DYNAMIC LOAD SHARING FOR CONVEYOR BELTS WITH MULTIPLE DRIVE STATIONS

FRANK T. CHURCHILL
MEMBER IEEE
Reliance Electric - Dodge CST Business
P.O. Box 8 1 0 8 5
Seattle, WA 9 8 1 0 8 U.S.A.

Abstract-The characteristics and limitations of multiple between idlers. For the purposes of this discussion, the
drive station belt conveyors are explored and a effects of idlers, skirt boards, etc., are ignored.
computational method is presented as a means of providing The power imparted by the drive is proportional t o belt
drive coordination which can skirt some of the limitations. speed (mostly a relatively constant speed) and t o the
This paper will describe some of the design parameters difference in belt tensions going onto (TI) and coming off
of concern for belt conveyors, including belt tension and of (T2) the drive pulley. According t o CEMA, the maximum
h o w it varies with load, h o w the unpredictably non-uniform power that can be delivered t o the belt by the pulley is
loading of the belt affects performance, and will explore the bounded by the relationship in equation 1 111.
relationship of friction between the belt and a drive pulley
and h o w this limits the amount of horsepower which can T2 <cw
be transmitted t o the belt. The w a y in which these T I- T I
parameters change in configurations with multiple drives
and multiple drive stations is outlined. The t w o main In this equation, the wrap factor, ,C, is relatively
means of employing booster drives are presented; tripper constant for unvarying belt conditions and is a function of
booster and linear booster. the wrap angle of belt contact, the materials of belt cover
The tripper booster technology is amenable t o tension and pulley lagging, and the presence of water. The exact
control, in w h i c h the torque provided by each drive station relationship among all these parameters has been explored
is controlled by the tension in the belt measured but, as yet, is not well understood (21. Practical values for
immediately downstream of the drive station. This a single pulley drive might be 0.5 for a lagged pulley. In
approach has proved successful in belt systems with as other words, the slack side tension T2 must be at least
many as four booster drive stations, operating at belt loads one-third the tight side tension. This implies that high
in excess of 3200 tons per hour over 12,000 f t of belt. power inputs will require high values of takeup tension,
The linear booster arrangement does not allow a which in turn implies relatively high stresses in the takeup
convenient method of measuring belt tension, thus a components, tail pulley idlers, bearings and structure.
computational algorithm w a s devised to take advantage of For economic and reliability reasons, it is desirable to
the geometry of the belt system t o control tension of the minimize stresses in all components other than the drive
belt within certain limits without the use of tension pulleys usually by reducing the wrap factor. The wrap
feedback. This approach has proven useful in 10,000 foot factor can be improved by increasing the effective wrap
belt system w h i c h has a single booster drive station, angle with the use of a snub pulley (Figure 2A). Geometry
operating a t loads up t o 4000 tons per hour. limits this approach t o a wrap angle of 240' or so
(C,=0.3). The next step of evolution is t o employ t w o
BACKGROUND drive pulleys (Figure 2B). This arrangement can produce a
C,=0.08. This is equivalent t o allowing the tight side
Conveyor belt technology has changed enormously over tension TI t o be as much as 13.5 times the slack side
most of this century. Starting with the contributions of tension T2.
Thomas Edison and others in the early days, the Other limitations t o increasing power input t o the belt
transportation of bulk materials has ever increased in include belt speed being limited by loading and unloading of
capacity and distance of transport. Today it is not unusual bulk material a t high speeds, and the strength of the belt
t o speak of transporting thousands of tons of material per itself. Belt strength has improved over the years but
hour over distances measured in miles on a single belt. economics plays a big role in belt selection. In some
Reliability of belt systems has improved t o the point that applications, belting must be fabric based for splicing
belt availability approaching 100% is not unusual. considerations.
Of course, all of these improvements bear a fairly high Economics and some physical limitations have led t o the
cost. The cost of belting alone for single belt can be in the design of belt conveyors with multiple drives which are
millions of dollars. This is largely traceable t o the increase spatially separated along the conveyor at booster stations
in belt tension as a function of increasing load on the belt, which provides a means of reducing the peak belt tension
requiring ever stronger belting. Commensurate with higher for a given amount of total belt horsepower [31.
belt tensions are increases in load capacities for pulleys, Figure 3 demonstrates h o w this works. Assuming a
bearings and structure. uniformly loaded conveyor, concentrating all of the drive
To illustrate some of these limitations, consider the power at the head pulley produces a large maximum
simple uphill conveyor in Figure 1 consisting of a head tension at the head pulley (solid curve). If instead w e split
pulley, tail pulley and takeup mechanism. The head pulley the drives into t w o pieces, with one a t the head and the
is driven to impart conveyor movement, the tail pulley idles other halfway d o w n the belt, the effective maximum
along and backstops takeup tension and the takeup tension is split in half (dashed line).
mechanism keeps the belt taught t o prevent excessive sag

