Ground Support 2023 – J Wesseloo (ed.

)
© 2023 Australian Centre for Geomechanics, Perth, ISBN 978-0-6450938-5-8
doi:10.36487/ACG_repo/2325_0.02

Laboratory-based drop testing of rock reinforcement

G Knox University of Toronto, Canada

Abstract
The requirement for resources has resulted in mining activities moving into more challenging conditions, from
conventional, gravity driven ground conditions to highly stressed rock mass. In highly stressed, burst-prone
rock masses, mining-induced seismicity presents a challenge to most ground support systems. The capacity
of conventional rock reinforcement elements such as grouted rebar rockbolts and friction rockbolts is often
found to be inadequate when subjected to large deformations resulting from mining-induced seismicity.
The requirement to sustain large loads over large deformations has led to the development of several
energy-absorbing rock reinforcement elements. The performance of an energy-absorbing element is typically
determined through a laboratory-based drop test. During a laboratory-based test, the kinetic energy of a
known mass, released from a known height, is transferred to the rock reinforcement element installed in a
steel tube. There are two primary drop test methods, impact testing and momentum transfer. Although there
are arguably differences between the two methods, both share common limitations. This paper provides a
summary of recent investigations conducted to understand the effect of the test parameters on the
performance of rock reinforcement elements determined through laboratory-based drop testing. The purpose
is to provide a high-level overview rather a detailed review.
Keywords: drop testing, impact testing, momentum transfer, rock reinforcement elements

1 Introduction
As mining resources accessed through surface and low stress mining environments are consumed, miners
are accessing ore bodies located in increasingly more challenging conditions. Consequently, mining methods,
mine design philosophies, and ground support elements have evolved to allow for the safe and efficient
extraction of ore in these challenging, extreme conditions. In severe stress conditions, the resultant failure
of the rock mass manifests as sudden and violent releases of energy in burst-prone rock masses. It was
observed that conventional rock reinforcement elements lacked the capacity to maintain the confining forces
in these conditions. To ensure the excavations remain operational for the anticipated service life, the
application of energy-absorbing rock reinforcement elements, such as paddled energy-absorbing rockbolts,
has become a best practice (Potvin & Hadjigeorgiou 2020). Consequently, the requirement for a method of
quantifying the performance of energy-absorbing rockbolts prior to the implementation of a field trial was
sought.
Ortlepp (1969) and his collaborators observed that a system comprising of ductile rock reinforcement
elements was able maintain the confinement of the rock mass; first during blast testing, illustrated in
Figure 1, and secondly during in situ applications. The success was attributed to the elements’ capacity to
sustain large loads over large displacements, hence the principal concept of an energy-absorbing rockbolt
was formed. Energy-absorbing rockbolts are now considered as a best practice in burst-prone ground
conditions (Cai & Kaizer 2018; Potvin & Hadjigeorgiou 2020).
Early blast testing trials conducted by Ortlepp (1969) successfully demonstrated that an energy-absorbing
rock reinforcement element was a potential solution to the support challenges encountered in burst-prone
rock masses. The test was a comparative test with one half of the tunnel supported with conventional support
and the other half with yielding rock reinforcement elements. The tests were successful; however, it was
clear an alternative method with a higher degree of control was required to suitably quantify the capacity of
the rock reinforcement. The blast tests were both costly and subject to the site-specific variables.
Repeatability through the control of the sample configuration and installation parameters is a valuable

Ground Support 2023, Perth, Australia 23

Laboratory-based drop testing of rock reinforcement G Knox

feature offered in laboratory-based testing that cannot be attained with in situ blast testing. In addition,
laboratory-based tests can be conducted at reasonable cost without affecting mining activities
(Hadjigeorgiou & Potvin 2011). The adoption of laboratory-based drop testing has facilitated the
development of energy-absorbing technologies such as the paddled energy-absorbing rockbolt (Li 2010)
which have become widely accepted for use in burst-prone conditions.

Figure 1 Tunnel profile after the second blast test conducted by Ortlepp (1969)

Current laboratory-based drop tests employ a mass free fall method. An impulse of energy is generated by
raising a quantified mass to a specific height. The mass is then released to free fall until arrested by the rock
reinforcement element coupled to the frame of the test rig. There are two methods of performing a drop
test: the impact test method and the momentum transfer method. It is arguable that momentum is
transferred in both methods, however, this convention has been adopted by the broader industry.
The Noranda test rig, as shown in Figure 2, was one of the early designs of a drop test rig which is now known
as the impact test method. The Noranda test rig was transferred to Canmet and the instrumentation
upgraded (Hadjigeorgiou & Potvin 2011). The design has since been adopted by the developers of both the
Dynamic Impact Tester (DIT) (Knox & Berghorst 2018) and Dynamic Testing Services (DTS) (Potvin
& Hadjigeorgiou 2020). The impact test method predates the momentum transfer method developed at the
Western Australian School of Mines (WASM) by Player et al. (2004). There are arguable differences between
the two methods; however, when the body of knowledge is examined, some similarities can be observed in
the results when the variations in instrumentation and data processing are accounted for. The primary
limitation of both methods is the fact that the loading applied is axial. Consequently, it does not represent
complex loading which could include tensile, shear, bending and torsional loading (Thompson et al. 2012).
This complex loading can result in failures illustrated by Stacey (2012) (Figure 3).

