Analysis and Design of Wire Mesh/ Cable
Slope Protection
Final Report
Report CA05- 0222
April 2005
Division of Research
& Innovation
STATE OF CALIFORNIA DEPARTMENT OF TRANSPORTATION
TECHNICAL REPORT DOCUMENTATION PAGE
TR0003 ( REV. 10/ 98)
1. REPORT NUMBER
CA05- 0222
2. GOVERNMENT ASSOCIATION NUMBER
3. RECIPIENT’S CATALOG NUMBER
4. TITLE AND SUBTITLE
ANALYSIS AND DESIGN OF WIRE MESH/ CABLE NET SLOPE
PROTECTION
5. REPORT DATE
April 2005
6. PERFORMING ORGANIZATION CODE
7. AUTHOR( S)
Balasingam Muhunthan, Shanzhi Shu, Navaratnarajah Sasiharan,
O. A. Hattamleh, Tom C. Badger, Steve M. Lowell, John D. Duffy
8. PERFORMING ORGANIZATION REPORT NO.
WA- RD 612.1
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Washington State Transportation Center ( TRAC)
University of Washington, Box 354802
University District Building; 1107 NE 45th Street, Suite 535
Seattle, Washington 98105- 4631
10. WORK UNIT NUMBER
11. CONTRACT OR GRANT NUMBER
SPR- 3( 077)
12. SPONSORING AGENCY AND ADDRESS
Washington State Department of Transportation
Transportation Building, MS 47372
Olympia, WA 98504- 7372
California Department of Transportation ( Study Partner)
Division of Research and Innovation, MS- 83
1227 O Street
Sacramento CA 95814
13. TYPE OF REPORT AND PERIOD COVERED
Final Report
14. SPONSORING AGENCY CODE
15. SUPPLEMENTAL NOTES
This research was funded through the Transportation Pooled Fund program with the following partners: AK,
AZ, CA, ID, NE, NH, NY, OR, PA, WA, WYDOT
16. ABSTRACT
Since the 1950s, heavy gage wire mesh has been used along North American highways to control rockfall on
actively eroding slopes. More robust fabrics, such as cable nets, have more recently been introduced to improve
the capacity of these rockfall protection systems. To date, however, the design of these systems has been based
primarily on empirical methods, engineering judgment, and experience. This report summarizes research that
characterized existing performance, tested critical system components, back- analyzed system failures,
evaluated typical loading conditions, and developed analytical models to refine engineering design of these
systems. Finally, guidelines were developed to support the design of these systems for a variety of loading
conditions. Specifically, the report provides design guidance on site suitability, characterizing external loads,
fabric selection, anchorage requirements, and
system detailing.
17. KEY WORDS
Rockfall, wire mesh, cable net, slope hazard mitigation,
snow load, anchor, interface friction
18. DISTRIBUTION STATEMENT
No restrictions. This document is available to the
public through the National Technical Information
Service, Springfield, VA 22161
19. SECURITY CLASSIFICATION ( of this report)
Unclassified
20. NUMBER OF PAGES
327
21. PRICE
Reproduction of completed page authorized
Final Research Report
ANALYSIS AND DESIGN OF
WIRE MESH/ CABLE NET SLOPE PROTECTION
Balasingam Muhunthan
Shanzhi Shu
Navaratnarajah Sasiharan
Omar A. Hattamleh
Department of Civil and Environmental Engineering
Washington State University
Pullman, Washington 99164- 2910
Thomas C. Badger and Steve M. Lowell
Washington State Department of Transportation
P. O. Box 47365
Olympia, Washington 98504- 7365
John D. Duffy
California Department of Transportation
50 Higuera Street
San Luis Obispo, California 93401
Prepared for
Washington State Transportation Commission
Department of Transportation
And in cooperation with
U. S. Department of Transportation
Federal Highway Administration
April 2005
TECHNICAL REPORT STANDARD TITLE PAGE
1. REPORT NO. 2. GOVERNMENT ACCESSION NO. 3. RECIPIENT'S CATALOG NO.
WA- RD 612.1
4. TITLE AND SUBTITLE 5. REPORT DATE
ANALYSIS AND DESIGN OF WIRE MESH/ CABLE NET April 2005
SLOPE PROTECTION 6. PERFORMING ORGANIZATION CODE
7. AUTHOR( S) 8. PERFORMING ORGANIZATION REPORT NO.
Balasingam Muhunthan, Shanzhi Shu, Navaratnarajah Sasiharan,
O. A. Hattamleh, Tom C. Badger, Steve M. Lowell, John D. Duffy
9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. WORK UNIT NO.
Washington State Transportation Center ( TRAC)
University of Washington, Box 354802 11. CONTRACT OR GRANT NO.
University District Building; 1107 NE 45th Street, Suite 535
Seattle, Washington 98105- 4631
12. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED
Research Office
Washington State Department of Transportation
Transportation Building, MS 47372
Final Research Report
Olympia, Washington 98504- 7372 14. SPONSORING AGENCY CODE
Kim Willoughby, Project Manager, 360- 705- 7978
15. SUPPLEMENTARY NOTES
This study was conducted in cooperation with the U. S. Department of Transportation, Federal Highway
Administration.
16. ABSTRACT
Since the 1950s, heavy gage wire mesh has been used along North American highways to control
rockfall on actively eroding slopes. More robust fabrics, such as cable nets, have more recently been
introduced to improve the capacity of these rockfall protection systems. To date, however, the design of
these systems has been based primarily on empirical methods, engineering judgment, and experience.
This report summarizes research that characterized existing performance, tested critical system
components, back- analyzed system failures, evaluated typical loading conditions, and developed analytical
models to refine engineering design of these systems. Finally, guidelines were developed to support the
design of these systems for a variety of loading conditions. Specifically, the report provides design
guidance on site suitability, characterizing external loads, fabric selection, anchorage requirements, and
system detailing.
17. KEY WORDS 18. DISTRIBUTION STATEMENT
Rockfall, wire mesh, cable net, slope hazard
mitigation, snow load, anchor, interface friction
No restrictions. This document is available to the
public through the National Technical Information
Service, Springfield, VA 22616
19. SECURITY CLASSIF. ( of this report) 20. SECURITY CLASSIF. ( of this page) 21. NO. OF PAGES 22. PRICE
None None
TABLE OF CONTENTS
CHAPTER 1 INTRODUCTION.......................................................................... 1
1.1 Problem Statement...................................................................................... 1
1.2 Background................................................................................................. 3
1.2.1 General Applications .................................................................... 3
1.2.2 System Elements............................................................................ 6
1.2.3 Loading Conditions ....................................................................... 7
1.3 Literature Review ....................................................................................... 8
1.4 Research Objectives ................................................................................... 9
CHAPTER 2 FIELD PERFORMANCE ............................................................. 11
2.1 Gaviota Pass, California ............................................................................. 12
2.1.1 Problem Description ..................................................................... 12
2.1.2 Installation Description.................................................................. 13
2.1.3 System Performance ..................................................................... 14
2.2 Malibu Highway, California....................................................................... 15
2.2.1 Problem Description ..................................................................... 15
2.2.2 Installation Description ................................................................. 16
2.2.3 System Performance...................................................................... 18
2.3 Rain Rocks, California ............................................................................... 18
2.3.1 Problem Description ..................................................................... 18
2.3.2 Installation Description.................................................................. 20
2.3.3 System Performance ..................................................................... 20
2.4 Franklin Falls, Washington......................................................................... 25
2.4.1 Problem Description ..................................................................... 25
2.4.2 Installation Description ................................................................. 26
2.4.3 System Performance ..................................................................... 27
2.5 West Snowshed, Washington ..................................................................... 28
2.5.1 Problem Description ..................................................................... 28
2.5.2 Installation Description ................................................................. 29
2.5.3 System Performance ..................................................................... 30
2.6 Tumwater Canyon, Washington ................................................................. 32
2.6.1 Problem Description ..................................................................... 32
2.6.2 Installation Description ................................................................. 32
2.6.3 System Performance ..................................................................... 34
2.7 State- of- Practice ......................................................................................... 35
2.7.1 Anchors.......................................................................................... 37
2.7.2 Support Cables............................................................................... 38
2.7.3 Fabric............................................................................................. 38
2.7.4 Load- Influencing Factors .............................................................. 40
CHAPTER 3 TESTING AND MONITORING DATA...................................... 42
3.1 Fabric Testing............................................................................................. 42
3.1.1 Objectives ...................................................................................... 42
3.1.2 Methodology.................................................................................. 44
iii
3.1.3 Test Specimens.............................................................................. 45
3.1.4 Results ........................................................................................... 49
3.2 Seam Testing for Double- Twisted Hexagonal Mesh ................................. 51
3.2.1 Objectives ...................................................................................... 51
3.2.2 Methodology.................................................................................. 52
3.2.3 Results ........................................................................................... 53
3.3 Tumwater Canyon Instrumentation ............................................................ 55
3.3.1 Objectives ...................................................................................... 55
3.3.2 Methodology.................................................................................. 55
3.3.3 Results ........................................................................................... 57
3.4 Anchor Testing ........................................................................................... 61
3.4.1 Objectives ...................................................................................... 61
3.4.2 Methodology.................................................................................. 62
3.4.3 Results ........................................................................................... 66
CHAPTER 4 SNOW LOADS............................................................................... 75
4.1 Snow Load on Avalanche Structures.......................................................... 75
4.2 Snow Load on Mesh Systems..................................................................... 78
4.2.1 Snow Load below Freezing ........................................................... 79
4.3.2 Snow Load above Freezing ........................................................... 80
4.3 Performance of Mesh Systems ................................................................... 82
4.3.1 Upper Tumwater Canyon Site # 1 .................................................. 84
4.3.2 Upper Tumwater Canyon Site # 2 .................................................. 86
4.3.3 Instrumented Tumwater Canyon Site............................................ 88
4.3.4 Daggett Pass, Nevada .................................................................... 89
4.3.5 Franklin Falls Site.......................................................................... 91
4.3.6 US 20 Rainy Pass Site ................................................................... 92
4.4 Summary of Performance Analysis............................................................ 95
CHAPTER 5 LOCAL STABILITY OF MESH SYSTEMS.............................. 97
5.1 Deformation and Strength Analysis ........................................................... 97
5.1.1 Free Sides ...................................................................................... 98
5.1.2 Fixed Sides .................................................................................... 102
5.2 Local Failure Analysis................................................................................ 103
5.2.1 Transverse Failure ......................................................................... 103
5.2.2 Seam Failure .................................................................................. 105
5.3 Puncture Failure.......................................................................................... 107
5.3.1 Field Test Data............................................................................... 107
5.3.2 Finite Element Analyses for BCMoT............................................ 109
5.3.3 Back- analyses of Impacts to Mesh Systems.................................. 110
5.3.4 Anticipated Performance from Impact Loads ............................... 113
CHAPTER 6 GLOBAL STABILITY OF MESH SYSTEMS........................... 114
6.1 Limit Equilibrium Model............................................................................ 114
6.1.1 Anchor Capacity ............................................................................ 116
6.1.2 Interface Friction ........................................................................... 116
6.1.3 Mesh Weight.................................................................................. 118
6.1.4 Debris Load ................................................................................... 118
iv
6.1.5 Parametric Study of Overall System Performance ........................ 119
6.2 Finite Element Analysis.............................................................................. 123
6.2.1 Overview ....................................................................................... 123
6.2.2 FE Model ....................................................................................... 125
6.2.3 Description of Field Tests.............................................................. 125
6.2.4 Verification of FE Analysis........................................................... 127
6.2.5 Parametric Studies ......................................................................... 131
6.3 Modeling Results........................................................................................ 136
6.3.1 Anchorage and Top Connection .................................................... 136
6.3.2 Support Cables............................................................................... 139
6.3.3 Verification of Interface Friction................................................... 142
6.3.4 Limiting Conditions on Global Stability ....................................... 143
CHAPTER 7 DESIGN GUIDELINES ................................................................ 148
7.1 Site Suitability and Characterization .......................................................... 151
7.1.1 Block/ Event Size ........................................................................... 152
7.1.2 Slope Conditions ........................................................................... 153
7.1.3 Interface Friction ........................................................................... 155
7.1.4 Debris Loads.................................................................................. 158
7.1.5 Impact Loads ................................................................................. 159
7.1.6 Snow Loads ................................................................................... 160
7.2 Design Methodology .................................................................................. 162
7.2.1 Fabric Selection ............................................................................. 162
7.2.2 Anchor Capacity and Spacing ....................................................... 165
7.3 Design Details and Specifications .............................................................. 168
7.3.1 Slope Coverage.............................................................................. 168
7.3.2 Anchors.......................................................................................... 170
7.3.3 Support Ropes................................................................................ 172
7.3.4 Fabric Seaming and Fastening....................................................... 174
7.4 Aesthetic Concerns and Mitigation ............................................................ 175
7.4.1 Limiting Coverage Area ................................................................ 176
7.4.2 Increasing Mesh Contact ............................................................... 177
7.4.3 Colorizing System Components .................................................... 179
7.5 Construction Considerations....................................................................... 179
7.6 Maintenance................................................................................................ 181
7.7 Future Work................................................................................................ 182
ACKNOWLEDGMENTS...................................................................................... 183
REFERENCES........................................................................................................ 185
APPENDICES
