WEBVTT
00:00:06.000 --> 00:00:07.519
Thank you very much, Jane.
00:00:07.599 --> 00:00:10.080
Thank you all for coming and people watching online.
00:00:10.640 --> 00:00:12.960
So we're gonna have fun today.
00:00:13.359 --> 00:00:16.000
Well, I'm going to have fun for sure.
00:00:16.480 --> 00:00:21.760
This year I want to uh talk about what I call the mathematics of surprise.
00:00:21.920 --> 00:00:36.719
What I will do is take a simple observation, simple phenomena, and see trying to understand it both on physical ground but also using mathematics and then see how that knowledge expand to other areas.
00:00:37.119 --> 00:00:40.880
And here's one of my favorite subjects, which is cracking the whip.
00:00:41.439 --> 00:00:51.439
But before we do that, I want to talk about something called the scientific methods, mostly organized by people who don't do science, as far as I can tell.
00:00:51.600 --> 00:00:54.560
And they are different versions, but they mostly look like that.
00:00:54.719 --> 00:01:05.599
You have an observation, a question, then you research a topic area, you formulate an hypothesis, you test it, you analyze the data, you report the conclusion and so on.
00:01:06.000 --> 00:01:08.239
And it feels a little boring, right?
00:01:08.400 --> 00:01:11.439
There's no like you wreak a moment, no excitement.
00:01:11.599 --> 00:01:19.760
And frankly, in my entire life, all the science I've ever done, and the people around me, nobody has ever followed this kind of wrong, right?
00:01:20.159 --> 00:01:24.159
So let me tell you about the other scientific methods.
00:01:24.480 --> 00:01:28.400
And usually it starts with observing something bizarre.
00:01:28.560 --> 00:01:32.560
You are in a field, and something is not quite what you expect.
00:01:32.719 --> 00:01:35.359
It doesn't match your understanding of that area.
00:01:35.439 --> 00:01:38.640
It's something extra, and you want to try to understand.
00:01:38.879 --> 00:01:40.239
So what do you do?
00:01:40.480 --> 00:01:43.280
Well, you gather useful and useless facts.
00:01:45.359 --> 00:01:49.599
Well, you don't know if they might be useful, you're just early days, right?
00:01:49.760 --> 00:01:53.359
So you go around and read books and watch TV, whatever.
00:01:53.519 --> 00:01:56.000
You do whatever you can to try to get.
00:01:56.159 --> 00:01:59.680
And then you turn to what I call back of the envelope computation.
00:01:59.760 --> 00:02:06.560
You say, okay, let's do some estimate to see if I get the right mechanism, maybe some order of magnitude to see how things work.
00:02:06.719 --> 00:02:10.400
And usually it's not enough, but that gives you an idea you're on the right track.
00:02:10.639 --> 00:02:16.879
And then you build a minimal mathematical model, what we call a toy model in the trade.
00:02:17.039 --> 00:02:23.039
Something that has all the components that is as simple as possible but not simpler, as Einstein would say.
00:02:23.199 --> 00:02:25.919
So it still captures something and you can learn from that.
00:02:26.319 --> 00:02:33.199
So here is a job both of modeling, and we talk about modeling, but also your job as a mathematician.
00:02:33.280 --> 00:02:37.120
No, you have a model, you have actual equation or something to that effect.
00:02:37.280 --> 00:02:50.479
You need to study the model, find solutions, see how the solution behave, maybe do numerical solutions on the paper, and so on, and you go around trying to get extract information of this simple equation.
00:02:50.719 --> 00:02:54.000
And after that, you go for a reality check.
00:02:54.319 --> 00:02:55.919
Does that match what you see?
00:02:56.080 --> 00:02:57.520
Does that match anything?
00:02:57.759 --> 00:03:01.199
And maybe it does, and then the big question, so what?
00:03:01.280 --> 00:03:02.159
What have we done?
00:03:02.400 --> 00:03:02.719
Right?
00:03:02.960 --> 00:03:07.039
And that you mostly don't by then, and you can write a paper, something like that.
00:03:07.199 --> 00:03:09.280
But of course, nothing is linear.
00:03:09.439 --> 00:03:13.599
And when you do your reality check, typically you start gathering other facts.
00:03:13.759 --> 00:03:20.560
Maybe you do and do little experiments or you try to find data, which forces you to maybe revise your model.
00:03:20.719 --> 00:03:24.319
Maybe you're missing an important mechanism in the model and so on.
00:03:24.639 --> 00:03:33.439
And after you think about it, you think, well, maybe there are other things that are like that and maybe a little different that follows the same pattern.
00:03:33.919 --> 00:03:38.879
And that forces you also to maybe write other models for other circumstances.
00:03:39.039 --> 00:03:46.159
And as soon as you gather facts, of course, you study the model, you have solutions, you have to compare them and so on, and it gets very messy.
00:03:46.240 --> 00:03:50.879
And every day you wake up and you're in one of these boxes and you jump from one to the other one.
00:03:51.840 --> 00:03:56.560
So today we start with something bizarre, right?
00:03:56.879 --> 00:03:59.919
And the something bizarre is the whip cracking.
00:04:00.080 --> 00:04:04.400
And it was bizarre to me as when I first thought about it.
00:04:04.639 --> 00:04:13.919
So I spent a long time in the southwest in Arizona, but that's not where I first uh realized that there was something interesting with the whip.
00:04:14.319 --> 00:04:21.439
It was in a meeting uh in Budapest in Hungary, and you know, uh there is usually a conference dinner in this meeting.
00:04:21.519 --> 00:04:25.839
It's a rather boring affair, big group taking a long time, drinking too much.
00:04:26.000 --> 00:04:39.279
But on this occasion, there was a folkloric band from Hungary coming and doing a little show with whip cracking, and I was there ignoring my fellows and looking and said, hmm, that's bizarre.
00:04:39.519 --> 00:04:40.720
Wonder what is this?
00:04:40.879 --> 00:04:45.519
You know, what is that so distinctive sound, a whip cracking?
00:04:45.759 --> 00:04:55.360
I knew a little bit of it before, but I got interested because at the time I was interested in dynamics of curve, and it didn't quite fit my understanding of that.
00:04:55.839 --> 00:04:58.000
So I started looking into that.
00:04:58.319 --> 00:05:15.360
But before we start, and to avoid any confusion, despite the movie Zoros and Indiana Jones, I want to let you know that the only, the one and only purpose of long whips, the one I'm going to uh talk about, is to make that sound, that crack.
00:05:15.600 --> 00:05:19.839
It's never been a weapon, it's never used for the purpose kinky otherwise.
00:05:20.000 --> 00:05:26.000
These whips, the one I'm going to show you, are only there to uh to create that whip.
00:05:26.399 --> 00:05:29.839
So I went home and it's time to gather facts.
00:05:30.399 --> 00:05:33.040
Okay, uh, or at least to try things.
00:05:33.279 --> 00:05:38.399
So the first thing I got, I got this uh big boy here, eight-footer.
00:05:38.560 --> 00:05:41.439
You know, it has a nice structure, we'll talk about that.
00:05:41.600 --> 00:05:45.759
And I started practicing, and I bought the bull whip book.
00:05:45.920 --> 00:05:49.279
Uh it was a time where you could learn skills from books.
00:05:49.839 --> 00:05:53.360
That was before YouTube, before even Google and stuff like that.
00:05:53.519 --> 00:05:59.680
And you learn in the books, very interesting, that there are three uh typical motions when you do that.
00:05:59.920 --> 00:06:04.000
There is three phases the setup, the turn, and the follow-through.
00:06:04.240 --> 00:06:11.360
So if you give uh a whip like this to somebody and you ask them to crack, they're probably gonna try to do something like that.
00:06:11.519 --> 00:06:14.079
So this one is too big to do here.
00:06:14.240 --> 00:06:19.279
So here I have a smaller one that I just got for for this purpose here tonight.
00:06:19.439 --> 00:06:20.879
It's only a four-footer.
00:06:21.040 --> 00:06:26.800
And so if you're given a whip, the first thing you'll try is what is called a snap, which is something like that.
00:06:27.120 --> 00:06:32.319
Yeah, you have to put a lot of force and you get a little crack, nothing impressive.
00:06:32.480 --> 00:06:32.800
Okay.
00:06:33.279 --> 00:06:36.800
And it turns out that yes, it works, you can do that.
00:06:37.360 --> 00:06:41.360
Um, but it's not the real way you really want to crack a whip.
00:06:41.680 --> 00:06:47.360
The way you want to crack a whip is to make a loop and let the loop propagate along.
00:06:48.480 --> 00:06:56.720
So I'm gonna show you have videos just in case it doesn't work here, but the the setup is very simple.
00:06:56.879 --> 00:07:02.240
You raise the whip up to a certain height and you turn it to make a loop that goes at the end.
00:07:02.480 --> 00:07:02.800
Right?
00:07:03.120 --> 00:07:08.800
And so if you do it like you see, right away, you create a crack with very little effort.
00:07:09.040 --> 00:07:11.040
And that gets very interesting, right?
00:07:11.279 --> 00:07:16.399
Because you send a loop and that loop accelerates and then crack at the end.
00:07:16.800 --> 00:07:29.600
Very distinctive, and so distinctive that it's not a normal sound, and that's why we use it, because animals usually never hear that kind of sound and they react to it, and that's what controls horses and cattle.
00:07:29.759 --> 00:07:40.480
And horses being the one method of transport to the 19th century, there was a huge, huge business associated with whip and cattle and ranching and all that.
00:07:40.720 --> 00:07:46.560
So, just in case it didn't work, I did my own uh experiment to show you.
00:07:46.720 --> 00:07:52.079
Uh, as you can see, I'm not only a scientist, I also play one on TV.
00:07:53.040 --> 00:08:02.639
So that's the whip, and here is a slow motion of the same action, and you can almost you can see the loop going through, right?
00:08:02.720 --> 00:08:06.560
It's high speed, and great, it works.
00:08:06.800 --> 00:08:14.639
Here is a uh younger version of myself in Arizona, and here is another way, it is this with this one.
00:08:14.879 --> 00:08:16.480
I'm not gonna do it here.
00:08:16.720 --> 00:08:20.319
I could almost maybe do it, but maybe I'll get in trouble.
00:08:20.560 --> 00:08:35.519
Uh and the idea is that you turn the whip around, so you create tension, that's a setup, and then you invert the motion to create that loop, and that loop zips to the end, and by the time it ends, it unfolds and cracks.
00:08:35.759 --> 00:08:37.679
So here is how it goes.
00:08:39.200 --> 00:08:42.559
You can see that the neighboring dog doesn't like it.
00:08:42.720 --> 00:08:48.240
And you can see here, you can see the loop forming there and going around.
00:08:48.480 --> 00:08:56.799
So you find in popular science book that the crack, that special cracks, occurs when the tip reaches the speed of sound.
00:08:57.600 --> 00:09:01.200
At the type where people were still looking for books and stuff like that.
00:09:01.519 --> 00:09:04.559
So this is supersonic motion.
00:09:04.799 --> 00:09:11.519
This is the first time humans manage to make supersonic motion, it's through the motion of the cracking.
00:09:11.679 --> 00:09:15.519
And we're gonna go and look at that in quite uh detail.
00:09:16.320 --> 00:09:21.440
But there are other things you can do, not with whip, but with chain or with whip actually.
00:09:21.519 --> 00:09:25.919
Uh, and I wanted to show you there's something else bizarre called the chain fountain.
00:09:26.080 --> 00:09:31.600
And that's something that's done with a regular chain, the one you use to turn on the light.
00:09:31.840 --> 00:09:44.559
And I would again try to do a little experiment, it's not sure to work, but it's not for a dramatic purpose, it works better if the height is uh actually higher, but also for dramatic purpose.
00:09:44.799 --> 00:09:49.919
If you want, I can put my glasses on so that look more serious and more responsible.
00:09:50.639 --> 00:09:52.159
Okay, so let's see.
00:09:52.320 --> 00:09:55.679
Let's get one and two and three.
00:09:56.000 --> 00:10:01.120
So all I'm going to do is I have a chain and I'm gonna release it.
00:10:01.519 --> 00:10:08.240
And you can see all yeah, the idea it goes, well, it's not very impressive, this one.
00:10:08.399 --> 00:10:09.759
But I have a high-speed move.
00:10:10.000 --> 00:10:13.519
I have a movie, and you'll see exactly what I mean with it.
00:10:16.399 --> 00:10:27.360
You see it's the same beaker, the same chain, and you have this chain going up, and very satisfied with myself, of course.
00:10:27.679 --> 00:10:33.679
I have my uh trademark thumbs up.
00:10:35.039 --> 00:10:43.440
Okay, so I also took the same chain and went into we have a little lab in the math department in in Oxford.
00:10:43.600 --> 00:10:51.519
We call it the uh mathematical observatory, where we have a very good high-speed camera, and I filmed that in high speed.
00:10:51.679 --> 00:10:55.600
And here you really see the effect, the chain going up.
00:10:56.240 --> 00:11:08.960
And what's amazing is that it doesn't touch the beaker, the whole thing goes and unravel, goes up and so this was popularized by Steve Mole in 2013 in a series of YouTube videos.
00:11:09.039 --> 00:11:13.279
You can really see the effect on popular science now of the arrival of YouTube.
00:11:13.600 --> 00:11:22.399
And the same year, uh John Biggins and Mark Warner from Cambridge wrote a first paper and called it the chain fountain, trying to give the first explanation for it.
00:11:22.639 --> 00:11:33.919
So that's also a very bizarre phenomena, it's quite beautiful, especially in slow motion, where you have this elevation of uh of the chain going up uh with no good reason.
00:11:34.000 --> 00:11:35.519
There is nothing pulling it.
00:11:35.679 --> 00:11:41.679
And so we want also to try to understand T and uh in the same context.
00:11:41.919 --> 00:11:47.200
So now it's time to turn and look at facts, you know, with practice, we've tried a little things.
