RECENT RESEARCH DEVELOPEMENTS IN BELT CONVEYOR TECHNOLOGY

A.W. ROBERTS1 and A. HARRISON2

1. DEAN, FACULTY OF ENGINEERING, UNIVERSITY OF NEWCASTLE, AUSTRALIA
2. PRINCIPAL RESEARCH SCIENTIST, DIVISION OF APPLIED PHYSICS, CSIRO,
AUSTRALIA.

ABSTRACT

This paper outlines recent research in the area of belt conveying highlighting the research being
conducted in Australia. A review of design trends associated with long overland conveyors is
presented indicating the influence of economic and technical considerations in the design
methodology. Aspects of conveyor dynamics are discussed and research and development work
concerned with specific subjects including belt drum friction, rolling resistance and bulk solid and
conveyor belt interaction. A test rig for examining stability between bulk solids and conveyor belts
during transportation is described.

1. INTRODUCTION

Throughout the world bulk materials handling operations perform a key function in a great number
and variety of industries. While the nature of the handling tasks and scale of operations vary from
one industry to another and, on the international scene, from one country to another according to
the industrial and economic base, the relative costs of storing and transporting bulk materials are,
in the majority of cases, quite significant. It is important, therefore, that handling systems be
designed and operated with a view to achieving maximum efficiency and reliability.

The theme embodied in the foregoing remarks is of direct relevance to Australia in view of the
heavy dependence on bulk solids handling operations. While these operations range across the
broad spectrum of the bulk handling field, a major emphasis is on the storage, handling and
transportation of coal, mineral ores and agricultural products, notably grain, in large tonnages. An
indication of the tonnages of principal minerals produced in Australia is given in Table 1, while
Table 11 summarises the mineral exports. Export earnings from minerals and agriculture are very
substantial, with those from minerals approximating 50% and agriculture 30%.

Of the various modes of conveying bulk solids, belt conveyors are of considerable importance in
view of their widespread use and proven reliability. Although their use is largely confined to in-
plant movement of materials, their application to long distance overland transportation and to the
conveying of large tonnages is gaining considerable interest. These applications are made
possible through the development of steel cord conveyor belting with much potential seen in the
newer light weight, high strength belt materials such as those reinforced with Aramid fibres.

Despite these developments, belt conveyor systems are often designed using static type analysis
based on manufacturers' handbooks or current design standards. In general, steady state
operation is assumed and belt tensions are computed on the basis of the drive requirements and
the need to limit belt sag. High factors-of-safety of the order of 7:1 are adopted to account for
unknown dynamic effects and uncertain belt splice efficiencies.

Fortunately, this situation is changing. Over recent years, significant improvements in conveyor
belt technology have occurred and more exacting design analysis procedures have arisen as a
result of the considerable research that has taken place. There is now a much better
understanding of the dynamic behaviour of belts during starting and stopping, as well as during

1
operation, and this has led to resonance free designs, lower factors-of-safety and increased
reliability.

In view of the importance of belt conveying to Australia's mining and mineral processing
industries, it is not surprising that the subject of belt conveying has received considerable
attention in terms of research and development. There has been wide acceptance by industry of
the advances being made and this is reflected in the new conveyor installations particularly in
regard to those associated with long overland transportation. The following list highlights some
recent and current advances:

I. Conveyor dynamics - starting and stopping characteristics and dynamic belt stresses.
II. Transverse vibrations of conveyor belts and the associated influence of idler spacing and
troughing configurations on such vibrations.
III. Conveyor belt idler resistance taking into account belt rubber hardness, sag, troughing
configuration, idler/belt indentation and ambient operating temperature.
IV. Conveyor transition geometry.
V. Stability of bulk solid on conveyor belts during motion on horizontal and vertical curves.
VI. Conveyor belt/drive drum friction taking into account rubber hardness, surface roughness
wrap angle.
VII. Steel cord splice design and analysis.
VIII. Conveyor belt monitoring as applied to steel cord and fabric belts
IX. Belt rip detection employing ultra sonic devices.
X. Belt tension monitoring during operation.
XI. Belt cleaning including carry back measurement and development of improved cleaning
efficiencies.
XII. Economic analysis applied to conveyor design.

A selection of research publications concerned with belt conveyor developments in Australia are
given by References [1-26]. While it is beyond the scope of this paper to discuss these
developments in any detail, an overview of some salient aspects of the research. is presented.
TABLE I
PRODUCTION OF SELECTED MINERALS IN AUSTRALIA
(Monthly summary of statistics Australia - January, 1988)
Mineral Unit of Quantity 1984 - 85 1985 - 86 1986 - 87 1st Quarter 1987 - 88
Coal '000t 145,173 163,669 182,255 39,406
Lignite '000t 38,500 35,469 42,656 10,268
Iron Ore '000t 91,411 97,660 96,770 26,670
Manganese '000t 1,858 1,966 1,585 4,710
Bauxite '000t 31,839 32,432 -- --
Aluminium '000t 8,120 9,080 9,838 2,573
Ilmenite '000t 1,264 1,272 -- --
Rutile '000t 191 212 -- --
Zircon '000t 452 476 -- --
Nickel t 80,267 80,528 74,654 18,800
Tungsten t 1,830 1,830 1,249 295
Gold kg 48,853 64,780 -- --
Copper t 251,782 241,706 239,803 56,388
Lead t 480,175 469,637 438,249 133,939
Zinc t 731,321 689,411 715,974 185,766
Tin t 6,430 7,391 8,524 1,873