Appendix A Survey of State- of- Practice for DOT’s ........................................ A- 1
Appendix B Fabric Test Reports ...................................................................... B- 1
Appendix C Anchor Load Test Data ................................................................ C- 1
Appendix D Anchor Spacing/ Load Charts....................................................... D- 1
Appendix E Plan Sheets ................................................................................... E- 1
v
LIST OF FIGURES
Figure Page
1- 1 Schematic drawing shows basic elements of a drape mesh system........... 2
1- 2 Typical rock anchors include ( A) deformed steel threaded bar, and ( B) a
wire rope tendon ........................................................................................ 6
1- 3 Typical soil anchors include ( A) hollow core drillable- groutable bars and
( B) MANTA RAY 􀂓 ................................................................................... 6
2- 1 Gaviota Pass, CA, slope condition before mesh installation ..................... 13
2- 2 Malibu Highway, CA, slope condition after the slope was covered with
a cable net .................................................................................................. 16
2- 3 Rain Rocks, CA, slope condition after the mesh installation .................... 19
2- 4 Slope condition after slope instabilities destroyed the hexagonal wire
mesh installation at Pitkins Curve adjacent to Rain Rocks chute.............. 22
2- 5 Franklin Falls, WA, slope configuration for the western portion of the
1998 cable net installation ......................................................................... 25
2- 6A Failure in the eastern portion of the 1982 chainlink installation ............... 27
2- 6B Localized rupture by a 3- to 4- foot- diameter boulder that initiated from
the bouldery talus shown at the top of photograph .................................... 27
2- 7A Failure of the cable net system in the western portion of the site in July
1999............................................................................................................ 28
2- 7B Deformation of the ground where loads exceeded passive earth pressure,
causing 8 inches of anchor deflection........................................................ 28
2- 8 Looking west of the West Snowshe, WA, slope, active raveling near the
crest of the cut slope and a large percentage of boulders within the debris 29
2- 9 Note the large accumulation of debris at the juncture of the underlying
support ropes and the midslope anchor..................................................... 30
2- 10 Rupture of the vertical seam and fabric; the midslope horizontal support
cable was also damaged. .......................................................................... 31
2- 11 Tumwater Canyon, WA, slope condition before cable net installation;
note the rockfall- related damage to the concrete barrier ........................... 33
3- 1 Testing apparatus with TECCO 􀂓 mesh ..................................................... 44
3- 2 Setup for Mac Double Galv1 mesh with bolts extended above the loading
plates .......................................................................................................... 47
3- 3 Setup for ( A) Geobrugg square grid cable net and ( B) Geobrugg diagonal
grid cable net.............................................................................................. 48
3- 4 Setup for Maccaferri diagonal grid cable net utilizing modified test
apparatus .................................................................................................... 48
3- 5 Setup for Maccaferri diagonal grid cable net utilizing modified test
apparatus .................................................................................................... 50
3- 6 Tested seams included ( A) butted seam with 6- inch fastener spacing, ( B)
single- cell overlap with 3- inch fastener spacing, and ( C) two- cell overlap
with doubled fasteners on 3- inch spacing.................................................. 52
3- 7 Testing/ clamping apparatus and failed specimen of bulk material ( no
seam) .......................................................................................................... 53
vi
3- 8 Dimensions and configuration of the cable net system and layout of the
instrumentation .......................................................................................... 56
3- 9 Normalized values of load for vertical ( longitudinal) strain gauges,
temperature, and snow depth ..................................................................... 60
3- 10 Normalized values of load for horizontal ( transverse) strain gauges,
temperature, and snow depth ..................................................................... 61
3- 11 A boring log depicts the subsurface conditions in which the anchors
were founded.............................................................................................. 63
3- 12 Layout of anchors and the depths to which they were installed ................ 63
3- 13 The tested anchors included ( A) Manta Ray 􀂓 ; ( B) single- strand cable;
( C) deformed steel threaded bar; ( D) HI- TECH ( drillable- groutable);
and ( E) Geobrugg double- strand cable ...................................................... 64
3- 14 Setup for vertical loading........................................................................... 65
3- 15 Setup for horizontal loading....................................................................... 66
3- 16 Load versus displacement plot for all vertical anchor tests ....................... 67
3- 17 The data from Anchor 7 in Figure 3- 16 are plotted as a hyperbolic
relation ....................................................................................................... 69
3- 18 Ground cracking ( white painted lines) that developed around Anchor 17
( double cable anchor) during vertical loading ........................................... 70
3- 19 Load- displacement plot for horizontal test of a Manta Ray 􀂓 anchor (# 4). 72
3- 20 Load- displacement plot for horizontal test of a double- strand cable anchor
(# 16)........................................................................................................... 73
3- 21 Load- displacement plot for horizontal test of a single- strand cable anchor
(# 20)........................................................................................................... 73
4- 1 Stress components of snowpack upslope of an avalanche structure.......... 76
4- 2 A cross- section of snowpack upslope of an avalanche structure illustrates
the distribution of drag force and velocity along the slope........................ 77
4- 3 The force components of the mesh and snowpack when the snow is
bonded to both to the mesh and ground ..................................................... 79
4- 4 The force components of the mesh and snowpack when the snow is
bonded to the mesh but uncoupled from the ground.................................. 81
4- 5 The force components of the mesh and snowpack snow, including the
contribution of interface friction................................................................ 82
4- 6 Upper Tumwater Canyon Site # 1 mesh installation ( yellow line) on a
planar and relatively smooth slope segment upslope of the steep bedrock
exposures at roadway level ........................................................................ 84
4- 7 Cross- section of installation with slope configuration and area of snow
accumulation.............................................................................................. 85
4- 8 Upper slope area before 1997 installation and the area where snow
accumulates................................................................................................ 86
4- 9 Slope area in ( A) 1997 before and ( B) 2002 five years after mesh
installation and just after repair of the system ........................................... 87
4- 10 Upper Tumwater Canyon Site # 2, cross- section of installation shows
approximate slope configuration................................................................ 88
vii
4- 11 Daggett Pass, NV, eastern half of the damaged installation still partially
suspended on the slope, held in place mostly through interface friction
and the few remaining intact anchors ........................................................ 90
4- 12 Cross- section through failed cable net shows the portion of the installa-tion
that accumulated snow........................................................................ 91
4- 13 US 20 Rainy Pass Site slope conditions; snow accumulation occurs on
the upper portion of the slope ................................................................... 93
4- 14 Cross- section shows upper and lower slope segments .............................. 94
4- 15 View looking downslope; note numerous boulders protruding into the
mesh ........................................................................................................... 94
5- 1 Displacement of single hexagonal cell at centerline of an impact cone .... 98
5- 2 Deformation and movement of a single hexagonal cell............................. 99
5- 3 Pullout displacements versus pullout resistance assuming free sides........ 101
5- 4 Deformation of single hexagonal cell with two sides fixed....................... 102
5- 5 Applied load versus displacement of a cell assuming fixed sides ............. 103
5- 6 Forces on a circular membrane.................................................................. 104
5- 7 Assumed circular cross- sectional shape of the mesh and the conditions
that would result in seam failure............................................................... 106
5- 8 US 12 White Pass Site # 1, modified cable net installation, upslope source
of rockfall, and typical range of block sizes that impact the installation in
a sub- perpendicular orientation.................................................................. 111
5- 9 US 12 White Pass Site # 2, ( A) Configuration of slope and installation.
( B) Enlarged view of top of mesh, downslope of chute. ......................... 113
6- 1 Mesh installation and loads that act on the system .................................... 115
6- 2 Effects of debris accumulation ( H = height of debris) on anchor load as a
function of slope angle with an assumed interface friction angle ( 􀁇 ) of 36 􀁱
and debris accumulation angle ( 􀁉 cs) of 35 􀁱 ................................................ 121
6- 3 The effects of interface friction angle ( 􀄯 ) on anchor load for 30- m ( 100- ft)
slope heights for a range of debris heights on a 50 º slope ......................... 122
6- 4 The effects of interface friction angle on anchor load for a 30- m ( 100- ft)
high slope with no external load on the system ......................................... 122
6- 5 Finite element discretization shows loads ( P) and forces ( F) acting on
mesh ........................................................................................................... 124
6- 6 Test setup and locations of load cells ........................................................ 126
6- 7 The direction of loading of the TECCO 􀂓 mesh by the steel beam and the
load transfer to perimeters ( support ropes and anchors) of the mesh pane 126
6- 8 Stress- strain relationship for TECCO 􀂓 G65 mesh ................................... 128
6- 9 FE model setup shows locations of anchors and beam ............................. 129
6- 10 S tress contours for Test 1 ......................................................................... 130
6- 11 Logarithmic strain contours for Test 1 ...................................................... 130
6- 12 Anchor arrangements that were investigated for a mesh system of 50 ft
( 18 m) wide by 100 ft ( 30 m) long............................................................. 132
6- 13 Stress contours for arrangement 1.............................................................. 134
6- 14 Strain contours for arrangement 1.............................................................. 134
6- 15 Stress contours for arrangement 6.............................................................. 135
viii
6- 16 Strain contours for arrangement 6.............................................................. 135
6- 17 Anchor load v. spacing for double- twisted hexagonal wire mesh for a
vertical slope ( no interface friction) ranging in height from 50 to 300 ft
( 15 to 90 m)................................................................................................ 137
6- 18 Anchor load v. spacing for TECCO 􀂓 G65 mesh for a vertical slope ( no
interface friction) ranging in height from 50 to 300 ft ( 15 to 90 m).......... 138
6- 19 Anchor load v. spacing for cable nets for a vertical slope ( no interface
friction) ranging in height from 50 to 300 ft ( 15 to 90 m)......................... 138
6- 20 Mesh without support ropes....................................................................... 139
6- 21 Mesh with top horizontal support rope ...................................................... 140
6- 22 Mesh with top horizontal and vertical support ropes................................. 140
6- 23 Mesh with top and interior horizontal support ropes ................................. 140
6- 24 Mesh with both vertical and horizontal support ropes............................... 141
6- 25 The arrangement for verification of assigned friction includes a top
horizontal cable and the inclusion of an internal point of support to
simulate a protrusion on the slope ............................................................. 143
6- 26 Load distribution in the top horizontal support rope ................................. 146
7- 1 Recommended design approach for wire mesh/ cable net systems ............ 150
7- 2 Cross- sections show typical ( A) concave and ( B) convex slopes and the
areas of mesh contact, debris accumulation, and rockfall impacts ............ 154
7- 3 Ongoing erosion threatens a wire mesh system installed in the late 1980s
in the North Cascades of Washington....................................................... 155
7- 4 Rough slopes exhibit a high degree of surface roughness with planar,
uniform profiles ......................................................................................... 157
7- 5 Undulating slopes exhibit profiles with ( A) somewhat uniform particle
distribution with limited overall roughness, and ( B) numerous localized
protrusions.................................................................................................. 157
7- 6 These planar slopes exhibit little surface roughness or slope irregularity 158
7- 7 Coverage area depicted by stationing and slope length ............................. 169
7- 8 Testing setup of anchors in a sub- horizontal direction .............................. 171
7- 9 ( A) The mesh was carefully installed to closely conform to this moder-ately
inclined slope. ( B) On a steep to overhanging slope where mesh
conformance is generally more difficult to achieve, the mesh can become
more visually apparent............................................................................... 178
ix
LIST OF TABLES
Table Figure
2- 1 Current DOT practice for wire mesh/ cable net system components ......... 36
3- 1 Description of tested specimens................................................................. 46
3- 2 Results from tension testing by WMEL..................................................... 50
3- 3 Anchors and test type................................................................................. 64
3- 4 Theoretical ultimate vertical load of anchors............................................. 70
3- 5 Test results for horizontal anchor load....................................................... 74
6- 1 Test results for various loading configurations.......................................... 127
6- 2 Summary of reactions obtained from FE analysis of field tests ................ 129
6- 3 Stress and strain for six scenario arrangements of anchors ....................... 133
6- 4 Shear stress on three anchor diameters for six arrangements .................... 133
6- 5 Moduli and yield strengths of mesh types ................................................. 136
6- 6 Summary of von- Mises stresses for support rope arrangements ............... 141
6- 7 FE analysis results for interface friction .................................................... 143
6- 8 Mesh yield states as function of height for a vertical slope with no inter-face
friction ................................................................................................ 144
6- 9 Mesh yield states as a function of debris load ........................................... 145
6- 10A Maximum length for top horizontal support rope v. slope height for
double- twisted hexagonal and TECCO ® mesh ( no interface friction)....... 146
6- 10B Maximum length for top horizontal support rope v. slope height for cable
net backed with double- twisted hexagonal mesh ( no interface friction) ... 147
7- 1 Recommended fabric usage as a function of block size ............................ 164
7- 2 Recommended maximum anchor spacing as a function of slope height ... 166
7- 3A Recommended maximum length for top horizontal support rope v. slope
height for double- twisted hexagonal and TECCO ® mesh ......................... 173
7- 3B Recommended maximum length for top horizontal support rope v. slope
height for cable net backed with double- twisted hexagonal mesh ............ 173
x
CHAPTER 1
INTRODUCTION
1.1 PROBLEM STATEMENT
Since the 1950s, heavy gage wire mesh has been used along North American
highways to control rockfall on actively eroding slopes. Within the last 15 years, small
diameter wire rope ( cable) nets have been employed as a more robust alternative to wire
mesh. To date, these systems have been designed primarily by empirical methods,
engineering judgment, and experience. With the exception of the anchors and support
cables, the basic design of these systems is comparatively similar throughout the U. S. It
consists of a top horizontal cable suspended by regularly spaced anchors, typically a
perimeter or widely spaced grid of support cables, and chainlink or double- twisted
hexagonal wire mesh fabric laced to the support ropes ( Figure 1- 1). This basic design has
been used by the Washington State Department of Transportation ( WSDOT) since the
late 1950s, where its use was originally limited to slopes less than 75 feet ( 23 m) high.