00:11:47.360 --> 00:11:48.799
What do we know about whips?
00:11:49.120 --> 00:11:54.320
Well, one of the interesting things is that whips were with us for at least 3,000 years.
00:11:54.559 --> 00:12:05.759
If you look at this stone panel from uh Mesopotamia, Neo-Assyrian gypsum panel, you'll see that right there you have the whip used to uh direct uh horses.
00:12:06.000 --> 00:12:17.039
And through the year, whips have evolved uh technologically to rather sophisticated objects that typically have a handle, depending on the size of the handle.
00:12:17.200 --> 00:12:22.639
You can have a stock whip, a bull whip, or a snake whip, the one you you need to put in your saddle.
00:12:22.879 --> 00:12:30.240
You have a swivel, the place where you can turn, then you have a long part, the tongue, which is tapered and very flexible.
00:12:30.480 --> 00:12:41.679
And then it ends up with this part, which is called the fall, that direct the motion essentially to be flat, and always a little piece called the cracker that you can replace easily.
00:12:41.840 --> 00:12:47.360
That's the one that takes all the abuse and turns around and gets the most forces.
00:12:49.279 --> 00:13:02.080
So we think that we were the first to invent that, but some scientists in 97 uh thought that actually uh dinosaurs were way ahead of us.
00:13:02.320 --> 00:13:28.159
So when I was a kid, the apatosaurus were called the brontosaurus, and that means the the dinosaur of thunder, which was quite quite uh convenient, quite appropriate, because they did a little computation using a chain and showed that as the as the dinosaur whips its tail, the end of it reached supersonic speed, and that it might have used it for some purpose.
00:13:28.320 --> 00:13:30.480
So now you can speculate what it was.
00:13:32.000 --> 00:13:36.320
Uh and that that got a lot of people excited at the idea.
00:13:36.480 --> 00:13:52.320
Uh however, more recently, in uh four years ago, people did a more careful analysis with more like the vertebrae and all that, rather than a simplified physical model, and showed there is an acceleration but much more modest, only like 30 meters per second.
00:13:52.480 --> 00:13:58.480
So we don't know, but maybe it's unlikely that the dinosaurs were supersonic.
00:13:58.879 --> 00:14:03.360
So now what about on the science side?
00:14:03.519 --> 00:14:05.840
You play around, you can crack and all that.
00:14:06.080 --> 00:14:07.679
What is the knowledge there?
00:14:07.919 --> 00:14:22.320
Well, you have to wait until 1927 to have the first set of experiments that actually demonstrated that the the whip cracking was a supersonic motion, and that's from Zeferin Carrier, a French physicist from Toulouse.
00:14:22.559 --> 00:14:32.320
And he reasoned that in the motion of the whip as you go, as you do a snap, there is a piece that doesn't move, essentially the middle piece here, where you go around.
00:14:32.480 --> 00:14:47.200
And so if you want to do a control experiment, what you do, you replace that by a pulley, and now you have a pulley around, and all you need to do is put a weight here and let it unravel so that it accelerates with constant acceleration.
00:14:48.159 --> 00:14:54.879
So that was a very good idea, and you can imagine 1927 he actually managed to do high-speed photography.
00:14:55.120 --> 00:15:00.320
Must have been extremely complicated from an experimental point of view.
00:15:00.559 --> 00:15:04.080
But he had series of results, fascinating.
00:15:04.320 --> 00:15:15.600
He in red you see the pulley, and you see the result, the different stage of the chain or the of the string that he has going around and flapping around.
00:15:15.840 --> 00:15:31.039
And he also measured the velocity, and what he showed for the first time at that time was that the tip of the whip, or the tip of his string, was 900 meters per second, which is about three times the speed of sound in the air.
00:15:31.120 --> 00:15:31.360
Right?
00:15:31.600 --> 00:15:35.200
Speed of sound is 340 meters per second.
00:15:35.600 --> 00:15:42.559
And that's at that point, this is where the little cracker at the end goes from one side to the other side.
00:15:43.039 --> 00:15:47.039
So a few years ago, we thought, hmm, that's very interesting.
00:15:47.120 --> 00:15:49.919
And could we redo this type of experiment?
00:15:50.480 --> 00:15:59.279
And we thought that we simplify something in between the chain front end that I show you and the uh and the whip.
00:15:59.679 --> 00:16:12.720
And so what we did, we take a beaker like that and a string like this one, and we just added a weight on the string and drop the weight and see how this part moves around.
00:16:13.120 --> 00:16:16.879
Very much like the original uh experiment of Carrier.
00:16:17.360 --> 00:16:18.639
This is a movie of it.
00:16:18.799 --> 00:16:27.759
What you see, the big black uh circle is the is the beaker, and you see the motion, the chain being pulled, and this is what it looks like.
00:16:27.919 --> 00:16:29.360
That was the work with P.T.
00:16:29.519 --> 00:16:32.159
Barr, Basile, and Dominique.
00:16:32.399 --> 00:16:36.720
It was a lot of uh fun problem also to try to see that.
00:16:36.960 --> 00:16:47.759
And the interesting feature here is that it leaves also the beaker, and that means that the acceleration at one part of the chain is lower than the acceleration at the other part.
00:16:48.480 --> 00:16:52.559
It goes faster, that's why it lifts.
00:16:52.720 --> 00:16:57.279
And so you can ask why is this acceleration, extra acceleration coming from?
00:16:57.360 --> 00:17:04.400
And that's very much like the chain front, and as it goes around, this part has faster acceleration than the chain going down.
00:17:04.799 --> 00:17:15.440
And the other feature of interest is to look at what happened right at the crack, because the the chain or the or the whip is actually very curved there.
00:17:16.160 --> 00:17:25.119
And right at that point, if you look at the radius of curvature, it goes flat at some point, the radius of curvature goes to infinity.
00:17:25.200 --> 00:17:33.200
And that's very interesting because in mathematics and in physics, when you have this kind of behavior, something goes infinity, a lot of things simplify.
00:17:33.279 --> 00:17:34.400
It's a good place to study.
00:17:34.480 --> 00:17:39.119
It's called a singularity, and singularities are always uh interesting.
00:17:40.640 --> 00:17:44.000
So we're going to think about that, we come back to that.
00:17:44.319 --> 00:17:53.440
Another interesting paper was the paper uh in 1998 from German physicists, and they have this figure in the paper.
00:17:53.519 --> 00:18:05.839
Uh, a guy with kind of a, it turns out he's a Black Forest whip cracker, and they brought him in the lab with his costume and everything in order to do uh high-speed imaging.
00:18:05.920 --> 00:18:26.400
And they do they did beautiful high-speed imaging, not only looking at the the full motion of the tip, but on top of that they have something called Schlieren imaging, which captures the variation in the density of air, and that allows you to see uh supersonic crack as it's a supersonic wave emanating from the whip.
00:18:26.720 --> 00:18:35.759
And so you can exactly pinpoint when the crack happens and where it happens, which turns out to be uh important for a problem.
00:18:36.160 --> 00:18:37.839
So it is looks like that.
00:18:41.519 --> 00:18:47.920
And each frame is a hundred, about hundred uh microseconds, the whole movie is two milliseconds.
00:18:48.079 --> 00:18:58.240
And if you extract a few frames, what you'll see right at the time it cracks when the supersonic boom arises, it's right at these frames.
00:18:58.480 --> 00:19:00.640
And here I uh enhance them.
00:19:00.799 --> 00:19:04.960
This is what you see, the the supersonic wave coming out of it.
00:19:05.359 --> 00:19:10.319
If you had a bullet in with the same time of imaging, you'll see the cone, the MAC cone emanating.
00:19:10.400 --> 00:19:12.880
We'll see that a little bit later.
00:19:13.839 --> 00:19:17.279
So that's very interesting because they also had very good data.
00:19:17.519 --> 00:19:36.559
And what the data showed is that if you look at the velocity as a function of time, so that's the point here, a velocity a function of time, the acceleration of the tip of the tip, the cracker, right, is about 50,000 times the acceleration of gravity.
00:19:36.799 --> 00:19:42.960
So you get amazing acceleration, and deceleration is even faster, about 70,000.
00:19:43.200 --> 00:19:48.880
And you, since you have imaging, you can exactly know when the crack occurs.
00:19:49.519 --> 00:20:11.920
And the interesting things that they show is that the crack occurs right there, and the maximum speed at that time of the tip is twice, about twice the speed of sound, which is very strange because I just told you that the popular explanation is that a whip cracks when the tip reaches speed of sound, not twice the speed of sound.
00:20:12.160 --> 00:20:15.599
And so that's also a bit of a puzzle for us.
00:20:15.839 --> 00:20:18.799
Uh try try to uh understand that.
00:20:19.279 --> 00:20:36.880
So now we have a few facts, and for me the the basic or most interesting thing is to see how starting with a very simple motion, which is my arm one millip uh one meter per second, you can get to this tremendous acceleration and velocity.
00:20:37.359 --> 00:20:38.559
How is that possible?
00:20:38.960 --> 00:20:40.160
How does that work?
00:20:40.640 --> 00:20:45.200
What are the important effects, tapering or the way you do it, and so on.
00:20:45.519 --> 00:20:53.599
So we are part of now trying to throw estimate at that, back of the envelope computation.
00:20:54.559 --> 00:21:04.400
And the you'll find more for more than a century, actually, Mack himself did computation of this type, uh, what I call a naive computation.
00:21:04.559 --> 00:21:14.799
You you're going to decide that you know you know the motion of the whip, and that initially you it's the whole whip moves at a velocity u.
00:21:15.039 --> 00:21:21.039
So as I go like that and move, the whole thing, as it moves forward, has a velocity u.
00:21:22.559 --> 00:21:29.599
So the energy is the mass of the entire things times the velocity square divided by two.
00:21:29.839 --> 00:21:32.960
That's kinetic energy, one basic quantity.
00:21:33.200 --> 00:21:37.519
So we have the mass of the whip and the little m is the mass of the cracker.
00:21:37.599 --> 00:21:52.160
So we have one half big m, the mass of the whip of the of the tongue moving, not the not the stick, and uh starting at velocity u, so it's one half m plus little m u squared.
00:21:52.559 --> 00:21:55.680
And then you're going to postulate that energy is conserved.
00:21:55.839 --> 00:21:59.759
Yeah, there is some friction and things like that, but overall it's not a bad.
00:22:00.640 --> 00:22:12.960
But idea, if energy is conserved, then as you turn, there is a part of the whip that's not moving anymore, and the rest of the whip that goes further that goes above.
00:22:13.119 --> 00:22:17.680
So now a fraction of the whip is only moving as you do this turn.
00:22:18.079 --> 00:22:21.680
And now the whole thing goes at velocity v, that I don't know.
00:22:21.759 --> 00:22:32.960
And the velocity v depends on the position of how much is moving, that's decided by the length x, a fraction of the mass of that related, and the little mass m.
00:22:33.519 --> 00:22:42.559
And now what you do in physics is you say, well, the energy is conserved, so e is the same for both, and I can solve for the velocity as a function of x.
00:22:42.799 --> 00:22:50.240
And so I have the velocity, depending on the initial velocity, the mass, the little mass, and how much whip is actually moving.
00:22:50.799 --> 00:23:05.359
And that's that expression is not that interest important, but the interesting thing is when it reached the end, because when it reached the end, the whole the whip is not moving anymore, it's fully extended, only the little part, the cracker moves.
00:23:05.519 --> 00:23:06.960
And that's little m.
00:23:07.519 --> 00:23:07.920
Right?
00:23:08.079 --> 00:23:17.279
And so you get an expression for the maximum velocity as a function of the initial velocity times the ratio of the mass plus one.
00:23:17.920 --> 00:23:33.440
So it gives you an idea, and it gives you the basic physical mechanism of what is going on in this case in this system, that you start, you give the system energy, and that energy gets concentrated in a small and smaller part that's moving.
00:23:33.599 --> 00:23:37.200
And so the velocity has to increase since the mass decreases.
00:23:37.920 --> 00:23:42.880
However, if you start looking a little bit and say, okay, but what is the mass of the cracker?
00:23:42.960 --> 00:23:44.079
Are we talking the whole thing?
00:23:44.240 --> 00:23:48.079
What if I have a continuous whip that's just tapered, for instance?
00:23:48.240 --> 00:23:51.359
There is no little cracker, so the mass goes to zero.
00:23:51.519 --> 00:23:55.759
And that's fine, the mass goes to zero, but then the velocity of the tip goes to infinity.
00:23:56.079 --> 00:23:57.599
So, hmm, that's bizarre.
00:23:57.759 --> 00:24:05.119
And if you decide which part is the little m and the big m, you realize you can have any velocity between u and infinity, right?
00:24:05.359 --> 00:24:14.000
So said, okay, we have the basic mechanism, but maybe it doesn't really help us to get a better view of what's going on.
00:24:14.400 --> 00:24:17.599
Okay, let's put that aside, we'll go back to that.
00:24:17.839 --> 00:24:23.680
Let me turn out to our other problem, which was the beaker or the chain uh fountain.
00:24:24.799 --> 00:24:30.559
And you can do another computation that's very popular uh among physicists also.
00:24:30.799 --> 00:24:33.839
It's called the Atwood machine from the 18th century.
00:24:34.000 --> 00:24:51.599
At Wood at his idea that you could have a little machine and a pulley with two weights, and if you vary the respective weight of one and the other, uh the acceleration would lower the big weight, and that would give you a way to actually compute the acceleration of gravity by changing the mass.
00:24:51.839 --> 00:24:57.119
It's very popular, it's textbook undergraduate physics the way you do that.
00:24:57.359 --> 00:24:58.799
So, what is the computation?
00:24:58.960 --> 00:25:01.440
What is the back of the envelope computation?
00:25:01.680 --> 00:25:08.079
Well, you need an envelope, and so I here is the computation of the envelope.
00:25:08.160 --> 00:25:13.440
Uh one of the theorem in mathematics is that every theorem is a back of the envelope computation.
