or measure an additional condition where open area changes in a known way.
3. Reconcile measured moisture
A survey reports 72 t/h retained solids at 12% wet-basis moisture. Feed contains 22 t/h water. Find retained and drained water.
Solution. Wₒ = 72 × 0.12/0.88 = 9.818 t/h. The water remainder to undersize is 12.182 t/h. A measured drained rate that differs should be compared with timing, moisture basis and water balance uncertainty before changing the drainage coefficients.
Reading
- Mwale, A. N. (2015). A mathematical model for predicting classification performance in wet fine screens. University of Cape Town. Thesis record.
- Ng (1990), experimental drainage and transport correlations for vibrating screens, as described with the worked engineering calculations in King (2012), section 6.4.