24 Ground Support 2023, Perth, Australia

 Keynote addresses

Figure 2 Noranda impact test rig (Simser 2007)

Figure 3 Complex loading of rock reinforcement elements captured by WD Ortlepp (Stacey 2012)

This paper servers to collate the current understand of laboratory-based drop testing from multiple research
programs. The scope of the paper will be limited to laboratory-based drop test method for rock
reinforcement methods and is therefore not an exhaustive review of all drop test methods.

2 Requirement for high strain rate testing

The change in response of materials to an increase in loading rate has been well-documented for both steel
(Cloete & Stander 2012; Malvar & Crawford 2008) and rock (Ryder & Jager 2002). The work of Cloete
& Stander (2012), shown in Figure 4, illustrates the increase in both yield and ultimate strength of a material
as the strain rate is increased. These tests were conducted on a series of homogeneous material which
repeatably demonstrate the relationship between an increase in strength related to an increase in strain rate.
Energy-absorbing rock reinforcement elements absorb energy through three primary mechanisms: plastic
deformation, slip, and extrusion. Typically, the principle of increased yield and ultimate strength can be
applied to rock reinforcement elements which rely on a method of plastic deformation as demonstrated by
Knox et al. (2018a) when comparing the impact load response to the quasi-static load response of a paddled
energy-absorbing rockbolt (Ø20 mm PAR1 Resin bolt). The converse, a decrease in the resisting force, was

Ground Support 2023, Perth, Australia 25

Laboratory-based drop testing of rock reinforcement G Knox

recorded by Player (2012), when a test was conducted on the Garford Dynamic Solid Bolt at the WASM test
facility. The Garford Dynamic Solid Bolt relies on an extrusion mechanism to absorb energy. An increase in
the strain rate resulted in a decrease of the resisting load resulting from change in the dynamic friction, as
illustrated in the load response shown in Figure 4.

(a) (b)
Figure 4 Effect of strain rate on the materials. (a) Change in response of annealed mild steel to varied
strain rates (Cloete & Stander 2012); (b) Effect of loading rate on the response of a Garford
Dynamic Solid Bolt (Player 2012)

3 Drop testing
There are two methods of performing high strain rate testing on full-scale rock reinforcement elements
within a laboratory: impact testing and momentum transfer. Both methods rely on a free-falling mass to
generate an impulse of energy which is dissipated through plastic deformation of the sample, hence the
common name of drop tests. The sample is defined as the rock reinforcement element installed within a host
tube. While there are arguable differences between the two methods, neither replicates the complex loading
cases to which a rock reinforcement element is subjected in situ (Stacey 2012). Consequently, both should
be considered as a comparative test method as opposed to a comprehensive simulation of what is observed
in the field.
The primary difference between the two high strain rate testing methods is the following:
• During an impact test (Figure 5b), the drop mass is released and impacts with a stationary impact
plate affixed to the proximal end of the samples. The distal end of the sample is coupled to the rigid
frame of the impact test rig. Consequently, the impact energy is dissipated as plastic deformation
of the sample.
• During a momentum transfer test (Figure 5a), the mass is coupled to the proximal end of the rock
reinforcement element. The distal end of the sample is coupled to a loading beam, with the entire
loading train (loading beam, sample installed within the host tube and drop mass) being released
from a known height. The loading beam impacts with a pair of dampers which control the rate of
deceleration as the momentum of the mass is transferred to the sample. The kinetic energy of the
loading train at impact is dissipated through a combination of the dampers and plastic deformation
of the sample.