2
 TABLE II
VALUE OF AUSTRALIAN EXPORTS OF PRINCIPAL PRODUCTS - 1986
(Aust. Mineral Industrial Quarterly, Vol. 39 (4), 1986)
Bulk Material Unit of Quantity Quantity Value of f.o.b. ($'000)
Alumina '000t 7,687 1,427,168
Coal '000t 92,717 5,364,142
Copper t 140,133 247,567
Diamonds carat 3,376,323 38,561
Gold kg 58,735 972,591
Iron Ore '000t 79,678 1,937,705
Lead t 412,213 361,122
Nickel -- -- 438,044
Rutile t 229,665 116,412
Silver kg 957,710 53,750
Tin t 7,494 58,229
Tungsten t 2,526 13,482
Zinc t 661,823 488,720
2. TRENDS IN CONVEYOR DESIGN
Recent research has shown that belts ranging in widths from 800 to 1200 mm are the most
suitable. In this way, belt tensions are kept to acceptable limits allowing longer, individual
conveyor lengths to be realised for the given range of belt SR values commercially available. For
large tonnages, the use of belts within the above mentioned range of widths, but running at higher
speeds provides the best solution. While these trends are supported on the grounds of
economics, they are also supported on technical grounds. With regard to belt speeds, the
economic evaluations clearly highlight the advantages of employing speeds greater than 4 m/s
where large tonnage throughputs are involved.
The foregoing observations may be demonstrated as follows.
2.1 Basic Design Considerations
The application of "steady-state" design theory provides a useful basis for the study of the first
order influences of such factors as belt width, strength rating, factor-of-safety and costs for
particular conveying tasks. While the design equations and theory are well known and
documented, for the purpose of the present discussion, it is useful to briefly review the basic
procedures. Referring to Figure 1, the following basic design equations apply:

Figure 1: Conveyor Design Model

2. 1. 1 Throughput Q
Q = ρAv (1)

3
where:
ρ - bulk density of bulk solid
v - belt velocity.
"A" represents the cross-sectional area of the bulk solid on the belt and is given by:
A = Ub2 (2)
where:
U - non-dimensional cross-sectional area shape factor
b - contact or "wetted" perimeter.
The belt width B > b to allow for edge effects.
Shape factors for various idler troughing configurations are given in Ref.[20]
2.1.2 Belt Resistance and Power
n
F = Σ Fi (3)
i=1
Where Fi are the various resistances such as empty belt frictional resistance, load resistance,
slope resistance, special and localised resistances such as those due to skirtplates and belt
cleaners.
As a first order approximation, it may be assumed that the friction coefficients in the idlers and
drive components are constant and independent of the velocity. It is then apparent that the total
resistance to motion in functional form is:
F = fi (B, L p) (4)
where:
B - belt width
L - total belt width
The conveyor power is given by:
P = Fv / η (5)
where:
η - drive efficiency
2.1.3 Peak Belt Tension
For a given arrangement, the tight side tension Fl based on the simplified tension distribution
around the drive drum(s) is given by:
µθ µθ
F1 = (e / e - 1) (6)
This assumes the well known tension relationship
F1 / F2 = eµθ (7)
where F2 is "slack" side tension.
The maximum belt tension under steady-state operation will be the larger one of that given by Eq.
(6) and that required to limit the maximum belt sag in the low tension zone of the belt. When
dynamic effects are taken into account, the maximum belt tension will also include the influences
of transverse and longitudinal belt vibrations. For a "first-order" approximation, it is clear that in
functional form:
Fmax = f2 (B, L, p) (8)
This equation indicates that the maximum belt tension for a given length conveyor is directly
related to the belt width. The narrower the belt, the lower the tension. However, the interrelation
of belt width and conveyor speed needs to be taken into account when establishing the most
appropriate parameters to meet the throughput requirements.
2.1.4 Belt Rating and Width
For a computed maximum belt tension, the required belt width is given by:
B = Fmax fs / SR (9)
where:
fs - factor-of-safety
SR - maximum strength rating of belt (in kN/m).
2.2 Conveyor Economic Analysis
Economic studies based on life-cycle cost analysis have been conducted to identify the most cost
effective operating conditions for conveyors [1,10,20]. Relevant aspects of these studies are

4
highlighted below. While the cost data are based on 1985 values, the relativity with respect to
current costs would be expected to be the same.
2.2.1 Belt Width and Velocity Considerations
Figures 2 and 3 show the annual equivalent costs per unit length for various throughputs for a
conveying distance of 5 km. In Figure 2 the costs as a function of belt width are presented with
belt velocities curves superimposed. The graphs show that the lowest costs occur in the range of
belt widths 0.6 to 1.2 m with lm being a "good" average value. In Figure 3 the cost data are
plotted as a function of belt velocity with belt width curves superimposed. As indicated, the annual
equivalent cost decreases significantly at low belt velocities but, as the velocities increase, the
cost is substantially constant or increases very slightly.

Figure 2: Annual equivalent cost per unit length as function of belt width
L = 5km, p = 0.85 t/m3, fs 7:1

Figure 3: Annual equivalent cost per unit length as function of belt velocity
L = 5km, p = 0.85 t/m3, fs 7:1
2.2.2 Belt Length Considerations
The potential of belt conveyors for long distance transportation is illustrated in Figure 4. Here the
annual equivalent cost per unit length is plotted as a function of conveyor length for various belt
widths. The corresponding conveying velocities are indicated and the belt SR curves are
superimposed.

