Professional Documents
Culture Documents
The Influence of Screw Feeders On Bin Flow Patterns: Y. Yu, P.C. Arnold
The Influence of Screw Feeders On Bin Flow Patterns: Y. Yu, P.C. Arnold
The Influence of Screw Feeders On Bin Flow Patterns: Y. Yu, P.C. Arnold
Abstract
Screw feeders are often used in bulk solid handling systems to control the feed from mass-flow bins with wedge or transition shaped
hoppers. This requires good geometric design of the outlet-feeder combination in order for the bulk solids to be entrained over the full length
of the outlet. Otherwise, unnecessary problems are likely to occur. Achieving an even flow pattern in the bin or hopper can minimize these
problems. There are several methods to increase screw capacity in the direction of feed, e.g. stepped pitch, tapered shaft, tapered screw
diameter; or a combination of these methods. In this paper a theoretical model for achieving a uniform flow pattern is proposed based on the
pitch characteristic of screws. The limitations of some methods for increasing the screw cape~ity are discussed. Experimental studies on the
flow pattern in a wedge hopper with a screw feeder are also included. Four screws of different configurations were investigated on a test fig.
All the screws had increased screw capacity in the flow direction. In the experiments a dividing grid was fitted above the screw to form a
central division over the axis of the screw to isolate each side into a number of divisions which matched the different pitches of the screw.
The results from the experiments are presented and compared with the theoretical predictions.
A =A PI+---P2
RearI Hopper i Front a2 al P2
x P(x)
In general,
Vi = V~_ l + P~Aal = Aal (Pi + P2 +"" + Pi)
0.8
L
A ~ = A,, p'-~L (5)
0.6
Eq. (4) is the required criterion for uniform draw-down
performance and Eq. (5) is the boundary conditiot~. From i 0.4
Eqs. (4) and (5) it can be concluded that the average effec-
tive area in the first pitch has a very important influence on
-~ 02
(b) The pitch is constant along th, vhole feed length as shown
in Fig. 3(b), i.e.
- ^ A A A P! =P2 . . . . . PL = P
I "it Y • y y Vl Ip The volume withdrawn in pitch i is
I- L
(C) V~= PAai
Fig. 3. Screw configurationsfor increasingcapacity: (a) stepped pitch,
uniformdiameter; (b) taperedshaft, uniformdiameter; (c) taperedscrew Similar to the case in Section 3.1, a constant volumetric
mamete~.~nifo.,'myltch efficiency is assumed. As an example, the profile coefficient
Y. Y., P. C. Arnold~Powder Technology 88 (1996) 81--87
hopper with vertical end walls which allows both the hopper
1.o.] ~ dl/D=0.9 ] geometry and the slot opening width to be changed. The four
g 1 ~ dr/o=0.8 I test screws (Fig. 8) are listed as follows:
°'8 ~ ' - - - - - - - . - - . ~ . . _ ~ _ ~ a ~ - ° . 7 I
No. 1, stepped pitch and uniform diameter,
No. 2, tapered shaft and stepped pitch,
0.4 No. 3, variable-width ribbon flight and stepped pitch,
No. 4, stepped shaft and stepped pitch.
,~ 0.2
A dividing grid was fitted over the screw to form a central
0.0 division above the axis of the screw and isolate each zide into
3 4 5 6
a number of divisions which matched the different pitches of
Rado L/O
each screw. The screw was driven by an electric motor via a
Fig. 5. Variation of f~,, with L/D and dn/D for a tapered-shaft screw
shaft-mounted gear box which connected a transducer for
(d/D = I/3 and P/D = 1).
recording the screw torque. The rotating speed of the screw
was varied by frequency control of the drive motor. The screw
1"0t ~ DI/D=-O.5 [ speed or the amount of rotation was measured by a counter.
°., 1 • I
o.o _ i
D~-o.~
,~ 0.2
0.0
3 4 5 6
Ratio L/D
Fig. 6 Variation offp~ with L/D and Dn/D for a tapered-diameter screw
(P/D o=0.7 and d/D = I/3).
The proposed assumptions are the same as in Section 3.2; Fig. 7. Test rig for screw feeders.
again the profile coefficient in the second pitch is considered.
The rel',~ionships between the geometric variables are shown 5o.3 7s.3 loo.~ m.~ (a)
in Fig. 6.
