Using IR Imaging to Diagnose Problems in Conveyors and Crushers
Infrared imaging can be very useful for detecting thermal anomalies caused by mechanical problems in mine and plant equipment
By Hennie Matthee, Kumba Iron Ore Ltd.


Figure 1: Thermograms were taken with a FLIR P60
with a 12° lens.
Companies today are under tremendous pressure to reduce costs while maintaining production. Infrared thermography is valuable for surveying electrical problems, yet some of its most important applications are in mechanical systems. Plants often contain thousands of low-speed bearings that are virtually impossible to inspect costeffectively using vibration monitoring. For example, conveyor system idlers—which have a direct influence on production when they fail—are quite easy to inspect with thermography. As a highly visual condition monitoring technology, infrared cameras present information clearly and effectively. The source of thermal anomalies can be identified and repaired before equipment fails, yielding multiple benefits:
• A better predictive maintenance program and overall maintenance and operational cost savings.
• A reduction in fire hazards in potentially flammable environments.
• More focused and cost-effective maintenance.
• Possible reductions in power required to drive equipment.

Thorough IR surveys must include all factors of operation systems for it to be effective. This article will demonstrate the need for this attention by using IR systems for Root Cause Failure Analysis to eliminate costly maintenance issues in mine conveyors and crushers.

Temperature Sensor Comparison
Routine inspection was done on ore crushers with the infrared camera, in this case a FLIR P60 with a 12° lens (See Figure 1), chosen for its superior thermal and visual image quality, spot size resolution and temperature measurement accuracy. The main objective of this IR inspection was to determine the accuracy of the Pt100 (a common platinum resistance thermometer) by comparing countershaft and oil temperature readings to the camera’s LCD display and to report any anomalies. Figure 2 shows that the placement of the sensors is crucial in order to report the correct temperature, and that thermography can aid in locating the best area.


Figure 2: In this particular application, placement of sensors was crucial in order to report the correct temperature,
and thermography can aid in deterring correct placement. Left: Thermogram of PT 100 at oil tank; maximum
temperature 51.8°C. Right: Thermogram of crusher counter shaft; maximum temperature 46.4°C.
 

Figure 3: Thermogram indicates residue at the bottom of oil reservoir.
It is second nature to a thermographer to be curious and to look at all possible effects of heat flow. The oil reservoirs on the quaternary ore crushers at Sishen mine were observed. The thermal image indicated temperature differences at the bottom of some of the oil reservoirs, as shown in Figure 3. The question this raised, as always, was: “What do the images tell us?”

To clarify the abnormalities indicated by the thermogram, oil samples were taken from the bottoms of all the reservoirs where the infrared images indicated temperature differences. The lowest suction point of the reservoir is located 100 mm from the bottom of the reservoir.

To ensure that samples were taken from the bottoms of the tanks, a company specializing in oil filtration used a one-way valve mounted at the end of a 20-mm PVC electrical pipe to take bottom oil samples (See Figure 4). When the PVC pipe was at the base of the reservoir the plunger of the valve opened the valve and the oil flowed to the inside of the pipe. The pipe was removed from the reservoir and the oil was drained into the sampling bottle. The oil samples were then sent to the Sishen mine oil lab to be analyzed. The oil analysis report indicated that the oil was very contaminated— so contaminated, in fact, that it clogged the filters in the lab instruments. The analysis, indicated in Table 1, showed that the buildup at the bottom of the tanks contained high concentrations of iron (Fe), copper (Cu), lead (Pb), silica (Si) and water (H2O). What the IR image actually revealed was the residue and build-up at the bottom of the tank.

The question then arose of how to prevent the oil pump from sucking in water and sludge. One way would be to raise the suction point above the level of the sludge, but that would do nothing to eliminate the sludge. The reservoir’s filtration system could not effectively remove it, and because there was no drain point to drain all of the oil from the tank, any new oil would be contaminated when refilling. Diagnostic Engineering proposed four potential solutions:
A.Clean the reservoir by hand—Cleaning by hand can only be done when major repair work is scheduled for a specific crusher. To do this the oil must be drained and the reservoir opened, flushed out and cleaned. This method would be effective but very time-consuming.
B.Make use of the filtration system—Stir the oil in the reservoir to force the residue at the bottom of the tank to move. The oil would flow through the existing filtration system and be cleaned to the filter specifications. This would take time and filters are expensive. Some of the contaminants might pass through the filter, causing unnecessary wear.
C.Redesign the oil reservoir—Redesign the oil reservoir so that the sludge and water can be drained at any time. The design would still protect the pump and filters, and there would be no need to drain all the oil, thus reducing costs.
D.Install a new filter system on all oil reservoirs—One of the reservoirs had a new filter system that was keeping it significantly cleaner than the others, as confirmed by both the oil reports and the IR images. That same filter system could be installed on all of the other reservoirs.


