Int. J. Mech. Sci. Vol. 20. pp. 561-570 0020-7403178/0901-0561/$02.

0010
© PergamonPress Ltd., 1978. Printedin Great Britain.

ANALYSING THE VIBRATING CONVEYOR

G. WINKLERt
School of Engineering, University of Bath, England

(Received 19 July 1977; in revised form 23 February 1978)

Summary--Several possible vibrating conveyor designs--the "sealskin" conveyor, the "jerk"

conveyor, the conveyor with inclined motion and the conveyor with out-of-phase motion--are
first discussed and their relative performances are compared. The most promising (and most
widely employed) design, namely the conveyor with inclined motion, is then analysed in
detail. The analysis is in all cases limited to conveyors with a horizontal track and with
vertical track accelerations not exceeding g, the gravitational constant. Since the analytical
performance prediction of the conveyor with sinusoidal motion is very difficult, an alternative
type of motion having a triangular velocity profile is introduced. This allows analytical
expressions for conveyor performance by simple methods to be derived. The results of this
analysis are compared with the theoretical predictions for sinusoidal motion obtained by
means of a simulation on the analogue computer. To simplify the presentation of the
performance results, the design variables are reduced to two non-dimensional parameters, one
for vertical acceleration and the other for horizontal sliding. The performance, too, is
described by a non-dimensional parameter, so that the completely general performance maps
presented for triangular and sinusoidal track velocity profiles need contain only three
parameters.

NOTATION
f frequency
g gravitational constant
u horizontal velocity of track
v horizontal velocity of particle
w vertical velocity of track
u amplitude of u
v mean value of v (transport velocity)
w amplitude of w
t~ amplitude of horizontal acceleration
¢¢ amplitude of vertical acceleration

A acceleration parameter ¢v/g

F friction parameter txg/fa
V transport parameter v/u
T time interval

c~ ratio of coefficients of friction ~b/t~f

/3 ratio of time intervals Ta/Td
~/ angle of vibration
d) phase angle
t~ coefficient of friction

1. INTRODUCTION
Vibrating conveyors are widely used in agriculture, food processing and manufactur-
ing industry to transport granulates, small parts and other loose goods between
processing machines or machine tools in a continuous flow. The most common type,
sketched in Fig. 1, consists of a track mounted on leaf springs that form an acute
angle with the vertical so that the track moves at the same angle to the horizontal, and
some drive mechanism to give it periodic motion.
The performance of vibrating conveyors has been analysed in textbooks such as
that by Boothroyd and Redford, I where further references to research publications
can also be found. More recently, research work has been published by Gaberson ~
and by Schofield and Yousouf; 3 the work by Rochell contains further references. 4
Most of the published work is of an empirical nature, emphasizing the analysis of
data obtained from experimental observations or computer predictions of conveyor
performance. By contrast, the author develops an understanding of the underlying
behaviour of the conveyor leading to analytical descriptions of its performance, so
tPresent address: Fachhochschule Flensburg, 239 Fiensburg, West Germany.

561
562 G. WINKLER

PorticleT~
W

• L
N I I
U

FIG. 1. Schematicof the vibrating conveyorwith inclined motion.

that criteria for the choice of conveyor type and of its design parameters can be
developed. The reason for the empirical nature of most of the previous work is that
the differential equations that describe the motion of a particle coupled by dry friction
to a track vibrating both horizontally and vertically are non-linear and parametric; if
the motion of the track is harmonic, there may not in general, exist an analytical
solution describing the motion of the particle.
Another difficulty associated with the non-linearity is that the methods of super-
position and scaling, widely used with linear second-order oscillating systems, cannot
be applied. As a result, the performance and design quantities are usually given in
some dimensional form, not necessarily a simple one, which makes the analysis and
the comparison between different workers difficult. It will be shown here that three
dimensionless parameters, one for the performance and two for the design quantities,
suffice to completely describe all operating conditions of the conveyor with inclined
motion.
To keep the initial analysis simple, some limitations of scope are required. Firstly,
this study is limited to conveyors with a horizontal track; the analysis of a conveyor
with an inclined track would not be much more complicated, but it would introduce a
third design parameter, Secondly, only sliding conveying will be analysed; hopping
conveying, where the particle leaves the track during part of the cycle because the
downward acceleration of the track is greater than the gravitational constant g, is not
considered here. Thirdly, since the analysis with a harmonic track motion is likely to
be difficult, an alternative motion has to be found. A parabolic displacement of the
track (see Fig. 2), as caused by a constant but alternating force on the track is not too
different from the sinusoidal displacement of equal amplitude; however, it is ac-

/ \
'= ~ \ \
\ .i i

Fro. 2. Comparisonbetween sinusoidaland triangularvelocityprofiles.

