MILL SELF-DISENGAGEMENT BARRING SYSTEMS

JOHANN J. van RENSBURG ENGINEERING MANAGER MET. PLANTS ANGLOGOLDASHANTI CONTINENTAL AFRICA REGION Jan 2013

CONTENT 1. ABSTRACT 2. INTRODUCTION 3. BACKGROUND 4. MILL DETAIL 5. PRINCIPLE and OPERATION OF SELF-DISENGAGEMENT BARRING SYSTEMS 6. METHODOLOGY 7. COST MODEL 8. ADVANTAGES OF SELF-DISENGAGEMENT BARRING SYSTEMS 9. DIS-ADVANTAGES OFSELF-DISENGAGEMENT BARRING SYSTEMS 10.INSTRUMENTATION and MEASUREMENTS 11.MEASUREMENT RESULTS 12.RESULTS 13.CONCLUSION 14.FINAL WORD 15.ACKNOWLEDGEMENT

1.ABSTRACT Grinding Mills are production critical machines in not just Gold mining Process Plants but in every plant where they are utilised in order to meet business objectives. They in a sense are the heart of plants and in order to achieve business objectives are required to be available well above 90 %. These machines have been in operation for close to a century and in this time with a diameter scale up of at least five fold. A typical or favoured drive arrangement feature a wound rotor induction motor driving a girth gear via a single or dual-stage gearbox. This drive train configuration in itself is about forty years in use. However numerous improvements and changes occurred during passed years at various operations on mills from various manufacturers. Large diameter Wrap-around motor designs with huge diameters are just one of these interventions. These designs eliminate the use of gearboxes, pinions and girth gears. Evolution of Mill technology

Conventional Milling

First generation SAG Milling

Latest Generation SAG Milling

14,7Meter diameter wrap-around around motor mill design.

Latest development in drives is sophisticated direct motor drives onto the girth gear without the use of Gearboxes and pinions. This latest test generation is a medium voltage frequency converter with exclusive DTC (direct torque control).Direct motor drives with frequency torque control.

This unfortunately does not exclude us improving and maintaining current plants with old designs and operating principles. It is appropriate to constantly apply comprehensive design and manufacturing quality assurance to achieve reliable mill drives. Further to this plant operating practices mill reliability should be taken into account. This paper summarises the investigation into some of these old features in design and operating principles. The findings indicate that some of our practices over years have negatively impacted on our reliability without almost realising it. Almost too small an issue to pay attention to. Another important factor to be considered is that these machines are all purpose-built and as such in many cases prototypes. It is necessary to undertake independent design audits and apply rigorous quality assurance and to assess operating practises. This abstract have focused on drives and their improvements as a result of operating practise audits that were conducted. One of the components of current drive trains is Barring and the paper summarises the investigation into Barring specifically and the effect thereof on the reliability and sustainability of Gearbox life.

2.INTRODUCTION
Focussing on the demands of grinding mills we can classify them into the following. Operating demands Maintenance demands Protective demands As per the abstract this paper will focus on the requirements of Operating demands. One of the critical tasks from an Operational demand identified in the start-up procedure of grinding mills is Barring. The Barring drive, also known as Inching drive, Sunday-drive or auxiliarydrive, is an important component of any mill installation. It is used for maintenance and inspection purposes, as well as an emergency auxiliary drive to keep the mill rotating when the main motor fails and it is required for the mill to rotate at certain intervals. Re-lining as a maintenance function cannot be done without a barring drive. Probably the most important function of the Barring drive is to dislodge Frozen or Lock charge thus as frozen or lock charge protection. Locked or Frozen charge is a common occurrence with mill applications and is therefore it is an advantage that the Barring drive can offer protection againstthis condition. Frozen charge can occur when the mill has beenstopped long enough for the product to solidify. If the mill is startedagain this solidified product may fall and damage the mill lining. This is achieved by the several slow rotations (1turn per minute) by the Barring drive to slowly dislodge the mill charge. The inching drive components include a prime mover, speed reducer, and a connection - engaged by hand or automatically, between the inching reducer and the main drive. Also included in the system, is a brake or backstop to hold the equipment when it is stopped in an unbalanced position as well as other appropriate safety devices. Starting a mill with a locked load will not necessarily causes visible damage to the shell or any other of the mill components, but can cause drive-shaft