00:25:13.519 --> 00:25:16.480
You just need a big envelope, big enough envelope.
00:25:16.799 --> 00:25:32.079
So in this case, you need a small envelope, and what you do, you essentially you say that the you have conservation of linear momentum, that the force acting in one side on one at the acceleration is balanced by the other one.
00:25:32.319 --> 00:25:32.720
Right?
00:25:32.880 --> 00:25:37.279
So you have two equations of motion, one for the little mass and one for the big mass.
00:25:37.519 --> 00:25:51.119
So conservation of linear momentum, you can get rid of the tension in between, that's t, and you get easily the acceleration as a function of the big mass and the little mass times acceleration of gravity.
00:25:51.359 --> 00:25:58.640
And you can see that if though if they are the same, big m and little m, you have zero acceleration as you would expect.
00:25:59.599 --> 00:26:02.960
Now, what happens if I get rid of the little mass?
00:26:03.119 --> 00:26:06.079
I, you know, you're not supposed to do that, but why not?
00:26:06.160 --> 00:26:06.480
Right?
00:26:06.720 --> 00:26:07.839
Does it still work?
00:26:08.000 --> 00:26:11.359
Well, the big mass accelerates at acceleration of gravity.
00:26:11.839 --> 00:26:14.079
It's just falling, nothing big.
00:26:14.400 --> 00:26:19.519
Well, that's strange because when I did the experiment with the B cut, this is not what happened.
00:26:19.839 --> 00:26:20.480
Right?
00:26:21.119 --> 00:26:29.839
The I had an acceleration, but without little mass at the end, you chain has no mass hanging there, the whole change was moving up.
00:26:30.079 --> 00:26:39.599
And you see that it's one of the problems with this computation in the previous one, is that you assume you know perfectly what the motion is and you just solve for one quantity.
00:26:39.839 --> 00:26:47.680
And so it's naive computation, gives you some intuition, but you have to be very careful about the assumption that you actually make.
00:26:48.400 --> 00:26:50.880
So, what about this chain font then?
00:26:51.039 --> 00:26:54.160
What is the standard explanation for that?
00:26:54.559 --> 00:27:09.920
Well, the idea there is that if you look at two of the beads in the chain, they're connected by a little bar, and one of the beads, like the one on the left side, is on a substrate, is on top of the other pack.
00:27:10.319 --> 00:27:14.720
And the one on the right is being lifted by the chain at any given time, right?
00:27:14.799 --> 00:27:16.880
So it's just at the time where it's being lifted.
00:27:17.039 --> 00:27:24.319
And as it's being lifted, what happens is that the rotation of that gives you an extra kick that lifts you.
00:27:24.480 --> 00:27:32.240
And that seems a little counterintuitive, but if you ever done any skateboard, you recognize that's exactly the only effect.
00:27:32.319 --> 00:27:36.000
So only is uh the only effect or the only move.
00:27:36.160 --> 00:27:38.400
It's named after Alan Gelfan.
00:27:38.480 --> 00:27:41.599
Uh he was a skateboarder that introduced this motion.
00:27:41.680 --> 00:27:51.119
And the idea is that this trick is that as you go along, you rotate the board, and that rotation provides an extra kick in the back and lift the whole thing.
00:27:51.200 --> 00:27:52.640
And I'm sure you've seen that.
00:27:52.880 --> 00:27:59.039
You look at skateboarders, they do that all the time with uh with great expertise.
00:27:59.279 --> 00:28:04.880
And so you have you might have the same effect on the chain, or the the chain, you have the two bead connected.
00:28:05.039 --> 00:28:12.400
If you lift one, the other one is going to be pushed out by an extra force, like that, but because of the rotation in the balance.
00:28:13.119 --> 00:28:20.720
So another way to look at that is uh a delightful experiment by uh my friend Andy Runa in Cornell.
00:28:21.119 --> 00:28:23.039
And what he did was very interesting.
00:28:23.200 --> 00:28:35.119
He took a ladder made of steps at different at different angles, two ladder exactly the same or symmetric with respect, and he let one fall on a table and the other one fall freely.
00:28:35.839 --> 00:28:42.319
And so the question is what's the effect of the table on the motion of the ladder?
00:28:42.559 --> 00:28:47.039
Well, it's very strange because there is no reason if something falls on a table, it changes.
00:28:47.200 --> 00:28:57.839
Except if you have the same effect, as it falls on the table, the each step in the ladder would accelerate and provide a force that pulls the ladder.
00:28:58.000 --> 00:29:00.240
And so you did this beautiful experiment.
00:29:00.559 --> 00:29:04.319
You have a free-falling one on the right and the table on the left.
00:29:04.480 --> 00:29:12.480
And what you see, they started exactly at the same point, is that each of these steps is pulling the ladder down and accelerating.
00:29:12.640 --> 00:29:17.599
So the one on the left goes faster than gravity by this effect.
00:29:17.759 --> 00:29:22.240
You have an acceleration due to the shock, which is quite surprisingly.
00:29:22.559 --> 00:29:24.960
Mechanics is infinitely surprising.
00:29:25.200 --> 00:29:32.720
Okay, so we have all these uh effects, some interesting, some strange, and we really want to make sense of that.
00:29:32.880 --> 00:29:43.200
And now we get to the next step, which is to formulate uh what in the trade we call a toy model, because we like to play with it, not just the toys that I show you.
00:29:43.440 --> 00:29:50.160
A toy model is a minimal uh mathematical model that has all the uh elements of that.
00:29:50.400 --> 00:30:05.519
You see, in the previous uh uh computation I did, I assume I knew everything about the motion, and in that case, whether you decide that linear momentum or energy is conserved, you get different answers because they cannot be conserved for the motion ahead.
00:30:05.759 --> 00:30:10.640
In reality, you really have to solve for the motion and see what the consequence are there.
00:30:11.119 --> 00:30:16.000
But solving for the motion is difficult because this is a rather complicated structure.
00:30:16.079 --> 00:30:19.119
So you want to idealize it, you want to model it.
00:30:19.279 --> 00:30:27.839
And the first thing I'm going to say is that, well, the motion is mostly in the plane, as I crack it, even if I go overhead, it's motion in the plane.
00:30:28.000 --> 00:30:34.400
So I'm going to look at a rod, a string in one dimension, uh in two dimensions in the plane.
00:30:34.480 --> 00:30:37.119
So a one-dimensional continuum in the plane.
00:30:37.279 --> 00:30:52.400
Now, this object is right away much more complicated than what we've seen, because now I look at each point, x, y, and I parameterize it with the lengths, the arc length from, let's say, the swivel up to the point that it reaches.
00:30:52.640 --> 00:31:03.519
So the position of, if I paint a little red dot like here, like I did on the on the slide, this position of x and y change as a function of time.
00:31:03.680 --> 00:31:05.759
But each dot change also.
00:31:06.079 --> 00:31:12.559
So x and y are not both function of the arc length, that gives me the position, and time.
00:31:12.880 --> 00:31:15.359
So it's a function of two arguments.
00:31:15.680 --> 00:31:36.960
But you can then express essentially f equals ma with this continuum, and you get uh you get the first set of equations, which is the balance of linear momentum, which tells you that an acceleration of a little part is balanced by the forces that are inside, uh that it feels inside the material.
00:31:37.200 --> 00:31:44.559
And that's these two equations, F and G, are associated with the force of inside the material that pushes it.
00:31:45.359 --> 00:31:51.599
Now you have another equation, which is the equation for the angular momentum, which is also important.
00:31:51.759 --> 00:31:58.160
What you need now, if I look at a little piece of my material, it can also change its orientation.
00:31:58.319 --> 00:32:05.039
That provides a torque, and the torque is related to angular momentum, so I need also to balance that.
00:32:05.359 --> 00:32:10.480
That gives me a second uh third equation, and that gives me three equations for that.
00:32:10.559 --> 00:32:12.000
Don't worry about equations.
00:32:12.160 --> 00:32:14.640
T's are actually very complicated equations.
00:32:14.880 --> 00:32:18.319
T's are three non-linear partial differential equations.
00:32:18.559 --> 00:32:24.880
So when I started this Gresham job, I was told don't show differential equations because this is too complicated.
00:32:25.599 --> 00:32:30.000
And they were talking about ordinary differential equations, which is the baby version of this.
00:32:30.319 --> 00:32:33.119
T's as as complicated equation as you can get.
00:32:33.279 --> 00:32:41.440
And they are complicated because they describe extremely rich all the possible motion of a chain or a rod in in the plane, right?
00:32:41.599 --> 00:32:46.480
So they have a lot of solutions, this extremely rich set of equations.
00:32:46.880 --> 00:32:55.279
So this is a minimal model because now we can solve for both velocity and the shape at the same time if we know to solve this equation.
00:32:55.440 --> 00:32:57.759
So forget a moment of this equation.
00:32:58.000 --> 00:33:07.519
Now I have my back of trick of methods from applied mathematics, and I know how to get that, uh, extract information of that.
00:33:07.759 --> 00:33:11.519
And the first result is that loops exist.
00:33:11.759 --> 00:33:17.440
They are solutions of the equation, and actually it's more or less something that Euler knew about.
00:33:17.599 --> 00:33:30.400
And what I mean by that is that if I look at a uh a loop like this one, I know that these equations support the same type of solution on an infinite domain, things that go to zero on both sides.
00:33:30.559 --> 00:33:41.680
And more important, the loop can have any height, and I have also a nice little formula for it that is not important for us, but I can actually write the solution exactly for that.
00:33:43.519 --> 00:33:46.400
Now, that's a loop being static.
00:33:46.640 --> 00:33:52.240
The second result, and a bit of a miracle, is that any of this loop can be made to travel.
00:33:52.400 --> 00:33:58.240
And that came more recently, this type of idea of looking now at a dynamical solution.
00:33:58.400 --> 00:34:09.440
That means if I start with something like that on an infinite domain with no variation of tapering, there's a solution that makes it roll further.
00:34:09.519 --> 00:34:11.519
And you're gonna see what I mean by that.
00:34:11.760 --> 00:34:18.079
This is the solution, and I can make it travel depending on the initial condition at any velocity.
00:34:18.239 --> 00:34:21.199
So an infinite light, a little loop like that can travel.
00:34:21.280 --> 00:34:23.599
It's part of the set of solutions of this.
00:34:24.159 --> 00:34:26.239
And why is that interesting?
00:34:26.400 --> 00:34:30.079
Well, for a problem, it's not that interesting because it's constant velocity.
00:34:30.159 --> 00:34:39.119
So in this case, if I send a loop with a velocity c, it would just go velocity c forever, if there is no friction or anything like that.
00:34:41.599 --> 00:34:43.920
However, the whip is stapled.
00:34:44.320 --> 00:34:51.599
And now we can look at the effect of the tapering on the whip, and and that becomes much more interesting.
00:34:51.920 --> 00:35:03.840
And now T's loop now is stapled, and you have to now solve the equation for T's problem where the mass is actually changing, the radius is changing as it goes along.
00:35:04.000 --> 00:35:11.679
And that's that's a problem that hadn't been solved before, and I did it with uh Tyler Macmillan, who was my uh PhD student at the time.
00:35:11.840 --> 00:35:15.519
And it's very much like a problem, uh like a wave coming ashore.
00:35:15.599 --> 00:35:21.519
And if the beach is a little slanted as it comes around, you have a variation due to the fact that the depth change.
00:35:21.760 --> 00:35:29.039
So it's part of this type of problem where you have propagation of a wave on a changing medium, and there are a lot of nice applications.
00:35:29.280 --> 00:35:34.480
But in the context of this problem, it hadn't been done, so we looked at that and we solve it.
00:35:34.639 --> 00:35:37.199
And the problem is the following, going back to energy.
00:35:37.280 --> 00:35:39.599
Let's think about energy for a second.
00:35:39.840 --> 00:35:49.920
It says the physicists tell us that uh the the motion is completely confined, but if I have a loop, there's plenty of ways I can conserve energy.
00:35:50.159 --> 00:35:59.199
The loop, there is less mass traveling, it's true, but the loop can get tighter and have more elastic energy by being more curved.
00:35:59.840 --> 00:36:05.039
That would conserve the and go at constant velocity, that would conserve the energy.
00:36:05.280 --> 00:36:18.880
The loop could get bigger as it travels, because there is more mass more mass moving, it is still a solution at constant velocity, or the loop could stay the same and increase in velocity.
00:36:19.039 --> 00:36:20.639
You don't know that a priori.
00:36:20.880 --> 00:36:25.119
Nothing is telling you what happens, and you have to do the computation.
00:36:25.360 --> 00:36:37.519
And what the computation showed us is that essentially there's small variation of height, but the loop actually accelerates but remains the same height essentially.
00:36:37.679 --> 00:36:45.840
And you can now, as a function of the tapering, you can actually find exactly what the velocity is in the system.
00:36:49.840 --> 00:36:59.280
Of course, you're gonna tell me, well, wait a second, this is not an infinite rod on which a loop travel, no, it's a finite piece, but we have this effect.
00:36:59.440 --> 00:37:10.480
And the other objection would be well, if I'm if I if I do it well enough, I can take and I can do it also, I I could take a non-taped uh whip and make it crack.
00:37:10.639 --> 00:37:16.000
It provides a little bit more energy and all that, but if you know what you're doing, you could also make it cry.
00:37:16.159 --> 00:37:21.280
So tapering is not the whole story, and you want to go to the next step.
00:37:21.440 --> 00:37:28.480
Let's see, in more realistic conditions, now uh what happens exactly in the system.
00:37:28.639 --> 00:37:40.000
The good thing is that you still have all the equations, but as far as doing them analytically, pure mathematics, getting formula and all that, that's probably the best you could do, is what I showed you.
00:37:40.239 --> 00:37:45.599
What you can do, however, is develop numerical methods to solve this equation.
00:37:45.840 --> 00:37:48.719
That's also part of the bag of tricks that we have.
