Designing the Perfect Cutter

From what I have seen on Youtube videos etc. nobody has yet perfected a cutter that does not produce a droopy nose or trailing drips. Even the famous Epcot leapfrog fountain streams have droopy noses. All the others, Oase, Pentair etc. as well as all the private ones I have seen have one or both problems to a greater or lesser degree.

A mechanical cutter like most people use is simpler than an air jet or a water jet like Wet uses but the problem is getting a clean cut and uncut. I believe most cutters, including the commercial ones, fail to address the basic issues that cause trailing drips and droopy noses.

The cut and uncut are slightly different and need to be considered separately.


Cutting the Stream

I believe there are four main causes of the trailing drips:

(a) Failure to cut the stream square

If a cutter moves at right angles to the stream it will be impossible to produce a square end on the stream no matter how fast the cutter moves. That is because between the time the cutter starts cutting and finishes cutting, the far edge of the stream will have moved. The result is an angled cut. The slower the cutter moves the longer the angled taper will be.

Where the stream tapers down to nothing there is very little surface tension to hold the stream together. The tail of the stream breaks apart, causing the trailing drips.

To achieve a square end on the stream the cutter must be angled so it cuts partly in the direction of the stream, not straight across at right angles. The actual angle will depend on the speed of the cutter and the speed of the stream.

(b) Cutter Too Slow

The cutter must be very high speed. With a slow cutter even a very angled cut will still not produce a square end on the stream. Also with a slow cutter some of the water will slide around the edge of the blade and mingle with the other water, causing turbulence.


(c) Drag from The Cutter Edge

A razor sharp exit orifice is used in most nozzles to reduce drag. The same applies to the cutter. If a blunt cutter is used it will cause drag on the stream as it cuts into it, resulting in turbulence in the tail and hence drips. A razor sharp cutter must be used to prevent that.


(d) Surface Tension from Cutter Face

As the cutter swings into the stream some of the water molecules will track around the edge of the cutter and will briefly cling to the back of the blade, then break free causing turbulence again.

To prevent that the blade must be angled slightly relative to the direction of travel such that only the razor edge slides across the back of the stream.

The following diagram explains the above ideas. The colours are merely to show the position of the stream at different points in time.





The desired result is a completely square end on the stream.

Assume that the cutter moves at a 45 degree angle relative to the stream. For a 1/2" stream the far side of the cut will be 1/2" further up the stream than the start of the cut. The length of the cut, which is a diagonal, will be square root of 2 or 1.414 times 1/2" = 0.7"

According to a separate posting a 1/2" stream that is making an arc 6ft high and 12ft long will be travelling at around 29ft/sec. That means that the stream travels 1/2" in 1.4ms. So for a square cut we have to move the blade 0.7" in 1.4ms or 500" per second
(When are you guys going to change to metric???)

If we change to a 60 degree cut angle relative to a plain right-angle cut the far side of the cut will be 0.866" further up the stream than the near side. The stream takes 2.4ms to move 0.866". The length of the diagonal cut increases to 1". So we have to move the blade 1" in 2.4ms = 417" per second (35 ft/sec).

If we change to an 80 degree cut the far side of the stream will be 2.8 inches above the near side. The stream takes 7.8ms to travel 2.8 inches. The length of the cut will be 2.9 inches. So the cutter has to move 2.9 inches in 7.8ms or 359" per second.

So the greater the angle of cut the slower the cutter needs to move, even though it has to move further. The disadvantage of a cut like 80 degrees is that the cutter blade has to be about 3 inches wide to block the entire stream. A 60 degree angle is more practical.

If we assume a rotating cutter driven by a stepper motor with the distance from the centre of the shaft to the centre of the stream being 2", the circumference of the rotating cutter at that point is 2 pi 2" or 2 x 3.142 X 2" = 12.6"

With a 60 degree angle cut the cutter has to move 1" to fully cut the stream. 1" / 12.6" = 0.08 of the circumference.

That means it has to travel 0.08 of the circumference in 2.4ms or a full revolution in 30ms. That is 33 revs per second or 2,000RPM. That is beyond the capability of a normal stepper motor starting from a stationary position, given that you have to slowly increase the speed of a stepper motor. You cannot instantly go to maximum pulse rate.