misalignment and will impact on the long term integrity of the mill. Locked charge start-ups will cumulatively reduce the overall mill life. As a general rule of thumb all tumbling mills are to be barred for at least 2 revolutions prior to start-up if they have been standing for more than 6 hours. How do they work:The inching drive for the mill motor consists of a small motor with reducer (gearbox) and clutch/brake package. During normal operation, the clutch/brake is disengaged. However, if forsome reason it becomes necessary to turn the motor slowly, or inch the motor, the main motor isstopped and the clutch is engaged on the inching package. Using the inching drive package, themain motor now can be turned or inched at a much slower rate. This task is accomplished by havingthe inching package supply a pulsing feature to provide rotational movement through the motorpinion to the main motor drive. To return to the main motor, the clutch is disengaged and the main motor is energized. Typical Barring System

3.BACKGROUND
At present, most mills are started by the main drive motor from an initial rest position. This results in large torque outputs and consumption of high torque dependant electric current. Even when Barring Systems are fitted this is the case as at the end of the traditional Barring operation the drive train comes to rest for the system to be dis-engaged. The main drive motor again is at rest and for milling needs to overcome initial inertias from its rest position. In a mill configuration with self-disengagement coupling systems this is not the case as the main drive motor does not start from initial rest position as the mil is rotated by the barring motor before start-up. At start-up the drive motor does not start from rest, but from the barring rotation speed.An automatic coupling ensures speed dependant uncoupling of the barring system after start-up of the main drive motor. Siguiri, one of the Anglogoldashantis Continental Africa Region business units have a Self-disengagement system installed on a large mill and due to its successful history and operational simplicity it was decided to evaluate this as possible implementation on other business units.

4.MILL DETAIL Ball Mill : 6.1m diam. 9.0m (Polysius) Motor Rating : kW 6000 , 994 r.p.m Gearbox Rating : 57kN.m Gearbox Type : Flender H3H12, Double Reduction Combiflex

5.PRINCIPLE AND OPERATION OF SELF-DISENGAGEMENT BARRING SYSTEMS

With the new baring gear coupling installed, the mill motor starts up whilst baring is taking place, and the mill load is in motion. The baring gear coupling slides back in a keyway and a limit switch simply stops the baring motor. The self-disengagement clutch is a directionally actuated freewheel clutch. Part A is mounted on the driver gear unit shaft with axial movement by means of a key and part B on unit to be driven and is fixed. This imply that A is driving B in the Barring mode. The overrunning clutch is engaged in the stationary condition by shifting partA axially to engage with B. Once the speed of clutch part B is higher than that of clutch part A (Starting point of main motor), independent dis-engagement caused by the angled faces of the engaging dogs on clutch part A and clutch part B takes place. Motor and driven machine unit are then dis-engaged mechanically. A is then locked in the dis-engaged position.

The overrunning clutch is suitable only for horizontal arrangement

A B

COUPLING ENGAGED

COUPLING DIS-ENGAGED

6.METHODOLOGY
The methodology was to develop a model and quantify the effect of Selfdisengagement barring systems on key performance indicators of a mill with special attention to life time reliability. The purpose of the investigation was to compare the following key performance indicators for operation of a ball mill with and without a selfdisengagement barring system. The following was taken as critical evaluation factors. Total cost. Fatigue life. At present, most mills are started by the main drive motor from an initial rest position. This results in large torque outputs and consumption of high torque dependant electric current. The following strategic questions were addressed: How can the effect of the barring & automatic smooth coupling be modelled on the mill life cycle cost? How does a model correspond to reality and why is the correspondence justified? What is the lifecycle cost saving realised by changing a typical existing system into this automatic coupling system? What is the effect of the barring system with smooth coupling on the fatigue life of the gearbox, motor and the rest of the drive train? How will it influence reliability and availability of the system? Is it feasible to modify an existing ball mill system into the automatic coupled system given remaining service life and other client specified parameters? This report is complementary to the Financial Excel model and contains the workflow of the design process, results and findings, and, user manual thereof.

