Development and Testing of a Modular Rockfall Protection Wall to Mitigate Earthquake-Induced Slope Hazards

Prepared for the 69^th Highway Geology Symposium, September, 2018

ABSTRACT

The November 2016 M7.8 Kaikoura earthquake resulted in excess of 40 landslides that directly impacted the key road and rail corridor on New Zealand’s South Island. Within two months, the New Zealand Government formed the North Canterbury Transport Infrastructure Recovery (NCTIR) alliance, a team of more than 1700 workers who were tasked with restoring road and rail service by the end of 2017.   

The work has involved a wide variety of landslide hazard mitigation measures that have included source treatment, installation of passive rockfall protection measures and relocation of sections of road further away from the base of the slope onto new seawalls.  One of many challenges facing the geotechnical design team is space limitations along the narrow coastal corridor. 

A modular rockfall protection wall has been developed to add to the suite of permanent rockfall protection structures in use on the project.  The wall comprises interconnected concrete blocks with an upslope energy-absorbing layer of sand-filled and rock-filled gabions.  The key advantages of the wall are a narrow footprint and a relatively fast installation time.  

It was necessary to demonstrate the performance and capacity of the wall before it could be approved for use on site.  Full-scale physical testing was performed at a vehicle impact testing facility.   Six tests were undertaken to investigate sliding and overturning failure modes; impact energies were 250 and 750 kJ.  Data collected during testing includes multiple high-speed videos and pre- and post-test laser scans.  

The wall performed successfully, and it has been approved for use on site.  The first installation is anticipated by mid-to-late 2018.  

INTRODUCTION

The November 2016 M7.8 Kaikoura earthquake caused significant damage to transportation infrastructure located in the northeast of New Zealand’s South Island.  Part of the damage was due to nearly 1 million cubic metres of rock falling onto the Main North Line (MNL) railway and State Highway 1 (SH1) from more than 40 primary landslides, cutting off a major transportation corridor and isolating the town of Kaikoura and surrounding rural communities.

By the end of 2016, the New Zealand Government made the decision to form an alliance to undertake work to restore the coastal transportation corridor.  NCTIR, the North Canterbury Transport Infrastructure Recovery, is an alliance partnership between the New Zealand Transport Agency (NZTA), Kiwirail and four major construction contractors (1).  The alliance team has consisted of up to 1700 people from more than 100 organisations.  They were given the challenge of re-opening the corridor by the end of 2017.    

One of two NCTIR geotechnical design teams was tasked with works related to characterization and mitigation of slope hazards along 28 km of coastal corridor affected by landslides.  The work involved design and construction of landslide hazard mitigation works, from mapping and characterization of landslides to design and construction of protection structures.   

A key part of the work involved finding robust rockfall protection solutions for fragile, earthquake-damaged slopes that could be constructed relatively quickly within a narrow corridor.  To this end, a modular rockfall protection (MRP) wall has been developed and tested for use on the NCTIR project.

Overview of Earthquake and Damage

The 14 November 2016 M7.8 Kaikoura earthquake was a complex event that involved rupture along multiple faults. Figure 1 shows the area most affected where significant ground shaking occurred.  The event was felt throughout most of New Zealand.  Fault rupture propagated northwest from the epicenter; surface ground rupture was observed along at least 20 faults spanning a distance of about 100 km.  

Due to the significant ground shaking, more than 10,000 landslides were generated over an area of about 10,000 km² (3).  The area affected by landslides is shaded red in Figure 1.

Figure 1: Area affected by November 2016 M7.8 Kaikoura earthquake (2)

More than 80 landslides either directly affected or occurred upslope of the transportation corridor.  The main part of the transportation corridor affected by landslides extends for a distance of about 7 km south and 21km north of Kaikoura; the location of Kaikoura is shown by the KIKS station in Figure 1.   Figures 2 and 3 show the transportation corridor and provide some general context for the project setting and scope.

Figure 2: Panoramic views of SH1 / MNL north (upper) and south (lower) of Kaikoura;bare areas along the lower slopes are where landslides have occurred

Figure 3: Two of the largest landslides affecting the corridor; for scale, the landslide on the right is up to about 250 m high.

MODULAR ROCKFALL PROTECTION WALL

An additional rockfall protection solution with a relatively narrow footprint and low deformation under impact loading is needed along several areas of the corridor where there are space constraints between the slope toe and road/rail alignment. The slopes in many of these areas contain varying quantities of potentially unstable material and are expected to generate multiple rockfalls over time.   Flexible barriers such as rockfall fences are not considered a practical option in some areas given the anticipated barrier deformation and the amount and frequency of rockfalls.