I would be interested to know how fast a typical solenoid moves. My personal belief is that they would not move at that speed either (35 ft/sec).

I believe the only way of achieving the necessary speed may be to use a spring-loaded mechanism a bit like the hammer on a revolver.



Un-cutting the Stream

Maybe the nose of the stream will always droop no matter how cleanly the cutter releases the stream, due to the air pressure slowing and causing turbulence in the front of the stream.

I personally believe however that a relatively clean nose can be achieved. The issues are similar to those for the cut explained above:

(a) Failure to un-cut the stream square

The un-cut must produce a square end on the stream just like the cut.

Note that when using an angled cut the blade cannot move in one direction for the cut and reverse direction for the un-cut or the un-cut will have an extremely long tapered nose, far worse than with a right-angled cut. That is because the start of the un-cut will be further upstream than the end of the un-cut.

The blade must move in the same direction for the un-cut as it moved for the cut. That implies the need for a rotary cutter not a push-pull solenoid type cutter.

(b) Cutter Too Slow

Same issues as with the cut.

(c) Drag from The Cutter Edge

A razor sharp trailing edge must be used on the cutter.

(d) Surface Tension from Cutter Face

This is a bigger problem than for the cut. If the cutter blade is simply "dragged" out of the stream the water molecules will cling to the the surface of the blade and be given a sideways motion just before they escape around the edge of the blade. That will cause localised turbulence.

The other problem is that the blade is slightly angled relative to the direction of motion of the blade. For the uncut that will mean that the last of the molecules that do not get released will bounce off the angled blade and mix with the ones that are released, again creating turbulence.

(e) Vibration from Cutter

A solenoid knocker is used on many nozzles to destroy the laminar flow so as to allow light to escape. I believe the cutter does the same. As it activates it causes turbulence in the beginning of the stream which contributes to the droopy nose.

The only way to avoid that is to mount the cutter separately from the nozzle. That should not be a problem with most installations that are installed underground. It is definitely required when a very high speed cutter is used due to the inevitable jarring when it starts and stops.


The following diagram explains the blade configuration that I believe is needed for the un-cut. Just as you have to have a razor edge to cut off the back of the stream, so I believe you have to have a razor edge that "scrapes" across the front of the stream to cut it off cleanly, rather than just dragging the blade out of the way.

Obviously a very fast cutter is required so that the water does not catch up with the lower blade and run into it.



So using the above blade the cut sequence looks like this:




Cutter Acceleration Issues

No matter what type of mechanism is used it needs time to accelerate before cutting or uncutting. To achieve that the blade needs to be able to swing for a while before cutting the stream. When uncutting it needs to be wide enough so it can accelerate while still blocking the stream, before the trailing edge of the blade starts uncutting.

One mechanism that would avoid that inertia problem would be a continuously spinning flywheel and a clutch arrangement, but that starts to get a bit too complex.


3D Model of Cutter

Here is a possible configuration that incorporates all of the above ideas. It is a 3D model in Google Sketchup format. Sketchup is a free download. It shows an exploded view of a motor that tensions a spiral spring that drives the cutter, the cutter, a solenoid that releases/stops the cutter and a water diverter.

http://www.schoolzonelights.com.au/fountain/cutter.skp

Video of the 3D model:

https://www.youtube.com/watch?v=S7ezEwO0IY0

Note the following features of the cutter:

- Two wide blades with wide space between. Gives the cutter time to accelerate before starting the cut/uncut.
- Geared motor which tensions the spring.
- Solenoid that releases the cutter and allows it to turn one quarter turn each time.
- Upturned leading edge on the cutter blade so only the very edge scrapes across the back of the stream when being cut.
- Razor sharp leading edge.
- Second cutter blade on the bottom with razor edge to shave off the start of the stream during the un-cut.
- A diverter screen that directs the water into a collection pipe and prevents any drips affecting the stream
- Because the cutter is angled relative to the stream all water will be deflected in one direction only making capture simpler

That looks like it took a lot of thought to get that right.

Excellent idea.