However, even some of the earliest installations were successfully installed on slopes
over 150 feet ( 45 m) high. Now, both wire mesh and cable net slope protection systems
are routinely being installed on slopes far in excess of 75 feet. The basic design has been
modified to address a variety of slope and loading conditions, so that numerous design
variations now exist.
1
Figure 1- 1. Schematic drawing shows basic elements of a drape mesh system.
While these basic designs have not been supported by quantitative design
methodology, with one noted exception ( Sandwell, 1995), overall, these systems have
functioned very well. Recently, some consensus has developed among geotechnical
specialists and contractors that certain system elements may be over- designed or even
unnecessary. In addition, system failures under a variety of loading conditions have
occurred within the last several decades, indicating that certain design elements may in
fact be under- designed for their desired application. Although incomplete site
characterization and inappropriate applications have been factors for some system
failures, a general lack of understanding regarding load and energy transfer, as well as
system capacity, remains a fundamental design obstacle. Furthermore, little quantified
knowledge exists about two primary causes of system failures, debris accumulation and
snow loading, and practical design guidance is needed for these loading conditions.
2
Given the design unknowns and observed performance of wire mesh and cable net
slope protection systems, there is a substantial need to improve upon existing design
methodology. The larger goals of this research are to
􀁸 􀀃 develop a rational and broadly applicable design methodology
􀁸 􀀃 make appropriate design revisions
􀁸 􀀃 ensure optimal system performance
􀁸 􀀃 where possible, construct more economical systems.
1.2 BACKGROUND
The origin of and first transportation- related application of draped wire mesh for
mitigating rockfall hazards is uncertain. Within the last 50 years, it has achieved
widespread use in the transportation industry in the United States and Canada, due in
large part to its effectiveness in controlling raveling type rockfall and its relatively low
cost per unit area of treatment. Fortuitously, considerable benefit has accrued to the
current research because of the large number installations and wide variety of
applications that now exist in North America.
1.2.1 General Applications
A number of factors influence the effectiveness, and thus appropriateness, of
draped mesh systems to mitigate rockfall. These include
􀁸 􀀃 orientation, length, irregularity/ roughness of slope
􀁸 􀀃 source, size, and frequency of rockfall
􀁸 􀀃 trajectory of rockfall
􀁸 􀀃 external loads such as debris and snow
􀁸 􀀃 intrinsic design elements.
3
Draped mesh systems are commonly installed on slopes ranging from as flat as
35 􀁱 to overhanging; however, most systems have probably been installed on steep rock
slopes in the range of 60 􀁱 to 80 􀁱 . For North American highway applications, the
maximum slope heights where these systems have been installed and function with
minimal damage reach about 400 feet ( 120 m); more commonly, slope heights range
from 50 to 150 feet ( 15 to 45 m). Mesh systems have been successfully applied to very
uniform slopes and highly irregular slopes. The degree to which mesh contacts the slope
is infinitely variable, since slope orientation and roughness, fabric type, and installation
procedures influence it.
Draped mesh systems are most typically used to mitigate raveling type rockfall
that involves small volume slope failures (< 10 cubic yards) comprising small block sizes
[< 2 feet ( 0.6 m) in diameter for lighter weight wire mesh and < 4- 5 feet ( 1.2 – 1.5 m) for
heavier weight cable nets], where other containment measures ( e. g., ditches, rockfall
barriers) are not available or provided. Failures of much larger volume and block size
have also occurred without resultant damage to these systems. Slope raveling is a
common occurrence; hence, systems are often installed on oversteepened, coarse surficial
deposits ( e. g., colluvium, alluvium, residual soils) and highly fractured rock masses.
Many systems that are now several decades old have been installed on slopes with a high
frequency of rockfall, and they exhibit little to no damage from rockfall.
Two generalized design approaches of draped mesh systems have evolved:
secured and unsecured systems. The design philosophy of an unsecured system entails
only anchoring the system along the top, allowing rockfall to occur between the rock face
and the mesh, and controlling the trajectory into a containment area at the base of the
4
slope/ installation. In effect, it seeks to minimize the external loading caused by
accumulating debris. This design approach has further evolved in the last decade to
elevated/ suspended systems that contain rockfall originating upslope of the installation.
Use of unsecured systems is predicated on having a suitable containment area at the base
of the installation and accounting for the transient but possibly large impact load. To
date, there is not a widely used methodology for designing for these transient loads, and
current practice is dictated by experience. The North American transportation industry
most commonly employs this unsecured design approach. Such an approach results in
lower installation costs and simplified maintenance than secured systems.
Secured mesh systems incorporate anchors within the field of the mesh, often on a
patterned spacing, and attempt to either stabilize the slope face ( e. g., TECCO 􀂓 system) or
hold the debris between the mesh and the ground. This design approach is widely used in
underground workings and is commonly seen along highway and rail slopes in Central
Europe and Japan, where there is little tolerance for minor instability or containment area
for debris accumulation. These systems are typically more costly, and those not
appropriately designed may require frequent maintenance to minimize damaging debris
loading.
Because many installations are located in mountainous regions in North America,
many systems are exposed to snow loading. Recently, snow loading has caused several
partial and entire system failures in Washington and Nevada. It is evident that, to date,
the transfer of snow loads onto draped mesh systems has been poorly understood.
5
1.2.2 System Elements
Draped mesh systems consist of three primary elements: anchors, support cables,
and mesh. While these common elements are shared, system components and installation
details vary considerably.
Anchors can be grouped into those for intact rock or soil conditions. Rock
anchors most commonly consist of a solid core, deformed steel, continuously threaded
bar ( Figure 1- 2A) or, more recently, a wire rope tendon ( Figure 1- 2B) placed in a fully
grouted hole. Some of the more common soil anchors include deadman- type, wire rope
tendons, driven and/ or grouted steel bars, hollow core drillable- groutable bars ( Figure 1-
3A), and MANTA RAY 􀂓 ( Figure 1- 3B) anchors.
Figure 1- 2. Typical rock anchors include ( A) deformed steel threaded bar, and ( B) a wire rope
tendon.
Figure 1- 3. Typical soil anchors include ( A) hollow core drillable- groutable bars and ( B) MANTA
RAY 􀂓 .
6
A variety of support cable configurations are currently employed, varying from no
support cables, to only a top horizontal cable, to an interior grid of horizontal and vertical
cables.
Originally, chain link mesh was most commonly used for North American
installations. In the 1980s, double- twisted hexagonal wire mesh started replacing the use
of chain link mesh for slope protection systems, due in part to its higher strength
( Agostini et al., 1988). In the late 1980s, cable nets were first used in North America on
a 250- foot ( 75- m) high rock cut in the North Cascades of Washington State, after two
previously installed, hexagonal wire mesh systems failed because of frequent, high-energy
rockfall and severe ice loading. Around 2000, high tensile steel wire mesh
( TECCO 􀂓 ) was introduced in North America as another high- strength mesh alternative.
1.2.3 Loading Conditions
Load sources on draped mesh systems include the following:
􀁸 􀀃 self- weight
􀁸 􀀃 rockfall impact
􀁸 􀀃 debris accumulation beneath the mesh
􀁸 􀀃 snow/ ice accumulation on top of the mesh.
Self- weight is the summation of system component weights, which includes the
fabric, support ropes, lacing wire, and related appurtenances. Load transfer occurs in a
complex manner through the mesh into the anchors. Support of the system is achieved
through interface friction where the mesh is in contact with the slope and the anchors.
For unsecured systems, the primary design objective is debris containment rather
than slope stabilization. It is, therefore, anticipated that most unsecured systems will be
7
repeatedly exposed to transient impact loads. The orientation and degree of loading is
defined by the rockfall trajectory between the mesh and the slope. For typical
installations installed against a steep slope, this impact loading is directed obliquely or
near parallel to the mesh. The resultant load is then transferred in a complex manner
through the mesh to the anchors.
Debris accumulation between the mesh and the slope can occur on unsecured
systems and is a common cause of observed local and global system failures.
Accumulation most commonly occurs where horizontal seams or support ropes
inadvertently trap debris, or when the bottom of the mesh is pinned or buried, often by
snow. Less commonly, protrusions on detached rock and abrupt slope
convexities/ irregularities can cause debris accumulation. Because of the sizable weight
of earthen debris, its accumulation can rapidly impart damaging, unintended loads onto
the system. Vegetation that grows through the mesh could also be considered as a debris
load, particularly if the substrate is creeping or otherwise unstable.
Snow and ice accumulation is another source of loading in some geographic areas
and has been a source of several recent system failures in Washington and Nevada. It is
notable that the weight of a relatively thin snowpack of 1 to 2 feet on even a short length
of slope is very great. Both the system and the ground potentially carry this load. Yet the
degree to and manner in which load is transferred to the system have been largely
unknown.
1.3 LITERATURE REVIEW
While experience with these various designs has grown considerably, especially
in the last decade, only one known study ( available for citation) has attempted to
8
quantitatively evaluate the system components and overall performance of a mesh slope
protection system ( Sandwell, 1995). The Sandwell study, which was commissioned by
the British Columbia Ministry of Transportation ( BCMoT), used finite element modeling
to evaluate the structural strength of their double- twisted, hexagonal wire mesh system to
resist specified impact energies at several scenario locations. The modeling was used as
a basis to make structural refinements to BCMoT’s designs. Officine Maccaferri S. P. A.
published a technical manual in 1988 ( Agostini and others) for use in the design of
rockfall protection systems. The publication includes some information on mesh
properties, as well as general details and design guidelines for its products.
1.4 RESEARCH OBJECTIVES
The overall objective of this research was to develop design guidelines and
generalized plans and specifications for unsecured wire mesh and cable net slope
protection systems that can be applied by a geotechnical specialist to a broad range of
field conditions. Toward this end, the research
􀁸 􀀃 summarized the experiences of numerous designers, contractors and suppliers
specializing in rockfall control
􀁸 􀀃 evaluated contributing factors of numerous system failures
􀁸 􀀃 instrumented a large cable net system to evaluate snow loading
􀁸 􀀃 performed strength testing on various fabrics and seaming configurations for
hexagonal mesh
􀁸 􀀃 performed extensive structural and finite element analyses of these systems to
better understand the performance of these systems
􀁸 􀀃 compared the vertical and horizontal capacities of various anchors.
The specific goals of the project were as follows:
9
( 1) Develop methods to evaluate design loads from debris accumulation and snow, and
study the global stability of the slope protection system.
( 2) Develop methods to analyze the structural capacity of system components.
( 3) Develop methods to describe load transfer characteristics.
( 4) Establish the resistance contributions from interface friction and anchors.
( 5) Refine the available methods of snow loading analysis with instrumentation and
back- analysis of failed systems.
( 6) Develop methods to evaluate the local stability of mesh.
( 7) Develop an analytical method to assess the anticipated energies from impact loads
The results of the above objectives were used to develop design guidelines for
slope protection systems for a variety of field conditions.
10
CHAPTER 2
FIELD PERFORMANCE
Rockfall initiation and trajectory are difficult to predict and quantify. Geology
and climate are the principal causal mechanisms of rockfall, factors that include intact
condition of the rockmass, discontinuities within the rockmass, weathering susceptibility,
ground and surface water, freeze- thaw, root- wedging, and external stresses ( Smith and
Duffy, 1990; Hearn and Akkaraju, 1995; Hearn et al., 1995). Trajectory is a function of
slope and rock geometry, as well as the slope and rock material properties ( Ritchie, 1963;
Pfieffer, 1989). All of these factors that affect initiation and trajectory can be variable
within and between slopes. The performance of rockfall control measures such as wire
mesh/ cable net systems is largely dictated by proper characterization of these factors and
understanding of the function and limitations of the applied mitigation.
By examining the field performance of a select number of existing systems, this
chapter seeks to summarize experience gained from more than five decades of
application. This review had three principal objectives. The first was to characterize the
limitations of these systems. Hence, the examples represent the more extreme range of
loading conditions with regard to frequency, block/ event size, impact energy, and
external ( snow) loads. Second, the review sought to identify the salient features of the
design guidelines, installation, and performance. Last, these data were used to verify and
calibrate analytical methods and support design recommendations that are presented in
subsequent chapters.
The examples were selected from a number of sites visited by the investigators
with input from the Technical Advisory Committee ( TAC). For each site, the
11
presentation has been organized to provide a description of the rockfall hazard,
installation, and performance.