00:37:48.960 --> 00:37:58.559
And using these methods, you can start with an initial loop, as I show you, and now look at the effect of having an open end at the end rather than an infinite one.
00:37:58.719 --> 00:38:13.920
And then you can vary tapering, you can vary the length, you can vary the force, the initial velocity, and you can exactly point out what are the different effects that matters in order to get an acceleration at the tip.
00:38:14.800 --> 00:38:35.519
And you can you can tabulate and you can show that all these effects are somewhat important, both tapering, uh, force, initial motion, and all that, exactly like you would if you practice the whip, you would have the right coordination to send the right wave and pull it, put the right force at the right time, and you get very quick acceleration.
00:38:35.679 --> 00:38:45.280
And since you have the whole motion, the whole motion of the whip as it as it goes around, you can now look at the velocity at each point.
00:38:45.599 --> 00:38:53.119
And the velocity, look at the tip speed, for instance, follows very much the type of experiment that you saw.
00:38:53.599 --> 00:39:13.760
And so if now you look at uh the chain experiment on the beaker, the one that's nap, I don't know if you can see it well on the screen, there is a yellow curve, the the black dot is the experiment, and the yellow curve is the numerical solution of the same type of equation that I show you.
00:39:14.000 --> 00:39:17.920
And as you see, the the fit is actually extraordinary.
00:39:18.000 --> 00:39:20.159
It's not a fit, there is no free parameter.
00:39:20.320 --> 00:39:26.559
This is just a perfect prediction of the motion of the chain as you go along.
00:39:26.800 --> 00:39:35.760
And because you have that, you can now know from the numerical solution, the one the computer gives you, you have all the possible information.
00:39:35.920 --> 00:39:38.639
In particular, you can look at the curvature.
00:39:38.800 --> 00:39:51.679
And when you look at the curvature, you can look at the place where the radius is maximal, uh, the so the smallest possible curvature, the way the place where, as it unfolds, the chain is most curved.
00:39:52.400 --> 00:39:56.320
And its inverse is the radius, is the best-fitting circle at that point.
00:39:56.480 --> 00:40:04.880
And as it goes through this transition, there is a critical time where it becomes flat, and that means the radius of curvature goes to infinity.
00:40:05.119 --> 00:40:09.920
And that's what is called a singularity, and that's where the snap occurs.
00:40:10.320 --> 00:40:14.800
And this type of singularity is very interesting, it hasn't been completely resolved.
00:40:14.880 --> 00:40:22.159
It's a good place to study, but exactly how it goes to infinity is still a bit of an open, open question for us.
00:40:22.800 --> 00:40:32.320
But so far, all I've shown you, it was a toy model, did not include the motion, but no interaction with the air.
00:40:32.480 --> 00:40:37.599
And the crack sound that we hear is really a change in air pressure.
00:40:37.760 --> 00:40:51.119
So I want to go back to uh supersonic motion, and as you can see in Arizona, not only we have WIP, we have road runners, and maybe they're fast enough to be supersonic, who knows?
00:40:52.000 --> 00:41:03.920
So the first thing that I mentioned was that little puzzle, a little surprise of uh the tip loop, uh the velocity of the tip as it cracks.
00:41:04.639 --> 00:41:12.079
So remember that when you look at the data, you can show that the whip cracks when the tip reached twice the speed of sound.
00:41:12.400 --> 00:41:26.079
At first that seems a little bizarre to us, and then we realize the following is that in a motion where you have a loop, the motion of the loop, the velocity of the loop, is different than the velocity of the point on the loop.
00:41:26.400 --> 00:41:29.199
So let me give you a much simpler example.
00:41:29.360 --> 00:41:40.400
If I take a bicycle wheel and I roll it along a wall at a velocity V, you can ask what is the velocity of the material point right at the top of the wheel?
00:41:40.719 --> 00:41:46.480
Well, the velocity at the top of the wheel must be greater because that's why it goes in front of the wheel.
00:41:46.719 --> 00:41:47.119
Right?
00:41:47.360 --> 00:41:59.599
And if I roll it at constant velocity, you can show by nice geometric argument that the velocity of that point at the top is always twice the velocity of the wheel.
00:42:00.079 --> 00:42:00.639
Perfect.
00:42:00.800 --> 00:42:02.880
And that's true for any shape, actually.
00:42:03.039 --> 00:42:09.760
If you have a shape that travels at constant velocity and like that, the top will be always twice.
00:42:09.840 --> 00:42:13.039
So it's not a property of the physics, it's really a property of the geometry.
00:42:13.119 --> 00:42:20.719
So if that shape doesn't change, the loop goes at velocity c and the top of the loop goes at velocity 2c.
00:42:21.199 --> 00:42:43.920
So as it unfolds, you have C prime, it goes faster and faster because of tapering or other external extra tension or things like that, up to the point where the loop itself, now the loop itself, so you have a loop, and the loop itself reach the velocity of the sound, but the tip that's is being protected by the loop.
00:42:44.000 --> 00:42:46.960
The tip does not go against the air.
00:42:47.199 --> 00:42:53.119
So it's the loop that breaks the sound barrier, not the tip that breaks the sound barrier.
00:42:53.280 --> 00:42:55.440
The tip comes after and unfolds.
00:42:55.920 --> 00:43:04.480
And the way to see that is to know back to the data when you realize that you say, hmm, let's see if we can actually see that on the data.
00:43:04.639 --> 00:43:21.599
And if you look at the experiment where you see it is supersonic motion, here's the speed C, and you see this is where the supersonic shock is being created, the supersonic wave, it's the loop, not the top of the tip that at velocity 2c as it unfolds.
00:43:21.760 --> 00:43:25.920
And then you have further uh uh expansion.
00:43:26.480 --> 00:43:36.559
Now, supersonic motion is very interesting, very difficult, because you have to understand the physics of air and compression and jump and all that in very extreme conditions.
00:43:36.719 --> 00:43:38.800
But there are very simple things you can do.
00:43:38.880 --> 00:43:42.400
Uh, one of the simplest constructions is the MAC cone.
00:43:42.639 --> 00:43:59.599
So if you have a motion in the air, just by friction in the air, uh the that that point is going to uh send sound wave, just like you hear something going in inside a room if you throw it fast enough, is the sound that the air makes at its expense.
00:44:00.320 --> 00:44:11.360
But if that point is going supersonic, the sound waves do not have time to expand fully before they all coalesce on something called an envelope.
00:44:11.599 --> 00:44:33.519
And so if I have this sound wave emanating constantly from that moving point due to interaction with the air, if the velocity of my point is faster than the speed of sound, they all accumulate with an opening, the Mac Con opening, which is given, the sign of that is given by the ratio of the velocity of the sound by the velocity of the object.
00:44:33.760 --> 00:44:41.360
And this is what you see in Schlieren imaging for a plane or for a for a supersonic bullet or something like that.
00:44:41.519 --> 00:44:46.400
So for a plane, you might experience that you sometimes hear more than one boom.
00:44:46.559 --> 00:44:53.840
So the supersonic sound, the supersonic boom, is not when the plane gets past the sound barrier.
00:44:54.000 --> 00:44:57.679
It's something that's constantly emitted by ulti-sound wave gathering.
00:44:57.920 --> 00:45:05.840
So at that point, if you sit right at the intersection there, it's when that red line crosses you that you hear the sound.
00:45:06.159 --> 00:45:27.760
But a plane is an object of finite size, and so there are a lot of different waves being emitted by different pieces of the plane, and you have one main one that's from the tip of the plane that creates the first one, and you have a change in air pressure, and then you have the opposite change in air pressure at the end of the plane.
00:45:27.840 --> 00:45:28.639
You have a change.
00:45:28.800 --> 00:45:36.320
And so that's why you can hear a double double boom or multiple boom as supersonic plane goes by.
00:45:36.800 --> 00:45:46.960
But there is a huge set theory about uh supersonic motion, or bullet or planes or space and all that.
00:45:47.199 --> 00:45:56.719
But almost all of them, as far as I know, all of them were for uh rectilinear motion, because usually people, plane goes go straight, or bullet goes straight.
00:45:56.880 --> 00:46:06.639
And there is no equivalent what if the motion actually uh is not straight but circular, just like is as the whip cracks, it unfolds.
00:46:06.880 --> 00:46:08.400
But it's not too hard to do.
00:46:08.480 --> 00:46:12.400
You can play the same geometric trick and you can have a point.
00:46:12.480 --> 00:46:15.920
Now think about the the cracker, the loop as it unfolds.
00:46:16.159 --> 00:46:35.599
Now it rotates rather than going straight, so it follows more or less a piece of a circle, and that piece of a circle also emanates waves, and you see now that you have a curved macode depending on the trajectory, and you can now compare that to the data.
00:46:36.079 --> 00:46:50.400
And so by putting all these um different pieces of physics and mathematics and all that, you start having a very good picture of all the different effects and a good understanding of whip cracking.
00:46:50.559 --> 00:46:52.719
Uh that's that's quite nice.
00:46:53.039 --> 00:47:00.880
Of course, you can say, well, that seems a little bit frivolous, putting all your effort into uh into that kind of problem.
00:47:01.039 --> 00:47:02.320
What is it good for?
00:47:02.960 --> 00:47:12.239
Uh well, first I want to say that it's not a question that people ask to people who study black hole or or the origin of the universe and so on.
00:47:12.559 --> 00:47:14.800
But we do have a responsibility.
00:47:14.960 --> 00:47:16.079
What is it good for?
00:47:16.480 --> 00:47:26.559
So this example is of course very uh catchy because it's very extreme, but it's actually part of a whole body of problem for which we develop methods.
00:47:26.639 --> 00:47:31.199
It turns out that this one is one that that captivates the attention.
00:47:31.519 --> 00:47:34.719
So we get to the problem or the question, so what?
00:47:35.199 --> 00:47:38.480
So what else can you do with that in a way?
00:47:38.719 --> 00:47:45.199
Just explain the cracking of the whip or chain, this this is good, but that's not what we pay you for, right?
00:47:45.360 --> 00:47:47.440
We pay you to help society.
00:47:47.679 --> 00:48:07.039
Well, there are plenty of other problems when a cable snaps or something like that, but if you look at the smaller scale, there are microorganisms, for instance, any kind of flagellate organism that uh beat the flagellum and in doing so provide linear momentum that's expelled for the motion.
00:48:07.119 --> 00:48:10.480
So this one is not a little uh space lunar module.
00:48:10.559 --> 00:48:20.239
If you look at the scale, this is three micron, it's a very uh nice uh uh organism called a coanoflagellate that has this specific tail that flips.
00:48:20.400 --> 00:48:27.440
And these are very close in some aspects to actual sperm and the flagellum of the sperm.
00:48:27.679 --> 00:48:36.559
And indeed, sperm do create this whipping motion by making loops and pushing them in order to propagate.
00:48:36.719 --> 00:48:49.599
And all of a sudden, you went from the whip to understanding something about motility of the sperm, which is a very important problem view both on physiology and biology.
00:48:49.840 --> 00:48:53.119
And we've applied the same type of problem techniques.
00:48:53.280 --> 00:49:04.480
Again, you completely have to refine your model because nautics operate in very viscous environments, so you have to take into account the interaction with the fluid, the viscosity on one, the other one, and so on.
00:49:04.639 --> 00:49:22.079
But using that we manage with uh Emman Gaffney, which is in Oxford, and and uh Hermes Gadea, who is now in Bristol, uh, to explain some effects that are observed in in human sperm uh related to uh what is called the counterband phenomena.
00:49:22.400 --> 00:49:30.000
Another cool example uh that uh I that that I found is uh prey catching in octopus.
00:49:30.800 --> 00:49:39.840
And here what you have is the octopus very quickly unfolding and doing this nap in order to catch a prey that is uh indicated.
00:49:40.000 --> 00:49:47.039
And that as you can see from the scale, this this happens in about uh 400 milliseconds.
00:49:47.920 --> 00:50:00.000
And one of the uh last examples that you that you uh can see also, like nice application, is a much bigger scale.
00:50:00.159 --> 00:50:02.400
Uh for instance, the space tether.
00:50:02.719 --> 00:50:13.199
So satellites have these tethers, which are long rods that uh uh have a weight at the end, and they use that either to stabilize or change their orbits.
00:50:13.440 --> 00:50:17.920
It's the the the mechanics of that is very nice in terms of orbital motion.
00:50:18.159 --> 00:50:22.880
But there is one problem is that these tethers sometimes have to be retracted.
00:50:23.280 --> 00:50:35.280
And when you retract the tethers, all the shape and the motion gets concentrated as you retract the cable, you call you coil the cable in smaller and smaller portion.
00:50:35.440 --> 00:50:48.800
And now in space there's very little friction, so that you don't have viscosity and all that, and so what happens is that as you coil it, the motion of the end gets wider and wider.
00:50:49.039 --> 00:51:06.239
Very much like when you eat spaghetti and you slurp a spaghetti, and what happens is that all the energy that you do for the velocity of slurping is being transferred, or the energy is being transferred in a smaller and smaller portion.
00:51:06.400 --> 00:51:10.639
And if you have that with uh pasta saus, you know what happens.
00:51:10.880 --> 00:51:13.119
It oscillates and it splashes on you.
00:51:13.280 --> 00:51:15.599
And that's exactly the same phenomena that happens.
00:51:15.679 --> 00:51:26.800
I've never worked on space tethers, but they're very interesting because in in space the problem is that you it's very hard to uh to stop things from happening.
00:51:26.960 --> 00:51:32.159
The motion of objects in space is very much uh inertial, that means not damp.
00:51:32.400 --> 00:51:37.199
There is very little dissipation that you can, but and that creates uh other problem.
00:51:37.519 --> 00:51:55.360
So I hope that I've given you a global picture, starting with a very simple problem and bringing up to a more complicated one, but one for which we have no answers and uh different application to the rest of the world on that.
00:51:55.519 --> 00:51:56.800
Thank you very much.
