Alternatives to Strain Gauging for Accurate Determination of Torque Required when Extending Belt Conveyors 

Alex Harrison, USA
Courtesy : Trans Tech Publications  Bulk Solids Handling Journal
Summary
common problem arising in the open pit industry is the determination of motor torque required to drive a belt conveyor when belt length increase is proposed, belt lift is changed, or tonnage increases are sought. Strain gauges attached to the drive shafts of the conveyor may be unreliable for many reasons and therefore reliable alternatives have been developed and tested that give insitu values of running torque and conveyor friction. Three alternative nonstrain gauge methods are available that can be used as a crosscheck against each other, including belt vibration to tension analysis, belt sag distribution to tension and dynamic direct belt tension measurement of T_{1} , T_{3} and T_{2} values. Additional valuable information on belt quality (loadsharing of stress) and takeup dynamics is a byproduct of these testing procedures. Permanently installed tension monitors will indicate increases in torque requirement due to conveyor component wear, seizure, or gradual misalignment. Application to design upgrading is discussed in this paper.
1. Introduction
Conveyor design has evolved to the point where belt tension computations are readily modelled by a variety of calculation methods such as ISO 5048, CEMA and DIN 22101. A particularly noteworthy trend in modern, high capacity mines is the need to upgrade conveyors to higher capacities, longer lengths or to change the lift; characteristics as the mine terrain advances.
There are technical considerations related to the continually varying mine conveyor, including:
belt safety factor characteristics [1]
belt dynamic response to starting or stopping [2]
belt quality and conditionjoining new belt to old belt [3]
drive torque capability evaluation.
The last subject involves quantifying the existing conveyor in terms of drive power and torque requirement particularly if there are nonlinear torque losses at the drive coupling interface that result in drive speed slip on starting or running.
Three distinct methods are available to determine the torque levels required by existing conveyors in relation to modelled torques, namely:
Direct strain gauging of the drive shafts
Direct measurement of belt tension by load cells
Indirect measurement of belt tensions T_{1} and T_{2}.
In this paper, a brief overview of methods 1 and 2 will be presented. The method of indirect tension determination (method 3), however, is of greater interest due to its accuracy and practical implementation on site.
2. Conveyor Torque Problem
When a conveyor is required to carry material over longer distances by increasing lift or by extending the belt, the question of torque required becomes an important factor due to the possible need to install new motors. The installation of new motors may be critical to the operation of the modified conveyor. Motors installed at an incorrect location of the new profile, or motors that have insufficient torque, could result in an inability to start the belt.
Similar comments apply to underground coal mining belts where the undulations of the seam, together with extending or retreating paneis, may result in a conveyor that has inadequate belt tension rating, carrying capacity or drive torque capacity. Measurement of conveyor torque (M_{c}) of an existing conveyor with a single equivalent drive drum (radius r) is governed by the equation
M_{c} = f_{1} (r, T_{1}, T_{2}, Q) (1)
= T_{e}. r = (T_{1}  T_{2}) . r (2)where T_{1} and T_{2} are the tight and slack side tensions and the temperature Q is in C. The function f_{1} is a multiparameter by nature.
The motor torque M is not necessarily transferred to the drive drum with linearity during startup if there are viscous couplings in the circuit, and so the motor torque required to move the conveyor may be a complex function of speed v
M_{m} = f_{2} (v, 
. 
) . M_{c} (3) 
v 
which involves a dimensioniess function f_{2}. For a locked coupling f_{2} = 1, approximately.
Measurement of M_{c} by a variety of methods will permit a reasonable understanding of the operation of the present system. Extension to longer systems requires modelling of the changes in T_{1} and T_{2} tension followed by comparison of "old" conveyor torque (measured) to "new" conveyor torque (predicted). Comparison with motor currents ("old") with what will be expected ("new") can also be made. A basic equation for determining the "new" torque from present torque measurements is
M_{c}(required) = (M_{c}/r + ΔT_{e}) . r (4)where ΔT_{e} is the change in the effective tension due to the proposed modification. Measurement of M_{c} and prediction of the additional T_{e} is therefore required to determine future torque requirements from the drive. A compound drive (primary, secondary or separated drive) requires torque to be split between drives.
3. Review of Methods
3.1 Strain Gauging
Strain gauging and telemetry methods require accurate application of strain sensors to ensure proper bonding to the steel shaft, proper orientation to ensure bending stress is removed from the torsional strain measurement, proper location of the gauge in relation to shaft diameter changes, accessibility and multichannel telemetry. Microstrain variations need to be calibrated accurately, and the steel modulus is required to calculate the stress, at shaft. The torque needs to be computed based on assumptions about the actual radial arm from the belt line to the strain gauge. Zones of influence also need to be taken into account well in advance by preferably Finite Element modelling of the enddisc/ shaft to determine zones for placing the gauge in a uniformly strained region.
