About science; About elastic water

This is about science produced by the California Institute of Technology in cooperation with station K PPC Pasadena California. The programs are made available to the station by national educational radio. This program is about elastic water with host Dr. Albert Hibbs and his guest Mr. David James. Here now is Dr. hips most of us when we think about water have a very clear idea of what it is it's one of the most common things in our every day life. But can you imagine a kind of water that will climb up hill over the edge of a glass pitcher and flow which you can stop by cutting it with a pair of scissors. All this. Just because your liquid has been discovered actually rather accidentally by Cal Tech graduate student who was working on a mixture of polymers type of plastic and water actually working is very dilute mixtures of water polymer and water and we have him with us this evening David James a graduate student at Cal Tech

who's been there now five years working toward his doctor's degree before he came to Cal Tech he did his undergraduate work at Queen's University in Kingston Ontario. David before we go into the story of the elastic water maybe the best way to start is to say why it is you're worried about mixing plastics and polymers with water at all. You're referring to how I came upon this particular for what your research is what what is the interest in melting plastics in the water what's what's the purpose of this sort of study find the particular phenomena that you referred to but to the syphon involves concentrations of this ball are of the order of one half of one percent or roughly 5000 parts per million. That's a very dilute. Only relative to ordinary things for these particular polymers that's a concentrated solution. I see my own research has been concerned with dilute solutions of these polymers.

This means we're down in the the range of 50 parts from 100 as much then as this. Exactly exactly and you'll get interesting results now and there I suppose less spectacular as well. That's true the water is still elastic there but the elasticity is much much smaller scale correspondingly because we have much less powder and or tolerance. When you say lastic does this mean you can stretch it and it will snap like a rubber band. And why do you use the term elastic. It's very hard actually to conceive of a liquid being elastic because the usual property that's assigned to liquids is the Scotty of course. And these particular kinds of liquids are both a viscous and elastic. Now you can stretch it like a rubber band. If you take a rubber band and stretch it and hold it there the the force remains the same as long as you keep it stretched. If you try with one of these liquids you'll feel some restraining force as you stretch it. But then when you stop

the the force will die down. What's really happening is that these liquids are elastic only while they're flowing. In fact the amount of elasticity depends on the rate even of you longer or the rate of stretching and the faster you can stretch and the more you last the city. You can undo some of them. Well how does this kind of elasticity show itself in any sort of measurement of any sort of phenomenon. OK when we have one of these lasting liquids in which the concentrations say is of the order of one half of 1 percent it's very visible when we try cutting the the stream that you referred to in the tube the FE from there you can readily see it because if you cut the string you'll see a definite snap or recoil of the fluid. You mean one is when you're pouring a fluid from one container to another. That's right. You can actually cut this pouring stream so it will behave like a snapping. In fact the best way I've found is the cut oval two were four inches

below the rim of the upper beaker and the liquid that was above the scissors will snap back into the upper container. There's that much less to see involved in it saves a lot of spelling. Well as a matter of fact if you don't do that you'll find this liquid is very peculiar in that it's quite filler metal. If you didn't cut up by scissors the stream would become smaller and smaller and smaller. It would never form into blobs the way water does. It would come as small as say the filaments on a kind of cup and make itself into very thin thread instead of acting like arrays of droplets. That's right and you stop the stream by blowing on them because you throw it away. Most peculiar water it is. Well let me give you one other good illustration that I've used of the waters or the water any more of the liquid elasticity. If you have a rotating flow involving this would say you have

something in there to indicate the fluid motion say bubbles or trace or particles of some sort with an ordinary viscous flow all these bubbles would would come to rest. That's all you would see the flow would become slower and the bubbles would stop. With these kind of fluids the bubbles come to rest and then start going in the opposite direction as if they were wound up. Exactly and this is another consequence of another demonstration of the fluid elasticity so that in this case however the elasticity persists even though the motion has stopped. That's right it doesn't. Doesn't stop suddenly as you say it dies down. That's why it looks and stop it looks like a more viscous fluid than it really is comes to rest and then flows for a slight duration in the opposite direction how does it feel. It's very slippery or slimy and in fact in the lab where I work I'm sometimes referred to as the slimy water King.

So you can actually feel the same amount of the same change in properties that you observe in these. That's that's correct unusual experiment but I so meant I assume with the small concentrations it still looks just like water that's right in the. In these very concentrated solutions is when it's extremely slimy stuff in the dilute solutions in which I'm doing my research. You'd have a very hard time telling that there was anything dissolved in the water. To all appearances and to everything you could do to water it would behave the same as water. I would feel the same as well. That's right a pretty limey feeling would be gone. That's right. I think with enough ink with enough experience you could put your fingers in there and detect it and I think I could but I'm not sure you could. My wife can help me but it's one of a bit of experience that in these very dilute concentrations you can detect the dissolved polymer. What does it do in the dilute concentrations that brings up your interest in it and keeps you working at night.