26 Ground Support 2023, Perth, Australia

 Keynote addresses

(a) (b)
Figure 5 Simplified illustrations of drop test methods. (a) Momentum transfer method (Player 2012);
(b) Impact test method (Li et al. 2021)

Both methods rely on potential energy generated by raising a known mass to a known height. For simplicity,
the potential energy ( ), or input energy, is considered to be the energy imparted to the sample. In practice,
system efficiencies result in a loss of energy and consequently, the kinetic energy ( ) of the drop
mass (impact testing) or loading train (momentum transfer) is less than at the point of impact. Assuming
that the free-falling mass is arrested during the test, the primary component of the kinetic energy is
dissipated through the plastic deformation of the sample. Minor components of the energy are dissipated
through a combination of thermal energy, acoustic energy and frictional energy losses between the guide
rails and mass. For practical purposes, the minor components of the energy dissipated are considered as a
singular component known as the ‘system losses’ which can vary significantly between different test rigs.
During the period over which energy of the free-falling mass is dissipated, the displacement and resistive
load generated by the rock reinforcement element are recorded. The period of the impact is isolated, and
the energy dissipated ( ) is calculated through the integral of force ( ) and displacement ( ) (Equation 1).
This is determined by calculating the area under the load–displacement curve using trapezoidal integration
(Equation 2). The area is discretised using the width of the trapezoid defined by the displacement occurring
during a single sampling interval. What is frequently observed in impact tests is that the calculated dissipated
energy is greater than the input energy. This is because the input energy does not represent the total change
in potential energy of the mass, but merely the potential energy of the mass relative to the impact location.
The total potential energy of the sample mass is relative to the location at which the mass is arrested, which
is unknown at the initiation of the test. Hence, the total potential energy (Equation 3) is calculated using the
final location (drop height (H) + deformation ( )) which includes the additional deformation which occurs
during the dissipation of the energy. This phenomenon is not observed in the momentum transfer method,
as a component of the potential energy is consumed by the buffer system.

(1)
∑ 0.5 (2)
! (3)
The load–displacement response and resultant energy dissipated is calculated from the data recorded by the
load and displacement sensors. Consequently, to ensure reliable results, significant effort is directed to the
selection, maintenance, and verification of both the load and displacement sensing systems. The violent

Ground Support 2023, Perth, Australia 27

Laboratory-based drop testing of rock reinforcement G Knox

nature of the drop test and subsequent impact which occurs after the rupture of the sample increases the
probability of damage to the instrumentation. Therefore, a robust QA/QC verification program is required to
maintain the integrity of the results. For the same reason, optical displacement measurement systems are
typically employed at most testing rigs.
Due to a number of reasons, including complexity of testing and differences in data processing, the two
systems cannot be fully compared. In addition, there are few instances where the same rockbolt have been
tested using both methods with similar sample preparations and input parameters. One example is the
results of a Ø22 mm D-Bolt testing at both WASM (Villaescusa et al. 2015) and at Canmet (Li & Doucet 2012).
The load–displacement responses of the rockbolt are shown in Figure 6. It should be noted that the length
of the loaded segment was 1.2 m for the sample tested at WASM and 1.5 m for the sample tested at Canmet.
Both samples arrested the drop mass and neither test represented the ultimate displacement capacity.
During the data processing and analysis, a filter was applied to the data recorded at WASM, whereas the raw
data is presented by Li & Doucet (2012).

(a) (b)
Figure 6 Drop test results for a Ø22 mm D-Bolt. (a) Momentum transfer method (WASM) after Villaescusa
et al. (2015); (b) Impact test method (Canmet) after Li & Doucet (2012)

3.1 Drop testing rigs

Globally, there are currently six laboratory-based rock reinforcement drop testing rigs and a single
instrumented field testing rig. Figure 7 illustrates the locations of the rockbolt testing laboratories. Four of
the laboratory testing rigs are owned by a government or educational institute: Canmet (1)
(Plouffe et al. 2008), WASM (2) (Player et al. 2004), SWERIM (5) (Vallati et al. 2022) and the GIG (6)
(Pytlik 2020) which provide testing services to the industry. The remaining two laboratory-based drop testing
rigs, DIT (3) (Knox & Berghorst 2018a) and DTS (4) (Potvin & Hadjigeorgiou 2020) are privately owned by
rockbolt producers. Consequently, many of the tests are conducted on the respective companies’ products.
In-field testing is conducted using the Sandvik in situ rig enabling testing to be conducted at a mine site, as
described by Darlington et al. (2018, 2019). As the rig is privately owned, much of the data published to date
is on a mechanical hybrid rockbolt, the Sandvik MDX bolt.

