Is sheer thenardite attack impotent compared with cyclic conversion of thenardite–mirabilite mechanism in laboratory simulation tests?

The precipitation of mirabilite during the wetting stage of a cyclic total immersion test has widely been implicated as the main damage mechanism of sodium sulfate. We, however, hypothesized that the damage contribution from thenardite cannot be discounted in laboratory tests. To verify the supposition, a series of laboratory experiments was undertaken by subjecting three types of rocks to the sodium sulfate attack. To unambiguously demonstrate the potency of thenardite, in addition to cyclic total immersion tests, continuous partial immersion tests were also performed under different environmental conditions, which allowed either sheer thenardite precipitation or reversible thenardite–mirabilite conversion. Although rock breakdown occurred only during immersion, the results clearly indicated that thenardite alone could induce significant damage in both cyclic total immersion and continuous partial immersion tests. In most cases, tests that involved mirabilite precipitation, however, were more destructive than sheer thenardite tests, regardless of the type of salt supply technique. With the same duration and temperature of the immersion stage, the difference in drying temperature influenced the degree of damage, revealing the damage contribution from thenardite even in the tests with cyclic phase conversion. The reason why mirabilite is generally credited to be more destructive than thenardite and the effect of test duration on the damage potency of thenardite are also discussed.