Designing and Failure of Monolithic Refractory Structures – Part 2
Part 1 of this paper has highlighted the errors in refractory design methodology and a general methodology for the design of monolithic refractory structures is presented. In particular it has been shown that traditional concepts or guidelines are erroneous which are not based on engineering science can leave open the risk for catastrophic failure.
Refractory lining design has posed significant engineering challenges but the understanding of failure mechanisms is enabling improvements in refractory lining life.
Previous research  has shown that failure of refractory linings due to creep rupture of the steel anchor at or near the interface zone is the major failure mechanism. This is due to the fact that thermal loads induced low stress (<10MPa) at high temperatures. However, the thermal strain loads on steel anchors occurs at low strain rates, έ s-1, which is about 10-7s-1 depending on the heating rate. Research has shown that the yield stress of a stainless steel is much lower than the published yield stress (which is at a much higher strain rate) and will be < approximately 50MPa. However, once the refractory stops expanding the strain has reached its maximum value and will start to decrease due to creep at temperature. This creep is a relaxation process, stress relief, and occurs very quickly typically less than 30 minutes at temperatures >550C. Unlike creep rupture strength tests which are carried out at constant loads. This means once a refractory system is at steady state the anchor stress will not be maintained and start to decrease due to creep.Some hypotheses on anchor failure have proposed that anchor stresses is due to If the design of the refractory structure is not properly considered the refractory anchors can fail in a short amount of time (e.g. hours).
The current approach to anchor design and spacing which have been developed from experience and applied “rules of thumb” are considered inadequate and fundamentally incorrect. This paper discussed the analysis of failed refractory structures and the numerical analysis used to predict anchor stress. Our research has shown that numerical analysis techniques can be used to predict and design refractory structures. The numerical analysis when compared to real structures shows that the results are in-line of observed failure modes. It is also clear that non-linear numerical analysis is required when designing refractory structures.
This paper shows how numerical analysis should be used in the design of refractory structures.