 Analysing the vibrating conveyor 563

companied by a triangular velocity profile and half-periods of constant acceleration

which make an analysis easy. This triangular velocity profile can be seen either as an
alternative motion, or an approximation to sinusoidal motion.
Within these limits, the performance of the conveyors will be established by a
combination of geometrical and analytical methods; emphasis is placed on graphical
evaluation before proceeding to the mathematical relationships, and the figures form a
major part of thi-s presentation. Only for the purpose of comparison will computed results
for harmonic motion be introduced again.

2. TYPES OF VIBRATING CONVEYORS

Whereas the transport surface of the well-known conveyor with an endless belt returns empty during a
half-cycle to its starting point, the surface of the vibrating conveyor moves forward and backward with a
net velocity of zero in a short cycle during which it must support the load. To those not familiar with the
design of vibrating conveyors, this may seem paradoxical; however, common experience produces two
working principles for a horizontally vibrating conveyor---one with unequal coefficients of friction in the
two directions (the "sealskin" conveyor) and another in which there is slow advance and a quick return (the
"jerk" conveyor).
The "sealskin" conveyor will always operate when there is any relative motion between track and
particle, and it is easily analysed even for harmonic motion. As regards the "jerk" conveyor, it may be
interesting to demonstrate by means of the linear differential equation for the motion of the particle that no
net transport velocity can be produced with linearily speed-dependent (viscous) friction; vibrating convey-
ing requires some non-linearity in addition to some kind of non-symmetry.
Further thought, or a familiarity with the design of vibrating conveyors, will lead to appreciating that
friction forces and normal forces may differ for the forward and return strokes of the track, and thus to
considering vertical acceleration and motion of the track. In general, the vertical and horizontal components
of the track motion could have different frequencies, or be out of phase. The most common vibrating
conveyor has an inclined motion--the frequencies of both components are the same, the components are in
phase, and the design of the conveyor mounting is simple. The performance of the out-of-phase conveyor
will also be analysed, but hopping conveying, where an out-of-phase motion may be advantageous, is not.
The graphical-analytical approach can be introduced by studying the motion of a particle on a
horizontally oscillating track. If the absolute velocity of the particle is greater than that of the track, it will
decelerate at the constant rate/~g ; if it is less than that of the track, it will accelerate at the same rate. If the
velocities of track and particle are equal, and if the acceleration of the track is less than -+/~g, the particle
will follow the track. Thus the jagged or vee-velocity profile of the track (see Fig. 3), divides the plane into
an upper half in which the slope of the velocity profile of the particle is a negative constant, and a matching
lower half where it is a positive constant. The velocity profile of the particle can thus readily be
constructed, both for triangular and sinusoidal track velocity profiles, even when the static and dynamic
coefficients of friction differ. The only difficulty is to find the steady-state solution because the initial
velocity of the particle is unknown; the steady state is reached slowly when the acceleration of the track is
much greater than that of the particle produced by the coefficient of friction, but quickly when the slope of
'the particle velocity becomes big enough to produce sticking. However, after the transient has died away, there
is no net transport velocity in these cases.

3. OVERALL BEST PERFORMANCE OF FOUR TYPES OF CONVEYORS

Below, criteria for overall best performance of the "sealskin" and "jerk" conveyors, with horizontal
track motion, and of the conveyor with both horizontal and vertical motion without and with phase shift
will be established. Clearly, the particle cannot under steady-state conditions travel faster than the track at
its maximum horizontal velocity u; the performance will therefore be judged with reference to this velocity
by defining a transport parameter

V = v/u. (1)

Arcta,~ ~cj

~- t Arctan ~.g

- - Transient ~- Steady state ~--

FIG. 3. Transient and steady-state motion of a particle on a horizontally vibrating track.