7.COST MODEL
As total costs was one of the critical factors and a cost model was developed to determine this. The requirement was a model that can be used by the different plants to establish the total cost of ownership of its ball mill. The output of the model must also enable to quantify the savings achieved by installing a barring system on the ball mill. The financial model is a representation of all activities related to the ball mill that influence its cost of ownership. Such models provide clarity regarding the financial consequences of past activities and allow for educated decisions regarding future activities. The financial model will also serve as a knowledge base to capture important information and statistics. The following steps were followed in creating the model: The ball mill components and financial variables were identified and any relationships quantified. Historic and operating data were obtained for all costs associated with the identified components and variables. The data included sufficient information to determine: o Capital costs. o Operating costs. o Product costs. o Repair and maintenance costs. Lost production due to unplanned failures These user requirements imply that: The model must be able to project costs over a fixed period (for example 20 years) to enable valid comparison of results from different mines. There must be a function that allows the user to determine predicted costs for the system when a barring system is installed. The primary requirements of the model can be summarised as follows: The model should run on a platform that is available at all sites to avoid any unnecessary expenditure of new software. Excel would be a suitable platform. Data entry can be manual but must be user friendly.

 The model must be able to store data that can be easily retrieved for calculation purposes, i.e. database capability. The graphical user interface (GUI) must be user friendly and easy to understand. Report writing must be quick and easy and the report may be in Excel or Word format. The model must be able to calculate Net Present Value for all costs accrued to date. The model must be able to calculate predicted Cost of Ownership over a period of 20 years. The model must be able to calculate average operating cost of a ball mill per year. The model must allow for an option to calculate the effect of installing a barring system at a point in time. Some calculations will require user input for inflation and interest values. The model must be applicable to all sites. The model must be easy to install and maintain with the use of a user manual. The model must not interfere with any other software applications that are currently run by the client. The model must be stable. To quantify the financial consequence of a barring system, the following tasks were conducted: Key measured performance indicators were compared between the system during start-up with and without the automatic coupling system. The performance indicators were utilized to aid in the construction of a fatigue model for the ball mill system. A life expectancy model was constructed based on the fatigue model, to establish the relationship between the system with and without the automatic coupling. This relationship is used to give an indication of the expected life of a system without barring, which is then used in the construction of the financial model.

Examples of Cost model Input-sheets. General Mill detail

Gearbox Detail

Maintenance detail

Operational Detail

8.ADVANTAGES OF SELF-DISENGAGEMENT BARRING SYSTEMS

Smooth torque transfer through the drive train because backlash is taken up by the barring drive. Angular momentum at point of start-up of the main drive. The system does not start from rest that could result in lower torques. Bearings remain lubricated during barring. Small and cheap motors and drive keeps the mill in rotation. Safe and non-complex system to operate and operator friendly Very low and basic maintenance intensive system Non-complex retrofit on current systems Time saving operation Relative small capital layout Eliminates the possibility of bridging out limit switches as it form part of the Control Philosophy.

9.DIS-ADVANTAGES OF SELF-DISENGAGEMENT BARRING SYSTEMS

Capital requirements. Retrofit to current systems that could require a specific design. In certain cases space can be a problem to retrofit depending on existing design. Production loss time to install. Culture change to long existing practise. Resistance to change the not invent here syndrome

10.INSTRUMENTATION and MEASUREMENTS The gearbox fitted with the Self disengagement Barring system was instrumented as follow:

Top view of ball mill layout, showing the accelerometer positioning.

1. Two 90 degree bi-axial shear strain gauges on the barring gearbox output shaft. Wired to measure torque output One full-bridge channel (position 1) 2. Two 90 degree bi-axial shear strain gauges on the main gearboxs input shaft. Wired to measure torque output One full-bridge channel (position 2) 3. Three accelerometers on the main gearbox housing on the input shaft end measuring tri-axial acceleration of the main gearbox. 4. Three accelerometers on the main gearbox housing on the barring coupling end measuring tri-axial acceleration of the main gearbox. This totals to 2 strain signals and 6 acceleration signals. The position of the accelerometers and strain gauges is shown in the above figure.

Photo of instrumented barring output shaft and part of main gearbox system.

11.MEASUREMENT RESULTS

Acceleration in the orthogonal directions on the barring end with the barring system uncoupled 2 1 0 -1 -2 -3 60 80 100 120 Time [s] 140 160

ACCELERATION

Acceleration [m/s 2] in x direction

180

2 1 0 -1 -2 -3 60 80 100 120 Time [s] 140

Acceleration [m/s 2 ] in y direction

160

180

2 1 0 -1 -2 -3 60 80 100

Acceleration on Gearbox Casing atBarring end. UNCOUPLED(Fig 1)

Acceleration [m /s 2 ] in z direction

120 Time [s]

140

160

180

Acceleration on Gearbox Casing atMain drive end. UNCOUPLED (Fig 2)