28 Ground Support 2023, Perth, Australia

 Keynote addresses

Figure 7 Location of rock reinforcement drop testing facilities

3.2 Drop test rig instrumentation

Due to the nature of the drop test methodology, a robust instrumentation and data acquisition design is
critical to the integrity of the results. The energy of the free-falling mass is typically dissipated over a period
of 30–160 ms and consequently, the frequency response (bandwidth), resolution of the acquisition system
and location of the load sensors affect the recorded result.
There are two primary instruments to a drop test rig: the impact load cell and the displacement measurement
system. Due to the upper frequency limit of 50 kHz, piezo-electric load cells are commonly applied (Plouffe
et al. 2008; Player et al. 2004; Knox & Berghorst 2018a; Darlington et al. 2018; Vallati et al. 2022). Frequency
response of the load cell allows for the impulse to be accurately captured. The location/placement of the
load cells is of equal importance. This is demonstrated by Charrette & Plouffe (2008) and Li et al. (2022).
In both instances, impact test rigs were instrumented with two sets of load cells; the lower located at the
impact plate and the upper located at the coupling between the sample and the frame of the test rig. A delay
in the response of the upper load cell is observed in Figure 8 and attributed to the inertia of the sample.

(a) (b)
Figure 8 Delay in impulse propagation recorded between the frame and impact loads. (a) 4.9 kJ impact
on a Roofex rockbolt (Charrette & Plouffe 2008); (b) The difference in response between the
frame and impact load for an unspecificed impact on an unspecificed bolt tested in Rig 3
(Li et al. 2021)

The second parameter recorded during the impact is the displacement. The initial design of the Noranda test
rig included a linear displacement transducer used to record the deformation that occurred during the
impulse (Gaudreau 2004). This has subsequently been changed to an optical displacement sensor which has

Ground Support 2023, Perth, Australia 29

Laboratory-based drop testing of rock reinforcement G Knox

improved the durability of the sensor. The use of line scan and high-speed cameras facilitates a contactless
displacement measurement system capable of sub-millimetre accuracy once the lens distortion is accounted
for. The importance of synchronising the load and displacement acquisition systems is highlighted in Figure 9.
It can be observed that during the initial loading phase, no displacement is recorded, and only after the first
peak is reached does displacement occur. The inertia of the sample is affected by the mass of the sample.
Consequently, the delay between the load cell location (Figure 8) and the time elapsed between applying a
load and displacement occurring (Figure 9) will vary between testing rigs and sample configurations. As the
line scan camera is a digital system and the load cells an analog system, it is critical to ensure the two
acquisition systems share a common timing as post-processing data alignment will be incapable of correctly
representing the delay.

Figure 9 A load–displacement versus time plot for an impact test after Knox & Berghorst (2018)

A third consideration when reviewing the result of a drop test is the filtering applied to the data during
processing and analyses. Figure 10 illustrates results for an MP1 rockbolt subjected to a 47 kJ impact (Bosman
et al. 2018). A second order 150 kHz low pass Butterworth filter was applied to these data. The result is a
significant decrease in the magnitude of the first peak. In effect, when applying a filter, a delay is introduced
into the peak loading.

Figure 10 Effect of the application of a filter to an impact test result

3.3 Performance comparison between testing rigs

Due to the geolocation of the rigs, cost of testing and variations in construction and ownership, comparative
testing between rigs has historically been challenging. In 2018, at the 3rd Annual Rock Dynamics Conference,

30 Ground Support 2023, Perth, Australia

 Keynote addresses

Professor Charlie Li engaged multiple stakeholders to establish a framework for comparing the impact test
rigs. While some rigs declined to participate in the comparative testing campaign, the program was able to
successfully perform the same test on four test rigs: Canmet, GiG, Sandvik in situ and DIT. Three of the four
rigs employed the impact test methodology, while the fourth, the Sandvik in situ test rig, is designed around
a modified impact methodology. The piezo-electric load cell and accelerometer are located in the coupling
between the bar upon which the mass impacts and the proximal end of the rockbolt.
Li et al. (2021) reported the results of this comparative study between the participating rigs. In summary, a
fully threaded Ø22 mm bar supplied from a single batch of material was shipped to the laboratories and
tested using an identical test set up (direct impact), input energies (≈20 kJ) and impact velocity (4.4 m/s).
The load–displacement responses are reproduced in Figure 11. There are clear similarities between rigs 1, 2
and 3, but the difference in the construction and instrumentation is clearly highlighted by the difference of
the profile of the response recorded for rig 4. A variance in between the results recorded on the same rig is
also evident for rigs 1, 2 and 4.
The investigation demonstrated that there is a need for a standardised test method, however, due to the
differences in construction of the rigs, there is also a requirement for a universal calibration method.
Consequently, due to the bias of each rig, it is arguably impractical to compare the results of different rigs
without accounting for these differences.

Figure 11 Results generated off of the individual rigs for impact testing (Li et al. 2021)

4 Considerations when reviewing results

Of equal importance to the result is the sample preparation, configuration and finally the input parameters.
It is recognised that it is important to have a good understanding not only of what was tested, but how it was
tested.