They found some time ago that dilute solutions would flow very differently in pipes specifically in turbulent flow. The plate friction would we would be reduced by as much as 70 percent. So a tremendous reduction in friction. That's right that from an engineering point of view of course the economics of this thing are quite important the thing is known as the Tums effect. And the Navy of course is interested in this particular application because they're interested in the reduction of any kind of drag along the surface between surface and solid surface in a liquid. That's correct. So I can imagine that would be most most interested in your research where they are sponsoring your research. Any chances this is going to be. That's exactly my project. I see the Office of Naval Research is my sponsor. Well how does it come about that the friction can be reduced in a pipe is there then no turbulence is that what happens as it removes the turbulence completely. So it's completely smooth flow in the boundary.

Noel from dye injection pictures that I've seen there are still turbulence. To see anything they've had to make these solutions quite concentrate and I think it gives a distorted picture. I'm not sure what's really happening and I think there are perhaps two explanations of explanations of what's going on. All of this is important right next to the wall of course where your drag is determined in the boundary. That's right. The one way of looking at it is that the fluid being elastic can absorb. The fluctuations that the fluid experiences it absorbs the energy and then at a later time gives a black the same way as if you had springs in there doesn't dissipate in heat. That's right that's right. And the less dissipation of course it means you require less energy to pump the fluid through the pipe. The other explanation is that the fluid because it's different than water or because it's viscous and

elastic is not allowed in some sense to have these very rapid fluctuations. They're just not allowed to exist in the in the turbulent eddies in which case the same thing takes place you don't have the dissipation associated with these very high fluctuations with that main on that extremely fine scale turbulence is missing that you do have a large scale turbulence which you might be able to photograph with a dye that's a micro turbulence maybe where it may be altered. Altered radically That's correct. A fellow at Penn State bringing under Lumley tried to determine this and it turns out that these fluctuations are of such high frequency that he can detect any change as to whether the the smallest eddies were damp though or disappeared or absorbed energy or what. That's why I'm very unsure at this point which of these conflicting theories but which of these ideas is correct. The energy in the eddies could be there or

fluctuations and could be absorbed and the energy given back at a later time. So it doesn't last which one is right is hard to say at this point. Well how are you approaching the problem what kind of observations and measurements are you making. What we tried to do was answer this question of what is going on in the target a bounded layer and we wanted to measure the mean velocities and the fluctuating velocities and a turbulent boundary layer namely an on water flow. We want to do this using a hot film sensor and hot film anemometer ie. What if what does this look like. These are very small. Why yours are saintly. They're fabricated using a glass tube over which is laid out platinum film. And these tubes are very small now they're the order of thousands of an inch. We have several sizes 1000 2000 a 6 ounces and

Platinum a seed. The amount of my putting electricity through it that's correct. That's right and the amount of heat lost from this platinum film is a function of the velocity of the fluid going by and in that water in this case. So that one is able to use this type of sensor to measure velocities. If you can call relate or calibrate the heat loss with the velocity going by the sensor. And in this way we can use these type of sensors to measure and fluctuating velocities in the target abound. So you then put this sensor into the fluid and the faster the fluid goes the faster the heat is conducted away how does it show up as a change in the current of the resistance resistance of the wire What do you actually measure when you can stick a thermometer down and measure the temperature the wire what no one one measures a temperature change in the resistance resistance of the plot. Exactly inductors so this is then gives you a measure to measure the temperature.

Put this in a standard bridge and hook up a lot of electronics to get a stick. I can imagine you may spend a large portion of your time getting the electronics to work without worrying about the fluid. Well in ordinary water is it as you say an ordinary liquids the faster it goes the the more heat is lost so this gives you a way of measuring what happens in the when you were last in water. While we found we can do that it turns out in these dilute solutions in water that up to a certain point the sensor behaves exactly as though it were in water and other words the higher the velocity the higher the heat loss. If we increase the velocity even more we find the velocity after which the heat loss is independent of the velocity. Other words after this cut off a loss of your critical Lost City whichever you like to call it the it will matter how much of a

lost city or how much water is flowing by the sensor. One reads the same output for the same amount of heat loss. Essentially then the so-called sensor is insensitive. Sounds very strange is this because locally around the sensor the velocity isn't what it is in the rest of the flow. That's what I think is happening and that's the main problem my thesis is trying to explain exactly what is going on as well as taking measurements of course. What I think's happening is that the flow field because of the fluid elasticity has radically changed. Around these very small cylinders in some sense I think you could say that the flow field is frozen at a certain velocity so that no matter what the external velocity is relative to the cylinder the cylinder sees the same velocity all the time and accordingly sends off the same amount of heat.

