The ranges of a variety of geotechnical properties encompassed by the
ibnimbrite lithology are established. Ignimbrites are of low density (
1212-1928 kg m-3) and high porosity (18-51%), are very weak to weak in
compression (0.23-54 MN m-2), have low tensile strength (0.12-7.1 MN
m-2) and cohesion (0.06-9.0 MN m-2), yet the angle of internal frictio
n is comparatively high (27-degrees-38-degrees). The second-cycle slak
e durability index (I(d2)) ranges from very low (30%) to very high (99
.3%). Considerable softening (softening factors of 1.3-10.8) occurs on
saturation, and all ignimbrites undergo significant plastic deformati
on prior to failure. Extensive systems of open, continuous, vertical j
oints occur in many ignimbrites, typically forming a very widely space
d (3-5 m), irregular columnar pattern, while other ignimbrites are eff
ectively non-jointed, though occasional closed, continuous vertical jo
ints occur at extremely wide spacings (10-15 m). Large changes in stre
ngth and jointing may occur within a single profile. Two broad categor
ies of ignimbrite are recognised: durable ignimbrites with I(d2) great
er-than-or-equal-to 90%, and non-durable ignimbrites with I(d2) less-t
han-or-equal-to 30%. Durable ignimbrites behave as weak rocks for whic
h the rock mass characteristics exert the primary influence upon slope
stability and engineering behaviour. Steep slopes and precipitous gor
ges result from the predominance of vertical jointing, and high cleft
water pressures have contributed to historical engineering failures. N
on-durable ignimbrites are typically non-jointed and are the weakest m
aterials; they are classified as soft rooks, and the intact rock chara
cteristics are the dominant control on their behaviour. High friction
angles allow relatively steep slopes, but low durability makes them pr
one to gully erosion. They are typically very sensitive materials and
susceptible to piping. Second-cycle slake durability index and effecti
ve porosity together allow classification of ignimbrites and predictio
n of likely material characteristics. These tests require small specim
ens and simple laboratory equipment.