Ground Support 2023, Perth, Australia 31

Laboratory-based drop testing of rock reinforcement G Knox

Two test configurations are typically employed: a direct impact continuous tube test and an indirect impact
split tube test. When configured for direct impact continuous tube test, the loading is applied directly to the
rock reinforcement element at the proximal end. The loading configuration can be likened to a strainburst
event (Li et al. 2021). As the loading is applied directly to the rock reinforcement element, it is of critical
importance that the ancillaries are correctly configured. This was highlighted by Simser (2007), where it was
observed that the plate–nut or plate–washer interface was failing prior to the tendon. The profile of the nut
can be clearly seen in the plate recovered after the test in Figure 12. Similarly, Player (2012) observed the
failure of thread–nut interface when testing fully threaded bars at WASM. Consequently, due to the failure
of the fixture accessories, the full capacity of the tendon was not mobilised. In contrast, Li & Doucet (2012)
observed that the loads applied to the accessories of multipoint anchored rockbolts was lower than the
impact load during indirect impact split tube tests.

(a) (b)
Figure 12 Failure of accessories during direct loading. (a) Thread–nut interface stripping (Player 2012);
(b) Nut–plate interface failure (Simser 2007)

When configured for an indirect impact split tube test, the loading configuration is considered to represent
a block thrust ejection (Li et al. 2021). The location of the split is typically located at the approximate midpoint
between the two anchors points. This is a common test methodology for determining the performance of a
paddled, energy-absorbing rockbolt. In both cases (continuous tube and split tube), the loading is limited to
a single section bounded by an anchor pair. For point anchored rockbolts, the continuous tube test represents
the capacity of the rockbolt. However, for multipoint anchor rockbolts, the split tube or continuous tube test
represents the capacity of a single section of the rockbolt bounded by two anchors. For this reason, the direct
impact split tube test was developed by Knox et al. (2018b). A discontinuity in the host tube is placed between
each anchor pair and consequently, the entire length of the rockbolt is loaded and the load is transferred
through the distal anchor.
The effect of these loading configurations on the performance of a rockbolt was demonstrated by Knox
& Hadjigeorgiou (2022), where the position of the split location and the configuration of the instrumentation
were adjusted, as illustrated in Figure 13. While this investigation highlighted the effect of the split location
on the performance of a paddled, energy-absorbing rockbolt, the program suggested that the performance
of a rockbolt may vary along the length of the rockbolt. A variance of 90% was recorded between a sample
with the split located in the centre of the proximal paddle set and a split located between the two distal
anchor points. Typical results for each testing configuration are compared in Figure 14. These tests were
conducted using a rockbolt which dissipates energy through the plastic deformation of steel. The capacity of
the distal anchor is best captured by placing the split location at the distal paddle set. The capacity of the
anchor was sufficient to result in rupture of the bar without significant movement within the resin. The
location of the split is an important consideration when comparing the performance of a plastic deformation
and slip for energy-absorbing rockbolts. The location of the split relative to the anchor will define the
maximum displacement capacity for an energy-absorbing rockbolt employing the slip or ‘ploughing’ method
to dissipate energy.

32 Ground Support 2023, Perth, Australia

 Keynote addresses

Figure 13 Sample configuration (Knox & Hadjigeorgiou 2022)

Figure 14 Effect of split location and loading configuration on the performance of a paddled
energy-absorbing rockbolt (Knox & Hadjigeorgiou 2022)

The input testing parameters are the drop mass and the height from which the mass is released (dropped).
From these two variables, the theoretical velocity at impact and impact energy are calculated. The indirect
relationship between the input energy and the cumulative total energy absorbed has been observed by Li
& Doucet (2012) when testing the D-Bolt, by Player (2012) for the Garford Dynamic Solid Bolt, and Bosman
et al. (2018) when testing the MP1 Bolt. The D-Bolt and MP1 rockbolts are both designed to dissipate energy
through the plastic deformation of the steel tendon from which it is produced. Consequently, as shown in
Figure 15, the total cumulative displacement decreased with the increase in the magnitude of the impact
energy, while the average impact force remained constant. This resulted in a reduction in the cumulative
absorbed energy (Figure 16). The Garford Dynamic Solid Bolt, however, dissipates energy through an
extrusion mechanism. While a reduction in the cumulative absorbed energy was recorded (65–70 kJ down to
53 kJ) with the increase in impact energy, a reduction in the resistive force was observed. The reduction in
the resistive force is illustrated in Figure 4b.

