Automatic Conveyor Tension Control

In the late 70’s, winch manufacturer, Jegkurt Engineering offered an automatic adjustment system using the Piab tension switch. Gravity towers were the normal method of applying a single constant tension for starting and running the conveyor. Jegkurt initially used a single switch to provide the required tension level. However, it was realised that without live adjustments of the starting tension, the tension had to be increased to a very high level for start-up. This was unacceptably high for running and a second tension switch was added to provide an acceptable running tension. This composed a simple but crude automatic adjustment system with a timer and a couple of relays. The tension switch system enabled longer conveyors to be operated but it was clear that the type of control directly influenced the achievable result. This kind of system could only be classed as an automatic adjustment system without any dynamic control whatever.

In mid-1982, Numeritech, the forerunner of Iptron Technology was approached by a winch manufacturer to measure the displacement of their Bellville Washer tension gauge with a view to use automatic tension control. Iptron immediately responded with the recommendation that a strain gauge based load cell was the superior approach. Numeritech was then commissioned to build a prototype which was demonstrated very successfully at an exhibition. With a solid background in physics, mechanics including transmission line theory, Ian Plunkett approached the challenge along purely scientific principles. The task was set out very simply:
a. Tension of up to 6 tons
b. Running level of 2 to 3 tons
c. Starting level of 25 to 50% above running level

The entire control system and load cell was designed along theoretical lines with no preconceived ideas or following any existing practices. All the electronics and the load cell were designed from scratch and manufactured in-house. The EGT (Electronic Gravity Tower) was conceived and born.
Numeritech proceeded to design and build a production prototype called the Series 1 – EGT-2000, a fixed-speed winch controller. Some 150 EGT-2000 and variants were built from October 1983 to July 1986. The tension switch automatic adjustment system was completely superseded by the EGT control system due principally to its proportional active control.

Series 2 commenced in July 1995 and continues to this day. This was a complete microcontroller based design in the same format utilising a true 2-wire 4-20mA loop-powered load-cell. The 2-wire 4-20mA tension transducer was a very difficult development and was the first of its type in the world. Today it has been dramatically improved and is still the one and only successful 2-wire loop-powered tension transducer (load cell).

The control philosophy of the EGT-2420 was complex, sophisticated yet easy to use and to adjust with every conceivable practical logic, timing, interlock and safety feature that was possible for a single speed winch. It remains unsurpassed in sophistication to this day. No other fixed-speed tension control system has bettered it. It’s high reliability and conservative design gave it a life-expectancy of more than 20 years. Its software is unconditionally stable and completely bug free.

The friction barrier
In time, every single system failed. The winch brake was the first casualty, followed by seizure of the take-up trolley wheels; rope entanglement on the winch drum was endemic and even the rope sheaves succumbed due to bearing collapse; ropes became stiff from loss of lubrication; interlinking with central control was missing or lost. Dynamic problems disturbed the control; winches were too slow to control during start-up. The company hundreds of site visits (without any backup support from winch manufacturers) to find in the majority of cases that there was absolutely nothing wrong with the winch control panel.

Therefore the conflicting symptoms that could only be attributed to friction. Every problem was reported back to the winch manufacturer who totally disagreed with every one of Iptron’s findings.

They claimed that friction was unimportant because the system would ‘find its own way’, that sheave wheels never seized, and that there is no friction in the flexing of steel ropes. Despite clear proof to the contrary, they claimed that DC brakes were as fast as AC brakes.

To Iptron there were some shortcomings in the EGT-2420 system; further development was needed. Long conveyor belts require faster winches but these were not available. Frictional diagnostics and safety interlocks were needed, dynamic active damping of tension transients was essential for long conveyors.

Whichever way you look at it, the following statements are true:
1. Friction is the single biggest problem affecting all types of tension adjustment system and especially gravity towers.
2. A bad start on a conveyor is a sign that the tension control system does not match its dynamic characteristics.
3. The tension control system imposes a limit on the achievable length of a conveyor.
4. If the starting tension cannot be maintained steadily during conveyor acceleration, the system is inadequate and drive slip will result.
5. Dissipation of tension transients is the key to controlling the tension of long conveyors. This technology is applied to very large gravity towers and is the sole key to achieving longer conveyors.

The next level of technology
In 2006 Iptron realised that a better tension controller was needed. One that that would provide the following:
• Variable-speed as well as fixed-speed winch control
• 5-level dead band for enhanced response
• Automatic latch-off when target level is manually reached
• Dynamic matching to the time constant of the belt
• High integrity tension measurement
• DC earth leakage diagnostics
• -ve limit – open loop monitor
• +ve safety limit
• Power supply monitoring
• Tension output signal (mirror signal)
• Tension loss detection
• Frictional diagnostics
• In-line rope transport friction
• External rope transport friction
• Trolley wheel/obstruction friction
• Brake performance
• Comprehensive behavioural diagnostics
• Enhanced safety features
• Protection against over-tensioning
• Protection against rope-jamming
• Validation between winch motor current and tension
• Communication with central control
• Intrinsic safety
• Accurate winch motor and brake current monitoring
• Take-up trolley limit monitoring
• Interlinking with AC variable winch drive control
• Winch speed hyper cycling under tension control to effectively double the winch speed when deviating widely from target control levels
• Monitoring of the AC drive by the main controller via RS485 link
• Dynamic braking resistor to match type of gearbox of winch. For example spur/bevel gearbox has a very high regenerative component.
• Stored energy display.

