172
D. Markauskas and H. Kruggel-Emden
mesh particles which do not pass through the apertures and thus lead to symmetric separation curves. Geometric scaling (size reduction) or a lower mass flow rate
also leads to nearly ideal separation curves, resulting in small differences between
DEM results and separation curve models. The amount of small particles in the overflow increases in cases with a greater transport velocity induced by larger amplitudes,
frequencies or inclination angles. The simulation results do not lead to symmetrical S-shaped curves, because undersized near mesh particles not passing the screen.
This leads to larger deviations between the separation curve screening models and
the simulation results. For the base case configuration or when a higher mass flow
or a stroke angle with a larger horizontal motion component is employed, the near
mesh sized particles are intensively pegging the apertures. Because of it, many small
and even very small particles are discharged into the overflow. As a result, separation
curve screening models show larger deviations when adjusted to these simulation
results.
Non-spherical particles (Fig. 12b, c) having an equivalent diameter larger than the
orifice size can pass the sieve apertures into the underflow if their minor-diameter is
smaller than the aperture size. In turn, undersized near mesh particles are discharged
into the overflow. As a result, the simulation results form symmetric S-shaped curves
that are readily represented by separation curve screening models. However, the
passing rates of near mesh sized non-spherical particles are more sensitive than with
spherical particles, especially when a higher mass flow is applied or when particles
accumulate on the screen when the screen is operated at a small amplitude or frequency. In these cases, the separation curves increase unevenly and therefore, much
larger deviations occur. A faster particle transport, which is achieved by applying a
larger amplitude or frequency, also results in a few more small and oversized near
mesh particles in the overflow amounting to larger deviations.
Smaller deviations for separation curve models result in some cases when double
cones are screened instead of spheres (Fig. 12b) due to a lower pegging probability.
The largest deviations result from applying of lower amplitudes and frequencies
followed by stroke angles with small vertical motion components. The smallest
deviations for double cones arise for the base case configuration.
The summed up deviations for volume equivalent cylinders (Fig. 12c) show the
lowest deviations of the investigated shapes due to the lowest pegging probability.
The largest deviations with volume equivalent cylinders occur for low amplitudes
and frequencies due to the formation of unsymmetrical separation curves. On the
other hand, good results are achieved for the base case configuration independent of
stroke angle, and for cases with a slightly enlarged mass flow.
The lowest overall deviations, summed up over all investigations, can be obtained
using the model No. II by Plitt and the first revised model by Trawinski (No. V). In
both models, the term (d/d cut ), which increases with particle diameter, is influenced
by one or even two adjustable parameters in the exponent independent of additional
parameters [see Eqs. (21) and (24)] which gives a good adaptability for both models.
The largest deviations are found in the case of the second revised model of Trawinski
(No. VI). Here, both adjustable parameters are present in the exponent of the term
(d/d cut ), but both depend on the particle size d [comp. Eq. (25)].
D. Markauskas and H. Kruggel-Emden
mesh particles which do not pass through the apertures and thus lead to symmetric separation curves. Geometric scaling (size reduction) or a lower mass flow rate
also leads to nearly ideal separation curves, resulting in small differences between
DEM results and separation curve models. The amount of small particles in the overflow increases in cases with a greater transport velocity induced by larger amplitudes,
frequencies or inclination angles. The simulation results do not lead to symmetrical S-shaped curves, because undersized near mesh particles not passing the screen.
This leads to larger deviations between the separation curve screening models and
the simulation results. For the base case configuration or when a higher mass flow
or a stroke angle with a larger horizontal motion component is employed, the near
mesh sized particles are intensively pegging the apertures. Because of it, many small
and even very small particles are discharged into the overflow. As a result, separation
curve screening models show larger deviations when adjusted to these simulation
results.
Non-spherical particles (Fig. 12b, c) having an equivalent diameter larger than the
orifice size can pass the sieve apertures into the underflow if their minor-diameter is
smaller than the aperture size. In turn, undersized near mesh particles are discharged
into the overflow. As a result, the simulation results form symmetric S-shaped curves
that are readily represented by separation curve screening models. However, the
passing rates of near mesh sized non-spherical particles are more sensitive than with
spherical particles, especially when a higher mass flow is applied or when particles
accumulate on the screen when the screen is operated at a small amplitude or frequency. In these cases, the separation curves increase unevenly and therefore, much
larger deviations occur. A faster particle transport, which is achieved by applying a
larger amplitude or frequency, also results in a few more small and oversized near
mesh particles in the overflow amounting to larger deviations.
Smaller deviations for separation curve models result in some cases when double
cones are screened instead of spheres (Fig. 12b) due to a lower pegging probability.
The largest deviations result from applying of lower amplitudes and frequencies
followed by stroke angles with small vertical motion components. The smallest
deviations for double cones arise for the base case configuration.
The summed up deviations for volume equivalent cylinders (Fig. 12c) show the
lowest deviations of the investigated shapes due to the lowest pegging probability.
The largest deviations with volume equivalent cylinders occur for low amplitudes
and frequencies due to the formation of unsymmetrical separation curves. On the
other hand, good results are achieved for the base case configuration independent of
stroke angle, and for cases with a slightly enlarged mass flow.
The lowest overall deviations, summed up over all investigations, can be obtained
using the model No. II by Plitt and the first revised model by Trawinski (No. V). In
both models, the term (d/d cut ), which increases with particle diameter, is influenced
by one or even two adjustable parameters in the exponent independent of additional
parameters [see Eqs. (21) and (24)] which gives a good adaptability for both models.
The largest deviations are found in the case of the second revised model of Trawinski
(No. VI). Here, both adjustable parameters are present in the exponent of the term
(d/d cut ), but both depend on the particle size d [comp. Eq. (25)].