5
 Figure 4: Annual equivalent cost per unit length for horizontal steel cord
Belt Conveyors. Q = 1000 t/h, p = 0.85 t/m3
For conveyor lengths beyond 1 km, the cost per unit length increases as shown, this being due to
the need to employ belts with higher SR values. For this reason it would be nominally more
economical to use several shorter conveyors in series than fewer longer conveyors for conveying
over long distances. This could well be the case when wide belts are used. However, the cost per
unit length advantages associated with shorter belts would be offset by the cost of the transfer
stations. Hence it is recommended that, in general, longer individual belts be used and this
reinforces the need to employ narrower belts.
2.3 Dynamic. Considerations
The two most important dynamic considerations in the design of belt conveyors are.

i. Reduction of starting and stopping stresses in the belt.

ii. Running stability or resonance-free operation.

2.3.1 Dynamic Belt Stresses - Starting and Stopping of Conveyors

Numerous methods have been devised for starting belt conveyors, including eddy current drives,
switched resistance starts, fluid and fluid-scoop coupling drives and controlled starting for AC and
DC motors. Most of these methods have discontinuities in the drive acceleration at the instant
torque is applied to the belt, leading to travelling elastic waves in the belt [41. Measurement of
starting tensions in belts containing shock waves indicates that tensions in the belt may exceed
ten times the highest operating tension, and this has been the reason for the high safety factors
that have been adopted in the past. Large safety factors increase the cost of belting manufacture.
The control of belt speed during the start-up and running period is most important in higher speed
belts if shock is to be avoided and if lower belt safety factor "sf" is to be employed to reduce
installation and driving costs. During the braking period, similar considerations apply; however,
the concept of emergency stopping must be reviewed when belt speed is increased. Naturally, all
safety requirements must be much more closely observed near the running belt and at the
loading and discharge stations.

6
The criteria for minimising transient stresses in conveyor belts during starting and stopping
particularly applies to fast belts. The dynamics of starting and stopping have been examined in
some detail by Harrison [5]. He showed that an optimal "stop start" is given by the belt start-up
velocity.
v(t) = vb / 2 (1 - πt / T) for t < T
(10)
= vb for t > T
where:
vb -belt running velocity
T -run-up period.
The form of Eq. (10) is shown in Figure 5. With modern computer control systems, it is possible to
generate the desired starting characteristic.

Figure 5: Ideal belt velocity characteristic to minimise transient tensions at start and stop
A general formula for the factor-of-safety has been proposed.
f.o.s ≥ 3 (1 + Td / Tmax + Σa Ta / Tmax) (11)
where Td is the non-oscillatory dynamic belt tension during transient periods, ΣTa are additional
tensions due to transitions, impact and vertical curves. The peak tension Tmax along the belt is
determined from conventional design methods such as DIN 22101 or IOS 5048.
It needs to be noted that the above expression is somewhat idealised in as much as many other
factors need to be taken into account. These factors include splice efficiency, belt quality and
tracking characteristics.
2.3.2 Conveyor Design Program
A dynamic analysis program shown in Figure 6 uses the results of conventional handbook
procedures and modifies the design in the light of computed dynamic stresses.
The value of Td critically depends on the starting or stopping characteristic of the motor drive or
brake. An over-rated belt is progressively de-rated in this programme until a suitable f.o.s. is
achieved. Typically a f.o.s. ≥ 3.3 is obtained and the belt strength and hence cost is significantly
reduced. In the case where a steel cord belt is selected, the weight is reduced in the selection
process, leading to further reductions in tension and power.
The design cost is computed each time the static design is modified in the programme loop.
Though the flowchart of Figure 6 does not show a different path for calculating f.o.s. for starting or
stopping, the largest average acceleration in the transient period should also give the largest
dynamic acceleration in the belt. As a result, the greatest rate-of-change of belt velocity must be
employed in f.o.s. calculations.

7
Figure 6: Dynamic analysis flowchart for evaluating the starting and stopping safety factors in
conveyor belts, in conjunction with conventional design

8
2.3.3 Resonance Free Design
Elimination of belt resonances by the correct design of idler spacing, belt tensions and belt weight
will result in lower operating costs due to increased idler life and reduced belt failure caused by
vibration of the belt.
Variables that are available for design optimisation are the excitation frequency of the idlers fi and
the belt modal frequency fmn [2,3]. The condition for stable belt running is expressed by the
relation
Σq {(qfmn + Δfmn) < fi < (qfmn - ρfmn)} q = 1,2,.... (12)
where fmn is the natural frequency of the tensioned belt. Plate mechanics [5] are used to
determine these frequencies.
Figure 7 illustrates the application of Eq. (12), and shows the location of idler frequency in relation
to belt frequency.

Figure 7: Change of modal frequencies for a belt due to tension variation between loaded and
unloaded conditions
The running belt tension profile is required in order to design idler diameter d, idler spacing a, belt
width b and belt speed, so that fi < fmn or a multiple of this frequency in the case of multi-lobed
idler eccentricities. Usually only single lobe, idler eccentricities occur in manufacture.
Conventional static design, as distinct from dynamic design, only uses loading, tension and belt
sag (including takeup counterweight masses) to obtain idler spacing.
In dynamic design, the belt width is selected and then fmn and fi are calculated. The belt speed
and idler diameter, together with belt width and throughput, are adjusted and the dynamic design
is repeated until fi < f11 , and the design is complete. The belt frequencies are determined for both
loaded and unloaded conditions to obtain design ranges. Generally, for very high-speed designs
near 15 ms-1, belt width is b ~ 0.8 m, a ~ 2 m and idler diameters are d ~ 0.20 m.
It should be appreciated that while the power requirement to drive a belt increases with belt
speed, high speed belts are obviously narrow to prevent resonances, and so rolling losses are
less with narrow belts in a linear proportion to the width. However, belting and structure cost is
considerably reduced in narrow belt systems and so long-distance high speed narrow belts are
more economical. Figure 8 shows how transverse belt resonances may be eliminated by idler
spacing considerations between loaded and unloaded conditions.

