Underwater speed?

[Ramius]

Hi All! :)
I have found that some of the smartest people are here on this site. I wish to measure the speed of an R/C submarine using pressure sensor(s). I have seen the "wheel" types which seem very large. It would probably be very unsightly to have a long pitot tube sticking out the front. It just seems "logical" that if you measure the pressure of the water on the bow you could somehow calculate the speed? Yes I did look at the "eagletree" system which uses a pitot tube and I doubt since they sell sensors they would wish to provide any information. Any ideas would be very much appreciated.

Thanks in advance.

Ed

[Demon]

Look at liquid cooled PCs, some have a flow meter. You might be able to use that.

I can't see how you can avoid a pitot tube. I'd put it below the waterline when surfaced, centerlined with sub. You'd only see it when you take the sub out.

Robert

[mark_s]

Maybe you could use a flowmeter. They have a plastic impeller with a magnet and a hall effect switch to send pulses to the pic. Could be mounted inside the hall with external ports in the front and back to allow water to flow through. Just an idea.

http://www.ebay.com/itm/Water-flow-sensor-flowmeter-Hall-flow-sensor-Water-control-1-30L-min-2-0MPa-/200910300858?pt=LH_DefaultDomain_0&hash=item2ec72fe6ba

[kellyseye]

Have you considered the effect of tides/water movement?

One method in use is to generate a magnetic field then measure the voltage induced into a pair of sensors by the water flowing through the field (it being the conductor). Obviously using the sub in fresh water will produce some problems.

[Ramius]

Thanks Guys!
All are great ideas! I already have a depth sensor accurate to 1/2" and the thought was to use a second pressure sensor mounted in a torpedo tube to look at the water pressure on the bow. Where I am stuck is how to figure out how to conver the incomming pressure to speed? I also have temperature sensors. My first guess and I could be completely wrong is that the depth sensor would be the same as the "static" pressure and that the static pressure is subtracted from the incomming pressure? I just do not have any working knowledge of how speed is done.

Best, Ed

[Demon]

I'd try some sampling. Stick the sensor/tube on the front of a piece of wood (broom handle), drag the setup in water at a certain feet/second, record the sensor values, then convert to whatever unit you want.

It all sounds very good in my head anyways.

Robert

AND TAKE PICS!
:)

[AvionicsMaster1]

What would be the max speed?

If you did mount it to the bow would you have to use a pitot tube? I was always taught pitot tubes put the sensed inlet in front of disturbed air, water in your case, to give you a more accurate indication. Could you have a hole in the bow, torpedo tube might work, and measure the pressure on that face of the water? Would you have to worry about purging the tube of air to get an accurate indication? What is subs diameter and what power supply do you have?

You might have answered these questions in your previous thread but I couldn't find it. Just some questions I thought of and was curious.

[kellyseye]

How about doppler? Modern speed instruments (for deep sea vessels where you can't get an echo back from the bottom) use doppler (iirc at around 2MHz) - you reflect off the bottom. It might work using ultra sonics (think echos sounder pointing 45 degrees forwards).

[Ramius]

Wow!
So many great ideas! Thanks! I think that maximum speed would not be more than say 10 mph? I occured to me on the way home that if you look at the water pressure on the bow you would have to subtract the depth reading for as you go down the pressure on the bow sensor would be reading the water pressure at that depth! Doppler sounds good and wouldn't the return become shorter the closer to the bottom you get? Robert maybe great minds think alike? I put together basically a motor to turn a shaft and on the end of the shaft I can mount the pressure sensor a given distance from center to calculate distance travelled. There is also a timer to tell how long it takes for one revolution. The sub itself is 8" in diameter, 6' long, and 1:100 scale. Mounting a pressure sensor just inside one of the torpedo tube openings should not allow any air to accumulate.

Best, Ed :)

[AvionicsMaster1]

Holy macaroni! That's bigger than a few compact cars I've rented.

I'm more familiar with aircraft and they do reference airspeed to altitude(pressure density) but mostly to find Mach. They reference airspeed to temperature to find true airspeed. Temp would be easier than a differential pressure to sense I think. Would water density or (there has to be a term for the column of water above the item) induce much of an error? I guess I assumed a local pond or small lake not the ocean.

What depth is absolute maximum? Please don't tell me the bottom of the lake. (It's an old submariners joke.)