Stacked mass concrete blocks have been used in these areas as temporary rockfall protection walls, however the energy capacity, deformation and damage response of these of structures is unknown. The question arose as to whether a permanent rockfall protection wall could be developed using stacked concrete blocks that could be quickly erected at locations where the space requirements did not suit existing protection systems. Given the anticipated range of rockfall energies, the rockfall protection solution would need to be able to withstand moderate impact energies in the range of 300 to 400 kJ (or more, if possible).

Development

Work undertaken by others has been considered in the development of the MRP wall configuration.  The concepts of particular interest are the use of gabion baskets as an energy-dissipating layer and the performance of concrete in rockfall protection structures.  The work considered includes two separate PhD research projects involving cellular (gabion) structures (4,5) and concrete roadside barriers (6), and their use as rockfall protection structures.  Both studies included physical testing programmes; the findings related to the behaviour and performance of structures during physical testing were considered in the development of the MRP wall configuration and its testing.

Cellular Gabion Wall

Researchers in France have undertaken work to evaluate the use of gabion baskets in rockfall protection structures.  This work has included an evaluation of the deformation of individual rock and sand-filled gabion baskets, and small-scale and full-scale testing of cellular gabion sandwich structures composed of layers of rock-filled and sand-filled gabions (4, 5, 7, 8, 9).

Of particular interest for the development of the MRP wall was the performance of the rock and sand gabion layers in terms of deformation and dissipation of impact energy.   The researchers undertook full-scale testing of a 2m-thick, 4m-high cellular wall backed by an earthen embankment; and a 3m-thick, 4m-high cellular wall.  The cellular walls were formed by 1m-thick rock gabion and sand gabion layers.  The test energies ranged from 200 kJ to 2200 kJ (8). This work was the basis for selection of rock gabion and sand gabion layers as a composite energy-absorbing layer, and it also helped to guide the selection of impact energies used in testing the MRP wall.

Concrete Barriers

The Ohio Department of Transportation (ODOT) recently sponsored research aimed at better defining the performance and energy capacity of concrete roadside barriers used as rockfall protection (6). Part of work involved physical testing of precast and cast-in-place concrete barriers. The barrier designs were modified to investigate the effects of various energy-absorbing features on the energy capacity and resulting damage to the barriers. This included varying the reinforcing steel type, size and spacing; as well as using different types of fibre-reinforced concrete. Test energies were up to 160 kJ. ODOT used the results of the work to modify the design of concrete roadside barriers where they are used as rockfall protection.

Of particular interest for the development of the MRP wall is the improvement in energy capacity and performance with the use of steel fibre-reinforced concrete. The addition of steel fibres significantly reduced concrete spalling and increased the energy absorption capacity by 30 to 100 percent, depending on the barrier type and test impact location.

MRP Wall Configuration

The modular rockfall protection wall configuration selected for testing (Figure 4) utilises a modified configuration of sea-wall blocks developed for the NCTIR project together with an upslope energy-dissipating layer consisting of sand-filled and rock-filled gabion baskets. The blocks are 2 m x 1m by 1m (L x W x H); each weighs about 5000 kg; they are chamfered on the upslope side to allow for easier installation around curves.  The gabion baskets are 2 m x 0.5 m x 0.5 m (L x W x H).  The rock fill is as per the gabion manufacturer’s specification.  The sand fill is concrete sand with a maximum grain size of 5 mm, lightly compacted within a geotextile-lined gabion basket.

Figure 4: Cross-Section through Modular Rockfall Protection Wall

The concrete blocks are installed in an interlocking arrangement and are joined together using vertical steel shear bars. The 32 mm diameter steel bars are installed within a 100 mm diameter open duct, affixed with a plate and nut in the top block.  The system is able to dissipate energy on large-scale impact via deformation of the rock and sand gabion layers and sliding and rotation of the individual rigid blocks, while still remaining joined as a coherent barrier to further rockfall.

Development of the wall configuration was a collaborative effort amongst the NCTIR design team, Stahlton Engineered Concrete and Geofabrics New Zealand. Stahlton provided the modified sea-wall block design, including the steel shear bar connections within the concrete blocks. Geofabrics provided general information and advice on the gabion basket layers; this advice included input from Maccaferri who have expertise in rockfall protection solutions.