Ground Support 2023, Perth, Australia 33

Laboratory-based drop testing of rock reinforcement G Knox

(a) (b)
Figure 15 Effect of impact energy on total displacement. (a) Effect on a Ø20 mm MP1 rockbolt
(Bosman et al. 2018); (b) Effect on a Ø22 mm D-bolt (Li & Doucet 2012)

Figure 16 Effect of input energy per drop on cumulative maximum absorbed energy

In their investigations, both Player (2012) and Li & Doucet (2012) varied the impact velocity and impact
energy, however, Bosman et al. (2018) maintained the impact velocity and modified the impact mass.
To demonstrate the effect of the impact velocity on the performance of a rockbolt under impact loading,
Knox et al. (2018a) conducted a series of impact tests on a PAR1 Resin paddled energy-absorbing rockbolt.
During the investigation the impact energy was maintained and the impact velocity of the mass was varied.
Within the range of velocities tested the velocity was found to have a negligible effect on the cumulative
absorbed energy. The difference between the observations of Player (2012) and Knox et al. (2018a) can be
explained when the variation in the employed rockbolts energy absorption method is accounted for (plastic
deformation versus extrusion).

5 Future considerations
Drop tests provide valuable insights into the performance of rock reinforcement systems during high strain
rate events. In the past, significant emphasis has been directed at understanding the effect of the input
parameters, sample configuration and components included in the sample on the performance of the rock
reinforcement element. However, the effect of variances within a single grade of steel on the performance
of a rock reinforcement element have not received the same level of attention. An example of reporting the
full traceability for a testing program has been provided by Knox & Hadjigeorgiou (2022).

34 Ground Support 2023, Perth, Australia

 Keynote addresses

Several testing campaigns have demonstrated the performance of a rock reinforcement element produced
from a singular batch of steel. This represents the performance of a variant of the rockbolt design.
The approach does not capture the range of the anticipated responses from the rockbolt design when the
variances in the mechanical properties of a steel are considered. It is reasonable to consider steel as a
homogenous material. Consequently, maintaining production of a rockbolt from a singular grade of a
material mitigates the risk of large variations in performance. However, the range of mechanical properties
of a given material is magnified due to the dynamic increase factor (Malvar & Crawford 1998), therefore an
increase in the variability of the performance at high strain rate is expected. It is arguably not practical to
perform destructive testing on every batch of material from which a rockbolt is produced. However, a
well-documented annual QA testing would facilitate the generation of sufficient data to perform a statistical
analysis on the results. Due to the lack of a global standardisation of the testing programs and demonstrated
differences in the test rigs employing the same method, these tests should be conducted on a single drop
test rig.
The focus of this paper was directed at laboratory-based drop testing of rock reinforcement elements.
These test facilities have assisted in the accelerated development of innovative rock reinforcement elements;
however, attention to the interface between the rock reinforcement element and the surface support is
lacking. Some insights have been provided by Villaescusa et al. (2015) and Brandle et al. (2019); however, it
can be argued that the boundary conditions applied to the mesh do not accurately reflect an in situ
installation. The ability to test the interface between the surface support and the rock reinforcement
elements is a limitation of the current laboratory-based drop testing facilities.
Practically, an in situ rock reinforcement element will be subjected to a series of quasi-static and dynamic
displacements during its service life, as opposed to a singular loading to rupture experienced during a
laboratory-based test. Knox & Berghorst (2019) and Crompton & Knox (2022) explored the effect of first
applying a quasi-static loading to a rockbolt followed by a destructive impact test. The investigation
illustrated the negative effects of the quasi-static displacement on the energy capacity during an impact test.
These investigations demonstrated an increase in the total displacement to rupture, an observation which
will require further investigation. This demonstrates the need to further explore alternative loading
configurations.

6 Conclusions
There are two methods of performing laboratory-based drop tests: the impact test method, and the
momentum transfer test method. Although there are differences between the two methods in construction
and operation, both require a free-falling mass to generate an impulse of energy. Both approaches employ
impact and a transfer of moment. Even in the best documented comparative testing campaign by Li
et al. (2021), all participating rigs employed a variation of the impact method. Therefore, comparisons
between the two methods are limited. Due to the cost of operation and geolocation of the testing facilities,
there are few published results of identical rock reinforcement elements tested using both methods. It is
important to further explore how differences in data processing can influence testing data obtained by
different drop test methods.
In the past, significant emphasis has been directed at understanding the effect of the input parameters,
sample configuration and components included in the sample on the performance of the rock reinforcement
element. Several of these tests have been conducted on paddled energy-absorbing rockbolts which rely on
the plastic deformation of steel (Li 2012; Bosman et al. 2018; Knox et al. 2018a). The work of Player (2012)
demonstrated the challenge in interpretation of results from rock reinforcement elements which employ
alternative energy absorption mechanisms such as the extrusion method. However, it can be concluded
irrespective of the drop testing method that the cumulative absorbed energy is indirectly proportional to the
impact energy.
Laboratory-based drop tests provide an opportunity to conduct comparative testing under controlled
conditions and a relatively acceptable cost. The primary limitation is the fact that the test is an axial loading