The new EGT controllers are designated as the EGT-28AX for variable speed control and the EGT-24FX for fixed-speed winch control. During demonstration of the first system to various mines, there was the continually repeated statement that the take-up winches were too slow.

Reasoning
High speed was only required if the tension increased to a high level or if the tension dropped to a low level.
This meant that by using the original motor, the AC variable drive output frequency could be increased provided that:

a. The motor and brake RPM limits were not exceeded
b. The AC variable drive was not overloaded
c. The conveyor running level should be no greater than 50% of the winch rating

The difficulty in doing this was ensuring that the AC variable drive did not stall.
This was covered by creating an algorithm using the rope tension to control the speed during any tension increase.

In South Africa, and most of the world, there are few dedicated conveyor take-up winches. Haulage winches are offered for the purpose but they present some serious problems:

a. Conveyor take-up winches operate from zero tension up to full tension This means that if multiple layers of rope occur, the lower layers of rope are easily displaced by overlaid ropes. If the winch is too slow to keep up with tension variations during start-up, there will be large tension transients which will encourage rope jamming on the winch drum. Haulage winches are all 5 to 6 metres per minute, too fast for short belts and too slow for long belts. It is all very well having the rope tightly wound when delivered but it is likely to be almost completely unwound during installation. This problem is particularly aggravated by fully loading the drum to provide spare rope.

b. Multiple layers of rope on the winch drum increase the winch loading by a minimum of 10%. This leads very quickly to winch stalling on narrow winch drums.

Concept of hyper cycling
The winch motor running at 100Hz has turned out to be an unbelievably elegant solution to the winch speed being too low. Speed could be increased without a bigger motor to enormous advantage. The question was: Could the winch motor, gearbox and the brake etc be rotated faster? This required the approval of the winch manufacturers, who, when approached in 2013, were totally against this. They never actually approved it even though both the motor and brake were adequately rated. The fact that they have copied this development implies that they do approve. The company successfully installed the first system on 1st May 2014. It has operated 24/7 for over 4 years now and has not faltered in its superior performance.

Iptron alone continued developing conveyor tension control at its own expense and has been solely responsible for innovating all of the following in South Africa:

a. Use of strain gauge load cells for tension measurement
b. Invention of two-wire loop-powered 4-20mA tension transducers
c. Invention of DC earth leakage system for 2-wire loop-powered transducers
d. Invention of the control logic and timing and interconnection philosophy for tension controllers
e. Brake pre-release timer to speed up response and the reduction of brake wear
f. Invention of hyper cycling winch motors to restore tension to normality in the event of excessive deviation
g. Invention of 5-level dead-band system to cater for rapid loading
h. Invention of measurement algorithms for take-up friction
i. Classification of frictional components of take-up system mechanics
j. First application of proportional, integral, derivative control of conveyor tension to actively dissipate the energy of tension transients
k. First specification for take-up winch design in South Africa.

Various other companies have hooked onto conveyor tension control believing it to be a very simple task. In South Africa not one other company has undertaken any development or innovation in the field of control philosophy for take-up winches. Every ‘knock-off’ is blatant copying intellectual property.

Current status of conveyor tension control in South Africa
In recent years, there has been a regression to automatic adjustment systems using slow winches and a ‘locked take-up start’ found 35 years ago before Iptron introduced tension control by the EGT Systems. This has resulted in a major deterioration of the quality of conveyor start-up, setting back conveyor enhancements. Despite the EGT System offering live adjustments for over 3 decades, there are many conveyor designers who unbelievably still specify a locked take-up start.

A locked take-up start is equivalent to disallowing a gravity tower from controlling the take-up tension.

This retro-attitude results from the belief that electronic automatic tension control is unreliable. The fact of the matter is that the controls and all well-designed electronics are highly reliable. What many engineers do not understand and Flextronics refuse to accept is that every conveyor tensioning system ultimately fails due to mechanical deterioration in the form of friction. Strangely, the worst offenders are gravity towers. Their tension reference (the suspended mass) is located remotely with typically 8 rope sheaves between it and the conveyor belt.

Note that sheaves are budgeted to have 1½ % friction. The electronics is routinely and erroneously blamed for all problems. The missing link in all this is frictional diagnostics.

The old EGT-2420 controller – known as ‘The system that just works’ Iptron’s peak decade of sales occurred in the 1990’s. This is also when every system started failing. Major customers abandoned perfectly good systems.

