Research

Expansion sleeve rock anchors

This page shows some selected examples of axial tensile tests performed on the sorts of expansion rock anchors commonly used by cavers - M8 self-drill and drop-in styles. The table shows our results and for the 'interesting' cases we've included images and/or videos.

All the tests were performed using a calibrated Hydrajaws 2000 tensiometer and samples of good quality white limestone from the Cefn Mawr beds of North Wales. Bolt positions were at least 150mm from any bedding plane or visible cracking, and the surfaces were cleaned and prepared prior to placement to the same degree a caver would be expected to operate. Each style was tested a number of times and the range and mean (bold) figures are given.

Results

Bolt and fastener type	Peak load range mean	Images and videos
M8 generic drop-in sleeve Direct connection to tester SDS power drilled hole, smooth walls Sleeve level with surface	4.3kN - 9.6kN 6.1kN
M8 generic drop-in sleeve Direct connection to tester Rotary power drilled hole, rough walls Sleeve level with surface	5.8kN - 11.9kN 8.5kN
M8 SPIT MF8 self-drill sleeve Direct connection to tester Used to drill hole	4.0kN - 12.3kN 8.1kN
M8 SPIT MF8 self-drill sleeve Direct connection to tester Unused	4.5kN - 11.7kN 8.5kN
M8 Rawlbolt Direct connection to tester SDS power drilled hole, smooth walls Setting torque 10 ft-lbs	8.2kN - 13.2kN 11kN
M8 generic drop-in sleeve Petzl Twist alloy hanger including M8 bolt SDS power drilled hole, smooth walls Sleeve level with surface	2.8kN - 4.2kN 3.4kN

Discussion

M8 drop-in sleeves show an alarmingly low failure load in some tests, even though there is a wide range of values the anchor sleeves pull out of their holes gradually, with only a small amount of damage to the rock (plating). This demonstrates that the failure is due to low friction between the sleeve and the rock - as the limestone is very hard in compression the sleeve does not deform the shape of the hole in any way, so all the holding power is from friction between the metal and the rock. A power-drilled hole is smooth (modern SDS drills make it even smoother) and it can be seen that a badly-drilled hole with uneven ridges on the internal surface gives a far stronger fixing, with the failure being more jerky (as the metal of the sleeve has deformed into the grooves of the rock, and has to be sheared past them each time it slips).

The Rawlbolt-style sleeves tend to fail by snapping of the wedge sections rather than a gradual slip. The failures tend to happen in a few rapid bursts as the sleeve and the wedge fail. Under load the sleeve tends to withdraw by about 5mm under a load of 3kN - 4kN as the collet seats in position, with the final failure point following later.

With a hanger attached, the failure is from a levering action, and the generic M8 drop-in sleeves as used by many cavers fail by snapping or tearing level with the end of the bolt (where the tensile force is highest and the walls of the sleeve are thinned by the presence of the internal threads). Although the failure loads are still higher than a bodyweight, there is real concern for using these styles of hanger and bolt in axial load where any falls are possible - even one bodyweight can easily create an impact force higher than the 300-400kg failure point seen above. In shear loading these anchors will perform much better as the levering action on the sleeve is reduced.

[back to the research pages]