But this must be a very local region very small just like a cylinder right next to the cylinder. That's correct but wouldn't heat still pass through this region into the rest of the flow. Even though there were that it would seem that you're just maybe making a slightly bigger wire effectively but you'd still the rest of the ILO would still be carrying away heat very quickly. Yeah but it'll carry weight. Only the amount of heat that that layer next to the wire will allow and I see in other words the layer next to the wire is all important if it won't allow any more heat to be conducted away. It can't be. And is there any any property of insulation you soon connected with this or have we looked into that because there is the possibility that what's going on with these very small wires is that the polymer molecules are in some sense piling up on the cylinders piling up enough in fact to effectively insulate the cylinder from the external flow. And we did a few rough calculations to find out how thick this

layer would have to be. And our calculations are that the thickness would have to be as big as the wire itself which pretty well ruled out that possibility. I just wasn't that much of knowable. We know by absorption that the. The polymer molecules may be at least or most one thick around the diameter or around the periphery of the cylinders. It's hard to conceive them being any thicker than that. There's no good reason why they could be actually and this was just one be even if it trapped all the water and no one thickness. It wouldn't be enough to insulate the wire also reduces peculier from what is your explanation then of what's happening. You suggested a change in the flow field what is the change. That's a good question. If I do. That I have my thesis done I think.

It has been unfortunate part is we can't look at this very readily experimentally to find out exactly what what's going on in the flow field. One would like to be able to inject some dye and watch the streamlines and see exactly what's going on. Turns Out Of course when the wires one Thousands of an inch in diameter that this is just about impossible. We found it to be impossible because we're not trying to do that. Most banks say at this point is because of the stagnation point flow all that is produced by the cylinder. The flow field in that region becomes frozen doesn't change with the external flow. And in that and this of course is because the fluids elastic. What I think time is because you can't get the fluid to move any faster when they sense near the point when there's an obstruction that makes the stats right.

Stagnation point is a point of flow just ahead of it just a directly in the front side of this obstruction and I think that's correct. That's correct so when you have something like that set up in the flow and elastic fluid it sets up a frozen flow regime in some sense or in some sense right now this is reasonably vague but I think anybody working in this field. It was that way about the subject and it's hard for me to be any more definite about what's going on really at this point. But it must be rather frustrating to have the one instrument which you counted on to be able to make the measurements of boundary layer behavior of the instrument which all of a sudden shows up with it. And a peculiar behavior of itself. Where are you now you started out measuring what was going to happen in a boundary layer and now you're measuring what's going wrong with your instruments. What's the next step in this. Do you have. Well you knew instrumentation was still another problem where do you go from here.

No it turns out that because we found these wires to become or hot film sensors they should turn to becoming sensitive that we felt this was a very important problem also. In other words we found a flow regime in which which is radically affected by these dilute solutions. And we previously found that only target of souls had been affected. When you had this very small amount of pot or what we think we found a laminar flow. Which is radically altered when in the same dilute solutions that hadn't turned up before and know the flow over of a smooth plate for example never showed anything odd in these. That's correct as long as it was a laminar flow. We do viscosity measurements and everything else. Course of Scot's that increases a little because you dissolve Palmer. But we did nothing radically changed because of the dissolve problems in laminar flow. And this experiment which of course is likely to

flow because we are very low velocities and very small wire diameters produces remarkable changes and we thought this was very significant and we conducted a series of experiments measuring the heat loss from various sized wires and various polymer concentrations and various polymer molecular weights. And that in fact has become only the basis of my thesis now. What kind of molecular weights are we talking about by the way for these polymers of the order of millions. So these are really long chain high molecular weight polymer that's right the highest weight we've been using is of the order formation. Is that incidentally to go off the subject just for a moment is that a man a chance perhaps a reason for the last history the fact that these are long chain very long. No question about Dolly. One gets the elasticity only because the mold kills are extremely like long fibers that in some sense that's right up within a liquid. That's

right if you could look at it under a microscope you would see something like a coiled ball of string. Or an extremely long spaghetti noodle level randomly. Wound up say. Well going back again then you've tried other polymers besides this first one you started with and this other molecular weight. Well yes we haven't tried a different polymer although I have a couple that I've been so busy with the one polymer that we've been so successful at a different time to wait for that one. That's right. They're the people who make this make it in several molecular weights three mainly to get around any dependence on that. Oh definitely there is. It's roughly depends on the cue any effects we've found so far depends roughly on the cube of the molecular weight. I see with the slowest molecular weights having a smallest effect. That's correct. In other words your chain is is that much shorter. What about the size of the wire itself that you're that. Are you using for this the diameter of the wires or

