Empirical Breaking Strengths of Single Prusiks in Four

Empirical Breaking Strengths of Single Prusiks in Four Diameters on 11 mm
Static Rope
Thomas Evans
Western Washington University, Geology Department, 516 High Street, Bellingham, WA 98225,
[email protected]
Prusik strengths are ‘well known’, but accessing the data from preexisting tests is difficult
because the data are not archived. This hinders new riggers from viewing the primary data needed to
develop informed decisions. Consequently, the community needs archived freely-accessible testing data.
Prusiks can be used for many functions (rope grab, progress capture, belay, etc.), however here
only one function was tested to limit variables. Testing focused on individual prusiks experiencing a slow
pull on static rope. This is equivalent to use as a progress capture device or rope grab in a haul system.
In this configuration, prusik strength is often assumed as twice the rated cord strength. This assumption is
reasonable because a prusik has two strands running from the hitch to an attachment. Consequently, the
null hypothesis is that prusiks should break at or above twice the rated cord strength.
New, unused, rope and cordage was generously donated by Pigeon Mountain Industries (PMI).
The two static ropes utilized were PMI 11 mm EZ Bend and PMI 11.4 mm Isostatic, while 5, 6, 7, and
8mm PMI accessory cord was used to tie prusiks. The rope was cut in to 89 cm (35 in) lengths, fused on
both ends, and one end tied with a bowline with a Yosemite tie off. Bowline loops were tied as short as
possible, yielding a rope length of ~30 cm (12 in) for each sample. The 5, 6, 7, and 8 mm cord was cut in
to 79 cm (31 in), 91 cm (36 in), 104 cm (41 in), and 114 cm (45 in) lengths respectively. Cordage ends
were fused then prusik loops were tied using a double fisherman knot.
Three-wrap prusik hitches were tied on rope lengths with the fisherman knot out of the prusik
hitch, and then tightened hard by hand. Rope lengths were attached to a fixed anchor with a ½ inch
diameter screw link, and the prusiks were connected to a hydraulic ram and a load cell with a ½ inch
diameter screw link. The assemblies were photographed and slowly pulled (12.5 cm/min) to failure.
The average breaking strengths of prusiks on 11 mm EZ Bend rope was 15.46 kN (N=24) for 8
mm cord, 12.36 kN (N= 27) for 7 mm cord, 9.61 kN (N=30), and 8.07 kN (N=43) for 5 mm cord. On 11.4
mm Isostatic rope the average breaking strengths of prusiks were 15.59 kN (N=20) for 8 mm cord, 13.57
kN (N=20) for 7 mm cord, 9.73 kN (N=20) for 6 mm cord, and 8.29 kN (N=19) for 5 mm cord. Standard
deviations for all populations were around ~0.5 kN, and ranges between 1.78 kN to 3.35 kN.
Prusiks failed in one of three ways. In 196 out of 203 (96.6%), the cordage broke where the cord
entered the prusik hitch right under the bridge. In all cases the strand that broke was the strand closest to
the prusik anchor. Freshly broken cord ends were warm or hot to the touch and some were partially
fused, suggesting strand compression was involved in failure. Six out of twenty-four (25%) 8 mm prusiks
tied on EZ Bend rope never broke, but the rope mantle failed by breaking and slipping down the core.
Mantle failure suggests that the 8 mm prusiks were near the rope mantle strength. One prusik, tied from 5
mm cord on EZ Bend rope, broke where the prusik bight connected to the screw link.
Prusik strengths appear to be the strength of one cord strand including the additional
manufacturers engineered safety factor. One strand of new 8 mm cord is rated to 13.4 kN, 7 mm cord to
10.7 kN, 6 mm cord to 7.5 kN, and 5 mm cord to 5.8 kN. Generally, prusiks broke ~2 kN stronger than the
rated cord strength, but far short of twice the rated cord strength. In all cases, the measured breaking
strength was greater than the single strand rated breaking strength, and the weakest point in the prusik
was the most constricted strand, not the knot. This has consequences for static system safety factor
calculations, because we should be using the single strand cordage strength in calculations not double
the cord strength. For 8 mm cord on 11 mm rope, some prusiks are nearly as strong as the mantle. So
when pulling an 8 mm prusik on an 11 mm rope, it is possible for the rope mantle to fail before the prusik.
