Author Topic: Mechanical Engineers at MIT analyze knots  (Read 2374 times)

KnotMe

  • Sr. Member
  • *****
  • Posts: 580
    • The Dao of Silk
Mechanical Engineers at MIT analyze knots
« on: September 10, 2015, 03:23:12 AM »
http://news.mit.edu/2015/untangling-mechanics-knots-0908
As the paper is not yet published, the publicity is mostly about change in the forces required to pull closed knots based on the number of turns.  So, surgeons' knots vs reef and how the number of twists affect the forces involved.

Academic research into knot strength.  Some of you should be happy.   :D

xarax

  • Sr. Member
  • *****
  • Posts: 2781
Re: Mechanical Engineers at MIT analyze knots
« Reply #1 on: September 10, 2015, 10:49:56 AM »
   As one of the first steps towards a scientific theory of physical knots, it is much welcomed, of course.
   However, we should not overestimate the breath of this study : As I understand, it involves only systems with multi-overhand knots :
   " Excellent agreement is found between theory and experiments for overhand knots over a wide range of crossing numbers."
   " the researchers tied knots with multiple overhand twists, creating increasingly longer braids ".
   As WE know very well ( and I am sure THEY, too, will learn after some time...:) ), the force required to form, dress and tighten a knot, is NOT related to how "tight" this knot remains after that. So, we should pay attention to what the paper is all about : the tightening of the knots in the first place, not their intentional or accidental untightening after that :
    "The researchers carried out experiments to test how much force is required to tighten knots with an increasing number of twists. They then compared their observations with their theoretical predictions, and found that the theory accurately predicted the force needed to close a knot, given its topology and the diameter and stiffness of the underlying strand."
   We know very well that some knots ( the Clove hitch is the best example ) "close" around themselves very tightly, they are "self-locking", and some not, but that has NOTHING to do with their topology, it is NOT depending very much on the stiffness or the friction characteristics of the rope, but it IS depending of how much compressible and elastic is the material - a matter they have not studied at all.
       
   " We now understand how the knot locks itself up when more turns are added ".

   Noope ! You do not. You may understand how the stiffness and friction of the material is related to the force required to tighten a knot, but NOT how, and why, a knot "locks up" .

   " I can tell you what force is required to close it."

   Yes - but that is not my problem, because I had eaten  my breakfast, and practical knots have a few, only turns / crossing points, so they can be tightened with ease.
   My problem is how easy or difficult is to UNTIE a practical knot - which may "close" around itself, may "self-lock", may jam, or not.

   " We re describing the force it requires to close a loop, which is an indicator of the stiffness of the knot ".
   The "stiffness" of the knot regarding its tying, NOT its untying ! There is no relation whatsoever between those two things. The Clove hitch "closes" around itself, it is "seldf-locking", it jams, much more than the Girth hitch, which has the same number of "turns" .
   
   [ Is THIS the kind of research NASA is funding ? Then no wonder it uses the "scientifically correct" hitches it uses on the Mars rover... :) :) ]
« Last Edit: September 11, 2015, 01:01:28 AM by xarax »
This is not a knot.

KnotMe

  • Sr. Member
  • *****
  • Posts: 580
    • The Dao of Silk
Re: Mechanical Engineers at MIT analyze knots
« Reply #2 on: September 10, 2015, 11:11:28 PM »
Funded in part by the National Science Foundation.  Not NASA.  Actually read the material before you start picking at things.

xarax

  • Sr. Member
  • *****
  • Posts: 2781
Re: Mechanical Engineers at MIT analyze knots
« Reply #3 on: September 11, 2015, 12:00:00 AM »
  Actually read the material before you start picking at things.