564 G. WINKLER

The "'sealskin" conveyor with the coefficients of friction t4 (backward) and ~/ (forward) performs best
when the particle is just able to follow the track during acceleration, and therefore slides forward during
part of the cycle if a =/~h/~¢ > 1 (see Fig. 4). The greater the value of a, the greater the transport velocity v.
This relationship is readily evaluated by analysing the graphical solution of Fig. 4. The rules of geometry
give

V - ~2)
or+l

For example, if the ratio of the coefficients is 3, a best transport parameter V (a = 3) = I/2 can be achieved.
(In the limit, V(a ~ o ~ ) = 1). The condition for best performance is that the particle must just be able to
follow the track during acceleration i.e.,

~bg = 4 uf (3)

where I.tbg is the maximum acceleration of the particle; 4u[ = t) the alternating value of track acceleration.
If the condition defined in equation (3) is not satisfied, the performance is reduced below that given in
equation (2).
Equations (1)-(3) can be combined to show that the product of best transport velocity and frequency,
sometimes used in the literature to describe the performance, is a function of the coefficients of friction and
their ratio a only:

la-1
vf = ~ ~ - ~ ~bg. (4)

For example, with ~ = 0-6 a n d / z / = 0.2, a product vf - 0.75 m/s 2 is possible.

In the case of the "sealskin" conveyor, the slopes of the velocity profiles of the track are equal and
those of the particle unequal and for the "jerk" conveyor vice versa. Defining the ratio of the relative
periods of acceleration and deceleration as /3 = Ta/Tj, the best performance is found by analysing the
graphical solution in Fig. 5.

v= ~ ' l (5)
2/3 "

If for example, the track returns three times as fast as it advances, a transport parameter V(/3 = 3) = 1/3 is
possible. As before, the condition for best performance is that the particle must just be able to follow the
track during acceleration:

~/3+1
~g = - T uf. ~6)

The best product of transport velocity and frequency is in this case

i/3-1 (7)
v / = ~ b-£-~ ~g.

This result corresponds to the one for the "sealskin" conveyor (see equation 4), and a similar performance
can therefore be expected if a =/3 and p,b = it.
The conveyor with inclined vibrating motion is next analysed; it is really a special case of the conveyor
vibrating with a phase shift in two directions. With the aid of Fig. 6, one can readily be convinced that this
conveyor performs best when the particle just follows the track during part of the forward and upward
acceleration, hovers over the track, just not touching it during the downward acceleration, and decelerates
at the same rate at that at which it had accelerated. Under these conditions

v 3
V .... (8)
u 4'

i.e. the best transport velocity is 75% of the maximum horizontal track velocity. This is a markedly better

FIG. 4. F~G. 5.

FIG. 4. Best performance of the "sealskin" conveyor.

FiG. 5. Best performance of the "jerk" conveyor.

 Analysing the vibrating conveyor 565

best performance friction too great

FiG. 6. Best performance of the vibrating conveyor with inclined motion.

performance figure than those of either the "sealskin" or "jerk" conveyors. The conditions for best
performance are

4 w/= g (9)

and

4 uf = 2/zg (10)

taking into account the doubling of the normal force during the upward acceleration. The product of
transport velocity and frequency for best performance from equations (8) and (10) becomes

3
vf = ~ ~g. (11)

With ~ = 0.6, vf can rise to 2.2 m/s 2. More interestingly, equations (9) and (10) yield the vibration angle for
best performance as a function of the coefficient of friction, i.e.

tan 3' = w/u = I/(2/,Q. (12)

For /.~ = 0.5, the vibrating angle for best performance is 45°; it decreases as the coefficient of friction
increases. This relationship, even though valid only for a triangular velocity profile of the track and sliding
conveying, does not appear to have been formulated previously.
Finally for the out-of-phase conveyor, best performance is achieved as before, if the particle either just
follows the track or decelerates with the same slope or slides along it at constant velocity. (Hopping
conveying is again excluded from this analysis.) Fig. 7 illustrates the velocity profiles at best performance
for a range of phase angles; the results are summarized in the centre of Fig. 7.
In-phase conveying is most effective; the transport parameter V decreases twice as quickly when the
vertical component begins to lead the horizontal component of motion (clock-wise two-dimensional track
motion) as when it begins to lag on it (counter-clockwise motion). The pattern of the transport parameter V
as a function of the phase angle $ is itself periodic, though not symmetrical as explained above; the range
of positive transport velocities extends over a phase range of 180°.
The table below summarizes best performance of the conveyors analysed thus far:

TABLE 1.