PAGE
 Two different drive configurations are in use 81s booster b) For ONLY2 S KWSP(O1 < 100
stations, the tripper booster and the linear booster (Figure
4). In the tripper booster, the belt is wrapped around t w o KWSP(H) = 2 (ONLY2)+
+

drive pulleys such that material spills off the belt which is 2 (KWSPIO)-ONLY2) (50-ONLY2)
exiting the bottom pulley. In the linear booster, the main (100-ONLY21 (3b)
belt is piggy-backed o n top of an auxiliary belt such that KWSP(B) = 100 (KWSP(O)-ONLY21
+

the friction between the belts provides power transfer from (1 00-ONLY2) (4b)
the auxiliary belt t o the main belt. The effective tension
diagrams for these cases are also shown in Figure 4. c) For 100 IKWSP(0)

PROBLEM STATEMENT (3c)

(4c)
Both of these systems are practical and each has its
advantages and disadvantages. Both systerns have a where: KWSP(0) is total power setpoint
problem with non-uniform belt loading. Consider the KWSP(H) is head drive setpoint
diagrams of Figure 5. In part a, the load is all on the KWSP(B) is booster drive setpoint
upstream half of the belt. This works well, as the total belt ONLY2 is the booster drive power breakpoint
loading is only 50% with the tension between the head and All variables are in percent
booster reaching half its rated value. Part b shows a fully
loaded belt with peak tensions a t rated values. IPart c A computer simulation was devised t o verify that the
displays the belt with only the part between the head and algorithm would perform satisfactorily under a variety of
booster loaded. In this case, equal load sharing between load distributions. A belt configuration was chosen with a
the head and booster drives results in negative belt tension length of 10,000 feet with the linear booster situated
just downstream of the booster. The pratical between 4000 and 5000 feet. This approximated the
consequences of negative belt tension in front of the design of the physical conveyor system t o vvhich the
booster is gross belt sag in-between the idlers which algorithm was ultimately applied.
ultimately leads t o destruction of idlers andl icoriveyor Figure 8 shows a printout of the calculated belt tensions
structure along with damage t o the belt. In some systems using a breakpoint of 0% which implies equal load sharing
negative tension can be overcome by increasing belt takeup a t all times. The simulation predicts negative tensions
tension t o ensure that belt tensions may drop fairly low when most of the load is between the head and booster.
without actually becoming negative. Note that the column labeled LOAD shows the load
In other systems, however, manipulation 01 takeup distribution injected into the system.
tension is impractical. Over the past f e w years, a nlJmber A n identical input load distribution was used in Figure 9,
of systems utilizing tripper boosters have employed belt but with the breakpoint set a t 40%. In this case, the
tension measurement t o control the power contributions of tensions are well behaved and non-negative.
the boosters (Figure 6). Thus, the 'power sharing is Quite a f e w simulations were run with variable iinput load
manipulated t o suit the changing load distributiorl on the distributions which might approximate the real world.
belt. Tension control has been successfully used with belts Figure 10 is an example of one of these.
employing as many as three booster stations distributed
along the belt. IMPLEMENTATION
With linear boosters, the situation is more difficult as no
means of measuring belt tension has been developed. The algorithm was then implemented in the main slope
belt of an underground coal mine. In this application, t w o
PROBLEM SOLUTION 500 hp drives constituted the head drive station and t w o
additional 500 hp drives made up the booster drive station.
A computational algorithm was devised t o a1locai:e power Belt length was little over 10,000 f t with the booster
between the head drive and booster drive based upon the installed at 3,800 t o 4,500 f t from the head. Figure 11
total power required by the load. The design phiilosophy displays the controls implementation which involved
w a s simply t o utilize the head drives only for belt loads cascaded PID loops.
form zero up t o a breakpoint percentage. Above the The outer loop was a speed control loop whose setpoint
breakpoint, the power level of the booster is incriEased as a was set by the logic of the belt controller (accelerate, run
function of the total loading. For loads a t or above! 100%, at speed, decelerate). The speed control loop established a
the head and booster drives share load equally (Figure 7 ) . power setpoint for the entire belt. The distribution
The breakpoint percentage is affected by the position of the algorithm was used t o calculate power setpoints for the
booster relative t o the length of belt. individual drives. Each individual drive was controlled by a
The equations take the form: power control loop based upon the setpoint provid'ed by the
algorithm.
a) For KWSP(0) < ONLY2: This installation proved t o provide effective and stable
control of this particular belt.
KWSP(H) = 2"KWSP(O) (3a)
KWSP(B) = 0 (4al

