empirically what we know is #1 that no knot has tested statistically above 50%...
In
your mind, as you rightly said !
If I had seen two, or a dozen white swans all and all in my life,
empirically what I "know" is that no swan is black - and I might even claim,
sceptically, as you do, that there will never be any black swan anywhere in the Universe - because, as Leibniz said, this World is the best, so it is the only possible !
So, when you will test the 60-120 known simple bends I have suggested to you,
and the 60-120 centrally re-tucked versions of them,
then you will "know" something...
You have tested what you said is "THE Sheet bend", and "THE fig.9 knot", without even realizing that there are TWO Sheet bends, and THREE fig.9 knots, for KnotGod s sake... And I do not even count how many different dressings / loadings combinations "THE fig.8 knot" can have... So, I guess that, although you have done a wonderful job, you have still some distance to cover before you start claiming that "we know".
I would suggest it is impossible to design a knot that does not have a 1:1 bend radius in it.
? ? ? You would nt
bet on this, I suppose...
Because I know
dozens of such bends !
Does the retraced overhand knot = Water bend, for example, have a 1:1 bend radius in it ? Or, for that matter, most the fig.8 knots you had mentioned ? I guess I have not understood what you say/mean here.
I did not say we will not find other applications for loops. I said there were not other applications for bends, because an end for end splice is the better solution whenever it is possible.
So, because spices are the better solutions, we have to abandon all end-to-end knots at the ends of a loop, and/or apply only the particular hitch-to-hitch bend you propose - because :
the only real practical application (on a sailing boat anyway) for bends in bare dyneema is to make fixed loops which are too short for an end to end splice, and un-tieing is really not all that critical feature for those... So, practically speaking I am happy with the solution we have already found. I am skeptical that we will find a better practical solution for this application, but if we do, I expect it will be more along the lines of ...
Your reasoning is not very convincing here, I am afraid - and the retreat to the splices sounds almost as a desperate defence.
We have to test
all the closed, by end-to-end knots or hitch to hitch knots, loops we know, before we can "know". If you abandon knots, you may even abandon splices and search for a nice cheap glue, or a mechanical fastener.
Practically speaking regarding you, you do deserve to be happy. However,
practically speaking regarding our knowledge about knots, we are veeery unhappy / sorry we know next to nothing !
I don't know how much experience you have doing pull tests, but I now know what makes for important statistical variation and what does not. In this slipping case the important factor is pull speed.
1. NONE
2. So, the other factors I had mentioned do not contribute in an "important" statistical variation? In what sense ? How you know before hand what is "important" in a distribution, if you do not know what form this distribution should have ? If you say they will not vary the results more than, say, 1%, I will agree - provided that the differences of the pull speed factor would be, say, ten or fifteen times more ( 10% - 15%)- but is this the case ?
hmmm . . . first, you have no idea how much experience I have testing bends in this material. I would suggest it is 'adequate' to express some skepticism. And second, I said that an important criteria for success was a low profile result. That makes my following comment about looking at long fishing knots over rounder knots almost by definition.
I apologize for anything that have said and it sounded like I think it sounded to you. I am sure that you have dozens, if not hundreds of times more experience than me - but this is still not adequate, I am afraid !
I read your site, and I see the number and kind of bends you refer there - and I would suggest this reading is "adequate" to express scepticism about your knowledge of bends - which, I repeat, may well be much more extended than mine s. However, I happen to know that there is no One Sheet bend, One fig.8 and One fig,9 knot, and to know the 120 known bends - while you do not... So,
we(plural) have many knowledge to exchange, before
we can claim that
we "know".
First, you mentioned the "facts", now you mention the "definitions" ! My feet have not started trembling, though...
Have you
measured the maximum width of the cross sections of even the very few bends you have tested ? If yes, where are the NUMBERS ( in cm, of course ..., not in Hers Majesty s inches
). And of HOW MANY BENDS have you measured the maximum widths of the cross sections ? How have you strengthen your
belief that the minute sample of bends you have tested is "adequate" to jump into such broad conclusions ?
Of course, a
long-long-long fishing knot, tied by
spin-spin-spin entangling, would be sleek, strong and will not slip. So you propose to tie the dyneema bends with fishing knots, and fishing knots only ? If so, I have to pull out my fishing knots memories, which I had buried dozens of years ago. Now, I eat fish in the restaurants - I even would not allow cooking of fishes in my apartment, because I can not stand the smell of the fresh fish more than 1 minute.
Frankly, I would nt expect a sailor be a fun of
fishing knots !