2.1 GAVIOTA PASS, CALIFORNIA ( HEXAGONAL WIRE MESH)
2.1.1 Problem Description
The Gaviota Pass site is located on California State Highway 101 between
mileposts ( MP) 46.80 and 47.90 in Santa Barbara County. Situated within the Santa
Ynez Mountains, the Gaviota Pass is a steeply incised canyon of the Gaviota Creek
drainage. The northbound and southbound lanes of Highway 101 are on opposite sides of
the Gaviota Creek at an approximate elevation 200 feet. Steep slopes bound the roadway
corridor, with the highest elevations above 3000 feet. Short duration, high intensity rains
are common during the winter months, sometimes resulting in 3 to 5 inches per event,
and high winds with gusts of up to 50 mph are common during the summer months. The
rockfall that affects the highway primarily develops in the lower portion of the canyon,
predominantly within the cut slopes ( Figure 2- 1). The cut slopes comprise Gaviota
Sandstone, Holocene colluvium, and Quaternary landslide deposits. The cut slopes were
designed with midslope benches and slope ratios from near vertical to 0.5H: 1V.
Rockfalls develop in the landslide and colluvial deposits as the result of differential
erosion. Within the sandstone, rockfalls develop as planar and wedge rock block failures.
In addition, the midslope benches gradually fail and fill with debris, creating rock-launching
ramps. Small- scale rockslides, debris flows, and debris avalanches occur
occasionally. Although some rockfall catchment area is available at the base of the slope,
rockfalls 1 to 2 feet in dimension have reached the roadway.
12
Figure 2- 1. Slope condition before mesh installation; yellow line delineates coverage area
2.1.2 Installation Description
A hexagonal wire mesh system was constructed in 1992 to mitigate the rockfall
hazard at two locations. One installation was approximately 200 feet wide and covered
about 130 feet of slope length. The other site was approximately 260 feet wide and
covered about 150 feet of slope length. The Gaviota drapery system was typical of
California drapery design at the time. Along the top of the slope, 6- foot- long rock
anchors were installed, set back from the top of the slope 6 to 9 feet. Anchor spacing was
20 feet or at significant changes in topography. The anchors consisted of 1- inch- diameter,
threaded steel bar placed in a 2 ½ - inch- diameter hole in bedrock and/ or colluvium. A ½ -
inch cable ( support cable) was laced through rings at each anchor secured by a nut and
two washers. The wire mesh was attached to the support cable by folding the wire mesh
13
over the cable and securing the fold with hog rings spaced every 12 inches. Once
attached, the mesh was draped down slope to road level. As the mesh was unrolled down
the slope, workers pressed down on the mesh to conform it to the slope face. Vertical
cables were used in an attempt to pull the wire mesh closer to the slope. The mesh was
connected to the cables with a ¼ - inch- diameter lacing cable. The wire panels overlapped
a minimum of 12 inches and were connected with hog rings at 12- inch spacing. A
horizontal cable was placed along the bottom of the mesh to dampen the curling effect of
the wire mesh. Before installation, vegetation was pruned to ground level.
Approximately 29,063 square feet of wire mesh were installed at both locations.
An estimated 70 percent of the wire mesh is in contact with the slope. The system
was designed to allow for rocks to pass down slope in a controlled manner to a catchment
area at grade. No anchors were installed in the field area of the net. The net terminates
approximately 3 feet above the base of the roadway. The basis for the 20- foot maximum
anchor spacing was the performance of similar systems and engineering judgment.
2.1.3 System Performance
The overall performance of the wire mesh has been excellent and much better
than expected. There have been no reported incidents of rockfall reaching the roadway,
and vegetation is growing beneath the system. In some areas, vegetation cover has
doubled from the pre- installation condition. After 12 years, the system is in good
condition, with only minor damage requiring minor maintenance approximately every 5
years. The hexagonal wire mesh drapery systems, normally designed for controlling
rocks smaller than 2 feet, exceeded expected levels by successfully controlling 4- to 7-
cubic- yard rock slides. Although the drapery design controlled the small rockslides,
14
when individual block sizes within the slide mass exceeded 2 feet, the wire mesh was
damaged and/ or stressed. Once the drapery was installed, the interaction between the
wire mesh and the rock surface increased to the extent that there is no visible load on the
anchors. The horizontal cable placed along the bottom of the drapery, to prevent the
bottom of the mesh from curling at the bottom, traps the rock from moving behind the
mesh into the collection ditch. Trapping of rock was evident in other locations.
Unfortunately, this led to increased stresses on the wire mesh and caused the mesh to
tear. It was also observed that this increased stresses on the seams, causing some seams
to split open. The vertical cables have not improved the ground contact of the mesh.
Ripped sections have been patched with new pieces of hexagonal wire mesh, and
split seams have been re- fastened. In one section, a landslide undermined the anchor
foundations, causing failure of the drapery system. The mesh and anchors were replaced,
but the anchors were placed an additional 50 feet upslope to prevent further undermining.
2.2 MALIBU HIGHWAY, CALIFORNIA ( CABLE NETS)
2.2.1 Problem Description
The Big Rock Mesa Bluff site is located on California State Route 1 between MP
42.7 and 42.9 in Los Angeles County. The highway is situated along the coastal bluffs at
the base of the Santa Monica Mountains at an approximate elevation of 40 feet. The
roadway passes between the coastal bluffs and the Pacific Ocean. The bluffs are part of
the “ Big Rock Mesa Landslide.” Short duration, high intensity rains are common during
the winter months, sometimes resulting in 5 to 8 inches per event. Rockfall activity is
most prevalent during heavy rainfall periods as the slide advances down slope, causing
15
small- scale slope instabilities such as rockslides, debris flows, and debris avalanches
( Figure 2- 2). The material consists of fractured sandstone overlain by Quaternary
landslide deposits ( angular fragments of sandstone, gravel, and silt). The slide is active
and creeps during wet periods. The cut slopes have slope ratios of from near vertical to
1H: 1V. Although some catchment area is available, rockfalls 3 to 6 feet in dimension,
and small debris flows and rockslides 25 cubic yards in size have reached the roadway.
Figure 2- 2. Slope condition after the slope was covered with a cable net.
2.2.2 Installation Description
A cable net system was constructed in 1998 to mitigate the rockfall hazard at this
site. In addition, a soldier pile wall with concrete lagging was installed to stop debris
flows. The drapery installation is 400 feet wide and covers about 230 feet of slope
length. The cable net was installed over portions of the slope that contained large
boulders and outcrops of fractured bedrock. The cable net was constructed of 5/ 16- inch
16
cable woven into an 8- inch grid pattern with pressed steel clips. Along the top of the
slope, 10- foot long rock anchors were installed, set back from the top of the slope 100
feet beyond the actively eroding brow of the cut. Anchor spacing was every 23 feet or at
significant changes in topography. The anchor design called for a 1- inch- diameter,
threaded steel bar, founded in a 3- inch- diameter hole in bedrock and/ or colluvium. A
7/ 8- inch cable ( support cable) was then connected to rings at 23- foot spacing. A cable
tag line connected each ring to each ground anchor. The cable tag lines were secured at
each anchor by a nut and two washers. Cable nets were attached to the support cable by
lacing a ½ - inch- diameter cable through the cable net and around the support cable. Once
attached, the nets were draped downslope to road level. As the nets were unrolled down
the slope, workers pressed down on the nets to conform it to the slope face. The cable net
panels were connected to each other with a ½ - inch cable laced through each mesh
opening. Overlap was minimized as much as possible. No horizontal cables, vertical
cables, or wire mesh backing were included in the design. Before installation, vegetation
was pruned to ground level. Approximately 100,000 square feet of cable nets were
installed.
An estimated 65 percent of the wire mesh is in contact with the slope. The system
was designed to allow for rocks to pass down slope in a controlled manner to a catchment
area at grade. No anchors were installed in the field area of the net. The net terminates
approximately 6 feet above the base of the roadway. The basis for the 25- foot maximum
anchor spacing was the performance of similar systems and engineering judgment.
17
2.2.3 System Performance
The performance of the cable net system in controlling rockfalls has been good.
There have been no reported incidents of rockfall reaching the traveled way. Because of
the irregular surface profile, segments of the cable nets are not in contact with the rock
surface. As mentioned previously, the interaction between the cable nets and the rock
surface is in most cases sufficient to ensure stability of the system. Still, as observed in
the field, the tag lines are slack, indicating no load on the anchors. Several other
observations were made that are notable. The contractor built the nets and did not use
standard manufactured nets. In addition, not all the cable was uniformly galvanized, and
in many cases the cable started to corrode within months of the installation. The
contractor also used different fasteners, some coated with zinc, some with no coating, and
some were stainless steel. This resulted in two problems: corrosion and fastener
tightness. In the coastal environment of salt fog and salt spray, many of the zinc- coated
and non- coated fasteners corroded within months of installation. Furthermore, many of
the fasteners were improperly connected or not connected. It is remarkable that in spite
of very poor workmanship, this system has performed satisfactorily. The cable net, as
expected, has retained rocks up to 5 ft. in dimension. In fact, these rocks have moved
only minimally and have essentially been contained in place.
2.3 RAIN ROCKS, CALIFORNIA ( HEXAGONAL WIRE MESH AND CABLE
NETS)
2.3.1 Problem Description
The Rain Rocks/ Pitkins Curve site is located on California State Route 1 between
MP 21.1 and 21.4 in Monterey County. Known as the Big Sur Coast Road, Highway 1
winds along the base of the Santa Lucia Mountains hundreds of feet above the Pacific
18
Ocean. Short duration high intensity rains are common during the winter months,
sometimes resulting in 4 to 6 inches per event. High winds with gusts of up to 40 mph
are common during the spring. Rockfall activity is most prevalent during heavy rainfall
periods, when the slopes are heated by the sun, and, to a lesser degree, during windy
periods. Rockfall that affects the highway originates from the steep ( 0.25H: 1V to
0.75H: 1V) northwest- facing cut and natural slopes ( Figure 2- 3). Slopes in this area
consist of meta- basalts ( greenstone), sheared schist and phyllite with hard blocks of
greenstone embedded in the matrix. Rockfalls develop as the result of differential
erosion and as planar and wedge failures. Very little catchment is available, and prior to
the placement of the drapery system, rockfalls and small rockslides frequently reached
the roadway.
Figure 2- 3. Slope condition after the mesh installation ( yellow line) at Rain Rocks site.
19
2.3.2 Installation Description
In 1998, a hexagonal wire mesh system was installed along the Rain Rocks
section. This installation was 900 feet wide and covered about 400 feet of slope length.
Along the top of the slope, 6- foot long rock anchors were installed, set back from the top
of the slope 20 to 30 feet. Anchor spacing was every 40 feet or at significant changes in
topography. The anchor design called for 3/ 4- inch- diameter cable anchors located in a 2
½ - inch- diameter hole founded in colluvium. Manta Ray 􀂓 anchors with a cable
attachment were used on this project. The ends of the anchors had cable loops to which
the tag line was connected with a cable loop. Cable tag lines connected the anchors to the
top horizontal support cable, which was located 6 feet behind the top of slope. The
support cable passed through cable loops in the ends of the tag line. The wire mesh was
attached to the support cable by folding the wire mesh over the cable and securing the
fold with high tensile steel ( Spenax) hog rings spaced every 6 inches. Once attached, the
mesh was draped down slope to road level. As the mesh was unrolled, workers pressed
down on the mesh to conform it to the slope face. The mesh panels were overlapped a
minimum of 12 inches and connected with Spenax rings at 12- inch spacing. Before
installation, vegetation was pruned to ground level. Approximately 313,000 square feet
of wire mesh were installed.
2.3.3 System Performance
The overall performance of the Rain Rocks hexagonal wire mesh has been
excellent. There have been no reported incidents of rockfall reaching the roadway, and
vegetation is growing beneath the system. After 6 years, the system is in good condition,
requiring minimal maintenance. In one case, a hole in the mesh was patched. This was
20
in an area where the mesh did not touch the ground, and a rock free fell into the mesh.
Damage was not significant. Other maintenance has entailed clearing debris from the
bottom of the mesh. The double twisted wire mesh, as expected, retains rocks up to 2
feet in dimension. In fact, these rocks have barely moved and have been essentially
contained in place by the mesh weight and strength. Once the drapery was installed, the
interaction between the wire mesh and the rock surface increased to the extent that there
is currently no load on the anchors, evidenced by the slack in the tag lines.
In 2000, following severe winter rains, the Rain Rocks project limits were
extended northward, by 150,000 square feet of slope area, in an area referred to as the
Pitkins Curve Landslide. Expected rockfall sizes and rock avalanche volumes exceeded
2 feet and 20 cubic yards, respectively; however, because of cost constraints and
emergency conditions, hexagonal wire mesh was installed. The installation was identical
to the adjacent system except that half the ground anchors were 1- inch steel bar and half
were ¾ - inch cable anchors. This system worked effectively through the first year in
controlling rockfalls (< 2 feet in diameter) and small rockslides (< 10 cubic yards). One-year
later, however, increased rockfall and rockslide activity developed throughout the
Pitkins Curve slide area. Relentlessly over a 3- month period, slope instability progressed
upslope to the ridgeline. Every day, rockfalls 1 to 10 feet in dimension and rockslides 50
to 100 cubic yards in volume occurred. Initially, small rockslides accumulated at the toe,
stressing the entire system. As debris accumulated, the load increased on the mesh,
causing elongation of the wire. The Spenax rings held the mesh together, but in time the
mesh began to tear apart. Under loading, the steel bar anchors bent and were
compromised. The cable anchors, however, were not affected until they became
21
undermined. In three months, more than 20,000 cubic yards of slide debris were
generated at this site. Eventually, this active instability destroyed the hexagonal wire
mesh system. The drapery has not been replaced; instead, the roadway has been shifted
away from the hillside, and a large rockfall catchment ditch has been constructed.
