Introduction
As part of writing the Canyoning Technical Manual, I needed to find out more about flat bends. For my own peace of mind, I had to make sure what I was putting in the book was suitable.
We undertook some testing in 2015 including a control test and five flat (offset) bend variations.
In 2020, we redid all the above tests. We also included another 3 control tests and 7 more flat bend variations. We were interested to test the Figure-8 and joining smaller and larger diameter ropes.
Total number of tests = 80.
Control
- Figure-8 on a bight knot (9mm)
- Figure-8 on a bight knot (10mm)
- Figure-8 rethread bend (10-on-10mm)
- Figure-8 rethread bend (9-on-9mm)
Flat bend variations
- Overhand (9-on-9mm)
- Figure-8 (9-on-9mm)
- 1.5 overhand (9-on-9mm)
- Stacked overhand (9-on-9mm)
- Stacked overhand (9-on-10mm)
- Stacked overhand (10-on-9mm)
- Stacked overhand (10-on-10mm)
- Doubled overhand (9-on-9mm)
- Doubled overhand (9-on-10mm)
- Doubled overhand (10-on-9mm)
- Doubled overhand (10-on-10mm)
- Triple fisherman’s (9-on-9mm)
Context and purpose
In canyoning, as well as other disciplines, you often need to join two ropes for retrieval during a pull through descent.
For long pitches in aquatic canyons, the bend can take a single person load (≈100kgs) as we operate on one rope. For example, when you have a releasable system in place above the bend.
Unlike other sports, in canyoning, we often lower a bend over an edge on a single rope. The bend, therefore, needs to be robust enough to handle this.
There are several reasons for operating on a single rope in an aquatic canyon including instant rescue, setting the length of the rope and bleeding the rope for sharp edge management (see the post ‘Rig a big canyon pitch’ for more info HERE). These techniques manage risk appropriately for the challenging environment we encounter.
Reason for tying the bend in an offset way is there is a flat side that has less opportunity to snag or catch. The bend, as it is offset, will naturally rotate outwards from an edge as the rope is lowered or retrieved (see diagram below). While having an offset bend is a good idea, it is only one of several things you can do to assist an easy rope retrieval. Such as positioning of the anchor for reduced friction and keeping the ropes separated to avoid twists when retrieving.
The difference between canyoning and climbing: when you are undertaking a climber abseil, you are often on a double rope. Therefore, the bend is only going to see half a person (and equipment) load (≈50kgs) and fixed in position, i.e. is not going to be lowering over an edge while loaded.
What do we want to know?
Whatever flat bend we end up choosing, it’s going to be a compromise.
Our regular go-to bends for joining ropes load nicely in line to maximise performance, e.g., a figure-8 rethread. When we tie an offset bend, we are pulling it apart when we use it.
No one bend meets all the criteria perfectly. The question is, which compromises are you willing to make? These four things I thought were important in context with canyoning use.
- How hard is it to tie? Scale:
- easy to tie (using the Flat overhand bend as the baseline – EDK)
- a little harder to tie
- much harder to tie – the time it takes, and complexity/is it easy to get wrong
- Has an acceptable failure mode
- ideally breaks the rope
- doesn’t have major rolls at low loads (capsize events)
- doesn’t significantly roll, e.g., roll off the end
- Note: Major Roll = bend capsizes suddenly and decreases force 1kN or more at a time.
Minor Roll = bend capsizes slowly and decreases force less than 1kN at a time.
- It is low profile and compact (length of bend). Scale:
- compact (65mm or below)
- not as compact (over 65mm)
- Has sufficient breaking strength
- ideally 10kN+ for a 9mm canyoning rope
What about the flat (offset) overhand?
For many years, the flat (offset) overhand has been used. This bend is often known as the European Death Knot (EDK). The flat overhand is popular as it is easy to tie, easy to undo after normal loading, low profile, and offset, providing a flat snag-free surface when retrieving.