The measurements reported here are only applicable to single prusiks pulling on a rope relatively
slowly. These data are not applicable to prusiks in use as a belay in a dynamic event (tandem triple wrap
prusik belay), but are applicable to prusiks used as a progress capture device or rope grab.
Thomas Evans began caving in 2005 and was instantaneously hooked on vertical caving. As he studied
single rope technique for cave access he realized rescue skills are an important part of a rigger’s
repertoire, so he started studying rescue material independently and attended trainings through NCRC,
BCCR, and ESAR. It was clear through exposure to different rigging instructors that many ideas were
propagated in trainings that often had an unknown origin. During graduate school he realized that the
skills necessary in the discovery of knew scientific information could and should be applied to technical
rescue subjects to improve the quality and quantity of information available to a rigger (both student and
teacher alike), and provide a reference that could be cited for why different techniques are used. After
stating these ideas at ITRS 2010, he began his own testing program to further the science of technical
rescue through targeted analytically robust research.
Empirical Breaking Strengths of Single Prusiks of Four Diameters on 11 mm Static Rope
Thomas Evans
Western Washington University, Geology Department, 516 High Street, Bellingham, WA 98225,
[email protected]
Introduction:
Breaking strength data from slow pull or dynamic loading of prusiks are readily available through
internet searches (Bavaresco 2002, Gommers 2012, Heald 2009, Kharns 2008, Kovach 2009, Lyon 2001,
Moyer 1998, 2000, Sheehan 2004), and as such, prusik behavior should be well known to most riggers.
However, accessing the raw or processed data from many of these tests is difficult, or impossible, and
when such data are available, they are published in grey literature. Grey literature is composed of papers
that could be removed at any time and are not permanently archived for availability to future generations
of riggers. In addition, this literature often has small enough sample sizes that results are anecdotes that
do not inform the reader of the variability in component breaking strengths. While such testing is
incredibly useful, the lack of a permanent record as well as the small sample sizes (anecdotes) means
ideas about prusik behavior are largely unsupported lore (“old riggers’ tales”).
A rigorous understanding of prusik strength and behavior is crucial for making decisions upon
which human lives literally hang in the balance. As such, research is needed that breaks prusiks in
sufficient quantities to observe the variability in breakage strength and behavior. This research must be
coupled with publication in a permanent archive so the community has access to the data to form their
own opinions. This necessity was recognized by attendees of the 2012 International Technical Rescue
Symposium (ITRS), and when surveyed, there was a request for additional prusik testing (Evans and
Stavens 2012b).
Prusiks can be used in a variety of functions (rope grab, progress capture device, belay, personal
climbing system, etc.). Only one function was tested in this study to limit the number of variables involved
in the investigation. Here the focus is the properties of individual prusiks experiencing a slow pull on a
static rope. This is equivalent to the use of a single prusik as a progress capture device or rope grab in a
haul system. In this configuration, the prusik is often assumed to have twice the rated cord strength. This
assumption seems reasonable because, when tied, a prusik has two strands running from the hitch to an
attachment point (an anchor, haul system, etc.). If the system behaves this way, the prediction is that
when prusiks are pulled to failure they will fail at near twice the cord breaking strength. For example, a
prusik made from 8 mm cord with a rated strength of 13.4 kN should have a breaking strength of ~26.8
kN, which is greater than the knotted strength of an 11 mm static rope (about 27 kN without a knot and
about 19 kN with a knot assuming ~30% strength loss). This prediction can be tested by rigging prusiks in
the manner of function, then pulling them to failure on a static rope. Consequently, the null hypothesis
tested here is that prusiks should break at or above twice the rated cord breaking strength when pulled on
a static rope.
Given the lack of archived data, detailed reporting of test conditions, and ambiguity in the strength
of prusiks in their manner of function, this study has the following goals:
1. Measure and calculate the average breaking strengths of prusiks of different diameters on 11 mm
static rope
2. Report the range of breaking strengths of prusiks of different diameters on 11 mm static rope
3. Report prusik breakage mechanism(s) during a slow pull scenario
4. Record observations of the range of behaviors during prusik breakage
5. Test the null hypothesis that prusiks are approximately twice as strong as the single stranded
cord breaking strength.