The motivation for using the sea-wall blocks was two-fold.  First, they could be fabricated using the concrete molds developed for the sea-wall blocks, saving both cost and time.  Second, with over 7000 sea-wall blocks being planned for use on the project, considerable experience will have been developed in their fabrication and installation.  

Test Design

A testing standard specifically applicable to rockfall protection walls formed with rigid elements does not exist. Instead, researchers and manufacturers have undertaken numerical modelling, physical testing and back-analysis of actual rockfall impacts to evaluate the performance and energy capacity of these types of structures (10). Rockfall protection walls are typically designed on the basis of allowable or acceptable deformation, considering an impact by a design boulder with a specified impact velocity. 

Testing Programme

In order to demonstrate the performance of the wall sufficiently, so that it could be approved for use on the project, it was decided to use a European testing guideline developed for dynamic rockfall fences as a basis for developing the testing programme. This was discussed and decided upon by the NCTIR design team, Stahlton and Holmes. The document used is the ETAG 027 Guideline for European technical approval of falling rock protection kits, published by the European Organisation for Technical Approvals (11). 

The key aspect of the ETAG 027 considered in developing the MRP wall testing programme was the selection of two impact energy levels representing Serviceability and Maximum energy levels.  Under ETAG 027, these broadly are: 

•  Serviceability Energy Level (SEL): The barrier should be able to withstand two impacts with no repairs after the first impact.  SEL is typically used as a design criteria where multiple or frequent rockfall impacts are anticipated.
•  Maximum Energy Level (MEL):  The barrier should be able to stop a single impact. It is expected that the barrier will need substantial repair or replacement after an MEL impact.  MEL is typically used as a design criteria where infrequent rockfall impacts are anticipated.  MEL is defined as 3 x SEL energy level.

The SEL and MEL designations used in ETAG 027 are for dynamic rockfall barriers and are not terms that are typically used to designate energy levels for rigid-type barriers, such as an embankments or this wall. The terms have been used here to indicate the likely “frequent” (SEL) and “infrequent” (MEL) rockfall impact energies.

In addition to impact energies, the other key factor considered in developing the testing programme was failure mode, either sliding or overturning.  These failure modes were tested by varying the impact height.  Sliding was evaluated by impacting the wall at mid-height;    overturning was evaluated by impacting the wall in the upper third of the wall height. 

A total of 6 tests were planned to evaluate the energy capacity and performance of the MRP wall; the testing programme is summarized in Table 1 and illustrated in Figure 5.  

*no repairs to wall after first test

Figure 5: Test impact locations for sliding and overturning failure modes

Test Walls

Two 3m-high by 10m-long walls were constructed for testing (Figure 6).  The walls were constructed within a ditch in order to achieve the impact heights; the ends of the walls were not constrained.  The two test walls were designated A and B. Test Wall A was used to investigate the sliding failure mode using a 1.5m impact height (Tests 1 to 3). Test Wall B was used to investigate the overturning failure mode using a 2.25m impact height (Tests 4 to 6).  The test walls were substantially re-built following the 2 x 250 kJ impacts.   

Figure 6: Overhead and side view of test wall

The composition of the concrete blocks differed in Test walls A and B.  Test Wall A was constructed using 50 MPa (28-day strength) plain concrete; Test Wall B was constructed using 50 MPa concrete reinforced with steel fibres at a dosage rate of 20 kg/m³.  It was anticipated that spalling of the concrete blocks would potentially be an issue, and the option to add steel fibres to the concrete blocks for Test Wall B was included as part of the testing program.  

Test Set-up

Impact energies were delivered to the MRP wall via a rolling bogey fitted with a spherical impacting head.  The impacting head consisted of a 1-m diameter concrete-filled, steel-reinforced spherical steel dome (Figure 7).  The bogey was fitted with steel ballast to scale the weight up and down to achieve the target impact energies.  The bogey travelled along a guide rail and was propelled using a tow rope attached to a drop-weight system.  The drop-weight system was composed of a known mass of concrete blocks that were lifted with a crane and attached to the tow rope via a system of pulleys (Figure 8).  The mass was lifted to a specified height and dropped such that the bogey reached a target velocity on impact with the MRP wall; target impact velocities were in the range of 20 m/s, which is within the mid-to-upper range of possible rockfall velocities. The tow cable dropped off of the bogey immediately prior to impact so that it was travelling freely on impact.  

Testing was conducted at the Holmes Solutions testing facility in Christchurch, New Zealand.  Holmes Solutions are an ISO 17025 Accredited Testing Laboratory under the International Laboratory Accreditation Cooperation (ILAC) scheme audited by International Accreditation New Zealand (IANZ).  Holmes has substantial experience with full-scale dynamic impact testing of roadside safety barriers to US, European and Australia/New Zealand testing standards.