-- 121
PAGE
CONCLUSION

A n effective straightforward method of controlling the

drive power distribution of a belt with booster station in the
absence of tension control w a s successfully designed and
implemented.

REFERENCES

11 1 CEMA: Belt Conveyors for Bulk Materials; Fourth

Edition; The Conveyor Equipment Manufacturers
Association: 1994.

121 Harrison, A. and Roberts, A.W.; Mechanisms of

Force Transfer on Conveyor Belt Drive Drums; Bulk
Solids Handling; Volume 12; Number 4; Trans Tech
Publications; 1992.

[31 Harrison, A.; Power Oscillation Mechanisms in

Booster Conveyors; International Mechanical
Engineering Congress and Exhibition; Sydney,
Australia; 1991.

122
PAGE
FIGURE 1

123
FIGURE 2

124
 \
\
\
\
\

-
I- L d z: tn - 0 7

125
 n
Q
'W
I

I
c
9,
I VI
Q,
DL 0
4- 0
v)
0 W m
0 I- L *
m -0
L
Q, 0
n 0
.-CL
I m
I-
h
0

I-wzm-oz

126
 1

T
E
N
S
I
0
N

FIGURE 5

127
A
A
W
0
n

128
 n
0
U
a
s
Y

0 n
‘ 0 6\”
7
W

c
-
7
0
a,
I-
w
v,
E
0
>
W
>
(v 7
> 0
-I
Z 0
0

129
ONLY2= 0

LOAD BOOSTER HEAD

KWT 9000 8000 7000 6000 5000 4 0 0 0 3000 2000 1 0 0 0 0000 FEET
1 1 1 1 1 1 1 1 1 1 1
5 4 5 5 5 5 3 3 3 3 3
12 7 11 1 2 12 12 6 6 6 6 6
22 10 17 21 22 22 11 11 11 11 11
32 10 20 27 31 32 16 16 16 16 16
42 10 20 30 37 41 21 21 21 21 21
52 10 20 30 40 47 25 26 26 26 26
62 10 20 30 40 50 26 30 31 31 31
72 10 20 30 40 50 24 31 35 36 36
82 10 20 30 40 50 19 29 36 40 41
31 10 20 30 40 50 15 25 35 42 46
97 10 20 30 40 50 12 22 32 42 49
100 10 20 30 40 50 10 20 30 40 50
98 8 18 28 38 48 9 19 29 39 49
93 5 13 23 33 43 7 17 27 37 47
85 2 7 15 25 35 3 13 23 33 43
75 0 2 7 15 25 -2 8 18 28 38
65 0 0 2 7 15 -7 3 13 23 33
55 0 0 0 2 7 -12 -2 8 18 28
45 0 0 0 0 2 -15 -7 3 13 23
35 0 0 0 0 0 -15 -10 -2 8 18
25 0 0 0 0 0 -12 -10 -5 3 13
15 0 0 0 0 0 - 7 -7 -5 0 8
7 0 0 0 0 0 - 3 -3 -3 -1 4
2 0 0 0 0 0 - 1 -1 -1 -1 1
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0

