Squeezing Competence Factor

This ratio is the first question worth asking about any deep tunnel, and it takes about ten seconds: is the rock mass stronger than the stress it has to carry? Divide the rock mass compressive strength by the in-situ stress. Below about one, the ground cannot support its own overburden elastically and it will deform into the opening — slowly, continuously, and for a long time. Well above one, it can, and the problem becomes structurally controlled blocks instead.

The single most important word on this page is "mass". $\sigma_{cm}$ is the unconfined strength of the rock mass, not of the core you tested. The Hoek-Brown pages on this site work exactly this example: an intact limestone that tests at 50 MPa in the laboratory is worth about 2.2 MPa as a jointed mass at GSI 45. That is a factor of twenty-two. Enter the laboratory UCS here and a tunnel that is going to squeeze badly will report a competence factor of 4 and no problem at all. There is no warning, no impossible number, nothing to catch the eye. It is the commonest way this check is got wrong and it fails in the reassuring direction, which is the worst direction for anything to fail in.

The threshold is a smear, not a line. Different authors put the squeezing boundary in different places; some write the ratio the other way up, as stress over strength, which flips every inequality; and several fold in the tunnel radius and the support stiffness, because how much a tunnel squeezes depends on how big it is and how quickly you close it as well as on the strength ratio. Treat 1 as the middle of a grey band about half a decade wide. Anything within a factor of two of it means "monitor, instrument, and be ready to change the support philosophy", not "safe" or "unsafe".

Roughly, and with all the caveats above: below about 0.35 the case histories describe severe to extreme squeezing — closure of several percent of the diameter, support that fails and has to be replaced, a face that will not stand on its own. Between 0.35 and 0.7 is moderate squeezing, with convergence in the tens to hundreds of millimetres continuing for days or weeks after the face passes. Between 0.7 and 1 the answer genuinely depends on the tunnel, and that is where convergence monitoring earns its cost. Above 1 squeezing is not expected, and what remains is wedge and block instability, which this ratio cannot see at all because it has no joint orientation in it.

What squeezing actually asks of a design is a change of philosophy, not a stronger lining. A stiff lining installed early in squeezing ground attracts load faster than it can carry it and is destroyed. The established answers are the opposite: yielding support that accepts deformation, sliding steel sets with friction joints, deformable elements built into the shotcrete shell, a deliberate over-excavation allowance so that the finished profile survives the closure, and in the worst ground a fully mechanised approach that keeps the opening small and closes the ring immediately. The competence factor is what tells you to start having that conversation.

And this ratio is a snapshot of a process that has a clock in it. Real squeezing ground keeps moving after the face has passed — sometimes for months, occasionally for years. A convergence reading taken the day after excavation is not the closure the final lining will have to live with, and a competence factor computed once at the design stage says nothing about how long the movement will go on. Time-dependence is the part of squeezing that this and every other simple ratio leaves out entirely.

One inversion worth naming. Well above 2, the ground is comfortably competent in the squeezing sense — but if it is also strong, brittle and deeply buried, the concern flips over into strain-burst and spalling, which is a strength-to-stress problem running in the opposite direction. A very high competence factor at very high stress is not automatically good news; it just means the failure, if it comes, will be sudden rather than slow.