9
 Figure 8: Diagram showing the removal of resonance zones by the application of change of idler
spacing in the resonance zone
4. BELT IDLER RESISTANCE AND IDLER BEARING PERFORMANCE
4.1 Idler Resistance
The determination of idler resistance has been based, traditionally, on the application of "artificial
friction" coefficients which are quite empirical. More recently, research has been undertaken to
examine the nature of belt and idler rolling resistance from a more fundamental viewpoint. This is
of particular importance in the case of long overland conveyors where the cumulative influence of
idler resistance demands an accurate determination of the individual resistance contributions.
One study performed by Harrison et al [14] has examined the various factors influencing idler
resistance. Factors taken into account are:

i. Belt sag
ii. Idler diameter
iii. Indentation effects
iv. Belt hardness
v. Temperature
vi. Idler configuration.

The simulation test rig, constructed for this purpose, is shown in Figure 9.

10
 (a) Apparatus for Testing Empty Belts in their Operating Shapes

(b) Test Set-Up for Loaded Belt

Figure 9. Simulation Test Rig for Measurement of Idler Resistance
Experiments have shown that the idler resistance is somewhat insensitive to idler diameter; all
other parameters have an influence.
The total resistance due to the carry idlers has been shown to be
FRC = (µi + µs + µf) N + FBC (13)
where:
µi = indentation coefficient depending on rubber hardness
µs = sag coefficient
µf = flexure coefficient
FBC = bearing losses of carry side idlers
N = normal load on idler set.
The components of the total resistance are defined below:
µi = 0.022 - 0.0113 / 32 (H - 55) (14)
An appropriate linear expression for the hardness H as a function of temperature is
H(T) = 0.5T + 80 (15)
where:
H = Hardness - Shore A degrees
T = Temperature °C.
The belt sag friction factors µs for steel cord, fabric and PVC belts are 0.0033, 0.0066 and 0.005
respectively, for a belt sag of 1 %. The sag friction along the whole return conveyor from head (n
= 0) to tail (x = ℓ), is computed using
µs (x) = µs [1 - x / L]-1 A (16)

11
where
A = µ0 / µs L=ℓ/B B = (1 - A)
An average value of the sag for the whole length of the conveyor is µs/2. Experimentally the
combined friction coefficient
µisf = µi + µs + µf (17)
has been measured. The flexure coefficient gf has been measured indirectly by this approach. It
is obtained by subtracting (µi + µs) from µisf.
3.2 Idler Bearing Life
The rolling fatigue life of a properly lubricated bearing is expressed in terms of the B-10 life rating.
For a sample of nominally identical bearings, the B-10 life, expressed in millions of revolutions at
constant speed or hours of operation, is the life that 90% of the bearings in the sample will attain
or exceed before the first signs of fatigue are evident. The B-10 life is expressed by:
LB-10 = 106 / 60n (C /p)u (18)
where:
n - rotational speed (in rev/min)
C - basic dynamic load rating (as given in bearing catalog)
P - actual bearing load
u - u = 3.0 for bass bearings, u = 3.33 for roller bearings.
(The load units for C and P are the same.)
It is evident from Eq. (16) that, in the case of idler bearings for belt conveyors, the life is inversely
proportional to the belt speed. However, as indicated by Talks and Kenny [27] of the National
Coal Board (NCB), U.K., the limiting factor for idler life is the grease life rather than the B-10
fatigue life. For instance, Talks and Kenny estimated that the NCB has around 11 x 106 idler
bearings in operation at any one time with load ratios C/P ranging from 13.8 to 41.9; the
corresponding B-10 lives range from twelve years to 360 years, based on continuous running at
400 rev/min. Yet, the average grease life for idler bearings is five years for operation at 50°C and
two years at 70 °C.
Much progress is being made to extend idler bearing life, and the work reported by Herraty and
Bras [28] concerning a SKF seize resistant bearing is worth noting. The seize resistant bearing
was compared, under dry abrasive test conditions, with three other bearings, a standard deep
groove ball bearing (DGBB), a DGBB with increased clearance and a taper roller bearing. The
results are shown in Table 3. As indicated, the seize resistant bearings had a B-10 life ten times
that of the DGBB with increased clearance, 15.6 times that of the standard DGBB and 70 times
that of the tapered roller bearing. The mean power consumption during all tests showed the
seize, resistant bearing to have the lowest power, being 64% of that for the standard DGBB, 78%
of that for the DGBB with increased clearance and 46% of that for the tapered bearing. The test
failure mode for the seize resistant bearing was by eventual wear without seizure, whereas the
other three bearings all failed by seizure.
A more recent paper by Herraty and Hamblin [29] has reported on the service life calculations of
idler bearings. Continued testing of the seize resistant bearing has confirmed the superiority of
this type of bearing under arduous operating conditions in dirty environments.
TABLE III
COMPARISON OF BEARING PERFORMANCE -
DRY ABRASIVE TEST CONDITIONS
(Ref. Herraty and Bras [28])
Test Mean Power
LB-10 Life
Bearing Type Failure Consumption
Revolutions
Mode (w)
Seize Resistant
1.40 x 106 Wear 28.0
(SKF 420205)
Standard DGBB
0.09 x 106 Seizure 44.0
(SKF 6205)
DGBB -
0.14 x 106 Seizure 35.7
increased

