The Feynman Sprinkler Problem

I'll go ahead and prove that I never took a hydrodynamics class.

I'll guess that the density difference between water and air make this at least a little bit of a trick question. I'll guess that the water being sucked in creates a localized low pressure at the openings. I'll guess that this rotates the sprinkler head in the opposite direction as when used in the yard. I'll guess that the density difference between the water and yard air makes the sprinkler turn much faster in the yard than in this submerged test.

I'll also guess that since Rixter is posting this question, my "I think I am logical" guess is probably wrong.

I think the question should be posed with both spraying out and sucking in being carried out under water to remove the density factor. Being under water shouldn't affect direction, just the speed.

I don't think the sprinkler will move at all, this is the difference between positive and negative pressure...Airflow duct under positive pressure needs to be smooth and aerodynamic, but under negative pressure does not...that's why your furnace filter is always positioned in the return side of the system...

I think the sprinkler will rotate in the opposite direction very weakly when water is sucked in. The initial response will be stronger, and the continuous reponse weaker if it is not stopped by the friction of the sprinkler head.

Ideally, water would approach the inlet of each nozzle from every direction, resulting in no torque being imparted to the sprinkler. But due to its finite size, the orifice blocks water from behind. More water approaching from the front of the nozzle results in circulation being imparted to the surrounding water.

Initializing the circulation of the mass of water would result in an initial larger torque applied to the sprinkler on start up and a weaker torque maintaining the circulation against the losses due to viscosity.

Due to conservation of angular momentum, the circulation in the external water results in an opposite circulation of the water entering the sprinkler head. This circulation imparts torque internally to the sprinkler head causing the backward rotation.

This internal circulation, as shown in the video, I think, is just the final link in the chain. The initial cause is that the water cannot approach the nozzle orifices equally from every direction.

Well I'm thinking that in order to be a fair test the GPM should match, so if the nozzle when sprinkling is dispensing 30 GPM into the air, then the underwater, which will be under pressure, should be made to match...In order to create low enough pressure to move the sprinkler head, you would need to increase the velocity of the suction water beyond the point where the water would start to boil....at least that's how I see it....Now it would work with air because of the low viscosity...

But even with air you can see the diminished effect...

It occurs to me that the nozzle design might have a dramatic effect one way or the other... one must take into account the pressure, spray pattern desired and direction of flow, one optimised for one scenario, might be poorly suited for another...

It seems to me the best nozzle design for reverse underwater movement would be no nozzle at all, just an open pipe....but altering the design to make it work, might be putting your thumb on the scale a bit too much...maybe

To make it fair, you should do the spraying out and sucking in both underwater.

Richard Feynman did the test in a pressurized vessel, but that didn't work out all that well for him.

"In 2005, physicist Edward Creutz (who was in charge of the Princeton cyclotron at the time of the incident) revealed in print that he had assisted Feynman in setting up his experiment and that, when pressure was applied to force water out of the carboy through the sprinkler head,

There was a little tremor, as [Feynman] called it, and the sprinkler head rapidly moved back to its original position and stayed there. The water flow continued with the sprinkler stationary. We adjusted the pressure to increase the water flow, about five separate times, and the sprinkler did not move, although water was flowing freely through it in the backwards direction [...] The carboy then exploded, due to the internal pressure. A janitor then appeared and helped me clean up the shattered glass and mop up the water. I don't know what [Feynman] had expected to happen, but my vague thoughts of a time reversal phenomenon were as shattered as the carboy.^[7] "

https://en.wikipedia.org/wiki/Feynman_sprinkler

I agree: the relative size of the vortices is highly dependent on the exact angle of the pipes entering the central tube, indicated here in red:

<picture missing>

If the angles were deliberately offset:

<picture missing>

then, I bet you could get the thing to rotate "forwards".

Alternatively, if the build is perfect then the initial direction is controlled by the initial "jolt" starting reverse rotation, and this is maintained by the differential size of the vortices. I bet if you forced it to rotate the other way it would reinforce that same direction.

___________________________________________________________________

You have encountered an error on our site. We're sorry for the inconvenience.

A transcript of this error has automatically been sent to our quality control group and the problem will be addressed.

Please hit the "Back" button on your browser and try your request again.

If the problem persists, please try again later.

___________________________________________________________________________

I am now unable to post pictures:

If I try to select a picture from my hard drive the interface just ignores clicking on the OK button. If I copy and paste pictures I get the above message when I select submit after reviewing.

Then to rub salt into the wound, when you delete the pictures, review and submit you get this:-

"Could not process form: This form was already processed once, and it cannot be submitted again."

Hope you can understand what I'm getting at.