Iptron’s mistake was that of leaving the mechanical problems to the mechanical engineers. For a specialist electronics engineer to tell a mechanical engineer what his problems are is just not acceptable.

The winch manufacturers were disinterested in problems that were not attributable to the winch. They never went to site and were thus completely and conveniently ‘unaware’ of any problems. Anyone who has not spent the hundreds of hours observing take-up winches and take-up trolleys has no clue about the practical aspects of tension control. Iptron’sstudy of trolley wheel friction, sheave wheel friction, rope flexure friction and rope entanglement on winch drums is ‘inconveniently’ denied by winch manufacturers and mechanical engineers.

Extrapolating this to other tensioning systems paints a sorry story of engineering incompetence and complete lack of understanding of the basic physical principles.

This situation is responsible for letting conveyor users believe that tension control is a very unimportant task and does not justify any kind of structured engineering approach.

Iptron realised, eventually, that the mechanical problems were casting the whole field of automatic conveyor tension control in an extremely bad light. Iptron researched this problem extensively and came to understand the mechanical problems and took steps to develop algorithms to identify and quantify the frictional components.

The company also tackled the eternal dynamic problem of tension transients in conveyors. This alone presents the biggest obstacle to the advancement of conveyor technology.

Iptron has for years been recommending to winch manufacturers to produce purpose-designed take-up winches instead of haulage winches. These recommendations are based on our findings and analysis of problems encountered on site. Virtually every system is prone to rope entanglement and rope speeds need to match the conveyor dynamics. Some of this is at last being actioned.

Standards for mechanical aspects
Standards need to be compiled for conveyor tensioning systems. This would cover the following:
• Take-up winches
• Winch frames should be rated in kN in size intervals of .2 with acommon final planetary ratio of +/- 7,2 : 1.
• Gear motors would be selected to provide the speed and power.
• Gear sizes and materials should be standardised.
• Fixed speed winches should be limited to 100kN.
• Worm primary gearboxes should be used for fixed-speed winches.
• Bevel or spur gears should be used for all variable-speed winches.
• Winch drums should be grooved and wide enough to accommodate the working length of the take-up system on a single layer.
• If multiple layers are to be used, a ‘Lebus’ style groove should be provided but rope overlay should be limited to two layers.
• The winch should be rated at a minimum of 2 times the running tension.
• Winch rope speed should be a minimum of 15% higher than the expected rate of generation of slack during acceleration.
• All winches should have the option of fitting a slip clutch so that the maximum tension rating of the winch will be not be exceeded by more than 25%.
• Winch slip clutches should be free of dynamics. They should only be multi-disk type.
• Adjustment of the slip clutch must be provided for with a security lock to prevent unauthorised adjusting.
• Adjustment of slip clutches can be hydraulically adjusted.
• Fixed-Speed winches should be provided with a high-speed brake.
• Variable speed winches require a brake to lock the winch when adjustment is completed.
• Winches with slip clutches must have rotation sensing on the drum. These can be proximity sensors operating in quadrature sensing bolt or holes in the winch drum flange.
• Take-up trolley standards
• All rotating components should have rolling bearings.
• Minimise the number or falls of rope.
• Rope sheaves should rotate about the horizontal axis; i.e. they should rotate vertically. This is to minimise yawing effects
• Trolley wheels and the take-up trolley must not foul the structure and must move freely.
• Total friction of the take-up trolley movement should be less than 5% of the running tension.
• The trolley rope sheave/s must never be allowed to contact the rope clamps on the tension transducer. This causes total loss of control.
• Take-up trolleys should have normally closed limit switches.
• The tension transducer should be mounted stationary on the structure.
• Any friction in the rope sheaves between the In-Line ropes will cause errors in the measurement and additional load on the winch. This friction will also tend to cause over-adjustment of the conveyor tension.
• All and any friction in the external rope sheaves (not in-line) will cause additional loading of the winch and will not affect the accuracy of measurement.
• All and any friction in the movement of the take-up trolley will tend to cause under-adjustment of the conveyor tension.

Customer standards
Very few customers seem to have published standards. Such standards should include:
• Preferred switchgear
• Supply input breaker/fuses
• Preferred AC variable drives
• Preferred control voltages
• Inductive RC suppression
• Motor protection
• Core balance
• Wiring standards
• Conductor sizes
• Wire marking
• Insulation
• Electrical safety standards
• Enclosure specification
• Enclosure component density/spacing
• Front panel control required
• Metering
• External electrical links and functions
• RFI Screening for AC variable drives
• Communication specification
• Documentation

Conclusion
The development of conveyor tensioning by Iptron Technology clearly places them at the forefront of the field. Continued development is totally dependent on conveyor users and designers actively endorsing and supporting their dominant efforts.

Iptron Technology alone have originated all the technology and have unequalled hands-on experience with hundreds of systems. Their contribution to this field has been unparalleled and highly significant.

Iptron Technology
Ian Plunkett
Tel: (011) 534-1285
Email: iptron@mweb.co.za
www.iptrontech.co.za