12
 Clearance
(SKF 6205-64)
Taper Roller
LM11949 / 0.02 x 106 Seizure 61.0
LM11910
4. CONVEYOR BELT AND DRIVE DRUM INTERACTION
4.1 Introductory Remarks.
An important aspect of belt conveyor operation is the frictional resistance generated between the
belt and drive drum. The resistance controls the development of drive tensions. In accordance
with Figure 10, the well known tension relationship (referred to in section 2.1.3), which is widely
used in design, is
F1 / F2 = eµθ (19)
where
F1 = Tight side tension
F2 = Slack side tension
θ = wrap angle
µ = coefficient of friction between belt and drum
It is usual to assume that µ = 0.25 to 0.3 for a steel drum surface and µ = 0.35 for a lagged drum
surface. The principal assumptions embodied in equation (19) are:

i. the frictional resistance is strictly of Coulomb type

ii. the friction is developed around the contact surface of the drive drum
iii. slip (or pending slip) occurs around the contact surface
iv. eccentricity of the surface defining cord location (and hence tensions) and drum surface
is neglected

These assumptions are not strictly valid and values normally assigned to µ in equation (19) are
not actual friction coefficients but, rather, empirical values which are lower in magnitude than the
actual values. Although it is never stated, the reduced empirical friction coefficients used in
equation (19) provide for a conservative design; they compensate for the fact that unless slip
occurs around the contact surface, the friction may not be fully developed.

Figure 10: Tensions Generated on Drive Drum

It is quite clear that the eccentricity "ec" in Figure 10 causes internal shear to be developed in the
rubber as a result of the moment generated. It is also known that the angle of wrap may be
divided into combinations of active slip, creep and passive zones as depicted in Figure 11. This
causes variations in the tensions as indicated.

13
 Figure 11: Typical Belt Friction and Tension Diagram for Combinations of Slip, Creep and
Passive Zones.
4.2 Current Research
A program of research being conducted at the University of Newcastle is aimed at providing a
greater insight into belt and drive drum friction [21]. To date, the research has involved
experimental studies using two separate test set-ups.

i. Direct Shear Apparatus - (Figure 12)

ii. Drum Friction Apparatus - (Figure 13)

The Direct Shear Apparatus of Figure 12 is based on the Jenike Direct Shear Tester. For a range
of belt and drum samples records of shear load and shear deformation were obtained for a range
of normal loads. By way of example, the fully developed friction for steel on rubber has been
shown to depend on the surface roughnesses of the belt and steel surfaces and, in particular, on
the rubber hardness.

Figure 12: Direct Shear Tests

14
 Figure 13: Drum Friction Apparatus
Tests have shown that the fully developed friction is given by
µ = Co / 2 Ra exp [H - Ho / C1 Ra] + µo (20)
where
µo = C2 Ra2 + C3Ra + C4 (21)
where Ho= 66
Co = 1.0
C1 = 10
C2 = - 2.9
C3 = 4.37
C4 = - 0.8
Ra = Roughness of steel drum
The influence of rubber hardness is illustrated in Figure 14 which shows the comparison between
the measured and predicted values of the friction coefficient; the predicted values are based on
equation (20).

Figure 14. Comparison Between Predicted and Experimental Results for three Steel Surfaces S1,
S2, S3

15
The influence of wrap angle has been examined using the drum friction apparatus of Figure 13.
Preliminary results indicate that the tension relationship does not follow the exponential form of
equation (19); a linear relationship of the form given below has been indicated.
F1 / F2 = 1 + Cθ (22)
where C = Constant
On the basis of the above relationship, the equivalent friction decreases with wrap angle
indicating that the friction coefficient is inversely proportional to the tension. However, these
findings are somewhat tentative and further tests are necessary before any firm conclusions can
be obtained. More recent experiments carried out over a wider range of wrap angles confirmed
the general exponential form of equation (19) but with significantly higher values of the friction
coefficient µ than those currently used in design. Current research is concentrating on finite
element analyses of the stress conditions in belts wrapped around drive drums; the results are
expected to provide greater insight into the fundamental nature of belt and drum friction.
5. BULK SOLID AND CONVEYOR BELT INTERACTIONS
It is important that, during the design stage, careful consideration be given to the interaction
between the conveyor belt and the bulk solid being conveyed. Interaction effects relate to such
aspects of conveyor operation as the stability of the bulk solid during transportation and the
discharge characteristics. The latter concerns the design of the discharge chute and the belt
cleaning system.
5.1 Stability of Bulk Solid on Belt During Conveying
The loading of bulk solids onto conveyor belts is governed by the need to ensure effective
transport of a maximum payload without spillage. This requirement becomes particularly
important in the case of long, overland conveyors which execute a combination of vertical and
horizontal curves.
In order to gain some insight into the load stability during transportation, a conveyor simulation
test rig was designed and constructed at the University of Newcastle, Australia [24]. The test rig,
which is illustrated in Figure 15, comprises three main components:

i. Conveyor belt clamping and tensioning mechanism;

ii. Cam driven troughing idler unit to simulate the belt and material passing over consecutive
idler sets; and
iii. Main support frame incorporating a two-way tilting base to accommodate vertical inclines
and/or the superelevation of horizontally curved conveyor belts.