00:00:06.000 --> 00:00:07.519
Thank you very much, Jane.
00:00:07.599 --> 00:00:10.080
Thank you all for coming and people watching online.
00:00:10.640 --> 00:00:12.960
So we're gonna have fun today.
00:00:13.359 --> 00:00:16.000
Well, I'm going to have fun for sure.
00:00:16.480 --> 00:00:21.760
This year I want to uh talk about what I call the mathematics of surprise.
00:00:21.920 --> 00:00:36.719
What I will do is take a simple observation, simple phenomena, and see trying to understand it both on physical ground but also using mathematics and then see how that knowledge expand to other areas.
00:00:37.119 --> 00:00:40.880
And here's one of my favorite subjects, which is cracking the whip.
00:00:41.439 --> 00:00:51.439
But before we do that, I want to talk about something called the scientific methods, mostly organized by people who don't do science, as far as I can tell.
00:00:51.600 --> 00:00:54.560
And they are different versions, but they mostly look like that.
00:00:54.719 --> 00:01:05.599
You have an observation, a question, then you research a topic area, you formulate an hypothesis, you test it, you analyze the data, you report the conclusion and so on.
00:01:06.000 --> 00:01:08.239
And it feels a little boring, right?
00:01:08.400 --> 00:01:11.439
There's no like you wreak a moment, no excitement.
00:01:11.599 --> 00:01:19.760
And frankly, in my entire life, all the science I've ever done, and the people around me, nobody has ever followed this kind of wrong, right?
00:01:20.159 --> 00:01:24.159
So let me tell you about the other scientific methods.
00:01:24.480 --> 00:01:28.400
And usually it starts with observing something bizarre.
00:01:28.560 --> 00:01:32.560
You are in a field, and something is not quite what you expect.
00:01:32.719 --> 00:01:35.359
It doesn't match your understanding of that area.
00:01:35.439 --> 00:01:38.640
It's something extra, and you want to try to understand.
00:01:38.879 --> 00:01:40.239
So what do you do?
00:01:40.480 --> 00:01:43.280
Well, you gather useful and useless facts.
00:01:45.359 --> 00:01:49.599
Well, you don't know if they might be useful, you're just early days, right?
00:01:49.760 --> 00:01:53.359
So you go around and read books and watch TV, whatever.
00:01:53.519 --> 00:01:56.000
You do whatever you can to try to get.
00:01:56.159 --> 00:01:59.680
And then you turn to what I call back of the envelope computation.
00:01:59.760 --> 00:02:06.560
You say, okay, let's do some estimate to see if I get the right mechanism, maybe some order of magnitude to see how things work.
00:02:06.719 --> 00:02:10.400
And usually it's not enough, but that gives you an idea you're on the right track.
00:02:10.639 --> 00:02:16.879
And then you build a minimal mathematical model, what we call a toy model in the trade.
00:02:17.039 --> 00:02:23.039
Something that has all the components that is as simple as possible but not simpler, as Einstein would say.
00:02:23.199 --> 00:02:25.919
So it still captures something and you can learn from that.
00:02:26.319 --> 00:02:33.199
So here is a job both of modeling, and we talk about modeling, but also your job as a mathematician.
00:02:33.280 --> 00:02:37.120
No, you have a model, you have actual equation or something to that effect.
00:02:37.280 --> 00:02:50.479
You need to study the model, find solutions, see how the solution behave, maybe do numerical solutions on the paper, and so on, and you go around trying to get extract information of this simple equation.
00:02:50.719 --> 00:02:54.000
And after that, you go for a reality check.
00:02:54.319 --> 00:02:55.919
Does that match what you see?
00:02:56.080 --> 00:02:57.520
Does that match anything?
00:02:57.759 --> 00:03:01.199
And maybe it does, and then the big question, so what?
00:03:01.280 --> 00:03:02.159
What have we done?
00:03:02.400 --> 00:03:02.719
Right?
00:03:02.960 --> 00:03:07.039
And that you mostly don't by then, and you can write a paper, something like that.
00:03:07.199 --> 00:03:09.280
But of course, nothing is linear.
00:03:09.439 --> 00:03:13.599
And when you do your reality check, typically you start gathering other facts.
00:03:13.759 --> 00:03:20.560
Maybe you do and do little experiments or you try to find data, which forces you to maybe revise your model.
00:03:20.719 --> 00:03:24.319
Maybe you're missing an important mechanism in the model and so on.
00:03:24.639 --> 00:03:33.439
And after you think about it, you think, well, maybe there are other things that are like that and maybe a little different that follows the same pattern.
00:03:33.919 --> 00:03:38.879
And that forces you also to maybe write other models for other circumstances.
00:03:39.039 --> 00:03:46.159
And as soon as you gather facts, of course, you study the model, you have solutions, you have to compare them and so on, and it gets very messy.
00:03:46.240 --> 00:03:50.879
And every day you wake up and you're in one of these boxes and you jump from one to the other one.
00:03:51.840 --> 00:03:56.560
So today we start with something bizarre, right?
00:03:56.879 --> 00:03:59.919
And the something bizarre is the whip cracking.
00:04:00.080 --> 00:04:04.400
And it was bizarre to me as when I first thought about it.
00:04:04.639 --> 00:04:13.919
So I spent a long time in the southwest in Arizona, but that's not where I first uh realized that there was something interesting with the whip.
00:04:14.319 --> 00:04:21.439
It was in a meeting uh in Budapest in Hungary, and you know, uh there is usually a conference dinner in this meeting.
00:04:21.519 --> 00:04:25.839
It's a rather boring affair, big group taking a long time, drinking too much.
00:04:26.000 --> 00:04:39.279
But on this occasion, there was a folkloric band from Hungary coming and doing a little show with whip cracking, and I was there ignoring my fellows and looking and said, hmm, that's bizarre.
00:04:39.519 --> 00:04:40.720
Wonder what is this?
00:04:40.879 --> 00:04:45.519
You know, what is that so distinctive sound, a whip cracking?
00:04:45.759 --> 00:04:55.360
I knew a little bit of it before, but I got interested because at the time I was interested in dynamics of curve, and it didn't quite fit my understanding of that.
00:04:55.839 --> 00:04:58.000
So I started looking into that.
00:04:58.319 --> 00:05:15.360
But before we start, and to avoid any confusion, despite the movie Zoros and Indiana Jones, I want to let you know that the only, the one and only purpose of long whips, the one I'm going to uh talk about, is to make that sound, that crack.
00:05:15.600 --> 00:05:19.839
It's never been a weapon, it's never used for the purpose kinky otherwise.
00:05:20.000 --> 00:05:26.000
These whips, the one I'm going to show you, are only there to uh to create that whip.
00:05:26.399 --> 00:05:29.839
So I went home and it's time to gather facts.
00:05:30.399 --> 00:05:33.040
Okay, uh, or at least to try things.
00:05:33.279 --> 00:05:38.399
So the first thing I got, I got this uh big boy here, eight-footer.
00:05:38.560 --> 00:05:41.439
You know, it has a nice structure, we'll talk about that.
00:05:41.600 --> 00:05:45.759
And I started practicing, and I bought the bull whip book.
00:05:45.920 --> 00:05:49.279
Uh it was a time where you could learn skills from books.
00:05:49.839 --> 00:05:53.360
That was before YouTube, before even Google and stuff like that.
00:05:53.519 --> 00:05:59.680
And you learn in the books, very interesting, that there are three uh typical motions when you do that.
00:05:59.920 --> 00:06:04.000
There is three phases the setup, the turn, and the follow-through.
00:06:04.240 --> 00:06:11.360
So if you give uh a whip like this to somebody and you ask them to crack, they're probably gonna try to do something like that.
00:06:11.519 --> 00:06:14.079
So this one is too big to do here.
00:06:14.240 --> 00:06:19.279
So here I have a smaller one that I just got for for this purpose here tonight.
00:06:19.439 --> 00:06:20.879
It's only a four-footer.
00:06:21.040 --> 00:06:26.800
And so if you're given a whip, the first thing you'll try is what is called a snap, which is something like that.
00:06:27.120 --> 00:06:32.319
Yeah, you have to put a lot of force and you get a little crack, nothing impressive.
00:06:32.480 --> 00:06:32.800
Okay.
00:06:33.279 --> 00:06:36.800
And it turns out that yes, it works, you can do that.
00:06:37.360 --> 00:06:41.360
Um, but it's not the real way you really want to crack a whip.
00:06:41.680 --> 00:06:47.360
The way you want to crack a whip is to make a loop and let the loop propagate along.
00:06:48.480 --> 00:06:56.720
So I'm gonna show you have videos just in case it doesn't work here, but the the setup is very simple.
00:06:56.879 --> 00:07:02.240
You raise the whip up to a certain height and you turn it to make a loop that goes at the end.
00:07:02.480 --> 00:07:02.800
Right?
00:07:03.120 --> 00:07:08.800
And so if you do it like you see, right away, you create a crack with very little effort.
00:07:09.040 --> 00:07:11.040
And that gets very interesting, right?
00:07:11.279 --> 00:07:16.399
Because you send a loop and that loop accelerates and then crack at the end.
00:07:16.800 --> 00:07:29.600
Very distinctive, and so distinctive that it's not a normal sound, and that's why we use it, because animals usually never hear that kind of sound and they react to it, and that's what controls horses and cattle.
00:07:29.759 --> 00:07:40.480
And horses being the one method of transport to the 19th century, there was a huge, huge business associated with whip and cattle and ranching and all that.
00:07:40.720 --> 00:07:46.560
So, just in case it didn't work, I did my own uh experiment to show you.
00:07:46.720 --> 00:07:52.079
Uh, as you can see, I'm not only a scientist, I also play one on TV.
00:07:53.040 --> 00:08:02.639
So that's the whip, and here is a slow motion of the same action, and you can almost you can see the loop going through, right?
00:08:02.720 --> 00:08:06.560
It's high speed, and great, it works.
00:08:06.800 --> 00:08:14.639
Here is a uh younger version of myself in Arizona, and here is another way, it is this with this one.
00:08:14.879 --> 00:08:16.480
I'm not gonna do it here.
00:08:16.720 --> 00:08:20.319
I could almost maybe do it, but maybe I'll get in trouble.
00:08:20.560 --> 00:08:35.519
Uh and the idea is that you turn the whip around, so you create tension, that's a setup, and then you invert the motion to create that loop, and that loop zips to the end, and by the time it ends, it unfolds and cracks.
00:08:35.759 --> 00:08:37.679
So here is how it goes.
00:08:39.200 --> 00:08:42.559
You can see that the neighboring dog doesn't like it.
00:08:42.720 --> 00:08:48.240
And you can see here, you can see the loop forming there and going around.
00:08:48.480 --> 00:08:56.799
So you find in popular science book that the crack, that special cracks, occurs when the tip reaches the speed of sound.
00:08:57.600 --> 00:09:01.200
At the type where people were still looking for books and stuff like that.
00:09:01.519 --> 00:09:04.559
So this is supersonic motion.
00:09:04.799 --> 00:09:11.519
This is the first time humans manage to make supersonic motion, it's through the motion of the cracking.
00:09:11.679 --> 00:09:15.519
And we're gonna go and look at that in quite uh detail.
00:09:16.320 --> 00:09:21.440
But there are other things you can do, not with whip, but with chain or with whip actually.
00:09:21.519 --> 00:09:25.919
Uh, and I wanted to show you there's something else bizarre called the chain fountain.
00:09:26.080 --> 00:09:31.600
And that's something that's done with a regular chain, the one you use to turn on the light.
00:09:31.840 --> 00:09:44.559
And I would again try to do a little experiment, it's not sure to work, but it's not for a dramatic purpose, it works better if the height is uh actually higher, but also for dramatic purpose.
00:09:44.799 --> 00:09:49.919
If you want, I can put my glasses on so that look more serious and more responsible.
00:09:50.639 --> 00:09:52.159
Okay, so let's see.
00:09:52.320 --> 00:09:55.679
Let's get one and two and three.
00:09:56.000 --> 00:10:01.120
So all I'm going to do is I have a chain and I'm gonna release it.
00:10:01.519 --> 00:10:08.240
And you can see all yeah, the idea it goes, well, it's not very impressive, this one.
00:10:08.399 --> 00:10:09.759
But I have a high-speed move.
00:10:10.000 --> 00:10:13.519
I have a movie, and you'll see exactly what I mean with it.
00:10:16.399 --> 00:10:27.360
You see it's the same beaker, the same chain, and you have this chain going up, and very satisfied with myself, of course.
00:10:27.679 --> 00:10:33.679
I have my uh trademark thumbs up.
00:10:35.039 --> 00:10:43.440
Okay, so I also took the same chain and went into we have a little lab in the math department in in Oxford.
00:10:43.600 --> 00:10:51.519
We call it the uh mathematical observatory, where we have a very good high-speed camera, and I filmed that in high speed.
00:10:51.679 --> 00:10:55.600
And here you really see the effect, the chain going up.
00:10:56.240 --> 00:11:08.960
And what's amazing is that it doesn't touch the beaker, the whole thing goes and unravel, goes up and so this was popularized by Steve Mole in 2013 in a series of YouTube videos.
00:11:09.039 --> 00:11:13.279
You can really see the effect on popular science now of the arrival of YouTube.
00:11:13.600 --> 00:11:22.399
And the same year, uh John Biggins and Mark Warner from Cambridge wrote a first paper and called it the chain fountain, trying to give the first explanation for it.
00:11:22.639 --> 00:11:33.919
So that's also a very bizarre phenomena, it's quite beautiful, especially in slow motion, where you have this elevation of uh of the chain going up uh with no good reason.
00:11:34.000 --> 00:11:35.519
There is nothing pulling it.
00:11:35.679 --> 00:11:41.679
And so we want also to try to understand T and uh in the same context.
00:11:41.919 --> 00:11:47.200
So now it's time to turn and look at facts, you know, with practice, we've tried a little things.