Accumulating and reporting these data are important because it provides the raw information for
static system safety factor (SSSF) analyses, and enables informed system engineering by showing the
variability in the average values used. Ultimately SSSF analysis should include average component
strengths, the variability in their breaking strengths, and contextual information based on the system
geometry. In addition, these data can be used to determine if our assumptions of how system
components operate are valid, thus determining if we are rigging based on faulty assumptions.
This empirical data set was gathered by slow pulling prusiks made of Pigeon Mountain Industries
(PMI) cordage in four diameters (5, 6, 7, and 8 mm) on 11 mm EZ Bend and 11.4 mm Isostatic rope.
Materials and Methods:
All new and unused rope and cordage was generously donated by Pigeon Mountain Industries.
PMI 11 mm EZ Bend and PMI 11.4 mm Isostatic were the two static ropes used and 5, 6, 7, and 8 mm
PMI accessory cord used to tie prusiks. The rope was cut in to 89 cm (35 inch) lengths, fused on both
ends, and one end tied with a bowline with a Yosemite tie off. Bowline loops were tied as short as
possible, yielding a rope length of ~30 cm (12 in) for each sample. The 5, 6, 7, and 8 mm cord was cut in
to 79 cm (31 in), 91 cm (36 in), 104 cm (41 in), and 114 cm (45 in) lengths respectively. Cordage ends
were fused then prusik loops tied using a double fisherman knot.
Numerous short lengths of rope and cord were donated. To eliminate the effects of which piece of
rope and cord a sample came from, each piece of rope and cord was assigned a number. An attempt was
made to evenly distribute prusiks from a given piece of cordage across the different pieces of rope.
However, because it was not possible to obtain an even distribution of prusiks from each piece of cord on
pieces of each rope, these data may be biased based on the pieces of rope or cordage used. To see a
list of the characteristics of each rope donated, see Table 1a, and for a list of the characteristics of each
length of cordage donated, see Table 1b. In Tables 1a and 1b, each length of rope or cordage is given a
number so when breaking strengths are reported, the piece of rope or cordage used can be correlated
with which original length it came from.
Three-wrap prusik hitches were tied on lengths of rope with the fisherman knot out of the prusik
hitch, and then tightened hard by hand. Lengths of rope were attached to a motionless fixed anchor with a
½ inch diameter screw link, and the prusiks were connected to a hydraulic ram and a load cell with a ½
inch diameter screw link (Figure 1). The assemblies were photographed and pulled to failure with a slow
pull (12.5 cm/min) using the hydraulic ram.
The results were divided into populations with data divided by rope type and diameter of prusik
used. For each combination of rope (EZ Bend or Isostatic) and cord diameter (5, 6, 7, or 8 mm), which is
eight different combinations, descriptive statistics were calculated including: average, standard deviation,
maximum, minimum, and range of breaking strengths in kilonewtons (kN) and pounds (lbs). Average
breakage strength results with 1 sigma error bars were plotted versus cordage diameter and cordage
cross sectional area, and a linear correlation coefficient calculated for each data set.
Results:
Note: Everyone is welcome to the data and encouraged to utilize it. Data tables are posted on the
ITRS web page in pdf format, and available in Excel format via e-mail request to the author.
