Some experiences on tunnelling in swelling materials

Design methods for swelling materials are burdened with a great number of unknowns, as in the beha&not ; viour of swelling rock, so by a broad spectrum of rocks tendinlg to swelling. The complexity of the problems is illustrated by the experience gained during the construction of two tunnels in Algeria. The first example shows a case when swelling, predicted based on the laboratory tests, did not occur during the execution. The second example shows a case of rock swelling in a tunnel floor occuring by a completely different mechanism of swel&not ; ling than that established by prior testing. The experiences presented confirm the fact that the law of swel&not ; ling determined in a laboratory cannot be copied directly to in situ conditions.