A small portion of the Pitkins Curve section extended into the original Rain
Rocks installation. A small ( 500- cubic- yard) rockslide tore down the hexagonal wire
mesh at the northern end, ripping the mesh from the infrastructure while the infrastructure
stayed intact ( Figure 2- 4). Subsequent instabilities eventually undermined the
infrastructure to failure. Again, the anchors and infrastructure were not damaged until
the anchors were undermined. Within the slide scarp, a rock chute developed from which
3- foot- diameter rockfalls were regularly affecting the roadway. To mitigate this problem,
the chute was covered with 7200 square feet of cable nets.
Figure 2- 4. Slope condition after slope instabilities destroyed the hexagonal wire mesh installation at
Pitkins Curve adjacent to Rain Rocks chute.
22
The cable net installation was 260 feet wide and covered 450 feet of slope length.
The cable net was constructed of 5/ 16- inch cable woven into an 8- inch grid pattern with
pressed steel clips. The cable was pvc coated, and the fasteners were stainless steel for
corrosion protection in the harsh coastal zone. Along the top of the slope, 6- foot long
rock anchors were installed, set back 100 to 200 feet beyond the actively eroding brow of
the cut. Anchors were spaced at 25 feet or at significant changes in topography. Backup
anchors were installed an additional 50 feet beyond the primary anchors and were used as
directionals for cable tag lines. A ¾ - inch horizontal support cable was then connected to
the anchors via cable tag lines of similar size. The anchor design and tag line
connections were identical to those at the Rain Rocks installation. The cable net was
underlain with pvc- coated, hexagonal wire mesh with 12- inch overlap and fastened with
Spenax rings on 12- inch spacing. The wire mesh was first placed on the slope in one
operation, and the cable nets were placed on the slope in a second operation. The drapery
system was attached to the support cable by lacing a ½ - inch- diameter cable through the
cable net and around the support cable. Vertical cables were placed from top to bottom at
each anchor location. The cable net and wire mesh panels were connected to each other
and the vertical cables with a ½ - inch cable laced through each mesh opening, and the
wire was connected with Spenax rings at 12- inch spacing. Overlap was minimized as
much as possible.
An estimated 75 percent of the wire mesh is in contact with the slope. The system
was designed to allow rocks to pass down slope in a controlled manner to a catchment
area at grade. No anchors were installed in the field area of the net. The net terminates
approximately 50 feet above the base of the roadway.
23
To date, the system is functioning well, but maintenance has been necessary.
Small rockslides, 5 cubic yards with rocks as large as 3 feet in dimension, have been
caught in the mesh where the wire mesh and cable net were not secured tightly together.
Once caught, the creeping load of the rock mass has caused local cable net fasteners to
slide apart and the wire to tear in tension. resulting in an opening in the mesh panel.
Repairs have consisted of patching the wire mesh with new wire mesh fastened in place
with hog rings and re- establishing the cable net grid with cable clips. Other areas of
concern have been at the boundaries of the cable net panels, overlapped sections of cable
nets and wire mesh, and gaps in the connection between the hexagonal wire mesh and the
cable nets. These are areas where rock debris is accumulating. This accumulation is
imparting a load on the system, causing damage to the wire, cables, and fasteners. In
contrast, rocks 1 to 4 feet in dimension and small rockslides 5 cubic yards in size
creeping down slope away from the seams are causing little damage. To reduce this
problem, areas likely to entrap rock debris should be eliminated. This could be improved
by eliminating overlaps of the wire mesh and the cable mesh. Furthermore, the wire
mesh and the cable nets should be tightly secured together. This was not successfully
accomplished on the slope with this method of placement. The two fabrics should be
connected together on the ground with fasteners on each side of the square of the cable
net and then placed on the slope. The vertical cables also could have been eliminated.
Interestingly, even with the rock accumulating in pockets in the mesh, no load is being
transferred to the anchors, evidenced by slack tag lines.
24
2.4 FRANKLIN FALLS, WASHINGTON ( CHAINLINK, HEXAGONAL MESH,
CABLE NETS)
2.4.1 Problem Description
The Franklin Falls site is located in the central Washington Cascades adjacent to
the eastbound lanes of Interstate 90 at MP 51.3 just west of the summit at Snoqualmie
Pass. At elevation 3000 feet, the average maximum snowpack at the pass is 8 feet. The
site consists of a 0.25H: 1V ( 76 􀁱 ) cut slope in volcanic bedrock with heights to about 70
feet. Bedrock is overlain by about 20 feet of well- graded bouldery glacial till, which is
mantled on the surface with 5 to 10 feet of cobble- boulder talus ( Figure 2- 5). Boulders 1
to 2 feet are typical, but the overburden deposits include boulders to 4 feet in dimension.
Figure 2- 5. Slope configuration for the western portion of the 1998 cable net installation ( yellow line).
Extreme snow loads during the winter of 1998/ 99 caused most of the cable net anchors located along
the top of the installation to fail.
The overburden deposits along the top of the cut are oversteepened between 1.25H: 1V to
1H: 1V ( 38 􀁱 to 45 􀁱 ). These deposits are an active source for both raveling type rockfall
25
and small- scale ( 10 to 20 cubic yards) rotational failures. The talus slope above, which
extends more than 200 feet upslope, is oriented around 38 􀁱 to 40 􀁱 . Snow avalanches
originate above the cut, and regular avalanche control is required during the winter to
mitigate for unstable snowpack.
2.4.2 Installation Description
The original installation was installed in 1982 and covered a slope area of about
800 feet in length. This early installation consisted of chainlink mesh fastened to a
diagonal grid of ¾ - inch- diameter wire rope cable. The support ropes had a roughly 50-
foot spacing and were anchored to the slope face, where they intersected with No. 8
deformed, continuously threaded bars. Chainlink fabric was overlapped about 12 inches
and seamed intermittently with light gauge steel hog rings.
Because of poor system performance and local instability of the eastern portion of
the cut, the system was replaced in 1998 with sections of both hexagonal wire mesh and
cable nets backed with chainlink fabric, and portions of the cut were regraded to increase
the ditch width. Drillable- groutable anchors ( Ischebeck Titan 30/ 11) were installed at a
25- foot spacing in the talus about 30 feet behind the cut slope; anchor lengths were about
6 to 8 feet.
During the winter of 1998- 1999, snow accumulation was nearly twice the annual
average. On the east end of the site, plowed snow covered the lower portion of the cable
nets, which inhibited passage of debris behind the system. Snow accumulation on the
system and localized debris accumulation behind the cable nets resulted in the failure of a
several- hundred- foot section. On the west end, heavy snow accumulation had a similar
effect, although the cable nets remained on the slope. In 1999, where the cable nets
26
remained on the slope, anchors were reinforced with a second Titan anchor placed
approximately 10 feet upslope of the first anchor. The remainder of the failed nets was
not replaced.
2.4.3 System Performance
By the mid- 1990s, the eastern portion of the initial chainlink installation was in
very poor condition ( figures 2- 6A and 2- 6B). Erosion had exposed many midslope and
top anchors. Debris accumulation behind the support ropes and large boulders had
caused extensive punctures and ruptures of the system.
A B
Figure 2- 6A. Failure in the eastern portion of the 1982 chainlink installation. Active erosion
undermined the top row of anchors. Pockets of debris accumulated along intermediate support
ropes. Figure 2- 6B shows localized rupture by a 3- to 4- foot- diameter boulder that initiated from the
bouldery talus shown at the top of photograph.
Despite the replacement of the chainlink system with the more robust hexagonal
wire mesh and cable net systems in 1998, heavy snows during the winter of 1998- 99
caused extensive damage to the eastern and western portions of the system. Nearly all
the anchors either failed in shear or pullout, or loads exceeded the passive pressure on the
27
anchors, causing severe ground deformation ( figures 2- 7A and 2- 7B). The ultimate
capacity of the anchors in shear was around 35,000 lbs. Slightly lower- than- design grout
strengths and difficulties with grouting in the cobbles and boulders often void of matrix
may have contributed to the anchors failing in pullout. Some of the anchors where
passive earth pressures were exceeded had been installed vertically, rather than normal to
slope, resulting in diminished capacity.
A B
8”
Figure 2- 7A. Failure of the cable net system in the western portion of the site in July 1999. The
yellow line is the approximate location of the top of the cable nets as installed in 1998. Figure 2- 7B
shows deformation of the ground where loads exceeded passive earth pressure, causing 8 inches of
anchor deflection.
2.5 WEST SNOWSHED, WASHINGTON ( HEXAGONAL WIRE MESH)
2.5.1 Problem Description
The West Snowshed site is located in the central Washington Cascades adjacent
to the westbound lanes of Interstate 90 at MP 58.0 just east of the summit of Snoqualmie
Pass. At elevation 2600 feet, the average maximum snowpack exceeds 4 feet. The
oversteepened, south- facing cut slope is 100 feet high; the lower portion of the slope is
oriented between 40 􀁱 to 42 􀁱 and steepens to around 50 􀁱 in the upper portion ( Figure 2- 8).
28
The cut exposes very coarse colluvial and glacial deposits with boulders 1 to 4 feet in
dimension comprising 30 to 50 percent of the deposits, and heavy seepage is prevalent in
the cut slope. The upper, oversteepened portion of the cut is the source of both raveling-type
rockfall and small scale (< 10 yds3), surficial slumping.
Figure 2- 8. Looking west of the West Snowshed slope, active raveling near the crest of the cut slope
and a large percentage of boulders within the debris.
2.5.2 Installation Description
The original installation was probably installed in the early 1980s and covered a
slope area of about 600 feet in length. This early installation consisted of chainlink mesh
fastened to a diagonal grid of ¾ - inch- diameter wire rope cable. The support ropes had a
roughly 50- foot spacing and were anchored to the slope face, where they intersected with
No. 8 deformed, continuously threaded bars. Chainlink fabric was overlapped about 12
inches and seamed intermittently with light gauge steel hog rings.
29
Because of excessive damage primarily from debris accumulation beneath the
mesh, the eastern half of the original system was replaced in 1998. The new system
utilized hexagonal wire mesh fastened to a 50- foot square grid of ¾ - inch- diameter
support ropes placed on top of the mesh, which was attached to anchors installed on a 50-
foot spacing along the top of the installation. The bottom of the mesh was folded
outward to minimize debris accumulation. The mesh was overlapped 12 to 24 inches and
fastened with high tensile steel hog rings at about a 12- to 24- inch spacing.
2.5.3 System Performance
After several decades of severe slope erosion and heavy snow loads, the eastern
half of the first installation was badly damaged. Many seams in the chainlink had split,
and localized rupture and puncture had occurred in numerous locations near the base of
the installation. Support ropes also trapped large quantities of rock debris, imparting
significant debris loading on the system ( Figure 2- 9).
support ropes
midslope anchors
Figure 2- 9. Note the large accumulation of debris at the juncture of the underlying support ropes and
the midslope anchor.
30
Significant damage has occurred to the section that was replaced in 1998 with
hexagonal wire mesh. While a different configuration of support ropes was placed on top
of the mesh to better pass debris, a large quantity of debris has still accumulated beneath
the lower half of the mesh ( Figure 2- 10).
Figure 2- 10. Rupture of the vertical seam and fabric; the midslope horizontal support cable was also
damaged. The volume of debris associated with mesh failure is estimated to be 10 cubic yards, with
block sizes of up to 3.5 feet in dimension.
The reasons for the continued debris accumulation are believed to include the
concavity and flattening of the slope; the ongoing erosion and voluminous quantity and
large size of the debris generated; and the extended duration of a snowpack on the lower
portion of the mesh, inhibiting the passage of debris. Elsewhere, numerous seams have
31
ruptured because of opening of the hog rings, most of which are located in the upper
portion of the installation.
2.6 TUMWATER CANYON, WASHINGTON ( CABLE NETS)
2.6.1 Problem Description
The Tumwater Canyon site is located on highway U. S. 2 between MP 97.0 and
97.1 on the eastern slope of the Washington Cascades. The highway is situated on the
east side and at the base of the 3000- foot deep, steep- walled canyon around elevation
1500 feet. During winters with higher than average snowfall, a 24- to 30- inch snowpack
develops in the lower portion of the canyon ( Rick Woods, WSDOT Maintenance
Supervisor; personal communication). The rockfall that affects the highway primarily
originates from a 200- foot- high, 1H: 1V, west- facing cut slope adjacent to the westbound
lane. The cut exposes coarse- grained colluvial ( and glacial?) deposits with boulders of
up to 6 to 8 feet in size, discontinuously mantling intermediate to mafic intrusive bedrock
that exhibits an adversely dipping planar structure ( Figure 2- 11). Ongoing erosion of the
exposed colluvial deposits, and small planar failures to a lesser extent, produce regular
rockfall that is evidenced by the damaged concrete barrier.
2.6.2 Installation Description
A cable net system was constructed in 1997 to mitigate the rockfall hazard at this
site. The installation was 180 feet wide and covered about 300 feet of slope length. The
original anchor design called for a 5/ 8- inch- diameter steel rod with a welded eyelet,
located in bedrock and a minimum of 50 feet beyond the actively eroding brow of the cut
with a maximum spacing of 20 feet. During initial placement of the cable net, a number
32
of the anchors sheared and/ or failed in tension. All the anchors were replaced with either
¾ - inch wire rope or 1- inch deformed steel bar anchors. The nets were laced with 5/ 16-
inch wire rope to the top horizontal wire rope and the ¾ - inch vertical ropes that were
spaced at 20- foot intervals; there are no intermediate or bottom horizontal support ropes.