Quantitative Results
Table 2 reports the breaking strength for each sample, and reports the average, standard
deviation, maximum, minimum, and range of breaking strengths in kilonewtons (kN) and pounds (lbs) for
each sample population. Table 2 is divided into sub-tables so that each rope and cordage combination
are viewed separately (for a total of 8 different combinations). The average breaking strength of 8 mm
prusiks on 11 mm EZ Bend was 15.46 kN (N=24) with a standard deviation of 0.74 kN, maximum of 17.52
kN, minimum of 14.17 kN, and a range of 3.35 kN. For 7 mm prusiks on 11 mm EZ Bend the average
breaking strength was 12.36 kN (N=27) with a standard deviation of 0.47 kN, maximum of 13.48 kN,
minimum of 11.42 kN, and a range of 2.06 kN. For 6 mm prusiks on 11 mm EZ Bend the average
breaking strength was 9.61 kN (N=30) with a standard deviation of 0.52 kN, maximum of 10.50 kN,
minimum of 8.63 kN, and a range of 1.89 kN. For 5 mm prusiks on 11 mm EZ Bend the average breaking
strength was 8.07 kN (N=43) with a standard deviation of 0.61 kN, maximum of 9.11 kN, minimum of 6.50
kN, and a range of 2.61 kN. Broadly the results were similar for prusiks on Isostatic rope. The average
breaking strength of 8 mm prusiks on 11.4 mm Isostatic rope was 15.59 kN (N=20) with a standard
deviation of 0.54 kN, maximum of 16.76 kN, minimum of 14.77 kN, and a range of 1.99 kN. For 7 mm
prusiks on 11.4 mm Isostatic the average breaking strength was 13.57 kN (N=20) with a standard
deviation of 0.69 kN, maximum of 14.79 kN, minimum of 12.45 kN, and a range of 2.34 kN. For 6 mm
prusiks on 11.4 mm Isostatic the average breaking strength was 9.73 kN (N=20) with a standard deviation
of 0.49 kN, maximum of 10.72 kN, minimum of 8.94 kN, and a range of 1.78 kN. For 5 mm prusiks on
11.4 mm Isostatic the average breaking strength was 8.29 kN (N=19) with a standard deviation of 0.72
kN, maximum of 9.26 kN, minimum of 6.89 kN, and a range of 2.37 kN.
Table 3 is a synopsis of the cordage cross sectional area, rated breaking strength, average,
standard deviation, maximum, minimum, and range of measured breaking strengths of all prusiks tied on
11 mm EZ Bend. Table 4 is a similar synopsis of the cordage cross sectional area, rated breaking
strength, average, standard deviation, maximum, minimum, and range of measured breaking strengths of
all prusiks tied on 11.4 mm Isostatic rope. Both tables are supplied as summaries to ease comparison of
prusik breaking strengths of different diameters on different rope types.
Generally prusiks broke ~2 kN stronger than the rated breaking strength of the cordage, but far
short of twice the rated cord breaking strength. In all cases the minimum breaking strength was greater
than the rated cord breaking strength. Otherwise, the standard deviations were small, as were the
breaking strength ranges, suggesting that new cordage behavior is remarkably consistent.
To visually depict the comparison between rated cord breaking strength and prusik strength,
Figure 2 plots cord diameter versus breaking strength, including rated strengths and data from both
Isostatic and EZ Bend ropes. Figure 3 plots the same information but transformed by plotting the cross
sectional area of cordage versus breaking strength. It is readily observed that prusiks on both the
Istostatic and EZ Bend rope fail above their rated strength (by ~2 kN), but significantly less than twice the
rated cord strength.
Observational Results
It is noticeable that prusiks were, on average, slightly stronger on the 11.4 mm Isostatic rope than
the 11 mm EZ Bend (Figures 2 and 3), which may suggest that rope diameter also influences prusik
breaking strength. However, while there is a visual difference in the prusik breaking strengths on 11.4 and
11 mm ropes, the difference is small and doubtfully practically significant. It is unclear at this time if this is
an important effect, or not, however, it does suggest that prusik breaking strength is not only a function of
the cordage diameter but of the rope diameter as well. This means there is no one breaking strength for a
given prusik diameter.
Prusiks failed in one of three ways. In nearly all prusik failures (196 out of 203, or 96.6%), the
cordage broke in the same place: where the cordage entered the prusik hitch right under the bridge
(Figure 4). In all cases, the strand that broke was the strand closest to the prusik anchor (the upper strand
in Figure 1). When observing the freshly broken cordage ends they were warm or hot to the touch and
some were partially fused, suggesting some form of compression and heating was involved in failure. It is
unclear if this strand broke because it did not have a knot in it (presumably because knots stretch under
load thus reducing tension in the knotted limb of a prusik), or if the strand closest to the prusik anchor
bears more of the load, or that portion of the hitch is compressed the most (thus causing failure). All
prusik hitches were tied the same way around the rope (with the knot bearing limb toward the rope
anchor), so the present data set cannot be used to answer that question. Six out of twenty-four (25%) 8
mm prusiks tied on EZ Bend rope never broke, but the mantle of the rope failed by breaking and slipping
down the core (Figure 5). Mantle failure suggests that the 8 mm prusiks were near the strength of partial
rope failure. This means that the 8 mm prusiks may be nearly strong enough to make the rope the weak
link in the system when rigged in this configuration. One prusik, tied from 5 mm diameter cord on EZ
Bend rope, broke where the prusik bight connected to the anchor at the screw link (Table 2D).