16
 Figure 15. Schematic isometric drawing of the conveyor simulation test rig
6.1.1 General Specifications
The general specifications of the test rig are:

i. All belt types with belt widths up to 1050mm

ii. Carry idlers to any diameter and specification. Idler types include flat carry idlers, 3 roll in-
line and offset troughing idlers (20, 30, 35 and 45°), and 3 roll and 5 roll suspended
troughing idlers. Idler spacing is variable from 1000 - 2000 mm.
iii. Simulated belt speeds ranging from 0 - 15 m/s are possible via a 3ph A.C. geared motor
unit fed by a variable frequency electronic drive.
iv. Belt sag and/or belt tension is controlled through two hydraulic jacking cylinders. The
cylinders allow the belt to be adjusted to a present tension, or until a required percentage
sag (either for coasting or loaded) is achieved.
v. Frequency and amplitude of belt vibration is to simulate material on an actual belt
passing from a position on top of an idler set down to the maximum sag position (midway
between consecutive idler sets) and finally to the top of the next idler set.

The cam drive which controls the amplitude of vibration is adjusted so that the loaded belt moves
between the maximum sag position to level with the belt clamping and tensioning brackets (that
is, the top of the next idler set). The frequency of vibration is set to the idler passing frequency
which is a function of belt speed and idler spacing.
Although the rig does not currently allow the idler rotating frequency to be superimposed it is
believed that whilst the frequencies are higher, the amplitudes are so small that the effect of the
rotating idlers will be negligible compared to the idler passing motion described above.
5. 1. 2 Typical Test Results

17
A range of tests have been conducted over a period of time to simulate the motion of bulk
materials, notably coal and iron ore. In the case of the latter, one series of tests was concerned
with the stability of iron ore on the horizontally curved section of an overland conveyor. By way of
illustration, a typical set of results are shown in Figure 16.
The results relate to a 35 deg. idler set, the conveyor rig being set at a horizontal curve super-
elevation angle of 8 deg. The iron ore was at a moisture content 0.8% (wet basis) and of a
particle size of 90% minus 16 mm. As the results show, there was very little movement of the iron
ore.

Figure 16: Change in surcharge profile between "as loaded" and after prolonged simulated
motion for
(a) Horizontally curved conveyor at a superelevation angle of 8° with symmetrically positioned
belt on 35° troughing idlers.
(b) Horizontally curved conveyor at a superelevation angle of 8° with the belt offset 40° mm down
the 35° troughing idlers.
5.2 Bulk Solid Discharge and Carry-Back After Discharge
5.2.1 Condition Occurring at Discharge
An indication of the discharge trajectories and possible levels of bulk solid carry-over following
discharge may be ascertained by considering, in a simplistic way, the mechanics of separation of
bulk solids from the belt as it passes around the discharge pulley [13,19]. Figure 17 shows the
condition at the conveyor discharge.
The equilibrium condition for the position at which discharge commences is given by
2
v / rg = cosθ + σo / ρgh (23)
where v = velocity of bulk solid leaving belt (m/s)
r = radius of bulk solid measured from drum centre (m)
g = gravitational acceleration
θ = angle at which discharge commences
σo = adhesive stress (kPa)
ρ = bulk density (t/m3)
h = thickness of bulk solid stream (m)

Figure 17: Conveyor Belt Discharge

The adhesive stress, depicted in Figure 18, is obtained from the Surface Yield Locus measured in
the laboratory.

18
 Figure 18. Surface Yield Locus for Conveyor Belt
Referring to Figure 19, the average height of material on the belt is given by
h=Q/ρvb (24)
where
Q = throughput (t/h)
b = width of contact of bulk solid on belt (m)

Figure 19. Bulk Material on Belt at Discharge

Discharge will commence at the line of intersection of the belt and drive drum, that is, for θ = 0.
The minimum belt velocity for this to occur follows directly from (23) and is given by:
vmin = (r σo / ρ h)½ (25)
By way of example, consider the belt conveying of bauxite. While the bulk density varies with the
consolidation pressure, it is assumed that p = 1.2 t/m3 corresponding to the likely consolidation
condition (from flow property data). Measurements of surface yield loci of bauxite indicate that
adhesive stresses up to 1 kPa may occur. Assuming a value of h = 0.09 (m) and a radius of R =
0.6 m, then
vmin = 2.35 m/s
Hence, if the belt speed v ≥ 2.35 m/s, then discharge will commence at the line of intersection of
the belt with the drum.
5.2.2.Carry-Over After Discharge
Referring to Figure 20, assuming the bulk solid does not slide relative to the belt, discharge takes
place as discussed previously. For cohesive bulk solids, sliding is unlikely in view of the very high
friction angle that accompanies the low normal pressures between the bulk solid and the belt.

19
 Figure 20: Mechanics of Discharge and Adhesion
After the point of discharge defined by equation (23), some bulk solid will adhere to the belt as
carry-over. This is an adhesive condition that arises as a result of negative or tensile stresses σnt
which occur between the bulk solid and the belt surface, as indicated in Figure 18. While these
stresses are very small, they can have a significant influence on the amount of carry-over that
may occur. An indication of the thickness h of the carry-over may be obtained as follows:
For an arbitrary location defined by θ > θL, where θL is the angle at which discharge commences,
it follows that the thickness h of the carry-over is given by
h = σnt / ρ [1000 / (v2 / R) - gcosθ] (26)
where
h = carry over on belt (mm)
σnt = tensile stress (kPa)
ρ = belt density (t/m3)
v = belt velocity (m/s)
R = pulley radius (m)
g = gravitational acceleration
The critical location occurs when θ = 180°, that is at the bottom of the pulley. For this location σnt
= σo (see Figure 18); the thickness h is given by
h = σo / ρ [1000 / (v2 / R) - gcosθ] (27)
By way of illustration, values of σo as high as 1.4 kPa for iron ore at 11% moisture content on
conveyor belt surfaces have been measured. It is usual for values lower than this to occur.
Assuming, for example, that σo = 0.1 kPa and density ρ = 1 t/m3, then σo / ρ = 0.1. From the
foregoing, it follows that for a belt travelling at 3 m/s and assuming a 1 meter diameter discharge
pulley, bulk solid will adhere to the belt in thickness up to h = 3.6 mm. Values such as this give an
inclination of the amount of bulk solid to be handled by a belt cleaner.