00:11:47.360 --> 00:11:48.799
What do we know about whips?
00:11:49.120 --> 00:11:54.320
Well, one of the interesting things is that whips were with us for at least 3,000 years.
00:11:54.559 --> 00:12:05.759
If you look at this stone panel from uh Mesopotamia, Neo-Assyrian gypsum panel, you'll see that right there you have the whip used to uh direct uh horses.
00:12:06.000 --> 00:12:17.039
And through the year, whips have evolved uh technologically to rather sophisticated objects that typically have a handle, depending on the size of the handle.
00:12:17.200 --> 00:12:22.639
You can have a stock whip, a bull whip, or a snake whip, the one you you need to put in your saddle.
00:12:22.879 --> 00:12:30.240
You have a swivel, the place where you can turn, then you have a long part, the tongue, which is tapered and very flexible.
00:12:30.480 --> 00:12:41.679
And then it ends up with this part, which is called the fall, that direct the motion essentially to be flat, and always a little piece called the cracker that you can replace easily.
00:12:41.840 --> 00:12:47.360
That's the one that takes all the abuse and turns around and gets the most forces.
00:12:49.279 --> 00:13:02.080
So we think that we were the first to invent that, but some scientists in 97 uh thought that actually uh dinosaurs were way ahead of us.
00:13:02.320 --> 00:13:28.159
So when I was a kid, the apatosaurus were called the brontosaurus, and that means the the dinosaur of thunder, which was quite quite uh convenient, quite appropriate, because they did a little computation using a chain and showed that as the as the dinosaur whips its tail, the end of it reached supersonic speed, and that it might have used it for some purpose.
00:13:28.320 --> 00:13:30.480
So now you can speculate what it was.
00:13:32.000 --> 00:13:36.320
Uh and that that got a lot of people excited at the idea.
00:13:36.480 --> 00:13:52.320
Uh however, more recently, in uh four years ago, people did a more careful analysis with more like the vertebrae and all that, rather than a simplified physical model, and showed there is an acceleration but much more modest, only like 30 meters per second.
00:13:52.480 --> 00:13:58.480
So we don't know, but maybe it's unlikely that the dinosaurs were supersonic.
00:13:58.879 --> 00:14:03.360
So now what about on the science side?
00:14:03.519 --> 00:14:05.840
You play around, you can crack and all that.
00:14:06.080 --> 00:14:07.679
What is the knowledge there?
00:14:07.919 --> 00:14:22.320
Well, you have to wait until 1927 to have the first set of experiments that actually demonstrated that the the whip cracking was a supersonic motion, and that's from Zeferin Carrier, a French physicist from Toulouse.
00:14:22.559 --> 00:14:32.320
And he reasoned that in the motion of the whip as you go, as you do a snap, there is a piece that doesn't move, essentially the middle piece here, where you go around.
00:14:32.480 --> 00:14:47.200
And so if you want to do a control experiment, what you do, you replace that by a pulley, and now you have a pulley around, and all you need to do is put a weight here and let it unravel so that it accelerates with constant acceleration.
00:14:48.159 --> 00:14:54.879
So that was a very good idea, and you can imagine 1927 he actually managed to do high-speed photography.
00:14:55.120 --> 00:15:00.320
Must have been extremely complicated from an experimental point of view.
00:15:00.559 --> 00:15:04.080
But he had series of results, fascinating.
00:15:04.320 --> 00:15:15.600
He in red you see the pulley, and you see the result, the different stage of the chain or the of the string that he has going around and flapping around.
00:15:15.840 --> 00:15:31.039
And he also measured the velocity, and what he showed for the first time at that time was that the tip of the whip, or the tip of his string, was 900 meters per second, which is about three times the speed of sound in the air.
00:15:31.120 --> 00:15:31.360
Right?
00:15:31.600 --> 00:15:35.200
Speed of sound is 340 meters per second.
00:15:35.600 --> 00:15:42.559
And that's at that point, this is where the little cracker at the end goes from one side to the other side.
00:15:43.039 --> 00:15:47.039
So a few years ago, we thought, hmm, that's very interesting.
00:15:47.120 --> 00:15:49.919
And could we redo this type of experiment?
00:15:50.480 --> 00:15:59.279
And we thought that we simplify something in between the chain front end that I show you and the uh and the whip.
00:15:59.679 --> 00:16:12.720
And so what we did, we take a beaker like that and a string like this one, and we just added a weight on the string and drop the weight and see how this part moves around.
00:16:13.120 --> 00:16:16.879
Very much like the original uh experiment of Carrier.
00:16:17.360 --> 00:16:18.639
This is a movie of it.
00:16:18.799 --> 00:16:27.759
What you see, the big black uh circle is the is the beaker, and you see the motion, the chain being pulled, and this is what it looks like.
00:16:27.919 --> 00:16:29.360
That was the work with P.T.
00:16:29.519 --> 00:16:32.159
Barr, Basile, and Dominique.
00:16:32.399 --> 00:16:36.720
It was a lot of uh fun problem also to try to see that.
00:16:36.960 --> 00:16:47.759
And the interesting feature here is that it leaves also the beaker, and that means that the acceleration at one part of the chain is lower than the acceleration at the other part.
00:16:48.480 --> 00:16:52.559
It goes faster, that's why it lifts.
00:16:52.720 --> 00:16:57.279
And so you can ask why is this acceleration, extra acceleration coming from?
00:16:57.360 --> 00:17:04.400
And that's very much like the chain front, and as it goes around, this part has faster acceleration than the chain going down.
00:17:04.799 --> 00:17:15.440
And the other feature of interest is to look at what happened right at the crack, because the the chain or the or the whip is actually very curved there.
00:17:16.160 --> 00:17:25.119
And right at that point, if you look at the radius of curvature, it goes flat at some point, the radius of curvature goes to infinity.
00:17:25.200 --> 00:17:33.200
And that's very interesting because in mathematics and in physics, when you have this kind of behavior, something goes infinity, a lot of things simplify.
00:17:33.279 --> 00:17:34.400
It's a good place to study.
00:17:34.480 --> 00:17:39.119
It's called a singularity, and singularities are always uh interesting.
00:17:40.640 --> 00:17:44.000
So we're going to think about that, we come back to that.
00:17:44.319 --> 00:17:53.440
Another interesting paper was the paper uh in 1998 from German physicists, and they have this figure in the paper.
00:17:53.519 --> 00:18:05.839
Uh, a guy with kind of a, it turns out he's a Black Forest whip cracker, and they brought him in the lab with his costume and everything in order to do uh high-speed imaging.
00:18:05.920 --> 00:18:26.400
And they do they did beautiful high-speed imaging, not only looking at the the full motion of the tip, but on top of that they have something called Schlieren imaging, which captures the variation in the density of air, and that allows you to see uh supersonic crack as it's a supersonic wave emanating from the whip.
00:18:26.720 --> 00:18:35.759
And so you can exactly pinpoint when the crack happens and where it happens, which turns out to be uh important for a problem.
00:18:36.160 --> 00:18:37.839
So it is looks like that.
00:18:41.519 --> 00:18:47.920
And each frame is a hundred, about hundred uh microseconds, the whole movie is two milliseconds.
00:18:48.079 --> 00:18:58.240
And if you extract a few frames, what you'll see right at the time it cracks when the supersonic boom arises, it's right at these frames.
00:18:58.480 --> 00:19:00.640
And here I uh enhance them.
00:19:00.799 --> 00:19:04.960
This is what you see, the the supersonic wave coming out of it.
00:19:05.359 --> 00:19:10.319
If you had a bullet in with the same time of imaging, you'll see the cone, the MAC cone emanating.
00:19:10.400 --> 00:19:12.880
We'll see that a little bit later.
00:19:13.839 --> 00:19:17.279
So that's very interesting because they also had very good data.
00:19:17.519 --> 00:19:36.559
And what the data showed is that if you look at the velocity as a function of time, so that's the point here, a velocity a function of time, the acceleration of the tip of the tip, the cracker, right, is about 50,000 times the acceleration of gravity.
00:19:36.799 --> 00:19:42.960
So you get amazing acceleration, and deceleration is even faster, about 70,000.
00:19:43.200 --> 00:19:48.880
And you, since you have imaging, you can exactly know when the crack occurs.
00:19:49.519 --> 00:20:11.920
And the interesting things that they show is that the crack occurs right there, and the maximum speed at that time of the tip is twice, about twice the speed of sound, which is very strange because I just told you that the popular explanation is that a whip cracks when the tip reaches speed of sound, not twice the speed of sound.
00:20:12.160 --> 00:20:15.599
And so that's also a bit of a puzzle for us.
00:20:15.839 --> 00:20:18.799
Uh try try to uh understand that.
00:20:19.279 --> 00:20:36.880
So now we have a few facts, and for me the the basic or most interesting thing is to see how starting with a very simple motion, which is my arm one millip uh one meter per second, you can get to this tremendous acceleration and velocity.
00:20:37.359 --> 00:20:38.559
How is that possible?
00:20:38.960 --> 00:20:40.160
How does that work?
00:20:40.640 --> 00:20:45.200
What are the important effects, tapering or the way you do it, and so on.
00:20:45.519 --> 00:20:53.599
So we are part of now trying to throw estimate at that, back of the envelope computation.
00:20:54.559 --> 00:21:04.400
And the you'll find more for more than a century, actually, Mack himself did computation of this type, uh, what I call a naive computation.
00:21:04.559 --> 00:21:14.799
You you're going to decide that you know you know the motion of the whip, and that initially you it's the whole whip moves at a velocity u.
00:21:15.039 --> 00:21:21.039
So as I go like that and move, the whole thing, as it moves forward, has a velocity u.
00:21:22.559 --> 00:21:29.599
So the energy is the mass of the entire things times the velocity square divided by two.
00:21:29.839 --> 00:21:32.960
That's kinetic energy, one basic quantity.
00:21:33.200 --> 00:21:37.519
So we have the mass of the whip and the little m is the mass of the cracker.
00:21:37.599 --> 00:21:52.160
So we have one half big m, the mass of the whip of the of the tongue moving, not the not the stick, and uh starting at velocity u, so it's one half m plus little m u squared.
00:21:52.559 --> 00:21:55.680
And then you're going to postulate that energy is conserved.
00:21:55.839 --> 00:21:59.759
Yeah, there is some friction and things like that, but overall it's not a bad.
00:22:00.640 --> 00:22:12.960
But idea, if energy is conserved, then as you turn, there is a part of the whip that's not moving anymore, and the rest of the whip that goes further that goes above.
00:22:13.119 --> 00:22:17.680
So now a fraction of the whip is only moving as you do this turn.
00:22:18.079 --> 00:22:21.680
And now the whole thing goes at velocity v, that I don't know.
00:22:21.759 --> 00:22:32.960
And the velocity v depends on the position of how much is moving, that's decided by the length x, a fraction of the mass of that related, and the little mass m.
00:22:33.519 --> 00:22:42.559
And now what you do in physics is you say, well, the energy is conserved, so e is the same for both, and I can solve for the velocity as a function of x.
00:22:42.799 --> 00:22:50.240
And so I have the velocity, depending on the initial velocity, the mass, the little mass, and how much whip is actually moving.
00:22:50.799 --> 00:23:05.359
And that's that expression is not that interest important, but the interesting thing is when it reached the end, because when it reached the end, the whole the whip is not moving anymore, it's fully extended, only the little part, the cracker moves.
00:23:05.519 --> 00:23:06.960
And that's little m.
00:23:07.519 --> 00:23:07.920
Right?
00:23:08.079 --> 00:23:17.279
And so you get an expression for the maximum velocity as a function of the initial velocity times the ratio of the mass plus one.
00:23:17.920 --> 00:23:33.440
So it gives you an idea, and it gives you the basic physical mechanism of what is going on in this case in this system, that you start, you give the system energy, and that energy gets concentrated in a small and smaller part that's moving.
00:23:33.599 --> 00:23:37.200
And so the velocity has to increase since the mass decreases.
00:23:37.920 --> 00:23:42.880
However, if you start looking a little bit and say, okay, but what is the mass of the cracker?
00:23:42.960 --> 00:23:44.079
Are we talking the whole thing?
00:23:44.240 --> 00:23:48.079
What if I have a continuous whip that's just tapered, for instance?
00:23:48.240 --> 00:23:51.359
There is no little cracker, so the mass goes to zero.
00:23:51.519 --> 00:23:55.759
And that's fine, the mass goes to zero, but then the velocity of the tip goes to infinity.
00:23:56.079 --> 00:23:57.599
So, hmm, that's bizarre.
00:23:57.759 --> 00:24:05.119
And if you decide which part is the little m and the big m, you realize you can have any velocity between u and infinity, right?
00:24:05.359 --> 00:24:14.000
So said, okay, we have the basic mechanism, but maybe it doesn't really help us to get a better view of what's going on.
00:24:14.400 --> 00:24:17.599
Okay, let's put that aside, we'll go back to that.
00:24:17.839 --> 00:24:23.680
Let me turn out to our other problem, which was the beaker or the chain uh fountain.
00:24:24.799 --> 00:24:30.559
And you can do another computation that's very popular uh among physicists also.
00:24:30.799 --> 00:24:33.839
It's called the Atwood machine from the 18th century.
00:24:34.000 --> 00:24:51.599
At Wood at his idea that you could have a little machine and a pulley with two weights, and if you vary the respective weight of one and the other, uh the acceleration would lower the big weight, and that would give you a way to actually compute the acceleration of gravity by changing the mass.
00:24:51.839 --> 00:24:57.119
It's very popular, it's textbook undergraduate physics the way you do that.
00:24:57.359 --> 00:24:58.799
So, what is the computation?
00:24:58.960 --> 00:25:01.440
What is the back of the envelope computation?
00:25:01.680 --> 00:25:08.079
Well, you need an envelope, and so I here is the computation of the envelope.
00:25:08.160 --> 00:25:13.440
Uh one of the theorem in mathematics is that every theorem is a back of the envelope computation.