The main concern, however, with the flat overhand is the failure by rolling off the end. A combination of stiff ropes, poor dressing/setting, short tails could be contributing factors to possible failure in the field.
Materials used
Korda’s Dana 9
- Manufactured by Korda’s: http://www.sacidkordas.com/en/sport/barranquismo/dana-9.
- Type: Canyons rope semi-static B
- Diameter: 9 mm
- Weight: 54 g/m
- Breaking strength: 21.15 kN
- Static elongation: 3.8% (150kgs)
- Core proportion: 58.7%
- Sheath proportion: 41.3%
- Materials: Core/Sheath – Nylon/ Nylon
- Standards: EN 1891:1998, type B
Korda’s Dana 10
- Manufactured by Korda’s: http://www.sacidkordas.com/en/sport/barranquismo/dana-10.
- Type: Canyons rope semi-static A
- Diameter: 10 mm
- Weight: 63 g/m
- Breaking strength: 27.30 kN
- Static elongation: 3.8% (150kgs)
- Core proportion: 58%
- Sheath proportion: 42%
- Materials: Core/Sheath – Nylon/ Nylon
- Standards: EN 1891:1998, type A
Testing procedure
A repeatable test procedure was set up.
- Testing device: a 100kN vertical testbed at Aspiring Safety, Christchurch.
- Testing speed: 100mm/minute.
- Sample rate: 60Hz.
- Material: New rope used for the testing of diameter and type used in canyoning.
- Tested between: rope clamps (no knots) except for the figure-8 knot on a bight (control) which had a 12mm pin at one end.
- 20cm tails were used. These were measured and marked pre-test and measured post-test measured.
Note: this length of tail was chosen for testing purposes so that the testbed did not run out of travel. For the flat overhand variation you choose, I would recommend longer tails for field application (e.g. 40cm+). - All knots and bends had hand tension. All strands were pulled to the get bend tight and compact.
- Length of the bend was measured with hand tension pre-test.
- Five tests were undertaken on each variation.
Control testing
Figure-8 on a bight knot
(9mm, 10mm)
Tying
On one end of the rope figure-8 on a bight knot was tied. The loaded strand is on the top.
Testing
- Average max force 9mm (5 tests): 15.95kN (72%)
- Average max force 10mm (5 tests): 19.53kN (72%)
Figure-8 rethread bend
(9-on-9mm, 10-on-10mm)
Tying
Tie a figure-8 knot and rethread with the other rope, so the tails are in opposite directions. The loaded strand is on the top.
Testing
- Average max force 9-on-9mm (5 tests): 13.60 (61%), all tails 20cm start and finish.
- Average max force 10-on-10mm (5 tests): 16.37kN (60%), , all tails 20cm start and finish.
Flat bend variations testing
Overhand EDK
(9-on-9mm)
Tying
With both strands wrap a half rotation around and thread through the loop.
Testing and Analysis 9-on-9mm
- Ease of tying: Easy
- Length of bend: 45mm (measured with hand tensioned pre-test)
- Average max force (5 tests): 6.79kN (31%)
- Failure mode: Started to roll at 4.73kN. 11-13 major rolls (bend capsizes more than 1kN at a time) before rolling off the end
- Tail lengths (start 20cm): 0cm
- Recommendation: Not recommended for canyoning.
+ Compact.
– Max force is low well below 10kN, the bend rolls off the end, multiple major rolls.
Figure-8
(9-on-9mm)
Tying
With both strands wrapped a full rotation around and threaded through the loop.
Testing and Analysis 9-on-9mm
- Ease of tying: A little harder to tie (compared to the overhand bend-EDK)
- Length of bend: 65mm (not as compact)
- Average max force (5 tests): 12.79kN (58%)
- Failure mode: Started to roll at 3.17kN. 1-2 major rolls before breaking the rope
- Tail lengths (start 20cm): 14cm
- Recommendation: Not recommended for canyoning
+ Max force well above 10kN
– Major roll (capsize) started at a low load which then reduced the force to less than 1kN, a little harder to tie, not as compact, tails got sucked in.