20
 Figure 21: Belt Carry-Over at Discharge Pulley Expressed in Non-Dimensional Form
The carry-over analysis may be generalised by presenting the parameters in non-dimensional
form.
From equation (27) it follows that
NA = 1 / N F + 1 (28)
where
NA = ρgh / σo = Adhesion Number (29)
NF = v2 / Rg = Froude Number (30)
5.2.3 Belt Cleaner Operation
Effective belt cleaning depends on an appreciation, at the design stage, of the amount of likely
carry-over of bulk material after discharge. It is unlikely that a single cleaning device will be
sufficient and a multiple cleaner system is recommended.
Apart from the selection of the most appropriate cleaning device, it is important that sufficient
attention be given to the need to ensure effective removal of bulk material without build-up on the
cleaner components. As illustrated in Figure 22, build up can occur on bolt projections and
ledges, thus impairing the performance of the cleaner. Such projections should be avoided.

21
 Figure 22. Belt Cleaner System Showing Areas of Possible Build-Up
6. OTHER RESEARCH DEVELOPMENTS
While it is beyond the scope of this paper to discuss, in any detail, other areas of research, it is of
interest to note areas in which research is research is presently being conducted.
6.1 Belt Tracking and Dynamic Tensions
Two of the more important parameters needed by the designer of belt systems are the dynamic
tracking characteristics and dynamic belt tensions that occur on starting and stopping. Both of
these parameters are difficult to predict during the design phase. There are no theories to predict
belt tracking characteristics based on belting construction, particularly with a bonded steel cord
belt in which cord load sharing is not always uniform. Research being conducted by Harrison [22]
is addressing this problem. A dynamic sag measuring transducer has been developed and
monitoring of belt tracking and sag during starting and stopping has provided new ways of
proving that these parameters have significant importance. The tracking and dynamic
characteristics may be predicted in advance during belt manufacture.
6.2 On Line Control of Belt Dynamics
Using the sag transducer mentioned in 6.1 above, Harrison has provided a methodology for the
on line control of conveyor belt dynamic characteristics [241. The analysis takes into account the
design and location of belt tensioning devices.
6.3 Conveyor Belt Monitoring
Following the development of the steel cord belt monitor [5], a fabric belt monitor has been
developed by Harrison [17]. This device employs non-contact capacitive transducers and can
readily detect cuts in the cover material, edge damage, repairs, splices and a number of
manufacturing defects. The monitor is shown schematically in Figure 23. The monitoring device
has scope for introduction into manufacturing processes in order to provide improved quality
assurance.

22
 Figure 23. Monitoring System for Fabric Belts
6.4 Use of "Kevlar" Aramid Fibres in Conveyor Belts
Research at the University of Newcastle by Betz et al [11,18] has concentrated on the
improvement of splices in "Kevlar" reinforced belts. Various splice configurations have been
examined and it has been shown that splice strengths greater than the parent belt are possible.
Two and three dimensional finite element models have been developed to analyse the stress
distributions in splices.
7. CONCLUDING REMARKS
It is clear that significant developments are continuing to be been made in the field of belt
conveying. Like other countries in the world dependent on bulk materials handling, Australia has
been active in research and development work in the various facets of bulk handling, including
belt conveying. Same of the developments that have occurred in recent years in Australia have
been highlighted in this paper. It is particularly gratifying to note the wide industrial acceptance of
modern design methods and monitoring procedures. The interaction between industry and
research institutions is of great importance in guiding current research and indicating new
research directions.
8. REFERENCES

1. Roberts, A.W., Hayes, J.W. and Scott, 0.J.,"Optimum Design of Continuous Conveyors".
Bulk Solids Handling, Vol. 1, No. 2, 198 1.
2. Harrison, A., "Flexural Behaviour of Tensioned Conveyor Belts", Mech.Eng.Trans. IE
(Aust.), Vol. MES (1983), No. 3.
3. Harrison, A. and Roberts, A.W., "Technical Requirements for Operating Conveyor Belts
at High Speed", Bulk Solids Handing, Vol. 4 (1984), No. 1, pp.99-104.
4. Harrison, A., "Criteria for Minimising Transient Stress in Conveyor Belts",
Mech.Eng.Trans IE(Aust), Vol. ME8 (1983), No. 3, pp. 129-134.
5. Harrison, A, "Dynamic Measurement and Analysis of Steel Cord Conveyor Belts", Ph.D.
Telesis, The University of Newcastle, Australia, March, 1984.
6. Harrison, A. and Roberts, A.W., "Future Design of Belt Conveyors Using Dynamic
Analysis". Paper presented at Transportation 1984, IE(Aust).
7. Roberts, A.W. and Hayes, J.W., "Economic Analysis in the Optimum Design of
Conveyors", TUNRA, The University of Newcastle, Australia, 2nd Ed., 1980,. ISBN 0-
7259-3406.
8. Roberts, A.W., Hayes, J.W. and Scott, 0.J., "Optimum Design of Continuous Conveyors",
Bulk Solids Handling, Vol. 1 (1981), No. 2, pp.255-264.
9. Harrison, A., Hayes, J.W. and Roberts, A.W., "The Feasibility of High Speed Narrow Belt
Conveying for Bulk Solids Handling":, Mech. Eng. Trans. IE(Aust.), Vol. ME7, (1982), No.
3.
10. Roberts, A.W., Harrison, A. and Hayes, J.W., "Economic Factors Relating to the Design
of Belt Conveyors for Long Distance Transportation of Bulk Solids", Int. Jnl. of Bulk Solids
Handling, Vol. 5, No. 6, December 1985 (pp. 1 143-1149).
11. Betz, E., "The Use of "Kevlar" Aramid Fibres in Conveyor Belts, Part 1: Pull-Out Problem
and Splice Design", Intl. Jnl of Bulk Solids Handling, Vol. 6, No. 2, 1986 (pp.355-367).