dependence on that. Surprisingly no. That is surprising. Maybe we just happen to be you know the bigger obstruction would would make a bigger change in the flow regime and set up this situation at a lower speed or something. Yes why do we tried only three sizes of wires because that's all it produced commercially right now. It's one thousands to thousands and 6000 and we found this critical velocity of a loss at 8 independant of the wire diameter. And I'm not really sure why. Maybe because it's a stagnation point flow which to the first order is not dependent on the other. On the other hand we just may have happened to hit a regime the effective diameter does not. We would like to try some larger larger diameters and find out exactly the dependence like you have to know. In the mean time what about the effect time or the behavior in turbulent flow if you had any

other opportunities to measure that. No we haven't. I've given up since we can use the film sensor pretty well given up trying to measure velocities of the target and boundary layer. There are some other people I think. Coming in to the gods of program who may try that I'm not sure how they're going to do it but I think they're going to try to invent some brand to play some other type of instrument. That's right they make small particles neutrally buoyant. And there may be other ways of measuring it but I've pretty well given up measurements on the target a boundary layer. I'm interested in this the laminar flow around circular cylinders. Well it would seem to me just offhand without an I guess much more than vague intuition that if there is a stable flow regime set up around an obstruction in the flow that that might act to increase drag on a body in the flow.

And yet pointed out that for example flow down a pipe has less friction. Is it because. And one is simply a laminar flow phenomena has nothing to do with drag and the other is turbulent and they're just completely different things as you hit it right on the head I said that was going off in the wrong way. No you hit it right on the head. I'm doing drug measurements now on these very small diameter wires and in laminar flow and surely enough as far as I can tell the drag is increased quite a bit. I see. But when it's flowing turbulent laid down a pipe is when you get these drastic reduction. And that's the point of difference between when I'm going to have a phone really. Well of course if low const usually starts out over a surface as laminar And then if it goes on long enough it gradually turns into turbulent flow. Does the elastic property of your liquid make a change and that

is it changing laminar turbulent at a different speed or different distance down the pipe. Other researchers have been looking at that and. Some seem to find that the so the critical velocities increased I think some found and decreased the whole area concerning the stability of a line or boundary layer seems to be quite confused regarding these dilute polymer solutions. The most thing we said I think right now is that not greatly affected that particular transmission. That's correct not affected but what happens on both sides of it apparently is the laminar flow regime can be affected as you've noticed with the wires. The turbulent flow is definitely affected because as you notice with the decreased rate through a pipe is there any practical application of this by the way as it may be found any good uses for you pointed out earlier they're interested. Have they done anything so far to put their interest to work. Yes the the one place a few Z's

polymers is in the torpedo business. They inject or Egypt at the nose of a torpedo. A solution of the polymer which then coats the surface of the torpedo and then once you have the layer right next to the wall full of this polymer one gets these drastic reductions and drag so that one can then get the same power plant. So you know the torpedo great increases in speed. I seem to recall seeing a graph where they had a torpedo that was traveling at 60 feet a second. Then they injected this polymer at the nose and had the speed go up to something like 90 feet per second. So it's quite a change. You know there's always been some some sort of a mystery at least in my mind is how some

fish manage to swim at such very high speed. Just any possibility that that fish sweat a polymer that's a good question. I hadn't thought too much about that up but I think you have perhaps the right if you the same sliming as one feels on a fish. You would feel the walls of a vessel that had just prior to that and containing one of these a poleaxe solutions a very very slippery feeling and it could be you know I'm not sure that the fish are slimy because of the. Of a very high weight polymer you know one of the bottom line exactly what a little they do not the stuff that is not I don't but that is a real it's a very real possibility. I could be for the same reason. I just don't know a lot of the fishes to reply correctly. Well one thing we've never quite gotten out yet is what is this polymer we've been talking about is it a quite unusual chemical unusual only I think in that

it is a very high molecular weight. And they've only been able to produce it recently in such a high molecular weight. And the other character is that is that it dissolves nicely or reasonably well in the water. What's its name. The chemical name is polyethylene oxide. The the Union Carbide company manufactures that under the brand name of Polly ox Polly ox. I see so if I wanted to duplicate some of your spectacular experiments of pouring water nonstop I would get from Union Carbide a can of poly oxide. You would try to get from them. Well David perhaps with this those who are listening in can begin to think of more practical applications for elastic water and go into business directly with the Union Carbide. Thank you very much for talking to us tonight. You're welcome doctor has this was about science with host Dr. Albert here and his guest Mr. David James. Join us again for our next program when Dr. Hibbs will lead a discussion about technology in

India. About science is produced by the California Institute of Technology in cooperation with station KPCC in Pasadena California. The programs are made available to this station by national educational radio. This is the national educational radio network.