I did have one thought though. Why not build a cutter out of something with no Physical surface area? Like air? What would the effect be of using a jet of air to cut and uncut. The mechanism can be seperate from the nozzle and it's on/off could be made a lot faster than any mechanical cutter.

There would be nothing wrong with air apart from the size, cost and noise of having an air compressor as part of the setup. It also requires analysis as to how it is going to achieve the cut and uncut. Is it going to split the stream like Wet do with a water jet or is it going to blast the stream off course with a wide flat jet of air? Either way the issue still remains of getting a clean cut and uncut. Nobody has achieved it yet to my knowledge.

First and for most, use whatever units you want! It really stinks having to use English units, when the rest of the world uses metric. I am constantly running into issues converting units from metric to English and back. Perhaps if you did use metric we would get an understanding of these units.



As always pmolsen, you amaze me with your attention to detail. That was very well thought out and explained expertly as well. I would love to see you succeed in this! Are you an engineer? What do you do for your profession?

I had an idea of using two different blades to cover the distance faster, but was unsuccessful due to the fact that my machine shop is a drill press. LOL! These complex ideas take some time, patience, and manufacturing skill. My design was doomed due to the complexity and the lack of manufacturing technique and tools.

I wasn't able to see the locking mechanism to keep the spring from unwinding, was it in there? My other concern (which may not be yours) is I want to be able to quickly turn on and off the cutter. Do you foresee any issue with the windup time? Do you think you could have a slug of a couple inches, say 15mm? Or better yet, do you think you can uncut and reset and cut in 17ms?

The diagram did not show everything. There needs to be some sort of ratchet mechanism to prevent the spring driving the motor backwards and unwinding. Or else you keep the motor permanently energised.

Speed is not a problem. Once the spring is wound up it can operate the cutter several times in a row while the tensioning motor catches up and re-tensions the spring. The solenoid can be activated as fast as it will move. For a really short slug you can just let the cutter rotate half a turn without activating the solenoid.

Unfortunately the whole design is more complex than I would like. Any ideas on simplifying it whilst retaining the speed would be appreciated.

I'm a computer programmer.

I'll take a serious look at it and do some brainstorming and see if I can't help out.

I am planning out my Winter projects and I would like to build several nozzles for a water display into a swimming pool. Has anyone tried to build the theoretical cutter described here? I really appreciate all the trial-and-error work that is shared on this forum - it saves many of us many hours. My goal to to find the best cutter design and hopefully find some simple step-by-step instructions for building a programmable controller to run 5 to eights nozzles with cutter and LEDs. Many hours in the shop are ahead of me.

To my knowledge no one has attempted the theoretical cutter.

I've been thinking about the idea of a cutter in a different way. Instead of blocking the flow with a mechanical device, why not divert the flow of the water out of the nozzle back to the resivoir. I would use a normally closed solenoid valve connected to the top of the nozzle. The output of the solenoid is controlled with a flow control valve so that only the necessary amount of water excapes from the nozzle. When the solenoid is turned on, the water is still moving through the nozzle but exiting the side, when you wish to have water exit the nozzle, the solenoid is turned off and the water flow exits the nozzle. Just my thoughts.

That's a very interesting idea. I like it. One question do you think that the pressure inside the jet will be constant, and not have to ramp up. Having the cutter on the outside creates problems with having noses, and tails, but when you uncut or cut you know that there is going to be all the pressure already stored up in the jet. I wonder if that would be constant with the diverting the water.

My thoughts on pressure is that there may be a small ramp up due to not always having the total volume of the nozzle filled. I have a solenoid and a small 4" nozzle that I've helped my daughter build for her science fair project. What I don't have is a flow control valve but I could fake that out with a fixed orifice the same size as the nozzle outlet or slightly bigger.

You could just do a test and drill the hole the same size as the exit orifice, and just cover it with your finger to see if the theory will work.

I like the idea of a cutter that move in the same speed and direction of the water.
That have always been in my head.

btw: your detailed images\video is incredible !!!
I have an idea, but only got 10 minutes + paint
So do you understand this simple drawing:


Closed and open !
The idea is that there is 2 wheels spinning in the same speed as the water moves (spin when cut to close, and spin when open again) !
In the 2 wheels there is 2 cutter blade, one blade on each wheel.
The blade will always stay in horizon with the laminar (as you see on the picture) when the wheel turn.