Overhand 1.5
(9-on-9mm)
Tying
With both strands wrap around and thread through the loop. Take the inside strand and wrap/thread a second time.
Testing and Analysis 9-on-9mm
- Ease of tying: Much harder to tie
- Length of bend: 50mm – compact
- Average max force (5 tests): 10.66kN (48%)
- Failure mode: Started to roll at 5.00kN. 2-4 major rolls before breaking the rope.
- Tail lengths (start 20cm): 8-13.5cm
- Recommendation: Not recommended for canyoning.
+ Max force above 10kN, compact.
– Much harder to tie, tails sucked in, major rolls.
Stacked Overhand
(9-on-9mm, 9-on-10mm, 10-on-9mm and 10-on-10mm)
Tying
Tie one flat overhand and then tie a second on top wrapping in the same direction.
- On the 9-10mm variation, the first overhand had the smaller 9mm rope on the bottom (blue), closest to the ropes under tension.
- On the 10-9mm variation, the larger 10mm rope on the bottom (blue), closest to the ropes under tension.
Testing and Analysis – 9-on-9mm
- Ease of tying: a little harder to tie
- Length of bend: 80mm
- Average max force (5 tests) 9.37kN (42%)
- Failure mode: Started to roll at 5.38kN. Two major rolls before breaking the rope. The second bend did not roll. Broke rope inside the bend.
- Tail lengths (start 20cm): 18cm
- Recommendation: Recommended for canyoning.
+ The tails did not get sucked in. The second bend did not roll.
– A little harder to tie, the bend is not as compact. The max force is just below 10kN.
Testing and Analysis – 9-on-10mm
- Ease of tying: A little harder to tie
- Length of bend: 100mm
- Average max force (5 tests): 10.72kN
- Failure mode: Started to roll at 8.91kN. 1-4 major rolls before breaking the rope. The second bend did not roll. Broke rope on the 9mm inside bend.
- Tail lengths (start 20cm): 18cm
- Recommendation: Recommended for canyoning.
+ The tails did not get sucked in. The second bend did not roll, max force above 10kN
– A little harder to tie, the bend is not as compact.
Testing and Analysis – 10-on-9mm
- Ease of tying: A little harder to tie
- Length of bend: 100mm
- Average max force (5 tests): 10.64kN
- Failure mode: Started to roll at 6.25kN. 2-3 major rolls before breaking the rope. The second bend did not roll. Broke rope on the 9mm inside bend.
- Tail lengths (start 20cm): 18cm
- Recommendation: Recommended for canyoning.
+ The tails did not get sucked in, the second bend did not roll, max force above 10kN
– A little harder to tie, the bend is not as compact.
Testing and Analysis 10-on-10mm
- Ease of tying: A little harder to tie
- Length of bend: 100mm
- Average max force (5 tests): 11.37kN (42%)
- Failure mode: Started to roll at 6.94kN. 3-4 major rolls before breaking the rope. The second bend did not roll. Broke rope on the bottom 10mm inside bend.
- Tail lengths (start 20cm): 18cm
- Recommendation: Recommended for canyoning.
+ The tails did not get sucked in. The second bend did not roll, max force above 10kN.
– A little harder to tie, the bend is not as compact.
Doubled Overhand
(9-on-9mm, 9-on-10mm, 10-on-9mm and 10-on-10mm)
Tying
With both strands wrap two times around and thread through the middle as you would for a double fisherman’s. You end up with a bend resembling the double fisherman except with the tails coming out the same side.
- On the 9-10mm variation, the smaller 9mm rope on the bottom (blue), closest to the ropes under tension.
- On the 10-9mm variation, the larger 10mm rope on the bottom (blue), closest to the ropes under tension.
Testing and Analysis 9-on-9mm
- Ease of tying: A little harder to tie
- Length of bend: 55mm
- Average max force (5 tests): 11.38kN (51%)
- Failure mode: Started to roll at 4.71kN. 2 minor rolls (bend capsizes less than 1kN at a time), slowly rolling before breaking the rope inside bend.