___________________________________________________________________

I was struggling with one of these scenarios the other day, it kept saying, each time I tried a different approach, that I was missing a "parameter", but it wouldn't say what parameter I was missing, since I could paste the pic in other venues, I'm dubious about anything missing...at least from my end...

Pasting pictures off my PC seems to work for me, using Chrome...

Maybe size is the problem.

Pasting a screen capture off my Mac did NOT work using Safari, when I tried to reduce its size, but leaving it alone:

When I pasted the image with no text and no editing of the image, it worked.

...or I thought it worked, but it has now disappeared.

A little more information...

https://physics.aps.org/articles/v17/15

Really interesting, but I find it difficult to believe that the eddies inside the unit have a greater effect than the molecular collisions occurring at the inputs and the bends of the input tubes, both of which have a much larger "lever arm" than the eddies have.

In order for water to be "sucked in" to the tubes, there must be a pressure gradient along the length of those tubes, and all those molecules impinging on the bends must exert some tangential forces.

Maybe the sprinkler will not rotate at all given the max pressure that the suction can apply is 100KPa of 14.76PSI. That may not be enough to rotate the sprinkler.

I think that the water entering the holes will strike the inside of the tube, causing a force against the tube. The reaction would cause the sprinkler to rotate in the normal direction.

After watching the video, I find that I was wrong! Oh, dear! Not a good way to start the day.

Seems like the viscosity of the water in the tub should have a major effect on this question.

Imagine an analogy, that you are standing stably on the rim of a standstill but free roller-coaster and pulling a rope tied, say, tangential to a stationary post. As you pull (like water sucked into the nozzle in our case), the reaction would be the roller-coaster would start rotating. So will the sprinkler rotate in the clockwise direction (the opposite direction of the sprinkler in the air).

Here is an analysis of the reverse sprinkler that is well worth reading.

Sorry, bad link, try this...

https://sciencedemonstrations.fas.harvard.edu/presentations/reverse-sprinkler

Yes, that link worked.

Since air is compressible, and water is practically non-compressible, among other things, I don't accept this as a definitive answer to the water sprinkler question.

What I found interesting in the section "A FORTUITOUS DISCOVERY":

1. The sprinkler hub spins without the L shaped nozzles, or without any nozzles at all. This shoots down my theory that the angular momentum is being added to the air entering the nozzles because of the bends.

2. Without the nozzles, the direction can be reversed by shutting off the air pump and turning it back on. It seems to be some interaction between the hub speed and the vortex speed.

3. Putting a wire mesh in the hub prevented the vortex formation and the rotation.

I still am suspicious that most of the "sprinklers" rotate in the opposite direction as with exiting fluid. It seems the bent nozzles must influence the initiation of the vortex, and the flow down the axis of the sprinkler hub energizes the vortex, similar to how updraft energizes a tornado.

"The sprinkler hub spins without the L shaped nozzles, or without any nozzles at all."

Where did you find that? I watched the video another time, and followed a few links, but read/saw nothing about no nozzles.

Where did you find that?

Sorry, this is the link I meant in #25.

https://dash.harvard.edu/bitstream/handle/1/27413727/reverse_sprinkler.pdf?sequence=4

Go to section IV

Everyone is familiar with water making a vortex going down the drain. A slight amount of rotation gets amplified as the water gets closer to the axis of rotation, much as a spinning ice skater spins faster as she pulls in her arms.

It seems likely that the sprinkler hub would also have a vortex and friction between the swirling fluid and the inner wall of the sprinkler hub would transfer angular momentum. The sprinkler hub would spin faster if the nozzles were removed and its moment of inertia was reduced.

True. The article did mention that the holes were drilled with a machine, but hand tapped. Depending on the machine used, its precision, and the skill of the operator, there is almost undoubtedly at least a slight offset from perfect centering of the holes. But the more probable source of imperfection is the hand tapping of the threads in those holes. It is virtually impossible to get a a tap started in exactly the desired direction when tapping by hand.

Even a slight deviation from the correct centering and orientation of the pipes leading from the nozzles to the hub will clearly place the stream of fluid off-center, resulting in a consistent angular momentum.

I'll wager that if several hubs were made using the exact same procedure, including the hand tapping, that at each one would have a slightly different spin velocity under the same flow conditions, and likely some would rotate in the opposite direction.

The article did mention that the holes were drilled with a machine, but hand tapped.

Probably the only relevant data not recorded was whether the hand tapper was left or right handed...

If you stop the video at 14:30, you can see that the tubes do enter at slightly offset angles.

I tried to post a picture earlier but was unable to.

Let's see if this works:

[attempted image insert]

Nope!

Haven't looked or even at other replies...I'm guessing opposite way.