[Ioannis]

I wonder, how do you control the sub when it is under water line?

Ioannis

[languer]

The sub itself is 8" in diameter, 6' long, and 1:100 scale.
If you don't mind, I would love to see pictures of this. And of course the method you choose to perform the speed measurement under water.

[Ramius]

I think that the max depth is about 50'? There is a pressure sensor to prevent the sub from going too deep. R/C subs operate at 54 MHZ which has little difficulty penetrating water. The video camera system for the periscope operates at 1.2 GHZ. FYI 2.4 GHZ is the frequency for microwave ovens so to use this frequency would result in you only warming the water around your antenna! :eek: I work at Fox Television and have a ham radio license so a friend of mine and I designed an antenna that sends the video in an up direction only so the periscope will work underwater which has been tested. Every 33' the water pressure doubles from surface pressure so if you only used one pressure sensor you would be seeing the depth as speed when stationary. I think differential pressure is the way to go, water pressure on the bow as you move minus the pressure at depth. What I do not know how to calculate is velocity. Pictures are similar to mine.

Thanks to all and best wishes, Ed

6985


6986

[languer]

That is quite awesome. Could you have some sort of towed array to serve a GPS antenna? That way a GPS module can give you the speed, position, etc.

Another method I saw was to use two ultrasonic transducers; one on front and and one on back. The time for a pulse to reach the rear transducer through the water will be a fuction of the speed of the boat (think SONAR).

[Ioannis]

Very impressive! Like it a lot!

Ioannis

[Demon]

What about 2 pressure sensors front and back? The difference tells you velocity.

Robert

[Acetronics2]

Nice thing ...

@ 2000 Euros per piece ( 2500 US $ ) ... ;)

Alain

[AvionicsMaster1]

Wow! Seems like such a simple question but I think it might require a complicated answer.

This site http://www.engineeringtoolbox.com/dynamic-pressure-d_1037.html can get you close but you've got to know density. This wiki site http://en.wikipedia.org/wiki/Density has a chart that shows water density at temperature.

So if you subsitute the fluid density p(rho) with the equation m/V your equation would be pd = 1/2 (m/V) v2. Solving for velocity would be, well I can't figure out how to write that here but you get the idea. What all this is trying to say is you'll have to get temperature somewhere.

I agree with all in that I'd like to see how you solve this problem.

[Demon]

(Talking out my wazoo on this)

But would density have a huge impact?

Is the sub going to be used in extreme temps, or only between ~15C - 20C?

Personally, I'd use it in temps and depths that I could dive to get it back (got no diving suit). But if you want to set a record crossing the English Channel, that's different.

Robert

[tasmod]

Hmm, your driving it with a propeller. Why not a small free rotating prop on conning tower and measure rotations.

There's example code on here somewhere for wind speed anenometer, which is basically same.

Rob

[Demon]

Won't work when surfaced. :)

Robert

[tasmod]

Conning tower location was just a suggestion, it could just as easily be on hull etc.

KISS. Small prop with photo reflector for pulse count. Mounted in small tube.

Easier to get going, then maybe try sophisticated solutions.

Rob

[NickMu]

Hi Ramius,

Can you provide some more information on your radio controller?
I’m interested on range, maximum depth and more importantly where do you purchase it from?
Regards,


Nick

[Ramius]

Hi Nick!
Here is a link to the radio controller: http://www.conrad-uk.com/ce/en/product/384271?WT.mc_id=webgains&insert=8X It has two throttles, one for the port motor and one for the starboard motor and a ton of expansion modules! I originally bought it from Ships n' Things. Range and depth are a funtion of the both transmit and receiver antenna's and how they are configured. You basically are trying to simulate a "dipole" for receive and you can use a yagi for transmit. Part of the other factors are where you place all your electrical cables and how you have prevented EMI (Electrical-Mechanical) and RFI (radio frequency) interference. For example motors produce small sparks which are basically a radio signal. So you have to use bypass capacitors. How you ground the motor cases and how you run power all make a difference. The technique I use is called "single point" power and ground where all the power wires connect to one single point and well as the ground wires. In addition power and ground wires have been put in an electric drill and twisted together. There are a lot of technical reasons for doing this. If I strike out with a pressure sensor then the propeller idea will be next!