Because the narrow back et~d is prone to having arches
form over it and it is difficult to fabricate properly the screw
trough and the interface with the bin, the tapered-diameter _~o 6s.3s.~.j~,~ ton ~zs.6 ~b ~
screw is not recommended for most materials.
In the above discussion the volumetric efficiency of the
screw feeder is neglected. However, the volumetric efficiency
decreases with increase in the ratio of the pitch to the screw 50 .ca_/sol__ _.~s,Z.__l~.~ izs-6 ~. (C)
diameter. It also decreases with decrease in the ratio of the "i"ff'~V~; ~ V v ,i v ,,
shaft diameter to the screw diameter. These negative effects
on the profile coefficient are further exacerbated by the inclu- (d)
sion of volumetric efficiency.
4. Experimental
Fig. 8. Configurations of the test screws: (a) No. I, step~d pitch and uniform
The test rig shown in Fig. 7 consists of screws of four diameter; (b) No. 2. tapered shaft mid stepped pitch; (c) No. 3, ribbon flight
different configurations and a variable-geometry plane-flow and stepped pitch; (d) No. 4, stepped shaft and pitch.
Y. Yu, P.C. Arnold~Powder Technology 88 (1996) 81-87
5. Results
J g, --i- I
O.0 0.2 0.4 0.6 0.8 1.0
Ratio xlL
The volume withdrawn was different between the rising Fig. l i. Flow patternwithscrew No. 3.
side and the descending side of the flight. A detailed trial was
done to show the cyclic nature at each side of the screw during
a 360 ° rotation. It was clearly observed that in the first pitch
the volume withdrawn on the rising side was larger than that
withdrawn on the descending side. This observation was pre-
viously described by Bates [ 1 ]. In subsequent pitches, how-
ever, the situation was different. In the downstream pitches
the volume withdrawn on the descending side was larger than ~o-I • 6
that withdrawn on the rising side for all four screws, with the
exception of the last or exit pitch. Figs. 9-12 show the hopper
flow patterns with the four screws. The results presented in 0.0 0.2 0.4 0.6 0.8 1.0
the figures were the mean values for both sides. The upper Ratio xlL
boundary line in all the figures is used as the starting or datum Fig. 12. Flow pattt;~nwith screw No. 4.
level of the material.
Fig. 9 shows clearly that at the first pitch of a group of limit the volume withdrawn in the first pitch. The voiume
three equal pitches the volume is withdrawn incrementally, withdrawn at the front side or e:~it end :s greatly increased.
b~:t.~nch an increment is insufficient to pr~luce an even flow The reason for this is that the front vertical wall is beyond
pattern. It can be seen from Fig. 10 that screw No. 2 gives the length of the tapered shaft and located in the pitch with
better results because the ratios d I/D ~ 0.87 and P ~ / D = 0.3 larger length. Although screw No. 3 has variable-width rib-
bon flights and stepped pitch, for the tested material the ribbon
flights seem not to have had the desired effect in increasing
the withdrawn volume in the feed direction. The flow p a t t e r
(Fig. 11) is similar to a screw with full flights and stepped
pitch. Fig. 12 gives a good comparison of the effect on the
J:-7 - B 2 flow pattern between a screw with stepped pitch and a screw
with stepped shaft diameter. In this instance, stepping the
,~ :,oo.'1 "6
I, t O I • , - , • ,
*
:l'0
-
8
, •
shaft provides a more significant increment in withdrawn
volume. Strategies for deciding optimal shaft steps are being
considered.
0.0 0.2 0.4 0.6 0.8 1.0 Experiments without the grid in the hopper were also per-
Ratio x/L formed. Flow patterns with screw No. I and screw No. 2 are
Fig. 9. Flow patternwithscrew No.l. shown in Figs. 13 and 14, respectively. Compared with Figs.
E Yu. P.C Arnold~PowderTechnology 88 (1996) 81-87
6. Conclusions
. , . . . .,0
Based on the discussion of the limitations of several design
0.0 0.2
0.4 0.6 0.8 1.0 methods and the results from experiments, some conclusions
Ratio x/L
can be made.
Fig. 13. Flow patternwithoutgrid for screw No. I.
(i) For a screw feeder with stepped pitch, there is no
continuous function for expressing the volume withdrawn
~ e00 r. .= ~ _-_- _ along the whole feed length. The pitch characteristic of a
screw feeder is introduced to aid in the consideration of the
resulting hopper flow pattern.