Type of conveyor Transport parameter V = v/u Product vf

a-I lt~-I
"Sealskin" = 50% for a = 3
ot+l 4 a + l p'bg
"Jerk" fl - 1 1/3 - 1
2/3 - 3 3 % f o r / 3 = 3 4/3+1 ~g

Inclined motion 3 = 75%

4
g~g
3
3
Out-of-phase < ~ 25% at 4~ = 45 ° < ~ ~g

This table suggests that the conveyor with inclined motion is the most effective type (notwithstanding the
fact that for the sealskin conveyor a very large ratio a of coefficients of friction could produce a transport
parameter close to I), in addition to being the simplest one to design--which by itself .is its greatest
recommendation. Its performance is therefore further analysed over the complete range of sli.~;~:
conveying.
566 G. WINKLER

-45 ° ~o=0 ° +45 °

\/ < / hor. ~" i [ \

-90 ° ~. /~ + 90 °

/ "-\
oz5 --f, T ÷,8oo
1 ' ; :=~,0
--- 0"25

\ /
_~35o ,, ±~8o o / +155 °

~
i / \

"~ 050

FIG. 7. Best p e r f o r m a n c e of the out-of-phase c o n v e y o r for various p h a s e angles between

horizontal and vertical motion.

4. A N A L Y T I C A L P R E D I C T I O N O F T H E P E R F O R M A N C E O F T H E C O N V E Y O R
WITH INCLINED TRIANGULAR MOTION
The differential equation for the motion of the particle is

mi~ + mi.L(g + ~i,)sign (t, -- u ) = 0. (13)

To s t u d y the p e r f o r m a n c e analytically or experimentally four physical quantities could be varied in-

dependently, namely the velocity amplitudes u and w (or the corresponding displacement or acceleration
amplitudes, or one amplitude and the angle of vibration), the f r e q u e n c y f, and the coefficient of friction tz.
This gives e x c e s s i v e scope for experimental work; if each quantity is assigned only 5 different values, 625
operating conditions have to be investigated, and the results can be arranged in m a n y ways. To reduce the
a m o u n t of this work, two d i m e n s i o n l e s s parameters will be used that can be s h o w n to cover all operating
conditions, namely an acceleration p a r a m e t e r

¢v
A = -- (14)
g

and a friction parameter

ttg
_

F - --:-. (15)
U

T h e s e parameters arise from the d i m e n s i o n l e s s differential equation,

L"
+~g l + ~ signu -l =o. (16)

The transport p a r a m e t e r V is now only a function of A and F. Once a value of V for values of A and F is
known, values for two of the physical quantities can be c h o s e n freely (e.g. tt and f ) and the other found by
inverting the above equations; in this example the angle of motion is determined by tan 3' = AF/tt and the
amplitude of horizontal velocity by u = p,g/(4fF).
For the purpose of determining V as a function of F and A, the operating range is divided into two
parts, one in which the particle sticks during the acceleration, and the other in which it slides at all times.
The transition is defined by (1 + A)F = 1. In the a b s e n c e of upward acceleration (A = 0), the particle sticks
w h e n F -> 1, but with increasing A, the friction p a r a m e t e r at the transition point is reduced.
Algebraic analysis of the g e o m e t r y of the velocity profile s h o w n in Fig. 8(a) produces,

V = ~ A F ( I + A 2 F 2 - F 2) for (I+A)F-<I. (17)

The p e r f o r m a n c e according to this equation is plotted in Fig. 9. If either A is close to 1, or F is small,

equations (17) can be approximated by V = 1.5AF.
A similar analysis of the g e o m e t r y of the velocity profile s h o w n in Fig. 8(b) p r o d u c e s

V = (1 + A F ) 1 - (1 - A ) F
I+(I+A)F for (I+A)F-1, (18)
 Analysing the vibrating conveyor 567

(b) " ~
(I+,4)F<I (I -.I-,4)F>.,,I
RO. 8. Velocity profile of a particle sliding forward and backward (a) and sliding forward only
(b).