The net panels consisted of a 12- inch square grid of 5/ 16- inch wire rope joined at
intersections with pressed steel clips. The cable net was overlain with 9- gage chainlink
fabric with no overlap and fastened with hog rings on a roughly 24- inch square spacing.
Figure 2- 11. Slope condition before cable net installation; note the rockfall- related damage to the
concrete barrier.
Given the moderate slope inclination and overall surface uniformity, an estimated
75 percent of the cable net is in contact with the slope. While the weight of the net
33
arrests much rockfall on this flatter slope orientation, the cable net system was designed
to pass debris, and therefore, no anchors were installed in the field area of the net. The
net terminates about 6 feet above the base of the ditch to facilitate ditch cleanout and
minimize damage to the cable net system. The design also considered snow loads, but
the calculated load that assumed full transfer to the anchors ( no interface friction) was
judged to be unrealistic and over- conservative. The basis of the 25- foot maximum
anchor spacing was the performance of existing systems exposed to similar loads and
engineering judgment.
2.6.3 System Performance
The overall performance of the cable net system has been excellent. There have
been no reported incidents of rockfall reaching the highway, and some low- growing
vegetation is growing beneath the system. After six years, the system is in good
condition, with only minor damage within the bottom 20 feet of the nets. On the south
end, recent failures originating about 20 feet upslope, involving about six 5- foot-diameter,
angular, discoid- shaped blocks and totaling about 10 to 15 cubic yards,
deformed the cable net, broke one wire rope, and punctured a 12- inch- diameter hole in
the chainlink. Some hog ring fasteners also burst, and the chainlink separated from the
cable nets. On the north end of the installation, about 50 cubic yards of debris have
accumulated behind the system 20 to 50 feet upslope of the ditch. Several large angular
blocks have been caught, resulting in minor damage to the nets. The damage has
consisted of slippage of the pressed clips over an area of about 10 square feet. One 1-
foot- long x 6- foot- wide x 4- foot- thick block and one 7- foot- long x 3- foot- wide x 3- foot-
34
thick block originating from 40 to 50 feet upslope slid out beneath the cable nets without
damaging the system.
Buildup of debris behind the cable nets and angular rocks with some rotational
component of motion appear to have caused the only damage to the system. Overall, this
damage is minor, and the nets are functioning as designed. During winter, snow
regularly accumulates and slides off, failing within the snowpack or along the snow-chainlink/
cable net interface with no apparent damage to the system.
2.7 STATE- OF- PRACTICE
In North America, state/ province departments of transportation ( DOTs) probably
represent the largest users of wire mesh/ cable net slope protection for rockfall control.
While there are some variations to design approach and detailing of these systems, a
state- of- practice has evolved, mostly within the last 10 years, among DOTs. Table 2- 1
summarizes the system components and detailing specified by DOTs, as well as general
performance. A synopsis of experience and performance from a variety of transportation
agencies is provided in Appendix A. Overall, the performance of systems that have been
installed in North America has been good to excellent.
35
36
Table 2- 1. Current DOT practice for wire mesh/ cable net system components
AGENCY
Rock
Anchor
Diameter
( in)
Rock
Anchor
Spacing
( ft)
Rock
Anchor
Depth
( ft)
Soil
Anchor
Diameter
( in)
Soil
Anchor
Spacing
( ft)
Soil
Anchor
Depth
( ft)
Soil Anchors
Drilled ( DR)
Driven ( DN)
Deadman ( DM)
Hand Dug ( HD)
Auxiliary
Anchor
Diameter
( IN)
Support Cables
Suspension ( S)
Horizontal ( H)
Vertical ( V)
Lacing Rope ( LR)
Seam Fasteners
Hog Rings ( HR)
Spenax ( S)
Tiger- Tite ( TT)
Lacing Wire ( LW)
Performance
Excellent ( E)
Good ( G)
Fair ( F)
Poor ( P)
Alaska
DOT & PF
¾ ” to 1” 5’ ¾ ” to 1” 5’ NA
3/ 8– ¾ ” around
panels; some mesh
anchored on
10’ x10’ pattern
HR, LW
F – chain link
E/ G – double twist
and cable net
British Columbia
Ministry of
Transportation
1 ¾ ”
# 14- 75ksi
threadbar1
12’ 6’
1 ¾ ”
# 14- 75ksi
threadbar
12’ 5’
HD - typically
12 to 24”
diameter
7/ 8” # 7-
75ksi
and 5/ 8”
cable
S = ¾ ”
H ( Bottom) = ¾ ”
LR = 1/ 4”
no internal cables
S, TT
E - infrequent tears to
mesh; easily repaired
California DOT 1” bar
¾ ” cable
WM: 50’ 2
CN: 25’ B
WM: 6’
CN: 6’
1” bar
¾ ” cable
WM: 50’ B
CN: 25’ B
WM: 6’
CN: 6’
DR, DN, HD same
WM: ½ ” ( S)
CN: ¾ ” ( S)
no internal cables
S, TT, LW or equal
to stronger than the
wire
E – wire mesh
G/ E – cable net
seams/ overlaps cause
problems
Idaho
Transportation
Department
¾ ” 6’ 5’ ¾ ” 6 5 DR, DN 5/ 8”
5/ 8” ( H)
no internal cables
HR G
North Carolina
DOT
1”
to 1- 1/ 4”
5’
up to 15’
for cable
nets
? ? ? ?
variable
for cable
nets
1- 1/ 4” to 1- 5/ 8” for
cable nets
HR, LW ?
New Hampshire
DOT
1” 25’ 5’ 1” 12’ 5’ DN, DR NA
S/ H/ V= ¾ ”
LR= 9 gauge wire
HT, TT or LW
G- infrequent tears to
mesh; problems w/
debris collecting on
bottom cable
Nevada DOT ¾ ” 40’ 6.5’ ¾ ” 40’ 6.5’ DR, DN NA all 3/ 8” HR G
New York State
DOT
¾ ” 50’ 6’ NA NA NA NA 1”
S/ H = ¾ ”
no internal cables
HR or LW E
Oregon DOT
¾ ” loop
eye rock
bolt
40’
( h< 75’)
20’
( h> 75’)
3’
Same as for
rock
Same as for
rock
Same as for
rock
Typically, hand
dug. Minimum
hole size –
3’ X12”
NA
S/ H/ V = 3/ 8”
6X19 wire rope
LR not used
HR, S or TT
E - infrequent tears to
mesh; easily repaired
Washington
State
DOT
¾ ” cable;
# 8– 60ksi
def. bar
50’ for
h< 75’
25’ for
h> 75’
3
6’
DM or
Contractor
design
NA
for h< 75’,
S/ H/ V= 5/ 8”
for h> 75’,
S/ H/ V = ¾ ”
LR = 9 gage wire
S, TT, or LW
G – problems w/
debris on horiz.
cables and snow
loads on anchors
Wyoming DOT
1” epoxy
coated,
thread bar
5.5’ 5’ ( min)
1” epoxy
coated,
threaded bar
5.5’ 5’ ( min)
Predominately
DR ( DM, DN
allowed)
NA
Top Support
Cable= 1/ 2”
no internal cables
S, LW
G – few problems w/
failures along slope
crest and snow on
bottom of mesh
1 Anchors are raised 3 feet above ground surface.
2 Anchors added at topographic changes.
3 Spacings do not consider snow loads.
Currently, double twist hexagonal mesh ( WM) and cable nets ( CN) are being used by all surveyed DOTs.
2.7.1 Anchors
In North America, anchors are usually located only along the top of the system,
and debris is allowed to pass beneath the mesh. Midslope anchors are employed on
occasion by DOTs to achieve greater mesh contact with the slope. Generally, this has
been done to visually blend the mesh with the slope or to reduce slope erosion.
Numerous documented failures associated with debris accumulation around midslope
anchors have limited this practice to date. Ruvolum, a new design methodology for
pattern- anchored systems developed in Switzerland, is seeing some implementation in
North America.
DOTs are using two general anchoring designs for top- anchored systems, a close
spacing and a wide spacing. The close spacing design specifies a range of 5 to 12 feet
( 1.5 to 3.5 m). The wide spacing uses 40 to 50 feet ( 12 to 15 m) for slope heights of less
than 75 to 100 feet ( 20 to 30 m) and/ or for wire mesh systems, and a spacing of 20 to 25
( 6 to 7.5 m) feet for higher slopes and/ or cable net systems. Typical anchors in rock
include either a ¾ - inch ( 19- mm) steel cable loop or a 1- inch, deformed steel threaded bar;
anchor depths generally range from 3 to 6 feet ( 1 to 2 m). There are more variations in
soil anchors. Anchors are typically driven, dug, or drilled and often consist of either a
cable loop or a threaded bar.
In nearly all cases, documented anchor failures have resulted from external loads
associated with snow and debris accumulation and high- energy impacts. A rare case of
anchor failure due to static load ( mesh weight) occurred at the Tumwater Canyon site in
Washington. During the placement of the 300- foot ( 90- m) ( slope length) cable net
system, numerous anchors failed in shear. The anchors consisted of ½ - inch- ( 12- mm)
37
diameter steel bars ( utility- type anchor) with a capacity of around 10,000 to 15,000 lbf
( 44 to 67 kN).
Minimum anchor setbacks of 10 to 15 feet ( 3 to 4.5 m) from the slope brow are
typically specified. However, numerous system deficiencies have been documented
where the anchor setback from the brow of the slope was insufficient, and erosion
undermined the anchors. This condition is most often observed in oversteepened
bouldery deposits near the top of a cut slope.
2.7.2 Support Cables
The use and dimensioning of support cables is varied among DOTs. Cable
diameters range from 3/ 8 to 15/ 8 inch ( 10 to 41 mm), with galvanized ¾ - inch ( 19- mm),
6x19 wire rope being most typically specified. The lengths of top horizontal support
cables are generally limited to between 50 and 150 feet ( 15 and 45 m). Internal support
cables, when used, are commonly observed to be slack and not carrying load, suggesting
that these cables are not adding to the system capacity. Additionally, recurring problems
with debris accumulation along internal and bottom horizontal support ropes are well
documented, particularly if the cable is located between the slope and fabric. For these
reasons, a number of DOT’s no longer use an internal grid of support cables but specify
only a top horizontal and sometimes a bottom cable to facilitate cleanout behind the
system.
2.7.3 Fabric
In the last 10 years, double- twisted, hexagonal wire mesh has mostly replaced the
use of chain link fabric for rockfall control, primarily because of its greater strength and
perceived resistance to unraveling if a wire is cut. The hexagonal wire mesh most
38
typically used in North America consists of a 3- inch ( 8- cm) by 4- inch ( 10- cm) sized
opening ( referred to as 8x10 type); 0.12- inch- ( 3- mm) diameter galvanized wire; or a
0.11- inch- ( 2.7- mm) diameter wire for pvc- coated fabric. Hexagonal mesh is most often
limited to slopes producing rockfall with block sizes of less than 2 feet in diameter,
although on near- vertical slopes, double twist fabric has performed well for block sizes of
3 to 4 feet in dimension. On flatter slopes ( 􀁼 45 􀁱 to 50 􀁱 ) that produce large quantities of
rockfall, 2- foot- diameter blocks have caused considerable damage, which has been
evidenced at the West Snowshed site in Washington.
Most DOTs specify a 12- inch ( 300- mm) overlap of the hexagonal wire mesh but
have switched from light gage steel hog rings for seaming to using high tensile steel
fasteners ( i. e., King Hughes and Spenax 􀂕 ) or interlocking fasteners ( i. e., Tiger- Tite 􀂕 ).
The specifications for the spacing of fasteners are varied, but they commonly range from
6 to 12 inches ( 150 to 300 mm). Most DOTs also allow for the use of lacing wire of
equal or greater gage thickness for seaming. Rupture of seams is a recurring problem,
particularly when light steel hog rings are used or the spacing of high tensile steel rings
exceeds 12 inches ( 300 mm). Debris accumulation has even proven problematic with
spacings of 6 inches ( 150 mm). Because of recurring problems with debris accumulation
along overlaps, California DOT now prohibits overlapped seams.
Cable nets are typically constructed of either ¼ - or 5/ 16- inch ( 6- to 8- mm) wire
rope woven in a 6-, 8-, or 12- inch ( 150-, 200-, and 300- mm) square grid. Pressed cross-clips
have been used exclusively to bind cable intersections, although new connections
are forthcoming in the North American market. Net panels are butted and laced with
similar- sized cable. Because of the larger opening sizes, cable nets are normally backed
39
with either a chain link or hexagonal wire mesh to prevent smaller- sized rockfall from
passing through the cable nets. The backing fabric is typically placed between the slope
and the cable net. Where optimized slope contact is desired, chain link has proved to be
somewhat better than hexagonal wire mesh because of its greater flexibility. The greater
strength of 0.12- inch- ( 3- mm) diameter hexagonal wire mesh than the 0.11- inch- ( 2.7-
mm) diameter chain link, however, suggests somewhat better puncture resistance with the
hexagonal mesh when restrained by an outer cable net. A 24- inch ( 0.6- m) spacing of
fasteners to connect the backing fabric to the cable nets is typically specified. California
DOT has experienced problems with the backing fabric creeping beneath the cable nets
and now requires fasteners on each side of the cable net cell. It also requires that the
backing fabric be attached to the cable nets before placement. Properly fabricated cable
nets have proved effective with block sizes of up to 4 to 5 feet ( 1.2 to 1.5 m) in diameter.