Ground Support 2023, Perth, Australia 35

Laboratory-based drop testing of rock reinforcement G Knox

case and consequently does not account for the complex loading to which a rock reinforcement element is
subjected to in situ. Acknowledging this limitation, laboratory-based drop testing is a powerful tool for
performing quality assurance testing and verifying the performance of a new technology prior to a field trial.
When reviewing the result of a drop test, careful consideration should be given to the effect of the sample
configuration and test parameters as these have been demonstrated to have a dramatic effect on the
performance of the rock reinforcement element being tested.

Acknowledgement
The author acknowledges the contributions of multiple authors who have investigated and published the
results on using different drop tests and their effect on the performance of rock reinforcement elements.
The author acknowledges Epiroc for providing access to the testing facility and its comprehensive database
of test results, and Professor John Hadjigeorgiou from the University of Toronto for providing guidance and
useful feedback on the evolution of testing practices.

References
Bosman, K, Cawood, M & Berghorst, A 2018, ‘Relationship between energy per impulse and dynamic capacity of a rockbolt’, in CC Li,
XB Li & ZX Zhang (eds), Proceedings of Rock Dynamics – Experiments, Theories and Applications, Trondheim, pp. 379–384.
Brandle, R, Fonseca, RL & Hangartner, JR 2019, ‘Large scale testing of surface support’, in W Joughin (ed.), Proceedings of the Ninth
International Conference on Deep and High Stress Mining, The Southern African Institute of Mining and Metallurgy,
Johannesburg, pp. 149–160.
Cai, M & Kaiser, PK 2018, Rockburst Support Reference Book, MIRARCO, Sudbury.
Charette, F & Plouffe, M 2008, ‘A new rock bolt concept for underground excavations under high stress conditions’, in Proceedings
of the 6th International Symposium on Ground Support in Mining and Civil Engineering Construction, SAIMM, SANIRE and
ISRM, Johannesburg, pp. 225–240.
Cloete, TJ & Stander, M 2012, ‘Refinement of the wedge bar technique for compression tests at intermediate strain rates’, in DYMAT
2012 - 10th International Conference on the Mechanical and Physical Behaviour of Materials under Dynamic Loading, EDP
Sciences, Les Ulis.
Crompton, BR & Knox, G 2022, ‘Dynamic testing: determining the relationship between rockbolt diameter and the residual dynamic
capacity of an axially strained tendon’, in Y Potvin (ed.), Caving 2022: Fifth International Conference on Block and Sublevel
Caving, Australian Centre for Geomechanics, Perth, pp. 141–150.
Darlington, B, Rataj, M, Balog, G & Barnett, D 2018, ‘Development of the MDX Bolt and in-situ dynamic testing at Telfer Gold Mine’,
Rock Dynamics and Applications, CRC Press, Boca Raton.
Darlington, B, Rataj, M & Roach, W 2019, ‘A new method to evaluate dynamic bolts and the development of a new dynamic rock
bolt’, in W Joughin (ed.), Deep Mining 2019: Proceedings of the Ninth International Conference on Deep and High Stress
Mining, The Southern African Institute of Mining and Metallurgy, Johannesburg, pp. 205–216, https://doi.org/10.36487
/ACG_rep/1952_16_Darlington
Gaudreau, D 2004, ‘Performance assessment of tendon support systems submitted to dynamic loading’, in E Villaescusa & Y Potvin
(eds), Ground Support in Mining and Underground Construction, Taylor & Francis Group, London, pp. 548–573.
Hadjigeorgiou, J & Potvin, Y 2011, ‘A critical assessment of dynamic rock reinforcement and support testing facilities’, Rock Mechanics
and Rock Engineering, vol. 44, pp. 565–578.
Knox, G & Berghorst, A 2018, ‘Increased agility for the research and development of dynamic roof support products’, in CC Li, XB Li &
ZX Zhang (eds), Proceedings of Rock Dynamics – Experiments, Theories and Applications, Trondheim, pp. 373–384.
Knox, G, Berghorst, A & Crompton, B 2018a, ‘The relationship between the magnitude of impact velocity per impulse and cumulative
absorbed energy capacity of a rock bolt’, Proceedings of the 4th Australian Ground Control in Mining Conference, Sydney,
pp. 160–169.
Knox, G, Berghorst, A & de Bruin, P 2018b, ‘An empirical comparison between new and existing laboratory-based dynamic sample
configurations’, in Y Potvin & J Jakubec (eds), Caving 2018: Proceedings of the Fourth International Symposium on Block and
Sublevel Caving, Australian Centre for Geomechanics, Australian Centre for Geomechanics, Perth, pp. 747–758.
Knox, G & Berghorst, A 2019, ‘Dynamic testing: determining the residual dynamic capacity of an axially strained tendon’, in Y Potvin
(ed.), Ground Support 2019: Proceedings of the Ninth International Symposium on Ground Support in Mining and Underground
Construction, Australian Centre for Geomechanics, Perth, pp. 231–242.
Knox, G & Hadjigeorgiou, J 2022, ‘Influence of testing configuration on the performance of paddled energy-absorbing rockbolts under
impact loading’, Rock Mechanics and Rock Engineering, vol. 55, pp. 5705–5721.
Li, CC 2010, ‘A new energy-absorbing bolt for rock support in high stress rock masses’, International Journal of Rock Mechanics and
Mining Sciences, vol 47, 3rd edn, pp. 396–404.
Li, CC & Doucet, C 2012, ‘Performance of D-bolts under dynamic loading conditions’, Rock Mechanics and Rock Engineering, vol 45,
2nd edn, pp. 193–204.