Acceleration in the orthogonal directions on the drive end with the barring system uncoupled 2 1 0 -1 -2 -3 60 80 100 A c c eleration [m /s 2 ] in x direc tion

120 Time [s]

140

160

180

2 1 0 -1 -2 -3 60 80

A c c eleration [m /s 2 ] in y direc tion

A c c eleration [m /s 2 ] in z direc tion

Acceleration onGearbox Casing atBarring end. COUPLED(Fig 3)

100 120 Time [s] 140 160 180 2 1 0 -1 -2 -3 60 80 100 120 Time [s] 140 160 180

Acceleration in the orthogonal directions on the barring end with the barring system coupled 0.8 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 0 20 40 A cc eleration [m /s 2 ] in x direc tion

60 Time [s]

80

100

120

1.5 1 0.5 0 -0.5 -1 -1.5 -2 -2.5 A cceleration [m /s 2 ] in y direction 0

Acceleration onGearbox Casing at Main drive end.COUPLED (Fig 4)

20 40 60 Time [s] 80 100 120

Acceleration in the orthogonal directions on the main drive end with the barring system coupled 1

A c c elera tio n [m /s 2 ] in x direc tio n 0.5 0 -0.5 -1 0 20 40 60 Time [s] 1

A c c elera tio n [m /s 2 ] in y direc tio n

A c c elera tio n [m /s 2 ] in z direc tio n

COUPLED(Fig. 5)
80 100 120 0.5 0 -0.5 -1 0 20 40 60 Time [s] 2 1 0 -1 -2 -3 80 100 120 0 20 40 60 Time [s] 80 100 120

INPUT SHAFT TORQUE

140

140

140

Torque on main gearbox input shaft with barring system coupled 35 30 25 20 Torque [kN.m] 15 10 5 0 -5 -10 -15
X: 29.12 Y: 32.92

20

40

60 Time [s]

80

100

120

140

COUPLED(Fig 6)
Detail section of initial torque peaks on main gearbox input shaft 45 40 35 30 Torque [kN.m] 25 20 15 10 5 0 -5
X: 5.051 Y: 32.09 X: 5.152 Y: 28.13

4.2

4.4

4.6

4.8

5 Time [s]

5.2

5.4

5.6

5.8

INPUT SHAFT TORQUE UNCOUPLED(Fig 7)

Torque on main gearbox input shaft with barring system uncoupled 40 35 30 25 20 Torque [kN.m] 15 10 5 0 -5 -10 -15
X: 85.68 Y: 34.65

60

80

100

120 Time [s]

140

160

180

UNCOUPLED(Fig 8)
Detail section of initial torque peak on main gearbox input shaft 45 40 35 30 T orque [k N .m ] 25 20 15 10 5 0 -5 61.3 61.4 61.5 61.6 Time [s] 61.7
X: 61.71 Y: 9.866 X: 61.73 Y: 1.557 X: 61.7 Y: 24.37 X: 61.65 Y: 38.9

61.8

61.9

62

12.RESULTS Acceleration results

From Fig. 1 to Fig. 4 the accelerations measured on the main gearbox housing is shown. When comparing the two sets of acceleration data the initial acceleration magnitude is larger on the gearbox housing (clearly visible on the drive end accelerations) when the barring system is uncoupled, which supports the statement above. Comparing Fig.2 with Fig.4 the acceleration measured on the main drive end on the y direction recuses from 0.8 mm/s.s uncoupled to 0.25 mm/s.s coupled. Reduction of 68%.

Torque results
The rated power of the main motor output of the ball mill is 6MW, with a rotational speed of 994rpm.
T= P

where: P = rated power of the motor, [Watt]. = angular velocity, [radians]. T = rated torque, [kN.m]. Using this equation the rated torque for the gearbox input shaft is thus calculated as being 57kN.m. The average operational torque calculated from the measured data, over the operational section of the data for the main gearbox, was calculated as being 21.2kN.m. This operational torque of the main gearbox input shaft is 37% of the rated torque of the system. When considering the start-up of the ball mill with the barring system uncoupled, (Fig.7&8) impact loads is seen by the shaft and other components of the gearbox. Impact loads could create stresses that are significantly higher than the stresses created when similar loads are applied gradually. The torque data for the main gearbox input shaft shows two significant impact loads, one at initial start-up, as shown in the red section and the second at the sorting of the soft start system, shown in the green section of Fig. 5 and Fig. 7. These impacts will have no influence on the fatigue life of the shafts; however the compressive force seen on the gearbox internals is significant.