Discussion and Conclusions:
The measurements reported here are only applicable to single prusiks pulling on a rope relatively
slowly. In other words, these data are not applicable to prusiks in use as a belay in a dynamic event (e.g.,
tandem triple-wrap prusik belay), but are applicable to prusiks used as a component of a haul system or
as a progress capture device.
Prusik breaking strengths were higher than the rated strength of a single strand of cordage by
about 2 kN. However, the breaking strengths were not double the breaking strength of the cordage as has
been suggested by some. Rather, the breaking strengths appear to be the strength of one strand of cord
including the additional the manufacturer’s engineered safety factor. This means that for static system
safety factor (SSSF) calculations, instead of using double the cordage strength, as is common practice,
we must use the single strand cordage strength.
The vast majority of prusiks rigged in the configuration tested here break at a strand under the
prusik bridge. This location appears to be the most constricted portion of the hitch, suggesting this may
be the weakest point regardless of prusik diameter. Most notably, the knot in the prusik is not the weakest
point. The weakest point is the strand most compressed under the prusik bridge. This is a similar result to
those of Evans et al. (2012) which demonstrated that knots in the limbs of webbing anchors with four
strands were not the weakest point but webbing that was constricted or smashed was consistently the
weak point in the system.
For 8 mm cord on 11 mm rope it appears that some prusiks are nearly as strong as the mantle.
So when pulling an 8 mm prusik on an 11 mm rope, it is possible for the rope mantle to fail before the
prusik does. This means that additional prusik strength (e.g., by adding multiple prusiks) may not provide
additional system strength increase because the rope mantle will be the weakest link. Further
investigation is warranted to identify the factors involved in prusik/mantle interactions.
Prusik strength is proportional to cordage diameter and cross sectional area (as expected).
However, prusik strength increased slightly on ropes with a larger diameter, but the difference is probably
not statistically or practically significant. This indicates, however, that there is not a single reportable
prusik strength for a given prusik cord diameter. Ultimately this means that when calculating SSSF’s it is
necessary to factor in both variability and the specific conditions under which the prusik is rigged (e.g.
rope diameter).
A notable result is that although we expect prusiks to break at twice the rated cord strength, this
does not happen, which is a similar conclusion from recent webbing research (Evans and Stavens 2011,
Evans et al. 2012, Evans 2013). It appears that a paradigm shift is needed to encompass these new
results. When software (e.g., webbing or cordage) is doubled, but travels through a connector (e.g.,
carabiner or screw link) the strength is not doubled. This has significant ramifications for SSSF
calculations.
Generally we need more research in to how systems are operating, because recent testing has
shown that the results are not as expected. In addition to observational and small scale empirical studies
(Evans and Stavens, 2012a, Evans 2010), we need studies with larger sample sizes to partially constrain
and capture the variability in the systems and equipment. Within the confines of prusiks we need dynamic
testing of tandem triple-wrap prusik belays with large numbers of samples to constrain variability in
system behavior.
Acknowledgements:
This study would not have been possible without the generous donations from PMI, and for the
support of Loui McCurley who facilitated acquisition of rope and cordage on short notice. Dr. Mike Berry
(Montana State University) provided lab space, time on equipment, and the use of numerous pieces of
equipment (load cell, hydraulic ram, and data acquisition hardware and software). Dave Hutchinson
designed and built the testing rig used. Ladean McKittrick provided logistical and mechanical support, as
well as paying for the shipping for cordage. Sherrie McConaeghey purchased the screw links used in all
tests. Sarah Truebe provided valuable assistance in editing the final document, though any errors remain
solely my own.