Recently, high tensile steel wire mesh ( TECCO 􀂓 ) has been introduced in North
America by Geobrugg as an alternative to cable nets. Maccaferri, one of the primary
suppliers of hexagonal wire mesh, has also recently introduced a cable- reinforced, double
twist mesh. To date, there is little documented experience or performance history with
the first product and none with the second in North America.
2.7.4 Load- Influencing Factors
External loads on wire mesh/ cable net systems are dominantly influenced by the
slope configuration. External loads include debris and snow accumulation, as well as
impacts from falling rock. Important elements of slope configuration include orientation,
uniformity, and roughness; these elements in turn control the degree of contact that the
system has with the slope.
40
Slopes steeper than about 70 􀁱 typically do not accumulate debris because the
majority of the mesh usually hangs free of the slope, and therefore, only a minor
component of mesh weight acts against the slope. Mesh systems on flatter slopes exert a
greater weight component and, hence, have a greater tendency to accumulate debris. In
addition, slopes flatter than about 60 􀁱 may also be at risk for accumulating snow, which
can exert large loads on installations. Increased mesh weight and the degree of slope
contact can also have the positive effect of arresting or reducing erosion and the
movement of debris behind the mesh.
Slope uniformity is also an important factor affecting mesh contact and the
accumulation of debris. More uniform slopes are more efficient at passing debris than
slopes with inflections. A common location where debris accumulates is at abrupt
convexities, for example, where a flatter slope of overburden materials lies above a steep
rock cut. The slope- normal force acting at such an inflection can be quite large,
inhibiting the passage of debris. Abrupt convexities may also be subjected to
concentrated impact loads; puncturing of mesh fabric has been observed in these
locations. In contrast, concavities are common locations where the mesh is not in contact
with the slope. In this configuration, sagging in the mesh may actually impart an upslope
force component along the bottom of the installation, acting to entrap debris.
Slope roughness similarly influences system load, but to a lesser degree than
slope orientation and uniformity. In areas where the mesh is only in limited contact with
the slope, roughness may also influence rockfall trajectory. Impact forces normal to the
mesh are more likely on rough slopes than on smooth slopes.
41
CHAPTER 3
TESTING AND MONITORING DATA
The analytical models and finite element simulations that are presented in
subsequent chapters needed to be verified before they could be used to develop design
guidelines. Therefore, selected laboratory testing, as well as field instrumentation and
full- scale testing, were carried out as part of the research study. These included fabric
testing, seam testing, instrumentation of a cable net installation to evaluate load
distribution under snow loads, and full- scale testing of anchors. This chapter summarizes
these testing and monitoring efforts used in the verification of the analytical and
numerical models and in the development of design guidelines. Testing and monitoring
data are included in the appendices of this report.
3.1 FABRIC TESTING
3.1.1 Objectives
Current practice in North America utilizes two types of fabric for draped rockfall
protection systems; these include double- twisted hexagonal wire mesh and woven cable
nets with pressed cross clips. Within the last several years, high tensile steel wire mesh
( TECCO 􀂓 ) has been introduced in North America. Each of these fabric types has
distinctly different weight, strength, and elongation properties. Unfortunately, very little
published data exist on such properties, and some of the published results have varied
significantly. Furthermore, while some manufacturers have independently tested their
products, hexagonal wire mesh is the only fabric type for which there is a widely
accepted, standardized test method in North America to evaluate these properties ( ASTM
42
A 975). Additionally, there is no widely accepted test method for comparing the
engineering properties of a variety of fabric types.
To complete the analytical and finite element modeling necessary for this
research, the elasticity modulus and strength of the fabric types commonly used in mesh
systems needed to be determined. Solely for this reason, this research undertook a testing
program to determine these properties. Known North American suppliers were contacted
and solicited to supply their products. Of those contacted, Geobrugg provided cable nets
and high tensile steel wire mesh ( TECCO 􀂓 ), and Maccaferri provided hexagonal wire
mesh and cable nets. As expected, the strength and modulus values that were obtained
from this testing varied from the values reported by the manufacturers in some cases.
This may have been due to differences in the types of tests, specimen size, and the
boundary and loading conditions. Because the actual distribution of stresses that act on
the overall mesh system determined from the finite element analysis were much less than
the ultimate strength values of most of the fabrics, the effects of such variation would
have minimal impact on the design estimates of overall systems. It must be emphasized
that the testing performed for this research was not intended to compare similar fabric
types from different manufacturers. However, testing the different types of fabric in the
same manner did allow for some rough comparisons of their engineering properties and
an opportunity to understand their observed range of field performance.
The testing was performed at the Wood Materials and Engineering Laboratory
( WMEL) at Washington State University in Pullman, Washington. Sixteen tension tests
were conducted. Eight were performed on five configurations/ types of wire mesh fabrics,
and eight were conducted on three types of cable nets. The summary test report by
43
Carradine ( 2004) is included in Appendix B. Subsequent to this testing by WMEL,
independent test results were provided by Maccaferri for the hexagonal mesh and by
Geobrugg for the TECCO 􀂓 mesh. These test results are referenced in section 3.1.4 and
are available from the manufacturers.
3.1.2. Methodology
The test fixture shown in Figure 3.1 was designed, fabricated, and bolted to the reaction
floor inside the WMEL’s structural testing facility. The test fixture was fabricated to
handle fabric specimens of up to about 3.5 ft ( 1 m) square. The intent of the test fixture
was to load the meshes in tension at the two edges perpendicular to the direction of
loading while restraining the edges parallel to the direction of loading from constricting
as loads were applied. While it is recognized that it is not possible to entirely replicate the
exact conditions in the field in a laboratory test program, the text fixture was designed to
best represent the boundary and loading conditions in the field. The test fixture was also
guided by similar fabric tests on TECCO 􀂓 mesh performed by Geobrugg ( LGA, 2003).
Figure 3.1. Testing apparatus with TECCO 􀂓 mesh.
44
Testing was conducted by following ASTM A 975 Standard Specification for
Double- Twisted Hexagonal Mesh Gabions and Revet Mattresses ( Metallic- Coated Steel
Wire or Metallic- Coated Steel Wire with Poly ( Vinyl Chloride) ( PVC) Coating) as a
general guideline. Loads were applied by utilizing a 100,000- lbf ( 445- kN) capacity
hydraulic actuator with a stroke of 10 inches ( 250 mm). It was controlled with an MTS
407 Controller, which received actuator displacement feedback from a string
potentiometer. Load data were obtained by placing a 100,000- lbf ( 44- kN) capacity load
cell in line with the loading apparatus. Linear variable differential transformers ( LVDTs)
and string potentiometers were used to monitor displacement of the loading head with
respect to the base of the test apparatus in order to accurately record the distance that the
meshes moved through the first 2 inches ( 50 mm) of displacement. These data were used
in determining the elastic modulus of the meshes. Load data and displacement data from
the string potentiometer and the two LVDTs were recorded by using LabVIEW version
6.1 software.
3.1.3 Test Specimens
Table 3- 1 provides information on the different types of specimens that were
tested. All specimens were tested as delivered, although it was necessary to bend the
untested portions of the Mac Double Galv1 and the Mac Double Coat so that they would
fit into the testing apparatus ( Figure 3- 2). Inadvertently, the meshes described as double
consisted of two layers of hexagonal mesh ( only one layer was intended to be tested),
both of which were fixed to the testing apparatus and contributed to the load carrying
capacity. Meshes described as having a coating were made from wires that were pvc-coated
in either gray or brown. Meshes described as Narrow consisted of 6 x 8- type
45
( cm) hexagonal mesh and were approximately 34.5 inches wide ( perpendicular to the
direction of loading), while the remaining hexagonal mesh specimens were 8 x 10- type
( cm) and ranged from 40.0 inches to 71.0 inches wide. Fabric specifications were
provided by the manufacturers and are included in the appendix of the test report.
Differences in the dimensions of the meshes made it necessary to attach the
specimens to the test apparatus slightly differently. As shown in Figure 3- 1, the
TECCO 􀂓 mesh specimens were pinned to the loading plates and side restraints through
holes machined in the steel components. All of the hexagonal mesh specimens were
pinned through holes in the side restraints but were attached to the loading plates with
bolts that were placed in the holes in the plates and that extended far enough above the
plates to capture the mesh, as shown in Figure 3- 2.
Table 3- 1. Description of tested specimens
Specimen Description
Geobrugg Twist 1 TECCO 􀂓 G65 mesh ( Geobrugg)
Geobrugg Twist 2 TECCO 􀂓 G65 mesh ( Geobrugg)
Geobrugg Twist 3 TECCO 􀂓 G65 mesh ( Geobrugg)
Mac Double Galv1 8x10 hexagonal wire mesh, galvanized ( Maccaferri)
Mac Double Coat 8x10 hexagonal wire mesh, gray pvc coating ( Maccaferri)
Mac Double Narrow 6x8 hexagonal wire mesh, galvanized ( Maccaferri)
Mac Coat1 8x10 hexagonal wire mesh, brown pvc coating ( Maccaferri)
Mac Coat2 8x10 hexagonal wire mesh, brown pvc coating ( Maccaferri)
Geobrugg Square 1 5/ 16”, 12” square grid, cable net ( Geobrugg)
Geobrugg Square 2 5/ 16”, 12” square grid, cable net ( Geobrugg)
Geobrugg Diagonal 1 5/ 16”, 12” diagonal grid, cable net ( Geobrugg)
Geobrugg Diagonal 2 5/ 16”, 12” diagonal grid, cable net ( Geobrugg)
Mac Cable 1 3/ 8”, 12” diagonal grid, cable net ( Maccaferri)
Mac Cable 2 3/ 8”, 12” diagonal grid, cable net ( Maccaferri)
Mac Cable 3 3/ 8”, 12” diagonal grid, cable net ( Maccaferri)
Mac Cable 4 3/ 8”, 12” diagonal grid, cable net ( Maccaferri)
46
Figure 3- 2. Setup for Mac Double Galv1 mesh with bolts extended above the loading plates.
Attachment of the loading plates was done far enough in from the ends of the
meshes that the wires would not unravel before failure. Longer segments of threaded rod
were required to attach the lateral restraints to the meshes described as Narrow. The
Geobrugg square grid cable nets were attached to the end plates and lateral restraints in
the same manner as the TECCO 􀂓 mesh specimens, except that segments of steel plate
were attached to the lateral restraints to maintain the distance between parallel cables, as
shown in Figure 3- 3A. The diagonally woven cable net manufactured by Geobrugg was
attached to each end plate with two bolts, and long threaded rod segments were utilized
on the lateral restraints to maintain the shape of the nets, as shown in Figure 3- 3 B.
All specimens were placed in the fixture so that slack could be taken out of the
specimens to ensure that they would undergo enough deformation to cause failure. In
general, very little load was applied to these specimens as they were installed in the test
fixture. Cable net specimens manufactured by Maccaferri were approximately 75 inches
long, parallel to the direction of loading, which made it necessary to remove the loading
47
plate closest to the actuator and some of the steel linkages so that the mesh could be
directly attached to the loading apparatus connected to the load cable, as shown in Figure
3- 4. Four lateral restraints were used on each side because of the shape of the Maccaferri
cable nets.
Figure 3- 3. Setup for ( A) Geobrugg square grid cable net and ( B) Geobrugg diagonal grid cable net.
Figure 3- 4. Setup for Maccaferri diagonal grid cable net utilizing modified test apparatus.
48
The specimens were secured in the test frame, and then the LVDTs or string
potentiometers were installed so that the maximum amount of displacement data could be
recorded before the instruments ran out of stroke on the plunger or extension of the
string. After installation of the displacement measuring devices, the data acquisition
program and the hydraulic actuator were started. Load was induced by the hydraulic
actuator, which ran at 0.25 inches ( 6 mm) per minute under displacement control. All
specimens were loaded to the full stroke of the actuator, with the exception of the cable
net, which failed before reaching the available stroke distance on the jack. Testing results
and descriptions of failure for the various specimens are presented in the following
section. Following each test, specimens were removed from the test apparatus, and the
regions of failure were documented.
3.1.4 Results
Figure 3- 5 shows an example of a typical load versus displacement curve from
which elastic modulus values were obtained. Note that the data from the initial portions
of the load versus displacement curves were neglected when elastic modulus values were
determined; this is because the initial data were erratic for most specimens because of
settling within the test fixture as loads were applied. Table 3- 2 summarizes the tension
testing results, which include the dimensions, ultimate load, yield strength, and elastic
modulus for each specimen. Note that the different specimen types failed in different
manners; the descriptions of the failures are provided in Carradine ( 2004). Subsequent
test reports have been provided by the manufacturers for two fabrics and are included in
Table 3- 2.
49
Geobrugg Twist 2
0
5000
10000
15000
20000
25000
30000
0 1 2 3 4 5
Axial Displacement ( in.)
Applied Load ( lbf)
Figure 3- 5. A load versus displacement graph for determining elastic modulus is shown for the
TECCO 􀂓 mesh.