I
(>9
0'5 0.8
0.7
06
0.5

04
~. 0.2
E
0.3

0.2 .~
g. e,~

8
o
<
o 0.1
~- 0.05

(>02

(>01
J
Friction perometer, F
FIG. 9. Performance map of the vibrating conveyor with triangular motion.

which is also plotted in Fig. 9. The transition between continuous sliding and sliding during the return stroke
only forms a sharp ridge of optimum performance for any given acceleration parameter; a friction
parameter that is too small will produce back-sliding and one that is too large, excessive deceleration during
the phase of forward-sliding. Within the limits of sliding-conveying, overall best performance of V = 0.75 is
achieved for A = 1 and F = 0.5, as has already been shown above through equation (8).
The purpose of this more detailed analysis is not only to show what the performance of this type of
conveyor can actually be expected to be over a wide range of operating conditions, and where optimum
performance can be sought, but to illustrate that by assuming a certain type of uncommon but quite
possible velocity profile (i.e. of triangular form), the performance can actually be predicted by pencil-and-
paper analysis rather than by electronic computation. We now discuss how this performance compares with
that of the" more usual sinusoidal conveyor motion; the results of a simulation using an analogue computer are
given below.

5. COMPUTED PREDICTION OF THE PERFORMANCE OF THE CONVEYOR WITH INCLINED

HARMONIC MOTION
The performance of the conveyor with inclined harmonic motion may not be open to simple analysis for
reasons explained above, but can be based on the same parameters V, A and F as that of the conveyor with
a triangular velocity profile.
The velocity profiles, obtained by simulation on the analogue computer (a schematic of the basic
circuitry is given by Fig. 10), are shown in Fig. I l for a range of values of A and F. The results are plotted
in Fig. 12. As before the performance improves as A increases, and again there is an optimum value of F.
However, this optimum value does not coincide with the limit to back-sliding, (1 + A)F = 1, but lies at lower
values of F which vary between about 0.5 at low values of A, and 0.35 at A = 1; optimum performance is
thus accompanied by some back-sliding as illustrated in Fig. I1 for A = 1 and F = 0.4. Overall best
performance is at about V = 0.57--lower than the value of V = 0.75 at A = I and F = 0.5 obtained with a
triangular velocity profile of the track.
This difference is associated with optimum conditions such as hovering existing during a finite part of
the cycle, and not at an instant as in the case of harmonic motion.
Finally, the performance results for harmonic motion can be shown to fit the equation

V = 2.5AF(1 - F) (19)
568 G. WINKLER

-~~ I I ITra k e o ,
ComparaPar.c
tor eloci,

Gravity Friction force

FIG. 10. Basic analogue computer circuit for simulating the vibrating conveyor with sinusoida]
motion.

A--I

11. Velocity profiles of track and particle for a range of acceleration and friction
parameters.

in the mid-range of operation 0 . 2 < F < 0 . 8 and for A < ! to within 10%. Equation (19) is simple and
accurate enough for a quick check of experimental results, and for the design of harmonically vibrating
conveyors without hopping. In the lower range of operation (F < 0-1) the equation V = 2 A F describes the
performance accurately.
6. M A X I M I Z I N G T H E P E R F O R M A N C E
If the designer wishes to maximize the transport velocity v, he must maximize the product v = Vu
within the design limits. These design limits may be defined in various ways. It is obviously always
desirable to make A as large as possible; A = 1 is however the limit of the present analysis, and it would
also be one of the design limits if hopping were to be avoided.
If the other design limit were to be a maximum value of horizontal track velocity u, one would simply
choose the peak value of V for a given value of A from Fig. 9 or Fig. 12 in order to maximize the transport
velocity v; this is what is meant by "best performance" in the present paper.
However, if the other design limit was to be a maximum value of horizontal track displacement,
maximizing the product Vu may no longer be achieved by choosing the peak value of V. It can be shown
 Analysing the vibrating conveyor 569

I[ I t I l I I I I

0.5

0"2

01

g E
g
F-
0"05
._g
8
o

£}O2

OOI
0 01 0.2 03 0-4 0.5 0.6 07 08 09 I0
Friction parameter, F

FIG. 12. Performance map of the vibrating conveyor with sinusoidal motion.