Literature Cited:
Bavaresco, Paolo, 2002, Ropes and Friction Hitches used in Tree Climbing Operations,
http://www.paci.com.au/downloads_public/knots/14_Report_hitches_PBavaresco.pdf
Evans, Thomas, 2013, Empirical Observation of Anchor Failure Points in Old and Retired Webbing,
International Technical Rescue Symposium, Albuquerque, New Mexico, November 7-10, 2013, 11 pages
Evans, Thomas, Stavens, Aaron, McConaughey, Sherrie, 2012, Causal Mechanisms of Webbing Anchor
Interface Failure and Failure Modes, International Technical Rescue Symposium, Seattle, Washington,
November 1-3, 2012, 17 pages
Evans, Thomas, Stavens, Aaron, 2012a, Basic Research Methods for Technical Rescue Science,
International Technical Rescue Symposium, Seattle, Washington, November 1-3, 2012, 7 pages
Evans, Thomas, Stavens, Aaron, 2012b, Implementing Example Research Methods at ITRS 2012 and
Their Results, International Technical Rescue Symposium, Seattle, Washington, November 1-3, 2012, 14
pages
Evans, Thomas, Stavens, Aaron, 2011, Empirically Derived Breaking Strengths for Basket Hitches and
Wrap Three Pull Two Webbing Anchors, Proceedings of the International Technical Rescue Symposium,
Fort Collins, Colorado, November 3-6, 2011, 8 pages
Evans, Thomas, 2010, Science as Applied to Technical Rescue Research,
Proceedings of the
International Technical Rescue Symposium, Golden, Colorado, November 4-7, 2010, 7 pages
Gommers, Mark, 2012, Knots Study Guide – Knots used for life support, Professional Association of
Climbing Instructors, http://paci.com.au.downloads_public/knots/01_Knots.pdf
Heald, Dino, 2009, The Performance of Polypropylene Ropes During Static Applications Including
Tensioning and Hauling, http://www.pyb.co.uk/ropetesting
Kharns, Brandon, 2008, Nylon and Polyester Prusik Testing, Proceedings of the International Technical
Rescue Symposium, Albuquerque, New Mexico, November 6-10, 2008, 2 pages,
http://itrsonline.org/nylon-and-polyester-prusik-testing/
Kovach, Jim, 2009, Fall Factors: Do They Apply to Rope Rescue and Rope Access? Proceedings of the
International Technical Rescue Symposium, Pueblo, Colorado, November 5-8, 2009, 8 pages,
http://itrsonline.org/fall-factors-do-they-apply-to-rope-rescue-rope-access/
Lyon Equipmet Limited, 2001, Industrial rope access – Investigation into items of personal protective
equipment, Contract Research Report 364/2001, Health and Safety Executive,
http://www.paci.com.au.downloads_public/knots/09_Tests_Lyon_Friction-hitches.pdf
Moyer, Tom, 1998, SLCo SAR Pull-Test Session – November 1998,
http://user.xmission.com/~tmoyer/testing/pull_tests_11_98.html
Moyer, Tom, 2000, MRA Intermountain Recert Pull-Test Session – July, 2000,
http://user.xmission.com/~tmoyer/testing/pull_tests_7-00.html
Sheehan, Alan, 2004, Load Testing,
http://paci.com.au.downloads_public/knots/08_Tests_OberonSES_29July04.pdf
Figure 1: Pull test apparatus. To the left is a stationary anchor, and to the right is the hydraulic ram with a
load cell (8 mm prusik on 11 mm EZ Bend rope used as an example here).
17
EZ Bend
Average Breaking Strength (kN)
16
15
Isostatic
14
Rated Strength (One
Strand)
13
y = 2.5734x - 4.9302
R² = 0.9715
12
11
10
9
y = 2.4905x - 4.8153
R² = 0.9794
8
7
6
5
5
6
7
Accessory Cord Diameter (mm)
8
Figure 2: Cord diameter vs. breaking strength in kN for prusiks broken on EZ Bend and Isostatic rope.
17
EZ Bend
16
14
Breaking Strength (kN)
y = 0.0629x + 3.1972
R² = 0.9735
Isostatic
15
Rated Strength (One
Strand)
13
12
11
10
9
y = 0.0613x + 2.9984
R² = 0.9938
8
7
6
5
75
85
95
105
115
125
135
145
Accessory Cord Area
155
165
175
185
195
(mm2)
Figure 3: Cord cross section area vs. breaking strength in kN for prusiks broken on EZ Bend and
Isostatic.
This strand broke, right
under the bridge.
Prusik
Bridge
Figure 4: Breakage location for the vast majority of prusiks, the strand closest to the load under the
bridge (gently loaded 8 mm prusik on 11 mm EZ Bend rope depicted here).
205
Figure 5: Mantle failure. The mantle elongated, snapped, and slipped down the rope core (8 mm prusik
on 11 mm EZ Bend rope shown here).