Figure 7: Test bogey and guide rail system

Figure 8: Test set-up

Data Collection

Data collected during the tests consisted of high speed video (up to 300 frames per second) from multiple cameras positioned at the sides, front and above the wall.  An accelerometer was installed on the test bogey to record velocity before and during impact.  Horizontal displacement measurements were taken manually at discrete locations for all tests.  Additional displacement data was acquired via laser scanning to provide a comprehensive survey of the wall before and after each test. Laser scanning was undertaken by Eliot Sinclair surveyors for 4 of 6 tests.  

Test Results

A total of 6 tests were conducted for the planned testing programme.  The wall successfully stopped the bogey in all tests.  

Energy dissipation in the MRP wall was observed through the following mechanisms:

• Deformation of rock gabion layer
• Deformation of sand gabion layer
• Displacement of impacted concrete block(s), both translational and rotational 
• Displacement of adjacent concrete blocks; engaged through block-to-block contact and through steel connections
• Deformation of steel connections, both rebar and steel plates
• Deformation of foundation, including slight embedment of the toe of wall and rotation of the wall about the toe

A summary of the actual impact conditions, horizontal wall displacements and rotational displacements are presented in Table 2.  The horizontal displacements are measured at the front face of the concrete blocks; the rotation is measured about the front toe of the wall. 

Figure 9 shows a comparison of the horizontal displacements for selected tests; laser scan results are shown where available.  Of note is the displacement pattern and greater number of blocks engaged for the higher energy impacts.  Figure 10 shows the bogey penetration and gabion basket deformation for the same set of tests as in Figure 9.  Damage to the gabions was generally confined to the impact zone, however there was increased deformation of the gabions above the impact point. 

Figure 9: Comparison of horizontal displacements for selected tests (colour scales differ)

Figure 10: Comparison of bogey penetration and gabion deformation for selected tests

Figures 11 and 12 show the front face of the wall following the 750 kJ tests for sliding and overturning (Tests 3 and 6).  Of particular note is the difference in damage to the plain and steel fibre-reinforced concrete blocks.  Spalling occurred in the plain concrete blocks, with relatively large pieces of concrete being lost off the blocks due to contact between the blocks as they displaced during impact.    The damage to the fibre-reinforce blocks consisted of cracking and crushing; no spalling occurred.  The paint marks in both photos indicate damage that occurred following the 2 × 250 kJ tests; the blocks were re-arranged when each of the walls was repaired following the 250 kJ tests.   Minimising spalling is an important road safety consideration if the MRP wall is to be located adjacent to a roadway. 

Additional damage sustained by the wall consisted of bending of the steel bars and deformation of the steel plates (Figure 13).    

Figure 11: Test Wall A following Test 3 (sliding, 750 kJ); plain concrete blocks

Figure 12: Test Wall B following Test 6 (overturning, 750 kJ); steel fibre-reinforced concrete blocks

Figure 13: Damage to steel bars and plates following Test 5 (2 x 250 kJ, overturning)

Test to Destruction

A 7^th test was undertaken immediately after Test 6 in order to further investigate the failure mode of the MRP wall.  No repairs were made to the wall and the test was undertaken using the same target impact energy and height as for Test 6.   

The gabion layers were substantially damaged during Test 6 (Figure 14). The rock and sand gabions deformed and had a reduced thickness; additionally the sand gabion would have undergone some compaction. 

Figure 14: Test Wall B following Test 6

The impact velocity for Test 7 was 23.1 m/s and the impact energy was 771 kJ.  The failure mode was detachment of the upper central block with punching of the steel anchor plates and bending of the steel bars (Figure 14).  The 5000 kg block came to rest about 1.6m from the front of the wall with its top face resting on the ground.  

Figure 15: Test Wall B following Test 7

FUTURE USE

Based on the results of the physical testing programme, the MRP wall has been approved for use on the NCTIR project.  The recommended energy capacity limits that have been adopted are presented in Table 3.  

*for first impact

Additional work that was undertaken in the order to gain approval includes consideration of seismic and debris loading conditions, design life, and inspection and maintenance.   

Some of the advantages of the MRP wall for the NCTIR project are:

• Re-use of concrete molds developed for the sea-wall blocks.
• Leveraging of site experience with fabricating and installing sea-wall blocks.
• Reduced footprint width in comparison with a similar embankment-type structure.
• Anticipated reduced construction time in comparison with similar embankment-type structure;  this has the advantage of minimizing the time workers spend in potentially hazardous areas, and it reduces the road closure time.