Figure 4: One-way valve plunger mounted on a PVC
electrical pipe.
Maintenance personnel chose both C and D: redesign the oil reservoir and install the new oil filter system on all the reservoirs. Figure 5 shows the results eight months later.

With regular inspections of the oil reservoirs the IR image will indicate any buildup of residue at the bottom of the tank and maintenance personnel can be tasked to drain the sludge.

Improving Selection of Proper Conveyor Components
In another example, thermography was used to detect design shortcomings and select the best available conveyor idler for a specific installation. IR inspections on conveyor equipment helped SAPO (South African Port Operations) Saldanha to determine the best idler supplier for their specific conditions and redesign conveyor equipment to increase production and reduce maintenance costs.


Figure 5: Thermogram of oil reservoir eight months
after modification.
Kumba Iron Ore’s Sishen mine was asked to do an infrared inspection at SAPO, since the problems experienced there have a direct impact on the mine. Conditions are difficult in Saldanha: temperature variations, regular rainfall, wind and seawater spray contribute to a high rate of corrosion on equipment. SAPO was experiencing misalignment of its conveyor belts and continuous replacement of idlers, which often failed due to corrosion and the collapse of bearing ends.

The primary inspection, involving all conveyor drives, substations, conveyor equipment, hydraulic systems and the tippler, was to determine the reason for deterioration of the existing rollers, make recommendations on how to solve the problem and help with necessary modifications. Two companies offered competitive prices to manufacture new rollers to Saldanha’s specifications, and it was decided to use thermal inspections to determine which company’s idler would be best for the conditions.

Conveyor belt specifications included:
• Garlands consisting of five rollers,
• Belt speed of 4 m/s,
• Capacity of 6,000 – 7,000 t/h,
• Idler spacing: 1.5 m trough and 3 m return idlers,
• Bulk factor 2.75 x volume, and
• Carried material comprising iron ore (-27 mm to +0.2 mm).


Figure 6 (above): Load on garland conveyor rollers.
 

Figure 7 (right): Thermogram of idlers from suppliers A and B. Supplier A: SP01: 63.0°C,
SP02: 48.1°C. Supplier B: SP03: 16.8°C, SP04: 18.5°C, SP05: 19.8°C, SP06: 24.1°C.
Garland rollers are a chain of belt conveyor rollers that replace the conventional steel idler frame and separate rollers. These chains have loops or hooks at both ends for hanging on flanges that run parallel to the conveyor. As garland rollers move freely, they minimize large items’ impact load on the rollers and the load on the belt is spread over a longer section, as shown in Figure 6.

Thermography showed hot spots on the idler bearings at normal load conditions. Problems were detected on more than 400 garlands, each consisting of five idlers per garland. Most of the problems were found on one type of idler. The defects were reported to the quality manager at SAPO.

The garlands were changed at the next maintenance window. After they were installed a follow-up IR inspection showed hot spots on the newly installed idlers. Examiners at SAPO said that some of the new idlers failed within two to 13 days after installation. Examination of one of the idlers showed that the bearings had failed. Possible causes of failure could be excessive belt tension, overloading, improper lubrication, under-designed idlers, misalignment of idler or structure, eccentric or out-of-balance idler or quality of the idler.

The IR inspection (See Figure 7) indicated that the average working temperature of idlers from one supplier (B) under load was between 8°C and 12°C. The average working temperature of idlers from the other supplier (A) was between 27°C and 170°C.

One idler from each supplier was cut open to determine the cause of the temperature differences. The main differences between the two suppliers were the outer seal, inner seal and the thickness of the bearing cap. Supplier A used a cheap outer seal with no inner seal while supplier B used a labyrinth outer seal and an inner dust seal. Supplier A used a no-name bearing while supplier B used a well-known brand of bearings.

The bearing end caps of supplier A gave no support to the bearing under load conditions. This caused the end cap to deform under load and misalign the bearings, as shown in Figure 8. The result of misalignment was overheating and failure of the bearings.


Figure 8: Misaligned bearings due to load, shaft deflection and under-designed end caps.
Due to the variations in temperature, the air inside the idler heated up and cooled down, forming water inside the idler. This water entered the bearing causing the bearing to fail because there was no inner seal.

Acknowledgements
The author wishes to acknowledge the cooperation of Maintenance and Oil Lab personnel at Sishen Kumba Iron Ore, Filter Focus South Africa personnel and H Rohloff (PTY) Ltd., Johannesburg, South Africa.

To download technical reference materials or view webcasts on IR technology and non-contact temperature measurements, visit www.goinfrared.com/IRdiagnosis.

Hennie Mathee is a diagnostic engineering thermographer at the Sishen mine, Kumba Iron Ore Ltd., South Africa.

As featured in Womp 2009 Vol 02 - www.womp-int.com