3.2 Load Cell Application
Load cells need to be installed under idler supports or under pillow blocks of bend pulleys to monitor the T_{1} and T_{2} tension loads. Monitoring T_{1} and T_{2} of an existing conveyor requires
calibration of the load cells
installation in high tension areas
vector analysis of the belt forces around bend pulleys or over instru mented idlers.
There are practical considerations and limitations in the installation of load cells that are overcome by method 3 if onsite data is required from any conveyor by a general method.
3.3 Indirect T_{1} and T_{2} Measurement Methods
Considerable success has been achieved with the use of natural "belt quality" measurements of frequency and sag. Measurement of frequency of a belt under tension gives a direct ability to compute the belt tension using plate theory methods [3]. The essential processes used in the measurement of T_{1} and T_{2} tension based on frequency include
installation of noncontact displacement sensors at T_{1} and T_{2}
excitation of the natural modes of transverse plate vibration of the belt
collection of the platemode frequencies
computation of the belt tension T_{1} and T_{2}.
These methods can be easily crosschecked by other means without permanently installing hardware such as is the case in section 3.2. The tension of a belt, based on a measured modal frequency is
f = 1/2π (√af_{3} + bf_{4}) (5)where a is the belt stiffness/mass ratio, f_{3} is a nondimensioniess frequency parameter that depends on the shape of the measured mode, b is the tension (T) to mass (m) ratio, and f_{4} is a modal constant [3]. When b = T/m, the value of T can be determined (T_{1} or T_{2} depending on 1ocation of test).
The value of T_{1} or T_{2} can be readily crosschecked by a sag measurement process in which belt tension is given by a function of sag and belt mass "m" [4],
T = f_{5} (sag, m) (6)Both frequency and sag methods lend themselves to easy setup in the field and to internal crosschecking of conveyor torques.
Accuracy of the methods has been established over many years and typically a T_{1} or T_{2} value of 5% error is obtainable with a good measurement setup.
3.4 T_{2} Belt Tension Measurement
In many conveyors, the determination of T_{2} belt tension is facilitated by an ability to monitor takeup rope loads, by the vibration method applied to ropes. This gives an ability to measure T_{2} and crosscheck its value in three ways, namely, by rope frequencytotension, belt frequencyto tension and sagtotension measurements.
In addition to being able to monitor tensions in a rope by the vibration method, a number of rope falls in a typical takeup system may be tested, from which
takeup mass is determined directly
takeup friction can be determined in relation to dynamic sag measurements on start or stop
real T_{2} belt value is determined irrespective of rope/sheave friction losses.
The ability to determine T_{2} accurately has considerable value when predicting a system's performance requirements, since the takeup tension value forms the basis for setting the high tension in the beit and the effective tension around the drive drum.
4. Permanent Installation Applications
Advantages of these simple methods of monitoring belt T_{1} and T_{2} tension include longer term sensing of drift in installation friction due to shiftable structures and wear of idlers. Additional friction in the takeup that could predispose the system to slip can be monitored as can the effect of grade changes and tonnage variations on motor torque requirement. Permanently installed monitors can be periodically interrogated to monitor changes in operating torques in situations where motor ratings are close to maximum limits of motor operation.
5. Concluding Remarks
In this brief paper, methods have been described that are used to monitor belt T_{1} and T_{2} tension directly, from which conveyor torque values are determined for future upgrades. There are considerable practical advantages when using frequency and sag methods to determine belt tension, and the fact that permanent monitoring is available is a big advantage to operators that have shiftable conveyors with continually changing T_{1} tension. Permanent installation devices also have an ability to monitor the T_{2} belt tension, and aid PLC controllers in monitoring conveyor slip potentials as takeups become stuck or increase their friction levels.
References
HARRISON, A.: Safety factor calculation for inclined highstrength belts based on NDT signal analysis; Coal Handling & Utilisation Conference, Sydney, Australia, 1921 June 1990, pp. 289 295.
BARFOOT, G.: Computer modelling of belt conveyor systems; National Conf. on Bulk Materials Handling, IEAust., Australia, 1993, pp. 151157.
HARRISON, A.: Determination of the natural frequency of transverse vibration of conveyor belts with orthotropic properties; JS. & V., Vol. 110 (1986) No. 3, pp. 483493.
Harrison, A.: New concepts for evaluating belt tracking and dynamic tensions; Mech. Eng. Trans., IEAust., 1989, pp. 114118.
Prof. Dr. Alex Harrison, President,
Conveyor Technologies Ltd., 102 Inverness Terrace East #201, Englewood, CO 80112, USA. Tel.: +1 303 790 87 71; Fax: + 1 303 790 87 66. 