36 Ground Support 2023, Perth, Australia

 Keynote addresses

Li, CC, Hadjigeorgiou, J, Mikula, P, Knox, G, Darlington, B, Royer, R, ... Hosp, M 2021, ‘Performance of identical rockbolts tested on
four dynamic testing rigs employing the direct impact method’, Journal of Rock Mechanics and Geotechnical Engineering,
vol. 13, 4th edn, CSRME, pp. 745–754.
Malvar, LJ & Crawford, E 2008, ‘Dynamic increase factor for steel reinforcing bars’, in Twenty-Eighth DDESB Seminar, Orlando,
pp. 1–17.
Ortlepp, WD 1969, ‘An empirical determination of the effectiveness of rockbolt support under impulse loading’, in Proceedings of the
International Symposium on Large Permanent Underground Openings, Oslo, pp. 197–205.
Player, JR 2012, Dynamic Testing of Rock Reinforcement Systems, PhD thesis, Curtin University of Technology, Perth.
Player, JR, Villaescusa, E & Thompson, AG 2004, ‘Dynamic testing of rock reinforcement using the momentum transfer concept’, in E
Villaescusa & Y Potvin (eds), Proceedings Fifth International Symposium on Ground Support, Ground Support in Mining and
Underground Construction, Balkema, Rotterdam, pp. 327–339.
Potvin, Y & Hadjigeorgiou, J 2020, Ground Support for underground mines, Australian Centre for Geomechanics, Perth.
Plouffe, M, Anderson, T & Judge, K 2008, ‘Rock bolts testing under dynamic conditions at CanMET-MMSL’, Proceedings of the 6th
International Symposium on Ground Support in Mining and Civil Engineering Construction, Southern African Institute of
Mining and Metallurgy, Cape Town, pp. 581–596.
Pytlik, A 2020, ‘Comparative shear tests of bolt rods under static and dynamic loading’, Studia Geotechnica et Mechanica, vol. 42,
no. 2, pp. 151–167.
Ryder, JA & Jager, AJ 2002, Rock Mechanics for Tabular Hard Rock Mines, The Safety in Mines Research Advisory Committee,
Johannesburg.
Simser, BP 2007, ‘The weakest link - ground support observations at some Canadian shield hard rock mines’, in Y Potvin (ed.),
Proceedings of the Fourth International Seminar on Deep and High Stress Mining, Australian Centre for Geomechanics, Perth,
pp. 335–348.
Stacey, TR 2012, ‘A philosophical view on the testing of rock support for rockburst conditions’, The Journal of The Southern African
Institute of Mining and Metallurgy, vol. 112, pp. 703–710.
Thompson, AG, Villaescusa, E & Windsor, CR 2012, ‘Ground support terminology and classification: an update’, Geotechnical and
Geological Engineering, pp. 553–580.
Vallati, O, Darlington, B & Sandberg, L 2022, ‘Dynamic drop testing of Sandvik’s D47 and D39 MDX bolts at the Swerim’s testing
facility’, Proceedings of The Fifth Australasian Ground Control in Mining Conference Proceedings, Australasian Institute of
Mining and Metallurgy, Melbourne, pp. 428–440.
Villaescusa, E, Thompson, AG & Player, JR 2015, MRIWA Report No. 312: Dynamic Testing of Ground Support Systems, JR Player (ed.),
MRIWA, Perth.

Ground Support 2023, Perth, Australia 37

Laboratory-based drop testing of rock reinforcement G Knox

38 Ground Support 2023, Perth, Australia