Degree Name

Master of Welding Engineering


Faculty of Engineering


Rotary kilns fired by coal or gas burners are widely used to process mineral sands to extract pure minerals, most commonly titanium dioxide, (TiO2), in Western Australia. The kilns vary in size depending on refinery throughput, however most are very large items up to 70 metres long and 6 meters in diameter. The kilns are constructed from carbon manganese structural and pressure vessel plates and welded in the same manner a s large pressure vessels, however structural welding standards are applied.

The mineral sands are purified using a pyrometallurgical process where the internal operating temperature reaches approximately 1200°C. These temperatures are outside the operational limits of carbon manganese steel, so a protective internal layer of refractory approximately 250mm thick is required to reduce the shell temperature to between 100°C and 200°C. The refractory is poured in strips along the bottom of the kiln incorporating refractory anchors to stabilise the refractory upon drying. The refractory anchors are 253MA stainless steel, secured to the kiln shell via a welded lug incorporating a slot. This design is referred to as a “rotor lok” assembly.

Failure of the welded lug is relatively common during service, requiring costly kiln shutdown for remedial work and casting new refractory. The failures are mainly contributable to stress concentrations present at the fillet weld toes, with fatigue the failure mechanism in all cases. When failures arise, movement of the refractory lining over time causes a loss of lining where irreversible damage to the kiln shell occurs.

This dissertation researches and discusses alternative methods for securing the 253MA stainless steel refractory anchor to the kiln shell to provide a higher level of serviceability and reliability to prevent costly unexpected refinery shutdown. Over time, total replacement of the refractory is necessary, and renewing the anchor attachments using current methods is time consuming and expensive without consideration to lost production revenue.

Stud welding is proposed as a realistic option for the application of securing the refractory anchor. The author has attempted to persuade operators of rotary kilns that this option is viable, by applying mechanical testing regimes to selected studs which show the stud welded option out performs the welded lug in all testing applied.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.