- Tail lengths (start 20cm): 12cm
- Recommendation: Recommended for canyoning.
+ No major rolls, compact, max force above 10kN.
– A little harder to tie, slowly rolls, sucks in tails.
Testing and Analysis 9-on-10mm
- Ease of tying: A little harder to tie
- Length of bend: 60mm
- Average max force (5 tests): 12.87kN
- Failure mode: Started to roll at 4.21kN. 2-3 minor rolls (bend capsizes less than 1kN at a time), slowly rolling before breaking the rope inside bend.
- Tail lengths (start 20cm): 12cm
- Recommendation: Recommended for canyoning.
+ No major rolls, compact, max force above 10kN.
– A little harder to tie, slowly rolls, sucks in tails.
Testing and Analysis 10-on-9mm
- Ease of tying: A little harder to tie
- Length of bend: 60mm
- Average max force (5 tests): 10.99kN
- Failure mode: Started to roll at 6.69kN. 3-4 major rolls on 3 of the tests, slowly rolling before breaking the core of the rope at the end.
- Tail lengths (start 20cm): 1cm (3 tests), 12cm (2 tests)
- Recommendation: Not recommended for canyoning.
+ Compact, max force above 10kN.
– Nearly rolled off the end in 3 of the results, a little harder to tie, major rolls, sucks in tails, inconsistent results.
Testing and Analysis 10-on-10mm
- Ease of tying: A little harder to tie
- Length of bend: 60mm
- Average max force (5 tests): 14.59kN (53%)
- Failure mode: Started to roll at 2.89kN. 2-3 minor rolls, slowly rolling before breaking the rope at the end.
- Tail lengths (start 20cm): 10.5cm
- Recommendation: Recommended for canyoning.
+ No major rolls, compact, max force above 10kN.
– A little harder to tie, slowly rolls, sucks in tails.
Triple (T) Fisherman’s
(9-on-9mm)
Tying
Tie a double overhand knot in one rope. Thread the second rope through the centre another double overhand around the first rope. You end up with a bend resembling the double fisherman except with the tails coming out the same side. Choose the strand from the first to tie another double overhand around the second rope.
Testing and Analysis – 9-on-9mm
- Ease of tying: Much harder to tie
- Length of bend: 80mm
- Average max force (5 tests): 12.92kN (58%)
- Failure mode: Started to roll at 8.12kN. 1 major roll before breaking the rope.
- Tail lengths (start 20cm): 14-17.5cm
- Recommendation: Not recommended for canyoning.
+ Max force well above 10kN, tails not sucked in, 1 major roll.
– Much harder to tie, not as compact.
Testing Summary
Based on the criteria set down at the beginning of the post:
Green = Yes – Ideal, Orange = OK, Red = No – Not ideal
Conclusions
In canyoning, your choice of bend becomes critical as a person (plus equipment) may need to be lowered over an edge on a releasable system. Like anything we use in the field, we need to assess the risks in context, with the purpose in mind, so we manage appropriately.
As stated initially, there is a difference in how a climber uses the bend (double rope) to a canyoner (single rope releasable) as what loading may occur.
With the context and purpose in mind, either the Double Overhand or the Stacked Overhand bend are recommended for canyoning as they have the best all-around performance. The bends best meet the criteria set down at the beginning of the article.
For different sized ropes tested (9mm and 10mm):
- The doubled overhand must have the smaller rope on the bottom – closest to the ropes under tension.
- For a stacked overhand, it does not matter which way round you have the different sized ropes as the second overhand, in either case, does not move.
Whatever bend you choose to join your rope, make sure you:
- Tie with sufficient tails at least 40cm – approximately two hand spans
- Dress your bend so all strands are aligned.
- Set your bend pulling tight on opposing strands.
- Get someone else to double-check your bend before use.
- Check the dressing and setting of the bend between abseilers if you are loading the bend.