J,oo] :2 (ii) The volume withdrawn in the first pitch of a screw has
a great effect on the flow pattern. With the practical limits on
~ 200-1 • 6 pitch ( P I / D at the beginning and PL/D at the end) it is
impossible to achieve uniform withdrawal for screws with
,00t only variable pitch
0! - , - , - , - , -
r~ 0.0 0.2 0.4 0.6 0.8 1.0 (iii) Techniques such as stepped pitch and/or tapered shaft
can increase the capacity of a screw feeder to provide uniform
Ratiox/L
Fig.14.Flow patternwithoutgridforscrewNo. 2. withdrawal in the feed direction, but these methods have their
limits in application, especially in a wedge hopper with a long
slotted outlet.
(iv) Ribbon flights may be good for sticky or cohesive
o
materials where the gap between the screw and the shaft
prevents material build-up, but, from the experiments, this
No. 3
type of screw is not effective in increasing the capacity of the
~ o \~* ,~_..,___o . ~ screw in the flow direction when used with coarse free-flow-
ing materials.
~o (v) Because of cross-flow of material, the actual condi-
tions in a bin or hopper without a grid are modified. For some
[ l'~n No.l
very free-flowing materials such modification may be sub-
• i - i . i • i •
stantial. This phenomenon is receiving further consideration.
0.0 0.2
0.4 0.6 0.8 1.0
Ratio x/L
Fig. 15. Comparisonof pr,)filecoefficients: , theoreticalpredictions; 7. List of symbols
I-I,measurements.
A cross-sectional area of screw (m 2)
9 and 10, it can be seen that the profile of the material in the A~ average effective area of transportation (m 2)
hopper was modified significantly. Comparing Figs. 9 and d core or shaft diameter (m)
13, the fluctuations in the profile indicated in Fig. 9 have been D screw diameter (m)
eliminated, while comparing Figs. 10 and 14, an almost com- fp profile coefficient
pletely uniform draw-down has been achieved. This is con- L feed length (m)
sidered an interesting result which warrants further P pitch length (m)
investigation. V volume withdrawn per revolution (m 3)
Fig. 15 shows the results from the experiments and theo- Vt maximum theoretical volume withdrawn per
retical predictions of the profile coefficients of the four revolution (m 3)
screws. For screw No. 3 the comparison was made for a screw
having full flights and the same stepped pitches. It can be Greek letter~
seen that the variable-width ribbon flight did not have a strong
influence on decreasing the effective cross-sectional area of a¢ equivalent helix angle of flight (deg)
the screw. Because the bottom of the grid was set 30 mm over /3c equivalent helix angle of element motion (deg)
the top of the screw flights, some cross-flow may h~ve qSr angle of frction of material on flight surface (deg)
Y. Yu. P.C. Arnold~Powder Technology 88 (1996) 81--87
r/v volumetric efficiency [5] A.W. Roberts, Proc. Bulk 2000 Conf., Oct. 1991, Institution of
Mechanical Engineers, London, 1991, pp. I I 1-116.
pf coefficient o f friction o f material on flight surface
[6] A.W Roberts, K.S. Manjunath and W. McBride, Trans. Mech. Eng.
(Inst. Eng. Aust.). ME-18 ( I ) (1993) 67-73.
[ 7 ] K.S Manjunath and A.W. Roberts, Proc. PawderandBulk Solids Conf.,
References Chicago, IL~ USA, May 1994, pp. 171-188.
[ 8] G. Hanker, M.P. Poppelen. M.P. Jongejana and J.H. Bekhuis, Proc. Int.
[ I l L. Ba~.es,Trans. ASME, J. Eng. Ind., 91 (1969) 295-302, Syrup. Reliable Flow of Particulate Solids !1. Oslo. Norway, Aug. 1993,
[2] L. Bates, Bulk Solids Handl., 6 ( I ) (1986) 65-78. pp. 551-561.
131 L. Bates, Powder Handl. Process., 6 (2) (1994) 215-221. [9] Y. Yu. Interim Res. Rep. 1. Department of Mechanical Engineering,
[4l J. Carson, Powder Balk Eng., (Dec.) (1987) 32-36, 41-42. University of Wollongong. Australia, Feb. 1994.