Table 3- 2. Results from tension testing by WMEL
Specimen
Initial Mesh
Width/ Length
in ( cm)
Ultimate
Load
lbf ( kN)
Tensile
Strength
lbf/ ft ( kN/ m)
Elastic
Modulus
lbf/ in ( kN/ m)
Geobrugg Twist 1 40.0/ 40.0 ( 101/ 101) 25,500 ( 113) 7,650 ( 111) 9,880 ( 1730)
Geobrugg Twist 2 40.040.0 ( 101/ 101) 28,200 ( 125) 8,460 ( 123) 12,400 ( 2170)
Geobrugg Twist 3 40.0/ 40.0( 101/ 101) 27,800 ( 124) 8,340 ( 122) 11,600 ( 2030)
TECCO 􀂓 G65 3mm1 39.4/ 42.5 ( 100/ 108) 11,000 ( 160) 14,300 ( 2500)
Mac Double Galv1 45.0/ 42.0 ( 114/ 107) 13,000 ( 57.8) 3,470 ( 50.6) 3,820 ( 669)
Mac Double Coat 45.0/ 42.0 ( 114/ 107) 14,800 ( 65.8) 3,950 ( 57.5) 2,970 ( 520)
Mac Double Narrow 35.0/ 43.5 ( 89/ 108) 14,100 ( 62.7) 4,830 ( 70.5) 6,010 ( 1050)
Mac Coat1 35.0/ 42.5 ( 89/ 108) 8,700 ( 38.7) 2,980 ( 43.3) 3,070 ( 538)
Mac Coat2 35.0/ 42.5 ( 89/ 108) 7,040 ( 31.3) 2,410 ( 35.1) 1,670 ( 293)
8x10, pvc coated 2.7mm2 29.5/ 48.0 ( 75/ 122) 3,530 ( 51.5) 2,060 ( 360)
Geobrugg Square 1 40.0/ 40.0( 101/ 101) 21,400 ( 95.2) 6,110 ( 93.7) 15,900 ( 2780)
Geobrugg Square 2 40.0/ 40.0( 101/ 101) 22,300 ( 99.2) 6,370 ( 97.6) 19,900 ( 3490)
Geobrugg Diagonal 1 24.0/ 39.0 ( 61/ 99) 19,200 ( 85.4) 9,600 ( 140) 11,200 ( 1960)
Geobrugg Diagonal 2 24.0/ 39.0 ( 61/ 99) 18,800 ( 83.6) 9,400 ( 137) 12,000 ( 2100)
Mac Cable 1 31.0/ 75.5 ( 79/ 192) 33,800 ( 150) 13,100 ( 203) 11,800 ( 2070)
Mac Cable 2 31.0/ 75.5 ( 79/ 192) 32,300 ( 144) 12,500 ( 183) 12,600 ( 2210)
Mac Cable 3 31.0/ 75.5 ( 79/ 192) 35,200 ( 157) 13,600 ( 199) 14,900 ( 2610)
Mac Cable 4 31.0/ 75.5 ( 79/ 192) 33,000 ( 147) 12,800 ( 186) 9,540 ( 1670)
1 Subsequent test report for TECCO 􀂓 provided by Geobrugg from LGA Nuremburg ( dated 4/ 17/ 2003); the
modulus was not reported but calculated from the test data.
2 Subsequent test report for hexagonal mesh provided by Maccaferri from CTC- Geotek, Inc. of Denver, CO
( dated 5/ 16/ 2001) following ASTM A 975 test method.
50
3.2 SEAM TESTING FOR DOUBLE- TWISTED HEXAGONAL MESH
One of the more commonly observed failures of hexagonal mesh installations is
seam rupture. Mesh panels are probably most often seamed by rapidly fastening hog
rings with a pneumatic tool. Hog rings currently in use generally consist of two types: a
medium tensile strength, 9- gage ring, and a high tensile steel, 9- gage ring ( i. e, Spenax 􀂕
or King Hughes). Other available seaming alternatives include a hooked fastening ring
( Tiger- Tite 􀂕 ) and lacing wire/ cable. Current practice of a number of DOTs has
prohibited the use of the medium gage hog rings because of their known poor seaming
performance for the high loading conditions associated with these systems. Furthermore,
lacing wire is generally not used when other alternatives are allowable because of the
time- consuming fabrication of such seams.
Because of the observed frequency of seam failures and the unknown capacity of
the various seaming details, limited tensile strength testing was performed in support of
this research.
3.2.1 Objectives
A simple testing program was undertaken by WSDOT to determine the
performance limitations of the high tensile steel hog rings for seaming double- twisted
hexagonal mesh. Three seaming details using high tensile steel hog rings were also
tested ( Fig. 3- 6).
The primary objectives of the testing were to determine the tensile strength
differential between typical seaming details and intact hexagonal mesh, and to develop an
optimized design for high tensile steel fasteners.
51
Figure 3- 6. Tested seams included ( A) butted seam with 6- inch fastener spacing, ( B) single- cell
overlap with 3- inch fastener spacing, and ( C) two- cell overlap with doubled fasteners on 3- inch
spacing. The dark vertical lines represent the finished longitudinal edge of the fabric, and the
ellipses represent fasteners.
3.2.2 Methodology
The fabric specifications consisted of an 8 x 10- type, galvanized, hexagonal mesh
with an approximate mesh opening ( cell size) of 3.25 inches ( 83 mm) by 4.5 inches ( 114
mm). The wire diameter was approximately 0.12 inches ( 3 mm) with a minimum tensile
strength of 2985 lbs/ in2 ( 20.6 MPa).
To replicate field- loading conditions, tensile testing was oriented perpendicular to
the twist. The tensile tests were performed on a 600,000- lbf ( 2670- kN) steel tensile
testing machine. Given the constraints of the machine, fabric sample dimensions were
restricted to about 30 inches square, and the maximum extension during testing was
limited to 8 inches. A clamping apparatus was fabricated to test approximately a six- cell
width and to tension the cells oriented in the direction of the wires ( Figure 3- 7), which is
somewhat less than what is required in the ASTM A975 test procedure for gabions and
revet mattresses. To reduce edge effects and unraveling of the mesh during tensioning,
the fabric was cut and clamped to maintain a two- cell perimeter around the clamping
apparatus. Unlike the ASTM A975 test method for double- twisted hexagonal mesh, the
52
sides were not restrained. Consequently, some necking of the sample occurred during
tensioning.
Figure 3- 7. Testing/ clamping apparatus and failed specimen of bulk material ( no seam).
A series of bulk material ( no seam) samples were tested first to verify the testing
apparatus and methodology and to determine the tensile strength of the fabric for this test
setup. Tensioning was terminated when one wire broke. Samples of the three seam types
were then run. Tensioning was continued for either the full 8 inches ( 200 mm) of
machine travel or until no increase in load could be achieved.
3.2.3 Results
Four samples of the bulk material ( no seams) were tested. The spacing between
the clamping apparatus varied between three and four cells ( 9.75 to 13 inches / 250 to
330 mm). All wire breakage occurred at least one cell inside the connections points.
Yield strengths ranged from 1830 to 2070 lbf/ ft ( 26.6 and 40.0 kN/ m), resulting in an
average yield strength of 1990 lbf/ ft ( 29.0 kN/ m). These strengths exceeded the
53
minimum strength requirements perpendicular to twist of 1800 lbf/ ft ( 26.3 kN/ m)
specified in ASTM A975 test method.
Four seam tests were run on the three seams shown in Figure 3- 6. One test was
run on seam A, which effectively tested two high tensile steel hog rings spaced around 6
inches ( 150 mm). The seam failed at 850 lbf/ ft ( 12.4 kN/ m) by consecutively popping
each hog ring. Two tests were run on seam B, which effectively tested five hog rings
spaced around 3 inches ( 75 mm). The seams failed at between 930 and 1110 lbf/ ft ( 13.5
and 16.1 kN/ m) by either consecutively popping each hog ring or breaking wires
individually. One test was run on seam C, which effectively tested ten hog rings doubled
and spaced around 3 inches ( 75 mm). The seam failed at 950 lbf/ ft ( 13.8kN/ m) by
consecutively popping each hog ring.
Load transfer during tensioning occurs through individual wires of the mesh
resulting in point stress concentrations at the fasteners. Consequently, the testing was
more a demonstration of the strength of the fasteners than of a continuous seam. In this
sense, scale limited the effectiveness of the testing in determining seam strengths.
However, in field conditions, the load transfer would likely be similar, in that point stress
concentrations would occur at the fasteners. On the basis of this limited testing, it
appears that high tensile steel fasteners spaced between 3 and 6 inches ( 75 and 150 mm)
provide a seam that is only half as strong as the mesh. This result is in rough agreement
with the reported requirements of test method ASTM A975, which requires a seam
strength that is only 40 percent of the required longitudinal mesh strength. Furthermore,
lacing wire is reported to provide seam strength that is only 60 to 70 percent of the
longitudinal mesh strength ( G. Brunet, personal communication 2004). These results
54
demonstrate that seams are inherently the weakest areas of the mesh. Specifications
requiring seams to be as strong as the longitudinal mesh strength are not being achieved
with currently known seaming details, nor may they be practically achieved in
construction.
3.3 TUMWATER CANYON INSTRUMENTATION
3.3.1 Objectives
As detailed in Chapter 2, most of the anchor failures at Washington installations
were associated with snow loading. Snow load applied to a mesh system is a function of
depth and density of snowpack, but many factors influence its magnitude, such as
temperature and slope surface conditions. Unfortunately, very little information exists
about the mechanisms and magnitude of loads that are transmitted to the system.
Therefore, to understand the mechanism of snow load on mesh systems, a cable net
system was instrumented at the Tumwater Canyon site in Washington, which annually
develops snowpack. The specific objectives of the instrumentation were to determine
how snow load varied with snow depth, snowfall, and temperature and how load was
accommodated within the support cables and anchors.
3.3.2 Methodology
In early November 2001, strain gauges were installed by the WSDOT and were
continuously monitored from November 2001 through April 2002. The details of the
system and the location of strain gauges are as shown in Figure 3- 8. Twenty Phoenix
Geometrix vibrating wire strain gauges were installed at ten locations on the upper
portion of the installation. Strain gauges were installed in couples welded onto cable
55
1
2 4
3 5 6
7
8
9
10
5/ 8 􀁣 􀁣 5/ 8 􀁣 􀁣
310 􀁣
162 􀁣
5/ 8 􀁣 􀁣 5/ 8 􀁣 􀁣
Cable Anchor ( all are 3/ 4 􀁳 except four 5/ 8 􀁳 anchors)
Vibrating Wire Strain Gauges
Wire mesh ( intermediate)
Cable Netting ( lowest)
¾ ’’ Vertical Support Cable ( upper)
Figure 3- 8. Dimensions and configuration of the cable net system and layout of the instrumentation.
56
clamps on the wire ropes, with one gauge on the top of the cable and another on the
bottom. The values from the two gauges were averaged to decrease error due to
differential strain on the cable. The strain gauges were continuously monitored with a
multiplexer and a Campbell Scientific CR10x data logger. The instrumentation was
sampled twice daily, at noon, and at midnight.
The ¾ - inch ( 19- mm) cables had an elasticity modulus of 15 􀁵 106 lb/ in2 ( 106 MPa),
with a metallic cross- sectional area of 0.272 in2 ( 1.75 cm2).
The strain gauges were installed several years after the system was installed. The
cables were not slacked to install the strain gauges but were installed on cables that
already were sustaining the static load of the system. Consequently, the measured loads
reflected a change in load relative to the initial reading.
3.3.3 Results
Because of the variation of the topographic and ground conditions, the measured
loads and their trends at each location were not consistent. Furthermore, the times at
which the maximum loads were recorded by the different strain gauges were also
different. Therefore, appropriate averages of the readings were calculated to obtain an
overall trend of the load variation with temperature, snowfall, and snow depth.
Accordingly, the readings at locations 1, 3, 5, 6, and 8 were averaged to check the
variation of the loads on the vertical ropes ( Figure 3- 8). Similarly, the readings at
locations 2, 4, 7, and 9 were averaged to check the loads on the top horizontal ropes.
Figures 3- 9 and 3- 10 show the variation of the loads, temperature, snowfall, and
snow depth during the period of November 2001 to April 2002. The snowfall and snow
depth data were collected from the records at the Leavenworth 3S weather station,
57
located about 2 miles ( 3 km) east of the site. To compare the data, the loads,
temperature, and snowfall were normalized with respect to their individual maximum
values. Note that snowfall data were not available after February 2002. The maximum
and minimum temperatures recorded during this period were 51.3 􀁱 F ( 10.7 􀁱 C) and 18.8 􀁱 F
(- 7.34 􀁱 C), respectively, and the largest 24- hour snowfall and maximum accumulated
snow depth were 11 inches ( 279 mm) and 25 inches ( 635 mm), respectively.
It can be seen that the first snowfall occurred on November 28, 2001, and almost
all strain gauges recorded its accumulation on the mesh by an increase in load ( figures 3-
9 and 3- 10). The snow depth soon reached about 18 inches ( 457 mm) and fluctuated
around this value until early January 2002. Notice that during this period, there were at
least ten snowfall events, and the temperature was below 32 􀁱 F ( 0 􀁱 C) for most of the time.
However, also during this period, the snow load continued to increase because of the
snowfall, despite the approxima