  I happen to know what an "Abstract" is, or should be. I had read the abstract. I am sure that, in the paper, there will be even less material relevant to why a knot can not be UNTIED... because that is what I was talking about in my previous post.
  Physical Review Letters is the pinnacle of the world scientific journals, so I am sure that the article is of the most high quality. I had not questioned what is claimed in it, I questioned the relevance of the findings to what is a permanent problem of the practical knots : how easily they can be UNTIED, and if they "jam", or not.
  All knot tyers know very well that there are convoluted knots, with many "crossings", which can be untied easily, and simple knots ( like the Clove hitch, for example ) which can not - so there is no doubt that topology does not play any role on THIS. Now, for "braids" or systems of multi-overhand knots, which the authors had studied, it seems that there was a relation, and I am glad that they had discovered it - but that was NOT what is our main problem. Do you know many practical knots, which are made of multi-overhand knots ? Do you know many practical knot tyers who want to know how easily they can tighten such a convoluted overhand knot ? I do not.
   The authors use the word "stiffness" for whole knots. From what is written in the abstract, or in the announcement of the paper in the news, they use it to mean the resistance of a knot to TIE it, not to UNTIE it.
   I know that many people have an almost religious belief in their country s great scientific institutions ( as the MIT and the NASA are, no doubt about that, of course ), but they should also remember, or learn, that science has no relation to religion and beliefs whatsoever. The hitches used by NASA on Mars Rover, which were supposed to be chosen after careful scientific studies, were not even "bon pour l orient", and the "documentation" of the "research" which lead to this particular choice and not any other was laughable... However, after 116658 (!)  visits on the page of the Forum where the issue was discussed, I have not seen many people to notice/acknowledge/accept this. Perhaps their "patriotism" did not allow them to "pick" at NASA. :)  To me, patriotic is what improves the quality of what the institutions of your country delivers, not what hides mediocre things under the rug.
   Anyway, we will see - and if, after this paper, I will learn why the Clove hitch jams while the Girth hitches does not (1), I will start BELIEVING ! :) Perhaps it is about time... :)

   " For a successful technology, reality must take precedence over public relations, for nature cannot be fooled."
     Richard Feynman, on the Challenger disaster.
   
   1.  http://igkt.net/sm/index.php?topic=4347.0
« Last Edit: September 14, 2015, 02:02:51 PM by xarax »
This is not a knot.

KnotMe

  • Sr. Member
  • *****
  • Posts: 580
    • The Dao of Silk
Re: Mechanical Engineers at MIT analyze knots
« Reply #4 on: September 11, 2015, 01:20:33 AM »
  • I'm not an American and have no patriotic duty to protect NASA, you're the one who brought them up out of the blue.  Giving you the benefit of the doubt that you had mistaken NSF for NASA, it turns out that you just have some sort of irrelevant axe to grind and chose this non sequitur to do so.  :o
  • But, hoping to engage in productive and positive discussion, perhaps you should consider addressing your concerns about rigor and research topics to the researchers in question (lots of contact info on all the articles) in a polite and considered manner rather than monologuing at us. ::)  Encourage them by being, y'know, encouraging rather than hoping negative stimulus therapy in a forum they will likely never read does the job.  :D
  • Speaking of trying to stay on topic, here's a link to the pre-print version of the paper mentioned in the above article that inspired/preceded the multiple twist paper about to be published.  61 pages of hard core math to sort of start describing what happens in the single overhand knot.  ;)
  • Edited above to clarify.

« Last Edit: September 11, 2015, 07:17:48 AM by KnotMe »

xarax

  • Sr. Member
  • *****
  • Posts: 2781
Re: Mechanical Engineers at MIT analyze knots
« Reply #5 on: September 11, 2015, 01:55:18 AM »
       you just have some sort of irrelevant axe to grind and chose this non sequitur to do so.  :o[/li][/list]

      It is not my mistake if you do not see any relation... I was just joking, you had not understood it, you came back "correcting" me about the NSF ( please, do not do the same with NSA  next time  :) ), and then you dare to talk about grinding axes...

    ...addressing your concerns about rigor

    WHERE exactly did I say anything about rigor ?  It seems that I am grinding the axe, but you smoke the stuff ! :) :)  We are all happy Indians.

    61 pages of hard core math to sort of start describing what happens in the single overhand knot.  ;)

    Not even 62 ! :)
    I would prefer 1 ( one ) page, or even 1 single line, which would tell me something about practical knots I want to know. Those "hard core maths" may be good, but not because they require 61 pages ( the Abstract you had cited, is of an article in PRL, not PR, where it will occupy 2 or 3 only pages...), but because they have some relation with reality. And if they had some relation with the reality of practical knots, they would had been MUCH better - but, for the moment, at least, they do not.
    I want to know why some knots ( and NOT the multi-overhand knot...) are difficult to UNTIE, and not to TIE.
    That is what I was saying in my monologue - which, evidently, you have not read. Perhaps you were a little dizzy after the 61 pages of hard core math ... :) :)

    P.S. Or perhaps you were a little sleepy... because those 61 pages of "hard core math" you cite belong to another article, not to the article which will be published in the PRL ( of course...). However, I have to admit that you were 25% right : there is one (=1 ), out of the four (=4), authors of the two papers, who is the same on both.
    Moral of the story : Do not read many pages of "hard core math" at once ! :) :)

    « Last Edit: September 14, 2015, 02:05:32 PM by xarax »
    This is not a knot.

    xarax

    • Sr. Member
    • *****
    • Posts: 2781
    This is not a knot.