Standards The standard for the use and care of abrasive wheels [1] recognizes that abrasive wheels are easily damaged and liable to fail during operation, and recommends procedures to ensure that such failures should not cause injuries. These procedures are largely the responsibility of the operator of the equipment. Where relevant, the regulations are mentioned in the discussion following. There are no standards relating to the use of fiber reinforcing on abrasive wheels, or specifying the procedures which should be followed in the manufacture of a particulate composite abrasive wheel. For reasons that can only invite speculation, abrasive wheel manufacturers have not shown great willingness to develop fabrication standards. Abrasive Cutoff Wheel Specification The cut-off wheel in question is a 30” diameter abrasive wheel, ¼” thick, with a central arbor hole 1¾” in diameter. It is manufactured of particles of aluminum oxide embedded in a matrix of phenolic resin. This mixture is pressed between layers of loosely woven glass fiber cloth, resulting in a composite sandwich with one layer of glass cloth at the top and one at the bottom of the disk. The cloth is a 0°/90° weave with thicker fiber bundles in the warp direction than in the weft direction. The two glass fiber cloths are oriented at some random angle with respect to each other. The composite is subjected to a high pressure for a short time, and then cured for 48 hours at 170°C. Disks are speed-tested to 20% above the maximum rated speed before being shipped, as required by the standard [1]. Operation The abrasive cut-off wheel was used to cut bronze bars into lengths in a continuous casting operation. Cartridge brass bar drops vertically through the cutting station and the wheel is hand-fed through the bar. Since the bar is moving while the wheel makes the cut, some amount of bending load is generated in the wheel during the cutting process. The cut-off wheel is mounted horizontally over the motor shaft. A cupped mounting flange is placed on either side of the wheel, clamping the wheel along an annulus with an outer diameter of roughly 4”. The ANSI standard [1] recommends, but does not require, that a compliant blotter be used between the mounting flange and the wheel to alleviate stress concentrations at the flange/wheel contact. A blotter was not used on the wheel that failed. A nut is used to securely clamp the wheel between the mounting flanges. Since there is no established tightening torque, the nut is tightened at the discretion of the individual mounting the wheel. Prior to and after mounting, the user is expected to conduct a subjective ‘ring’ test: the wheel is lightly tapped to excite vibrations. A skilled user is said to hear the difference between an undamaged and damaged wheel. When the user is satisfied with the ‘sound’ of the wheel, the wheel is put into use. This procedure was not followed in the present case. To allow ease of manual cutting the cut-off wheel is mounted at about waist height. Unfortunately, this also enables a dangerous envelope for debris in the event of a wheel failure, with fragments flying horizontally at waist height. To reduce the threat posed by wheel fragments, a safety guard is required by the standard [1]. The guard in this case was designed to encircle the wheel, apart from an opening of 80°, allowing bar stock of diameter 10” to be cut. The subject manufacturer had ‘modified’ the guard to accommodate larger diameter bars, by increasing the opening. Additionally, the rotational speed of the wheel had been increased from the design speed of 1800 rpm. The Failure Event In the present failure, the wheel had been used for roughly one hour some time prior to the failure. Immediately before the failure the operator started the saw motor and allowed the wheel to come up to speed. The operator had initiated a cut when a ‘popping’ sound was heard, followed by wheel fragmentation. Unfortunately, an employee walking across the room, some 3 meters distant, was struck in the thigh by one of the wheel fragments, causing severe injury.
RkJQdWJsaXNoZXIy MjM0NDE=