So when it`s open the blade is away from water, when the wheel turn around both blades enter the water from each side in the same speed UP as the water travel.
When the wheel\blade is in closed position they STOP !

When you uncut, the wheel turn with high speed, and you UNCUT the water in the same speed it travel out from the laminar jet
If you dont understand the pictures let me try to make a better one.
But the pictures\blades\cutter moves like the wheels on a steam-train


This cutter is actually used in the industry to cut hot steal moving in speed of 100mph !
It`s incredibly amazingly to see it move in action IRL !

Your cutters seem very interesting, I'd love to see them in action. I believe that the cutter has everything to do in having a perfect tail, but to have a perfect nose, that's another story. I haven't built a nozzle myself but I have seen videos of your projects. I believe that after you have a perfect cutter, you'll still have an imperfect nose in your stream due to the air resistance. The nose is the only part of the stream that fights air in order to reach the other side. The water that comes after the nose, doesn't need to fight the air as the nose did. And once the stream reaches the other side, since it's a continuous stream, no water will ever have to fight that air again until the stream is cut again. Think of it as race cars, they position themselves behind the rival's car so that they don't have to fight the air, and they get a speed boost, a drastical boost.

I know this seems really obvious, but it's something that the cutter just won't solve, even if it's perfect. What I propose is something complicated and requires a lot of testing/calculation and most likely is not even possible but might be worth the shot. What I believe causes the nose, is the speed reduction of the nose due to air. But what if we were to increase the pressure in the nozzle just for a brief moment just as we uncut so that the nose has a higher speed. Then we would linearly decrease the pressure to the nominal value in the jet. The nose would travel higher in the beginning but then, due to the air resistance, it would slow down a little and place itself in the same path as the rest of the stream. To do this, we need a pressure sensor, a feedback circuit and control algorithm. We would probably have to measure the speed and direction of the wind as well. To change the pressure, we could use a fast electronic valve or something at the entrance of the nozzle. Controlling the speed of the pump would be harder since we have a low pass filter in play, the chance we do in the pump would be filtered by it, unless we remove the filter and create a digital low pass filter using the pressure sensor and controlling the speed of the pump. That way we would have a constant pressure the whole time and when we uncut, we would speed the pump a little (some milliseconds earlier of course so pressure reaches the nozzle) and then back.

Probably, this wont remove the nose but it could make it look better, at least I hope.

I like your input scarfy.

many factors and thing to think about here ....

I'm almost done with my first nozzle, I'm just missing the brass rings (ordered 2 from John but they haven't arrived, probably they got lost or customs decided that they were too good for me to have). Once I have everything working, I'll build a cutter and install a pressure sensor in my nozzle and low pass filter, to graph the behavior of the nozzle. Then I'll try and build a control circuit to work with the cutter to increase the pressure when uncutting and I'll post the results, however, it might take a while for me to do that since I barely have time for this.

That seems like an incredible feat! I would love to undertake that but alas there isn't enough money and time to do so in my life. Maybe there is a way we could pool resources and each do a part or something?

Giving it more thought, maybe there is no need to have a feedback oriented control system. Maybe by just increasing the pressure when uncutting by opening a valve a little just before uncutting then closing the valve again after some milliseconds to give the extra pressure needed to fix the nose. This can then be adjusted with a trial and error method.

How would you increase the pressure on the pump? What do you use to control the pump?

A valve at the entrance of the nozzle or if you remove the low pass filter you could install a pressure sensor and control the speed of the pump to filter out any pressure transients and at the same time, increase the pressure in the nozzle when needed.

Any new ideas about the cutter ?

I have some trouble with my simple cutter.
The cutter is 2 magnetic solenoids that push and pull the cutter into the water and away again.

Anyway, the trouble is the splashing water that moves all over the place when I "CUT" the water.
The splashing water goes around the face of my laminar nozzle, and when I "uncut" , the water drops down into the water output.
And that make the laminar stream unstable until all drops of water is away.

Is there any other solution on this that can work better ?
I use a metal blade to cut away the water, maybe I should have used a 90 degree small tube that will drain away the water to another place?