00:25:13.519 --> 00:25:16.480
You just need a big envelope, big enough envelope.
00:25:16.799 --> 00:25:32.079
So in this case, you need a small envelope, and what you do, you essentially you say that the you have conservation of linear momentum, that the force acting in one side on one at the acceleration is balanced by the other one.
00:25:32.319 --> 00:25:32.720
Right?
00:25:32.880 --> 00:25:37.279
So you have two equations of motion, one for the little mass and one for the big mass.
00:25:37.519 --> 00:25:51.119
So conservation of linear momentum, you can get rid of the tension in between, that's t, and you get easily the acceleration as a function of the big mass and the little mass times acceleration of gravity.
00:25:51.359 --> 00:25:58.640
And you can see that if though if they are the same, big m and little m, you have zero acceleration as you would expect.
00:25:59.599 --> 00:26:02.960
Now, what happens if I get rid of the little mass?
00:26:03.119 --> 00:26:06.079
I, you know, you're not supposed to do that, but why not?
00:26:06.160 --> 00:26:06.480
Right?
00:26:06.720 --> 00:26:07.839
Does it still work?
00:26:08.000 --> 00:26:11.359
Well, the big mass accelerates at acceleration of gravity.
00:26:11.839 --> 00:26:14.079
It's just falling, nothing big.
00:26:14.400 --> 00:26:19.519
Well, that's strange because when I did the experiment with the B cut, this is not what happened.
00:26:19.839 --> 00:26:20.480
Right?
00:26:21.119 --> 00:26:29.839
The I had an acceleration, but without little mass at the end, you chain has no mass hanging there, the whole change was moving up.
00:26:30.079 --> 00:26:39.599
And you see that it's one of the problems with this computation in the previous one, is that you assume you know perfectly what the motion is and you just solve for one quantity.
00:26:39.839 --> 00:26:47.680
And so it's naive computation, gives you some intuition, but you have to be very careful about the assumption that you actually make.
00:26:48.400 --> 00:26:50.880
So, what about this chain font then?
00:26:51.039 --> 00:26:54.160
What is the standard explanation for that?
00:26:54.559 --> 00:27:09.920
Well, the idea there is that if you look at two of the beads in the chain, they're connected by a little bar, and one of the beads, like the one on the left side, is on a substrate, is on top of the other pack.
00:27:10.319 --> 00:27:14.720
And the one on the right is being lifted by the chain at any given time, right?
00:27:14.799 --> 00:27:16.880
So it's just at the time where it's being lifted.
00:27:17.039 --> 00:27:24.319
And as it's being lifted, what happens is that the rotation of that gives you an extra kick that lifts you.
00:27:24.480 --> 00:27:32.240
And that seems a little counterintuitive, but if you ever done any skateboard, you recognize that's exactly the only effect.
00:27:32.319 --> 00:27:36.000
So only is uh the only effect or the only move.
00:27:36.160 --> 00:27:38.400
It's named after Alan Gelfan.
00:27:38.480 --> 00:27:41.599
Uh he was a skateboarder that introduced this motion.
00:27:41.680 --> 00:27:51.119
And the idea is that this trick is that as you go along, you rotate the board, and that rotation provides an extra kick in the back and lift the whole thing.
00:27:51.200 --> 00:27:52.640
And I'm sure you've seen that.
00:27:52.880 --> 00:27:59.039
You look at skateboarders, they do that all the time with uh with great expertise.
00:27:59.279 --> 00:28:04.880
And so you have you might have the same effect on the chain, or the the chain, you have the two bead connected.
00:28:05.039 --> 00:28:12.400
If you lift one, the other one is going to be pushed out by an extra force, like that, but because of the rotation in the balance.
00:28:13.119 --> 00:28:20.720
So another way to look at that is uh a delightful experiment by uh my friend Andy Runa in Cornell.
00:28:21.119 --> 00:28:23.039
And what he did was very interesting.
00:28:23.200 --> 00:28:35.119
He took a ladder made of steps at different at different angles, two ladder exactly the same or symmetric with respect, and he let one fall on a table and the other one fall freely.
00:28:35.839 --> 00:28:42.319
And so the question is what's the effect of the table on the motion of the ladder?
00:28:42.559 --> 00:28:47.039
Well, it's very strange because there is no reason if something falls on a table, it changes.
00:28:47.200 --> 00:28:57.839
Except if you have the same effect, as it falls on the table, the each step in the ladder would accelerate and provide a force that pulls the ladder.
00:28:58.000 --> 00:29:00.240
And so you did this beautiful experiment.
00:29:00.559 --> 00:29:04.319
You have a free-falling one on the right and the table on the left.
00:29:04.480 --> 00:29:12.480
And what you see, they started exactly at the same point, is that each of these steps is pulling the ladder down and accelerating.
00:29:12.640 --> 00:29:17.599
So the one on the left goes faster than gravity by this effect.
00:29:17.759 --> 00:29:22.240
You have an acceleration due to the shock, which is quite surprisingly.
00:29:22.559 --> 00:29:24.960
Mechanics is infinitely surprising.
00:29:25.200 --> 00:29:32.720
Okay, so we have all these uh effects, some interesting, some strange, and we really want to make sense of that.
00:29:32.880 --> 00:29:43.200
And now we get to the next step, which is to formulate uh what in the trade we call a toy model, because we like to play with it, not just the toys that I show you.
00:29:43.440 --> 00:29:50.160
A toy model is a minimal uh mathematical model that has all the uh elements of that.
00:29:50.400 --> 00:30:05.519
You see, in the previous uh uh computation I did, I assume I knew everything about the motion, and in that case, whether you decide that linear momentum or energy is conserved, you get different answers because they cannot be conserved for the motion ahead.
00:30:05.759 --> 00:30:10.640
In reality, you really have to solve for the motion and see what the consequence are there.
00:30:11.119 --> 00:30:16.000
But solving for the motion is difficult because this is a rather complicated structure.
00:30:16.079 --> 00:30:19.119
So you want to idealize it, you want to model it.
00:30:19.279 --> 00:30:27.839
And the first thing I'm going to say is that, well, the motion is mostly in the plane, as I crack it, even if I go overhead, it's motion in the plane.
00:30:28.000 --> 00:30:34.400
So I'm going to look at a rod, a string in one dimension, uh in two dimensions in the plane.
00:30:34.480 --> 00:30:37.119
So a one-dimensional continuum in the plane.
00:30:37.279 --> 00:30:52.400
Now, this object is right away much more complicated than what we've seen, because now I look at each point, x, y, and I parameterize it with the lengths, the arc length from, let's say, the swivel up to the point that it reaches.
00:30:52.640 --> 00:31:03.519
So the position of, if I paint a little red dot like here, like I did on the on the slide, this position of x and y change as a function of time.
00:31:03.680 --> 00:31:05.759
But each dot change also.
00:31:06.079 --> 00:31:12.559
So x and y are not both function of the arc length, that gives me the position, and time.
00:31:12.880 --> 00:31:15.359
So it's a function of two arguments.
00:31:15.680 --> 00:31:36.960
But you can then express essentially f equals ma with this continuum, and you get uh you get the first set of equations, which is the balance of linear momentum, which tells you that an acceleration of a little part is balanced by the forces that are inside, uh that it feels inside the material.
00:31:37.200 --> 00:31:44.559
And that's these two equations, F and G, are associated with the force of inside the material that pushes it.
00:31:45.359 --> 00:31:51.599
Now you have another equation, which is the equation for the angular momentum, which is also important.
00:31:51.759 --> 00:31:58.160
What you need now, if I look at a little piece of my material, it can also change its orientation.
00:31:58.319 --> 00:32:05.039
That provides a torque, and the torque is related to angular momentum, so I need also to balance that.
00:32:05.359 --> 00:32:10.480
That gives me a second uh third equation, and that gives me three equations for that.
00:32:10.559 --> 00:32:12.000
Don't worry about equations.
00:32:12.160 --> 00:32:14.640
T's are actually very complicated equations.
00:32:14.880 --> 00:32:18.319
T's are three non-linear partial differential equations.
00:32:18.559 --> 00:32:24.880
So when I started this Gresham job, I was told don't show differential equations because this is too complicated.
00:32:25.599 --> 00:32:30.000
And they were talking about ordinary differential equations, which is the baby version of this.
00:32:30.319 --> 00:32:33.119
T's as as complicated equation as you can get.
00:32:33.279 --> 00:32:41.440
And they are complicated because they describe extremely rich all the possible motion of a chain or a rod in in the plane, right?
00:32:41.599 --> 00:32:46.480
So they have a lot of solutions, this extremely rich set of equations.
00:32:46.880 --> 00:32:55.279
So this is a minimal model because now we can solve for both velocity and the shape at the same time if we know to solve this equation.
00:32:55.440 --> 00:32:57.759
So forget a moment of this equation.
00:32:58.000 --> 00:33:07.519
Now I have my back of trick of methods from applied mathematics, and I know how to get that, uh, extract information of that.
00:33:07.759 --> 00:33:11.519
And the first result is that loops exist.
00:33:11.759 --> 00:33:17.440
They are solutions of the equation, and actually it's more or less something that Euler knew about.
00:33:17.599 --> 00:33:30.400
And what I mean by that is that if I look at a uh a loop like this one, I know that these equations support the same type of solution on an infinite domain, things that go to zero on both sides.
00:33:30.559 --> 00:33:41.680
And more important, the loop can have any height, and I have also a nice little formula for it that is not important for us, but I can actually write the solution exactly for that.
00:33:43.519 --> 00:33:46.400
Now, that's a loop being static.
00:33:46.640 --> 00:33:52.240
The second result, and a bit of a miracle, is that any of this loop can be made to travel.
00:33:52.400 --> 00:33:58.240
And that came more recently, this type of idea of looking now at a dynamical solution.
00:33:58.400 --> 00:34:09.440
That means if I start with something like that on an infinite domain with no variation of tapering, there's a solution that makes it roll further.
00:34:09.519 --> 00:34:11.519
And you're gonna see what I mean by that.
00:34:11.760 --> 00:34:18.079
This is the solution, and I can make it travel depending on the initial condition at any velocity.
00:34:18.239 --> 00:34:21.199
So an infinite light, a little loop like that can travel.
00:34:21.280 --> 00:34:23.599
It's part of the set of solutions of this.
00:34:24.159 --> 00:34:26.239
And why is that interesting?
00:34:26.400 --> 00:34:30.079
Well, for a problem, it's not that interesting because it's constant velocity.
00:34:30.159 --> 00:34:39.119
So in this case, if I send a loop with a velocity c, it would just go velocity c forever, if there is no friction or anything like that.
00:34:41.599 --> 00:34:43.920
However, the whip is stapled.
00:34:44.320 --> 00:34:51.599
And now we can look at the effect of the tapering on the whip, and and that becomes much more interesting.
00:34:51.920 --> 00:35:03.840
And now T's loop now is stapled, and you have to now solve the equation for T's problem where the mass is actually changing, the radius is changing as it goes along.
00:35:04.000 --> 00:35:11.679
And that's that's a problem that hadn't been solved before, and I did it with uh Tyler Macmillan, who was my uh PhD student at the time.
00:35:11.840 --> 00:35:15.519
And it's very much like a problem, uh like a wave coming ashore.
00:35:15.599 --> 00:35:21.519
And if the beach is a little slanted as it comes around, you have a variation due to the fact that the depth change.
00:35:21.760 --> 00:35:29.039
So it's part of this type of problem where you have propagation of a wave on a changing medium, and there are a lot of nice applications.
00:35:29.280 --> 00:35:34.480
But in the context of this problem, it hadn't been done, so we looked at that and we solve it.
00:35:34.639 --> 00:35:37.199
And the problem is the following, going back to energy.
00:35:37.280 --> 00:35:39.599
Let's think about energy for a second.
00:35:39.840 --> 00:35:49.920
It says the physicists tell us that uh the the motion is completely confined, but if I have a loop, there's plenty of ways I can conserve energy.
00:35:50.159 --> 00:35:59.199
The loop, there is less mass traveling, it's true, but the loop can get tighter and have more elastic energy by being more curved.
00:35:59.840 --> 00:36:05.039
That would conserve the and go at constant velocity, that would conserve the energy.
00:36:05.280 --> 00:36:18.880
The loop could get bigger as it travels, because there is more mass more mass moving, it is still a solution at constant velocity, or the loop could stay the same and increase in velocity.
00:36:19.039 --> 00:36:20.639
You don't know that a priori.
00:36:20.880 --> 00:36:25.119
Nothing is telling you what happens, and you have to do the computation.
00:36:25.360 --> 00:36:37.519
And what the computation showed us is that essentially there's small variation of height, but the loop actually accelerates but remains the same height essentially.
00:36:37.679 --> 00:36:45.840
And you can now, as a function of the tapering, you can actually find exactly what the velocity is in the system.
00:36:49.840 --> 00:36:59.280
Of course, you're gonna tell me, well, wait a second, this is not an infinite rod on which a loop travel, no, it's a finite piece, but we have this effect.
00:36:59.440 --> 00:37:10.480
And the other objection would be well, if I'm if I if I do it well enough, I can take and I can do it also, I I could take a non-taped uh whip and make it crack.
00:37:10.639 --> 00:37:16.000
It provides a little bit more energy and all that, but if you know what you're doing, you could also make it cry.
00:37:16.159 --> 00:37:21.280
So tapering is not the whole story, and you want to go to the next step.
00:37:21.440 --> 00:37:28.480
Let's see, in more realistic conditions, now uh what happens exactly in the system.
00:37:28.639 --> 00:37:40.000
The good thing is that you still have all the equations, but as far as doing them analytically, pure mathematics, getting formula and all that, that's probably the best you could do, is what I showed you.
00:37:40.239 --> 00:37:45.599
What you can do, however, is develop numerical methods to solve this equation.
00:37:45.840 --> 00:37:48.719
That's also part of the bag of tricks that we have.