FIGURE 8

130
ONLY2= 40

LOAD BOOSTER HEAD

KWT 9000 8000 7000 60100 5000 4000 3000 2000 1000 0000 FEET
1 1 1 1 1 . 1 1 1 1 1 1
5 4 5 5 5 5 5 5 5 5 5
12 7 11 12 12 12 12 12 12 12 12
22 10 17 21 :2X 22 22 22 22 22 22
32 10 20 27 ;31. 32 32 32 32 32 32
42 10 20 30 337 41 41 41 41 41 41
52 10 20 30 40 47 41 42 42 42 42
62 10 20 30 40 50 39 43 44 44 44
72 10 20 30 40 50 34 41 45 46 46
82 10 20 30 40 50 25 35 42 46 47
91 10 20 30 414 50 18 28 38 45 49
97 10 20 30 40 50 13 23 33 43 50
100 10 20 30 40 50 10 20 30 40 50
98 8 18 28 38 48 10 20 30 40 50
93 5 13 23 33 43 9 19 29 39 49
85 2 7 15 :25 35 8 18 28 38 48
75 0 2 7 It5 25 6 16 26 36 46
65 0 0 2 7 15 5 15 25 35 45
5 5 0 0 0 2 7 3 13 23 33 43
4 5 0 0 0 0 2 3 11 21 31 41
35 0 0 0 0 0 2 7 1 5 2 5 3 5
25 0 0 0 0 0 0 2 7 1 5 2 5
15 0 0 0 0 0 0 0 2
' 7 1 5
7 0 0 0 0 0 0 0 0 2 7
2 0 0 0 0 0 0 0 0 0 2
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 r 3 0 0 0 0 0

FIGURE 9

131
ONLY2= 40

LOAD BOOSTER HEAD

KWT 9000 8000 7000 6000 5000 4000 3000 2000 1000 0000 FEET
3 3 3 3 3 3 3 3 3 3 3
7 4 7 7 7 7 7 7 7 7 7
11 4 8 11 11 11 11 11 11 11 11
19 8 12 16 19 19 19 19 19 19 19
28 9 17 21 25 28 28 28 28 28 28
38 10 19 27 31 35 38 38 38 38 38
49 11 21 30 38 42 39 42 42 42 42
58 9 20 30 39 47 36 40 43 43 43
66 8 17 28 38 47 34 38 42 45 45
75 9 17 26 37 47 27 35 39 43 44
81 9 18 26 35 46 22 31 39 43 47
87 10 19 28 36 45 17 27 36 44 48
94 11 21 30 39 47 11 22 32 41 49
92 6 17 27 36 45 10 19 30 40 49
87 4 10 21 31 40 10 18 27 38 48
81 4 8 14 25 35 10 19 27 36 47
73 3 7 11 17 28 11 20 29 37 46
69 5 8 12 16 22 9 19 28 37 45
65 4 9 12 16 20 6 17 27 36 45
60 4 8 13 16 20 8 14 25 35 44
59 8 12 16 21 24 13 17 23 34 44
59 10 18 22 26 31 19 23 27 33 44
57 9 19 27 31 35 26 29 33 37 43
60 9 18 28 36 40 28 33 36 40 44
64 8 17 26 36 44 28 32 37 40 44
69 9 17 26 35 45 29 33 37 42 45
76 10 19 27 36 45 25 33 37 41 46
79 8 18 27 35 44 21 31 39 43 47
80 5 13 23 32 40 16 25 35 43 47
80 4 9 17 27 36 11 20 29 39 47
76 4 8 13 21 31 10 18 27 36 46
66 0 4 8 13 21 10 19 27 36 45
57 0 0 4 8 13 7 17 26 34 43
48 0 0 0 4 8 7 15 25 34 42
40 0 0 0 0 4 8 13 21 31 40
31 0 0 0 0 0 4 8 13 21 31
21 0 0 0 0 0 0 4 8 13 21
13 0 0 0 0 0 0 0 4 8 13
8 0 0 0 0 0 0 0 0 4 8
4 0 0 0 0 0 0 0 0 0 4
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0

FIGURE 10

132
 Y
0
a
a2
FI
w
W
LL
>-
I-
U
0
U
J
W
>

@----

133