Table 1A: Data available for all lengths of rope donated.
Primary 1o Carrier 2o Carrier Diameter Length
Color
(mm)
(m)
Color
Color
White
Black
11
5.0
White
Black
11
5.8
White
Black
11
6.8
White
Black
11
7.3
White
Black
11
7.8
White
Black
11
8.7
White
Black
11
9.2
White
Black
11
10.2
White
Black
11
10.5
Blue
Black
11
7.7
Red
Black
11
9.8
Orange
Black
11
8.7
Orange
Black
11
10.3
Orange
Black
11
12.0
White
Black
Green
11.4
11.7
White
Black
Green
11.4
6.7
Blue
Black
Green
11.4
15.2
Blue
Black
Green
11.4
10.0
Black
Green
White
11.4
29.9
Make
White EZ
White EZ
White EZ
White EZ
White EZ
White EZ
White EZ
White EZ
White EZ
Blue EZ
Red EZ
Orange EZ
Orange EZ
Orange EZ
Isostatic
Isostatic
Isostatic
Isostatic
Isostatic
Table 1B: Data available for all lengths of accessory cord donated.
Primary 1o Carrier 2o Carrier Diameter Length
Make
Color
(mm)
(m)
Color
Color
Blue
Yellow
Green
8
15.0
CC080WB100S
Blue
White
Green
8
45.0
CC080WC1005
Purple
Yellow
Blue
7
13.8
CC0704E100S
Purple
Yellow
Blue
7
12.2
CC0704E100S
Purple
Yellow
Blue
7
11.5
CC0704E100S
Purple
Yellow
Blue
7
15.1
CC0704E100S
Purple
Yellow
Blue
7
10.2
CC0704E100S
Black
6
15.9
CC0605B001S
Red
Black
6
14.6
CC0605S001S
Red
Black
6
12.3
CC0605F001S
Purple
Blue
6
13.6
CC0605E40
Yellow
Black
Purple
5
20.2
CC055QE100S
Black
Purple
Green
5
19.3
CC055QC100S
Black
Purple
Green
5
16.0
CC055QC100S
Notes
Rated Breakage
Strength (kN)
10F13061D
27
27F13021D
27
15D13071D
27
01E13061H
27
29E13061G
27
16D13071H
27
17F13071D
27
10G13131D
27
10G13131D
27
19F13061D
27
28F13071D
27
03G13061H
27
01H13061K
27
01H13061K
27
08C413121D
31
19J12121G
31
26C12121D
31
03J12121D
31
26I12121G
31
Notes
20E13181G
29E13041K
09E13041K
01E13041K
26D13041K
17F13041K
26F13331S
17E13311D
29C13331D
30E13321K
08D13321K
07C13321K
67C13321K
Rated Breakage
Strength (kN)
13.4
13.4
10.7
10.7
10.7
10.7
10.7
7.5
7.5
7.5
7.5
5.8
5.8
5.8
Rope
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Cord
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Table 3: Summary descriptive statistics of prusik strengths on 11 mm EZ Bend rope, all in kN.
Cord Diameter Cordage Area Rated Breaking Average Breaking
Standard
Maximum Minimum Range Sample
(mm)
(mm^2)
Strength (kN)
Strength (kN)
Deviation (kN)
(kN)
(kN)
(kN) Size (N)
8
201
13.4
15.46
0.74
17.52
14.17
3.35
24
7
154
10.7
12.36
0.47
13.48
11.42
2.06
27
6
113
7.5
9.61
0.52
10.52
8.63
1.89
30
5
79
5.8
8.07
0.61
9.11
6.50
2.61
43
Table 4: Summary descriptive statistics of prusik strengths on 11.4 mm Isostatic rope, all in kN.
Cord Diameter Cordage Area Rated Breaking Average Breaking
Standard
Maximum Minimum Range Sample
(mm)
(mm^2)
Strength (kN)
Strength (kN)
Deviation (kN)
(kN)
(kN)
(kN) Size (N)
8
201
13.4
15.59
0.54
16.76
14.77
1.99
20
7
154
10.7
13.57
0.69
14.79
12.45
2.34
20
6
113
7.5
9.73
0.49
10.72
8.94
1.78
20
5
79
5.8
8.29
0.72
9.26
6.89
2.37
19