00:37:48.960 --> 00:37:58.559
And using these methods, you can start with an initial loop, as I show you, and now look at the effect of having an open end at the end rather than an infinite one.
00:37:58.719 --> 00:38:13.920
And then you can vary tapering, you can vary the length, you can vary the force, the initial velocity, and you can exactly point out what are the different effects that matters in order to get an acceleration at the tip.
00:38:14.800 --> 00:38:35.519
And you can you can tabulate and you can show that all these effects are somewhat important, both tapering, uh, force, initial motion, and all that, exactly like you would if you practice the whip, you would have the right coordination to send the right wave and pull it, put the right force at the right time, and you get very quick acceleration.
00:38:35.679 --> 00:38:45.280
And since you have the whole motion, the whole motion of the whip as it as it goes around, you can now look at the velocity at each point.
00:38:45.599 --> 00:38:53.119
And the velocity, look at the tip speed, for instance, follows very much the type of experiment that you saw.
00:38:53.599 --> 00:39:13.760
And so if now you look at uh the chain experiment on the beaker, the one that's nap, I don't know if you can see it well on the screen, there is a yellow curve, the the black dot is the experiment, and the yellow curve is the numerical solution of the same type of equation that I show you.
00:39:14.000 --> 00:39:17.920
And as you see, the the fit is actually extraordinary.
00:39:18.000 --> 00:39:20.159
It's not a fit, there is no free parameter.
00:39:20.320 --> 00:39:26.559
This is just a perfect prediction of the motion of the chain as you go along.
00:39:26.800 --> 00:39:35.760
And because you have that, you can now know from the numerical solution, the one the computer gives you, you have all the possible information.
00:39:35.920 --> 00:39:38.639
In particular, you can look at the curvature.
00:39:38.800 --> 00:39:51.679
And when you look at the curvature, you can look at the place where the radius is maximal, uh, the so the smallest possible curvature, the way the place where, as it unfolds, the chain is most curved.
00:39:52.400 --> 00:39:56.320
And its inverse is the radius, is the best-fitting circle at that point.
00:39:56.480 --> 00:40:04.880
And as it goes through this transition, there is a critical time where it becomes flat, and that means the radius of curvature goes to infinity.
00:40:05.119 --> 00:40:09.920
And that's what is called a singularity, and that's where the snap occurs.
00:40:10.320 --> 00:40:14.800
And this type of singularity is very interesting, it hasn't been completely resolved.
00:40:14.880 --> 00:40:22.159
It's a good place to study, but exactly how it goes to infinity is still a bit of an open, open question for us.
00:40:22.800 --> 00:40:32.320
But so far, all I've shown you, it was a toy model, did not include the motion, but no interaction with the air.
00:40:32.480 --> 00:40:37.599
And the crack sound that we hear is really a change in air pressure.
00:40:37.760 --> 00:40:51.119
So I want to go back to uh supersonic motion, and as you can see in Arizona, not only we have WIP, we have road runners, and maybe they're fast enough to be supersonic, who knows?
00:40:52.000 --> 00:41:03.920
So the first thing that I mentioned was that little puzzle, a little surprise of uh the tip loop, uh the velocity of the tip as it cracks.
00:41:04.639 --> 00:41:12.079
So remember that when you look at the data, you can show that the whip cracks when the tip reached twice the speed of sound.
00:41:12.400 --> 00:41:26.079
At first that seems a little bizarre to us, and then we realize the following is that in a motion where you have a loop, the motion of the loop, the velocity of the loop, is different than the velocity of the point on the loop.
00:41:26.400 --> 00:41:29.199
So let me give you a much simpler example.
00:41:29.360 --> 00:41:40.400
If I take a bicycle wheel and I roll it along a wall at a velocity V, you can ask what is the velocity of the material point right at the top of the wheel?
00:41:40.719 --> 00:41:46.480
Well, the velocity at the top of the wheel must be greater because that's why it goes in front of the wheel.
00:41:46.719 --> 00:41:47.119
Right?
00:41:47.360 --> 00:41:59.599
And if I roll it at constant velocity, you can show by nice geometric argument that the velocity of that point at the top is always twice the velocity of the wheel.
00:42:00.079 --> 00:42:00.639
Perfect.
00:42:00.800 --> 00:42:02.880
And that's true for any shape, actually.
00:42:03.039 --> 00:42:09.760
If you have a shape that travels at constant velocity and like that, the top will be always twice.
00:42:09.840 --> 00:42:13.039
So it's not a property of the physics, it's really a property of the geometry.
00:42:13.119 --> 00:42:20.719
So if that shape doesn't change, the loop goes at velocity c and the top of the loop goes at velocity 2c.
00:42:21.199 --> 00:42:43.920
So as it unfolds, you have C prime, it goes faster and faster because of tapering or other external extra tension or things like that, up to the point where the loop itself, now the loop itself, so you have a loop, and the loop itself reach the velocity of the sound, but the tip that's is being protected by the loop.
00:42:44.000 --> 00:42:46.960
The tip does not go against the air.
00:42:47.199 --> 00:42:53.119
So it's the loop that breaks the sound barrier, not the tip that breaks the sound barrier.
00:42:53.280 --> 00:42:55.440
The tip comes after and unfolds.
00:42:55.920 --> 00:43:04.480
And the way to see that is to know back to the data when you realize that you say, hmm, let's see if we can actually see that on the data.
00:43:04.639 --> 00:43:21.599
And if you look at the experiment where you see it is supersonic motion, here's the speed C, and you see this is where the supersonic shock is being created, the supersonic wave, it's the loop, not the top of the tip that at velocity 2c as it unfolds.
00:43:21.760 --> 00:43:25.920
And then you have further uh uh expansion.
00:43:26.480 --> 00:43:36.559
Now, supersonic motion is very interesting, very difficult, because you have to understand the physics of air and compression and jump and all that in very extreme conditions.
00:43:36.719 --> 00:43:38.800
But there are very simple things you can do.
00:43:38.880 --> 00:43:42.400
Uh, one of the simplest constructions is the MAC cone.
00:43:42.639 --> 00:43:59.599
So if you have a motion in the air, just by friction in the air, uh the that that point is going to uh send sound wave, just like you hear something going in inside a room if you throw it fast enough, is the sound that the air makes at its expense.
00:44:00.320 --> 00:44:11.360
But if that point is going supersonic, the sound waves do not have time to expand fully before they all coalesce on something called an envelope.
00:44:11.599 --> 00:44:33.519
And so if I have this sound wave emanating constantly from that moving point due to interaction with the air, if the velocity of my point is faster than the speed of sound, they all accumulate with an opening, the Mac Con opening, which is given, the sign of that is given by the ratio of the velocity of the sound by the velocity of the object.
00:44:33.760 --> 00:44:41.360
And this is what you see in Schlieren imaging for a plane or for a for a supersonic bullet or something like that.
00:44:41.519 --> 00:44:46.400
So for a plane, you might experience that you sometimes hear more than one boom.
00:44:46.559 --> 00:44:53.840
So the supersonic sound, the supersonic boom, is not when the plane gets past the sound barrier.
00:44:54.000 --> 00:44:57.679
It's something that's constantly emitted by ulti-sound wave gathering.
00:44:57.920 --> 00:45:05.840
So at that point, if you sit right at the intersection there, it's when that red line crosses you that you hear the sound.
00:45:06.159 --> 00:45:27.760
But a plane is an object of finite size, and so there are a lot of different waves being emitted by different pieces of the plane, and you have one main one that's from the tip of the plane that creates the first one, and you have a change in air pressure, and then you have the opposite change in air pressure at the end of the plane.
00:45:27.840 --> 00:45:28.639
You have a change.
00:45:28.800 --> 00:45:36.320
And so that's why you can hear a double double boom or multiple boom as supersonic plane goes by.
00:45:36.800 --> 00:45:46.960
But there is a huge set theory about uh supersonic motion, or bullet or planes or space and all that.
00:45:47.199 --> 00:45:56.719
But almost all of them, as far as I know, all of them were for uh rectilinear motion, because usually people, plane goes go straight, or bullet goes straight.
00:45:56.880 --> 00:46:06.639
And there is no equivalent what if the motion actually uh is not straight but circular, just like is as the whip cracks, it unfolds.
00:46:06.880 --> 00:46:08.400
But it's not too hard to do.
00:46:08.480 --> 00:46:12.400
You can play the same geometric trick and you can have a point.
00:46:12.480 --> 00:46:15.920
Now think about the the cracker, the loop as it unfolds.
00:46:16.159 --> 00:46:35.599
Now it rotates rather than going straight, so it follows more or less a piece of a circle, and that piece of a circle also emanates waves, and you see now that you have a curved macode depending on the trajectory, and you can now compare that to the data.
00:46:36.079 --> 00:46:50.400
And so by putting all these um different pieces of physics and mathematics and all that, you start having a very good picture of all the different effects and a good understanding of whip cracking.
00:46:50.559 --> 00:46:52.719
Uh that's that's quite nice.
00:46:53.039 --> 00:47:00.880
Of course, you can say, well, that seems a little bit frivolous, putting all your effort into uh into that kind of problem.
00:47:01.039 --> 00:47:02.320
What is it good for?
00:47:02.960 --> 00:47:12.239
Uh well, first I want to say that it's not a question that people ask to people who study black hole or or the origin of the universe and so on.
00:47:12.559 --> 00:47:14.800
But we do have a responsibility.
00:47:14.960 --> 00:47:16.079
What is it good for?
00:47:16.480 --> 00:47:26.559
So this example is of course very uh catchy because it's very extreme, but it's actually part of a whole body of problem for which we develop methods.
00:47:26.639 --> 00:47:31.199
It turns out that this one is one that that captivates the attention.
00:47:31.519 --> 00:47:34.719
So we get to the problem or the question, so what?
00:47:35.199 --> 00:47:38.480
So what else can you do with that in a way?
00:47:38.719 --> 00:47:45.199
Just explain the cracking of the whip or chain, this this is good, but that's not what we pay you for, right?
00:47:45.360 --> 00:47:47.440
We pay you to help society.
00:47:47.679 --> 00:48:07.039
Well, there are plenty of other problems when a cable snaps or something like that, but if you look at the smaller scale, there are microorganisms, for instance, any kind of flagellate organism that uh beat the flagellum and in doing so provide linear momentum that's expelled for the motion.
00:48:07.119 --> 00:48:10.480
So this one is not a little uh space lunar module.
00:48:10.559 --> 00:48:20.239
If you look at the scale, this is three micron, it's a very uh nice uh uh organism called a coanoflagellate that has this specific tail that flips.
00:48:20.400 --> 00:48:27.440
And these are very close in some aspects to actual sperm and the flagellum of the sperm.
00:48:27.679 --> 00:48:36.559
And indeed, sperm do create this whipping motion by making loops and pushing them in order to propagate.
00:48:36.719 --> 00:48:49.599
And all of a sudden, you went from the whip to understanding something about motility of the sperm, which is a very important problem view both on physiology and biology.
00:48:49.840 --> 00:48:53.119
And we've applied the same type of problem techniques.
00:48:53.280 --> 00:49:04.480
Again, you completely have to refine your model because nautics operate in very viscous environments, so you have to take into account the interaction with the fluid, the viscosity on one, the other one, and so on.
00:49:04.639 --> 00:49:22.079
But using that we manage with uh Emman Gaffney, which is in Oxford, and and uh Hermes Gadea, who is now in Bristol, uh, to explain some effects that are observed in in human sperm uh related to uh what is called the counterband phenomena.
00:49:22.400 --> 00:49:30.000
Another cool example uh that uh I that that I found is uh prey catching in octopus.
00:49:30.800 --> 00:49:39.840
And here what you have is the octopus very quickly unfolding and doing this nap in order to catch a prey that is uh indicated.
00:49:40.000 --> 00:49:47.039
And that as you can see from the scale, this this happens in about uh 400 milliseconds.
00:49:47.920 --> 00:50:00.000
And one of the uh last examples that you that you uh can see also, like nice application, is a much bigger scale.
00:50:00.159 --> 00:50:02.400
Uh for instance, the space tether.
00:50:02.719 --> 00:50:13.199
So satellites have these tethers, which are long rods that uh uh have a weight at the end, and they use that either to stabilize or change their orbits.
00:50:13.440 --> 00:50:17.920
It's the the the mechanics of that is very nice in terms of orbital motion.
00:50:18.159 --> 00:50:22.880
But there is one problem is that these tethers sometimes have to be retracted.
00:50:23.280 --> 00:50:35.280
And when you retract the tethers, all the shape and the motion gets concentrated as you retract the cable, you call you coil the cable in smaller and smaller portion.
00:50:35.440 --> 00:50:48.800
And now in space there's very little friction, so that you don't have viscosity and all that, and so what happens is that as you coil it, the motion of the end gets wider and wider.
00:50:49.039 --> 00:51:06.239
Very much like when you eat spaghetti and you slurp a spaghetti, and what happens is that all the energy that you do for the velocity of slurping is being transferred, or the energy is being transferred in a smaller and smaller portion.
00:51:06.400 --> 00:51:10.639
And if you have that with uh pasta saus, you know what happens.
00:51:10.880 --> 00:51:13.119
It oscillates and it splashes on you.
00:51:13.280 --> 00:51:15.599
And that's exactly the same phenomena that happens.
00:51:15.679 --> 00:51:26.800
I've never worked on space tethers, but they're very interesting because in in space the problem is that you it's very hard to uh to stop things from happening.
00:51:26.960 --> 00:51:32.159
The motion of objects in space is very much uh inertial, that means not damp.
00:51:32.400 --> 00:51:37.199
There is very little dissipation that you can, but and that creates uh other problem.
00:51:37.519 --> 00:51:55.360
So I hope that I've given you a global picture, starting with a very simple problem and bringing up to a more complicated one, but one for which we have no answers and uh different application to the rest of the world on that.
00:51:55.519 --> 00:51:56.800
Thank you very much.
