Some edits herein...
[ ] Added: #1085 (double figure 8 loop) is proven to work in a load sharing anchor configuration - that is, one loop is attached to an anchor point and the second loop attached to the other anchor point. Tests in an Australian facility (near Sydney NSW by B Proctor and R Delaney) demonstrate that when one of the loops is suddenly cut, it
does not trigger catastrophic failure of the remaining knot structure. The loops in #1085 can be adjusted to achieve 'equalisation' - Note: Shortening one loop has a corresponding
opposite effect on the other loop - meaning the other loop lengthens.
[ ] #1072 is not a practicable solution for building a load sharing anchor
[ ] Here's an interesting one for thought: You can use 'Purcell prusiks' to build a load sharing anchor system. This provides an in-built shock absorber on each leg of the multi-point anchor system. I personally have used Purcells to build load sharing anchors many times over the past few years - it works well. Obviously, you need to have the Purcells pre-tied. Note: Sterling USA make 'bound' prusik loops - these are great tools for climbers (and rope rescue technicians).
[ ] The Bowline is likely the first type of knot used by climbers/mountaineers - you can see photos of Hillary using it on Everest. (Note: They modified it by using a various tactics such as a 'double strangle').
...
roo is correct - ABoK#1047 is the most widely used and relied upon tie-in method in the world of climbing.
The main point is that the rope must be tied directly into the front of the harness - there must be no intermediate connector involved (eg a carabiner).
roo is also correct in his position that there are no recorded deaths due to arbitrary/random failure of #1047. A point worth noting is that EN892 dynamic climbing ropes don't just randomly fail either - in fact, there is not one single instance in the history of rope making where an EN892 rope has failed due to a defect/manufacturing error.
So what causes the deaths then? Answer = human error
(as an exercise, google the so-called 'EDK' knot and also 'tie-in knots for climbing' - the web is full of information and disinformation about knots - hopefully I wont add to the 'disinformation' in my post
)
Now, here is where the waters get a bit murky:
ABoK #1047 is very secure and stable in all loading profiles encountered in a climbing situation (ie rope is tied directly into the harness - with no intermediate connector).
However, as ropes have become thinner and thinner over time and, as some climbers have gained weight (eg myself!), after a free-fall, ABoK #1047 can be somewhat challenging to untie. Now, with modern sport climbing techniques, multiple repeated free-falls are the norm - several sequential falls can be taken by a climber while 'working' a route - and the climber is generally not lowered back to the ground after each fall. He/she remains hanging in mid-air and then after a brief rest, gets back on the rock and tries again. In this type of scenario, ABoK #1047 can be a real PITA (pain-in-the-arse) to untie - particularly when your fingers and hands in general are 'pumped' and you are maximally fatigued.
This has been the catalyst to find a 'better' method of tying the rope into the harness.
Enter the Bowline (#1010).
Experienced/knowledgeable/skilled knot practitioners will all tell you never to trust your life to #1010.
Why?
Because it is not stable and it is not secure under cyclic loading profiles (tension-on, tension-off, tension-on and then tension-off again). Cyclic loading leads to gradual loosening of ABoK #1010. And when that happens, its the old 'express elevator to hell' trick - also known as 'decking', 'hitting the deck', 'ground-fall', 'logging big air time', etc.
Okay - so what do we do about it?
Answer = We modify ABoK #1010 and make it more stable and secure.
This subject has been almost exhaustively examined in this forum and by many others in the wider climbing community. Opinions are like cars - some people like Toyota Land cruisers, others prefer Nissan Patrols, some prefer Jeeps, some prefer Landrover Discovery while others prefer a Hummer. Who is right? Who is wrong?
One popular way of securing "#1010" is to use the so-called 'Yosemite finish'. Does this trick work? Answer = yes it does. But roo has a point, it also depends on the
stiffness (see rope modulus article at the AMGA website... this link was working at time I wrote this post:
http://www.google.com.au/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0CDEQFjAB&url=http%3A%2F%2Famga.com%2Fresources%2Fvarious%2FSequential_Failure_Paper.pdf&ei=0LXDUcLAHsWViAen1YGIBA&usg=AFQjCNEhz7ISMQ4oidnlpehtwgA2hEq1rg&bvm=bv.48293060,d.aGc&cad=rja ) of the rope you are using. Stiffer ropes might not behave so well with the Yosemite finish. There are other tricks we can employ - eg my own 'EBSB' variant discovered while playing around with Dan Lehman's 'EBDB'. There are still more variants out-there-in-the-wild. Not all have been subject to rigorous scientific testing - so caution needs to be applied.
As for the Portugese bowline, ABoK "#1072", this is not by itself going to improve the situation.
ABoK "#1085" is
not a practicable method of tying a rope directly into a harness either!
So yes, the tried and tested ABoK #1047 is practicable, it is secure and it is stable. Arguably, it is also 'relatively' easy to learn and most importantly, is 'relatively' easy to remember.
I for one have no trouble trusting my life to ABoK #1047. I have climbed extensively since early 1983 and still climb frequently today (Himalaya, New Zealand, every major crag in Australia in all States and Territories and have established well over 100 first recorded ascents. I have also personally established a major new climbing area in Australia - Fredericks Peak). I have sustained literally over
one thousand falls in my lifetime - some falls as big as 20m (metric). Since I am still typing this post in this forum, I must therefore still be alive and kicking. So I have
extensive practical experience with ABoK #1047.
Most of my falls have been sustained on ABoK #1047. However, in relatively recent years, (since about 2009) I've been increasingly using and relying upon the 'EBSB bowline' variant. It is proven to work. As to whether it is 'better' than #1047 is debatable. It comes down to factors such as; ease of untying after repeated hard falls, the diameter of rope you are using, your body mass, and your level of knowledge and skill.
In summary, I do not believe that "ABoK #1072" is a practicable solution as a tie-in knot for climbing applications. I also don't believe that "ABoK #1085" is practicable either.
There has been some discussion about the Zeppelin loop as a possible solution and it has some strong supporters (roo?). In my view, this knot is harder to teach and harder to remember (ie memory retention) for novice climbers. before anyone starts jumping up and down and screaming...this is just my view. I hold this view because the zeppelin loop requires thinking in 3D - whereas "ABoK #1047" can be held flat (in a 2D plane) and its a simple matter of 'retracing' the path of the rope. Most learners can acquire the skill of tying ABoK #1047 'relatively quickly. It is
not so easy to force the Zeppelin loop into a flat 2D structure and then weave the tail through the correct areas of the knot to create the finished product.
Also, there is no large body of evidence of in-field testing of the Zeppelin loop that I am aware of...(I could be wrong though) - but I have not found any test data or field tested evidence-based data re the Zeppelin loop.
If there is, I would like to read and study it.
I believe that "#1047" and derivatives of "#1010" will be around for some time to come - and I don't see this situation changing in my lifetime.
Mark