0"8

"x

o6 dV/dF=O

O" 0"4
m . I ~[

X
0.2~ ~ 0.6g
0.4~
0 0-2
0 02 04 o.s 0.8
Friction parameter, F
FIG. 13. Performance map for sinusoidal motion: (l + A)F = 1 borderline for back sliding,
d V / d F = 0 optimum if horizontal velocity is limited, d V / d F = (I/2)V/F optimum if horizontal
displacement is limited, (V/F)m=~ = 2A at F = 0; maximizes vF.

that in this case the slope d V/dF at the operating point and the slope V/F of a straight line through the
origin have to be related by d V / d F = I/2V/F The points at which this condition holds true is best found
graphically in the V-F-diagrams. In Fig. 13, these points are slightly to the left of the peak value of V; in
Fig. 9, they still coincide with the peak value because of the discontinuity at this point.
It has also been suggested in the literature that in order to maximize the transport velocity v, one should
aim at a design that maximizes the product vf. This concept arose out of the realization that this product is
independent of the frequency f, given a value of A. However, it can easily be shown that if this quantity is
maximized, in general one is no longer free to choose f. For triangular motion, one obtains v f = ( WF)~,g/4,
for sinusoidal motion v f = (V/F)/~g/(2w). V/F reaches its maximum values of (3/2)A and 2A respectively at
F =0, which can be verified from equation (17) and Fig. 13. This means that the product vf will be
maximized when the frequency f is infinite and the vibration angle y is zero; because vf remains finite, the
transport velocity drops to zero. Maximizing vf will therefore not lead to a useful design, even though it
may be a suitable quantity for research purposes.
570 G. WINKLER

7. CONCLUSION
The reduction suggested here of four design variables (for example vertical
acceleration, angle of vibration, f r e q u e n c y and coefficient of friction) into two
non-dimensional p a r a m e t e r s , describing the effects of vertical and horizontal track
acceleration should not only aid in the understanding of the concepts involved in
vibrating conveying, but also simplify the acquisition and presentation of data and
improve c o m m u n i c a t i o n s between workers in the field. Other dimensionless
p a r a m e t e r s are of course possible, and have f r o m time to time been suggested.
The analysis of vibrating c o n v e y o r s with triangular velocity profiles shows that
best p e r f o r m a n c e in terms of the dimensionless transport p a r a m e t e r V is always
obtained when the particle is just able to follow the accelerating track. If the friction
p a r a m e t e r F is too small, back-sliding will occur; if it is too large, this will hinder
forward-sliding. For a given acceleration p a r a m e t e r A, the c o n v e y o r with inclined
inphase motion is shown to p e r f o r m best of all the different designs, giving a transport
p a r a m e t e r corresponding to 75% of the p e a k horizontal track velocity for an ac-
celeration p a r a m e t e r of 1. U n d e r these conditions, the dimensional p e r f o r m a n c e
product of mean transport velocity and f r e q u e n c y v/, frequently used in the literature,
reaches a value of (3/8)txg; the vibration angle has to be related to the coefficient of
friction by tan 3' = l/(2/x).
The p e r f o r m a n c e of the vibrating c o n v e y o r with inclined sinusoidal motion is
shown to be, on the whole, similar to that with triangular motion, although the transport
p a r a m e t e r s are lower o v e r the region of operation because optimum conditions occur
at shorter intervals during the cycle. The transport p a r a m e t e r reaches a peak value of
about 57%; at this point, some back-sliding occurs.
In general, the vibration angle is related to the coefficient of friction by tan 3" =
A F l l z . To the e x p e r i m e n t e r wishing to c o v e r a wide range of operational conditions
with a minimum of experimental w o r k this equation suggests that it is sufficient to
vary either 3' or /z. To the designer wishing to maximize p e r f o r m a n c e the equation
shows, together with Figs. 9 or 12, how the vibration angle can be determined as a
function of the acceleration p a r a m e t e r and the coefficient of friction.

REFERENCES
1. D. BOOTHROYDand A. H. REDFORO,Mechanized Assembly. McGraw-Hill, London (1968),
2. H. A. GABERSON,Particle motion on oscillating conveyors. Part I: The equation of motion and the rules
for predicting motion from transitions. ASME paper No. 7l-Vibr-15. Part 2: Practical solutions to the
equation of motion, Trans ASME Engineering [or Industry (Feb. 1972).
3. P. ROCHELL, Untersuchungen tiber den F6rdervorgang auf dem Schwingf6rderer _unter besonderer
Berticksichtigung der Reinigungsanlage im M~ihdrescher. Diss. Univ. Stuttgart (1975).
4. R. E. SCHOFIELDand M. YOUSUF, The design of-a linear "out of phase" vibratory conveyor, Trans
ASME Engineering [or Industry (Feb. 1973).