The results from the measured data for the ball mill with barring, included:

 When the mill drive motor is started, it uncouples the barring system clutch. To uncouple this clutch the torque on the barring system changes due to directional changes that will cause a peak (32.09kN.m).However, the peak shown in Fig.8 (38.9kN.m) where the mill is started without barring is caused by the backlash in the mill gearbox and is transferred through the gearbox. Therefore, for damage comparison on the mill gearbox, the peak on the system started with barring (32.09kN.m) may be ignored and the peak on the system started without barring (38.9kn.m) shall be included. The first torque peak to be considered for the main gearbox input shaft was calculated as being 28.13kN.m. This is about a 32.6% increase from the average operational torque (21.2kN.m) calculated for the shaft and is 49% of the rated torque of the gearbox. (57kN.m). The results seen in Fig. 7andFig. 8 for the ball mill with no barring included: The first initial torque peak for the main gearbox input shaft was calculated as being 38.9kN.m. This is about a83.4% increase from the average operational torque (21.2kN.m) calculated for the shaft and is 68% of the rated torque of the gearbox (57kN.m). Fig 8 showsthe torque peaks with two distinct high torque values. Also, a second peak is seen at 24.37kN.m. The initial torque peak was considered in the fatigue assessment as this is the highest value calculated and the second peak is similar to what was seen by the system with barring and would not change the fatigue life when comparing the two systems. From the above calculated results the second torque peak in both the barring data and the no barring data is almost the same magnitude and the difference would have no effect on the operational life of the system, when comparing the barring and the no barring data. The result that would have an effect on the operational life of the system is the difference in the first torque peak (38.9kN.m) in the system uncoupled and the second torque peak (28.13kN.m) in the system coupled. This difference equates to a 38% increase in torque experienced by the system compared to starting up from the barring speed.

13.CONCLUSION

During the assessment on the data measure Siguiri mine the following was determined: The torque on the main gearbox input shaft, included: o An initial torque peak on both the sets of data, with barring and no barring. The initial torque peak in the data with barring could be neglected as this torque is not transferred to the gearbox, as supported by the acceleration data. The acceleration data showed a peak not at the initial start-up. o The torque peaks on the input shaft were 38.9kN.m for the system with no barring and 28.13kN.m for the system with barring. o A second torque peak is seen when the soft start system is disengaged. This peaks was not considered in the life assessment of the system as this peak was similar in both sets of data and would have no effect during comparison of the data. The torque on the barring gearbox output shaft reached 7kN.m. The fatigue life assessment included the design of a fatigue model for the gearbox system along with a model for system availability. The results from the effect of barring on fatigue life included: The contact stresses in the pinions and gears increase by about 17% for no barring compared to barring. The fatigue life assessment showed that the shafts are designed for an infinite life time and no difference was seen between barring and no barring on the fatigue life of this component. There is a significant change in operational life cycles for the pinions and gears of the gearbox, which are most likely the components to fail. A ration of 9.5 x 10-4 for life cycles of no barring compared to barring was determined. A reduction in operational life of 31.5% for the gearbox is calculated when the barring system is not used on the ball mill system. A financial model was constructed according to user requirements to enable the comparative analysis of cash flows for a Ball Mill with a barring and automatic coupling mechanism and a Ball Mill without it. Utilising this model in general indicated a payback of less than a year on a typical 3 Mw Mill. 14.FINAL WORD

It is my absolute opinion that this is what is called a no brainer. Just go and do it. From a safety perspective and with specific attention to Stored Energy this system is almost in comparable to the normal practice. Just the elimination of force by hammer or lever to engage the wheel speaks for itself .If you consider sometimes the physical condition of the engagement wheel you would prefer not to be present at this operation during specifically at night time. This is also a system that as far as bridging-out safety systems is concerned can be promoted as not possible. This becomes part of the Mill Control Philosophy and wired into the starting sequence. It is my absolute belief that this is a Step Change in Mill starting procedures with many other benefits and advantages not quantified in this paper. This does not eliminate other factors to be considered that will hinder this change but in no doubt I believe we can overcome that. I sincerely hope this paper will inspire either an individual, a team or in whatever way to embark on this route.

15. ACKNOWLEDGEMENT Dr. Michiel Heyns from Investmech for the in depth study conducted on the field. AMRE for the request to write this paper.