23
12. Harrison, A., "Troughability Measurement of Fabric Reinforced Belting for the Powder
and Bulk Handling Industry", Intl. Jnl. of Bulk Solids Handling, Vol. 7, No. 3, June 1987
(pp.381-384).
13. Roberts, A.W., Ooms, M. and Bennett, D.J., "Bulk Solid Conveyor Belt Interaction in
Relation to Belt Cleaning", Intl. Jnl. of Bulk Solids Handling, Vol. 7, No. 3, June 1987
(pp.355-362).
14. Hanison, A., Teo, L.H. and Roberts, A.W., "Measurement of Belt-Idler Interactions and
Material Flexure Coefficients for Design of Troughed Conveyor Systems", Intl. Jnl of Bulk
Solids Handling, Vol. 7, No. 3, June 1987 (pp.367-371).
15. Harrison, A., "Future Design of Belt Conveyors Using Dynamic Analysis", Intl. Jnl. of Bulk
Solids Handling, Vol. 7, No. 3, June 1987 (pp.375-379).
16. Attwood, R. and Smith, B.D., "Assessment of Air Velocity Profiles in Belt Conveyor
Galleries and their Drying Effect on Brown Coal During Transportation", Inti. Jnl. of Bulk
Solids Handling, Vol. 7, No. 3, June 1987 (pp.385-390).
17. Hanison, A., "A New Development in Textile Belt Monitoring", Inti. Jnl. of Bulk Solids
Handling, Vol. 8, No. 2, April 1988 (pp.231-233).
18. Roberts, A.W., Betz, E., Goh, B.N., Law, S.C. and Wang, W.C., "The Use of "Kevlar"
Aramid Fibres in Conveyor Belts, Part II: Static Strength of Splices", Inti. Jnl. of Bulk
Solids Handling, Vol. 8, No. 4, August 1988 (pp.429-441).
19. Roberts, A.W. and Bennett, D.J., "Friction, Adhesion and Wear in Conveyor Belt Cleaning
Operations", AntiWear 88, The Royal Society, London, 1988 (pp.28.1-28.8).
20. Harrison, A. and Roberts, A.W., "Modern Concepts in Belt Conveying and Handling of
Bulk Solids", TUNRA, The University of Newcastle, Australia, 1988.
21. Roberts, A.W., Papaliski, D. and Harrison, A., "The Friction and Tension Characteristics
on Driving Drums of Conveyor Belts", Proceedings, 12th Intl. Power and Bulk Solids
Handling Conference, Chicago, U.S.A., May 1988.
22. Harrison, A., "New Concepts for Evaluating Belt Tracking and Dynamic Tensions", Third
Inti. Conf. on Bulk Materials, Storage, Handling and Transportation, IE(Aust), Newcastle,
June 1989 (pp. 114-118).
23. Harrison, A. and Barfoot, 0., "Modelling the Effect of Take-Up Ucation on Conveyor Belt
Performance", Third Intl. Conf. on Bulk Materials, Storage, Handling and Transportation,
IE(Aust), Newcastle, June 1989 (119-123).
24. Bennett, D.J. and Roberts, A.W., "Bulk Solid and Conveyor Belt Interaction During
Transportation", Third Intl. Conf. on Bulk Materials, Storage, Handling and
Transportation, IE(Aust), Newcastle, June 1989 (pp.345-358).
25. Hardson, A., "Stress Distribution in Steel Cord Belts with Cord Plane Defects and Inlaid
Repairs", Intl. Jnl. of Bulk Solids Handling, Vol. 8, No. 4, Aug. 1988 (pp.443-446).
26. Morrison, W.R.B., "Computer Graphics Techniques for Visualizing Belt Stress Waves",
Intl. Jnl. of Bulk Solids Handling, Vol. 8, No. 2, April 1988 (pp.221-227).
27. Talks, M.G. and Kenny, P., "The Evaluation of Some Fire-Resistant Greases as
Lubricants for Conveyor Idler Bearings", Proc. Tribology in Mineral Extraction War on
Wear Conf., Sept. 1984, I.Mech.E., University of Nottingham.
28. Herraty, A.G., and Bras, J.C.M., "Rolling Bearings for Hostile Environments", Proc.
Tribology in Ifineral Extraction War on Wear Conf., Sept. 1984, I.Mech.E., University of
Nottingham.
29. Herraty, B. and Hamblin, R.F., "Idler Rollers - Observations and Predictions", AntiWear
88, The Royal Society, London, 1988 (pp.27.1-27.8).

24