Potential for staged installation of MRP wall, with concrete blocks installed in advance of gabion baskets; this may allow for use of concrete blocks as temporary rockfall protection.

Potential for further reduction of construction time if gabions are pre-filled and lifted into place.

REFERENCES

• New Zealand Transport Agency.  Kaikoura Earthquake Response.  https://www.nzta.govt.nz/projects/kaikoura-earthquake-response/.  Accessed May 2018.
• GNS Science.Kaikoura quake produced strongest ground shaking in NZ, new research shows.https://www.gns.cri.nz/Home/News-and-Events/Media-Releases/strongest-ground-shaking-in-NZ. Accessed April 2018.
• C. Massey, D. Townsend, E. Rathje, K. E. Allstadt, B. Lukovic, Y. Kaneko, B. Bradley, J. Wartman, R. W. Jibson, D. N. Petley, N. Horspool, I. Hamling, J. Carey, S. Cox, J. Davidson, S. Dellow, J. W. Godt, C. Holden, K. Jones, A. Kaiser, M. Little, B. Lyndsell, S. McColl, R. Morgenstern, F. K. Rengers, D. Rhoades, B. Rosser, D. Strong, C. Singeisen, M. Villeneuve; Landslides Triggered by the 14 November 2016 MwMw 7.8 Kaikōura Earthquake, New Zealand. Bulletin of the Seismological Society of America. doi:  https://doi.org/10.1785/0120170305.2018.
• Bertrand, D. Modélisation du comportement mécanique d’une structure cellulaire soumise à une sollicitation dynamique localisée – Application aux structures de protection contre les éboulements rocheux.2006.
• Heymann, A. Approche expérimentale du comportement mécanique des géos-ouvrages à technologie cellulaire. Application aux ouvrages pare-blocks. 2012.
• Musa, A. Evaluation of Concrete Barrier as Rockfall Protection. PhD dissertation, University of Akron, Ohio, 2015.
• Lambert, S., P. Gotteland and F. Nicot.Experimental study of the impact response of geocells as components of rockfall protection embankments.Natural Hazards and Earth System Sciences, European Geosciences Union, 9, pp. 459-467.https://hal.archives-ouvertes.fr/hal-00455337.2009.
• Lambert, S., A. Heymann, P. Gotteland and F. Nicot.Real-scale investigation of the kinematic response of a rockfall protection embankment.Natural Hazards and Earth System Sciences, European Geosciences Union, 14, pp. 1269-1281.www.nat-hazards-earth-syst-sci.net/14/1269/214.2014.
• Bourrier, F., S. Lambert, A. Heymann, P. Gotteland and F. Nicot. How multi-scale approaches can benefit the design of cellular rockfall protection structures.Can. Geotech. J. 48 : 1803-1816, 2011.
• A. Volkwein, K. Schellenberg, V. Labiouse, F. Agliardi, F. Berger, et al. Rockfall characterisation and structural protection – a review. Natural Hazards and Earth System Sciences, European Geosciences Union, 2011, 11, p. 2617 – p. 2651.
• European Organisation for Technical Approvals.ETAG 027 – Guideline for European technical approval of falling rock protection kits. 2013.
• NCTIR. Detailed Design Report – Modular Rockfall Protection Wall.  Document number 100001-DD-GT-RP-0001.  Internal report.  2018.

This paper was originally presented at the 69^th Highway Geology Symposium in 2018 in Portland, Maine, USA and is re-published by permission. Since the paper was prepared, two installations have been completed by NCTIR.  Both the New Zealand Transport Agency and KiwiRail are looking to utilise this design in future works.

Acknowledgements

The authors would like to thank the following individuals/entities for their contributions to the design and physical testing of the modular rockfall protection wall, and for the additional work that contributed to its approval for use on the NCTIR project:

Stuart Finlan – New Zealand Transport Agency
Daniel Headifen – KiwiRail
Eric Ewe – Geofabrics New Zealand Ltd.
Michal Tutko – Eliot Sinclair
Clive Anderson – NCTIR
Greg Saul – NCTIR
Ryan Jones – NCTIR

Disclaimer

Statements and views presented in this paper are strictly those of the author(s), and do not necessarily reflect positions held by their affiliations, the Highway Geology Symposium (HGS), or others acknowledged above.  The mention of trade names for commercial products does not imply the approval or endorsement by HGS.