Best to All, Ed

[NickMu]

Hi Ed,

Thank you for your detailed reply.
I had an application that needed to read the thrust of a jet of water and before our order of force sensors came in we used a paddle wheel to determine the speed of the jet.
The results were within 5% of the real sensors readings.
Depending on your precision and accuracy requirements any of the suggested methods should work but with some penalties.
IMHO your biggest problem will be the changes in water density due to various reasons (salt, temperature…..)

Thank you again and good luck.

Regards,

Nick

[Ramius]

Hi Nick!
Thank you and I would not mind a paddle wheel type except they all seem to be really big! I do not believe water density will be much of a problem since I will only be operating in fresh water and for a sensor I am using an MS5541C's which are temperature compensated. What I really do not know is the actual formula for velocity. Water density values seem to be all over the map. Do you know of or have a simple formula I can program into a PIC (PIC Basic)?

Antenna work is almost a "Black Art". For a transmitter your main concern is what is called SWR (standing Wave Radio). Basically you are sending power up a coax and when the power hits the antenna if you have the correct antenna and correct impedence (AC resistance) then none or a very minor amount of power will be reflected back to the transmitter or used to "heat" the coax. On a receiver you have the same problem in reverse where you want all the received signal you can get! A dipole (two wires 180 degrees from each other) has an impedence of 300 ohm. When the two wire are at 90 degrees you are looking at about 50 ohms. It even becomes more complicated with the lenght of the antenna wires! That is full wave lenght, half wave lenght, 5/8 wave lenght, 1/4 wave lenght. It can boggle the mind! ;)

Best, Ed

[Demon]

I was thinking of these. Just need 2 fittings to connect hoses in/out.

http://koolance.com/ins-fm16-coolant-flow-meter

You should be able to make your own controller with PIC.

Robert

[Demon]

Formula for velocity:

Follow sub in a boat at max speed, look at knots on gps, note reading from flow meter.

Half velocity = 50% knots, etc.

Or is that too simple?

Robert

[NickMu]

Hi Ed,

I knew it is not an easy task to get such system going especially when in my application the transmitter will be above and bellow water at times. I will try some other avenues first and if no success I might bother you with some more questions.

I will be away from my office for a few more days so I do not have access to my notes.
For our test we just calculated the speed of the jet and the volume of the water passing through in a time period. With the water density as a constant we were able to calculate the thrust.
In your case I’m quite sure you will need at least your differential pressure, the surface of the sensor (maybe) and the water viscosity. I will look into getting you a formula combining all the factors from which you can extract the speed as the only unknown.

What will be an acceptable size for the paddle wheel? The one I used was a little over 1” diameter and it is used on Sea-Doo PWC for speed indicator. It gives 4 pulses per revolution and it is a completely sealed unit wit three wires interface (power in and signal out). If you are interested I can send you one as soon as I get access to it. Just PM me your address and you will have a possible back up solution.

For the differential pressure method how hard will be for you to take a couple of readings at half and full speed to see if your numbers will make some sense. From that it will only be a matter of coding.

Regards,

Nick

[NickMu]

What you are dealing with is the case of dynamic pressure in fluids. One of your sensors will read the static pressure and the other one will read the total pressure. The difference will be your dynamic pressure. Using the formula: P = q*V² you can extract your speed.
P= dynamic pressure
q= density of water
V= speed
So you have your readings for dynamic pressure and you have your water density (I’m not sure if you have to compensate for temperature again since your pressure readings are already compensated) and you can calculate your speed.
I had the chance to do some Google search and I found this:

http://www.engineeringtoolbox.com/dynamic-pressure-d_1037.html

HTH.

Regards,

Nick

[NickMu]

One mistake on my formula:
The left side should be multiplied by 2 or the right side should be divided by 2.

Regards,

Nick

[AvionicsMaster1]

Just sayin', that's what I said. Even used the same link. Compensation for density in same post. Number 18 if you'd care to peruse.

[NickMu]

Avionics,

I can not explain the duplicates of your suggestions in your earlier posts other than I started my input to this topic late and I did not carefully read the posts before my question to Ed about the RC system.
I felt that I should share the little information I had but the last thing in my mind was to take credit for something that you already posted. Ed asked for a possible formula and, without reading the earlier posts, I gave him the best I could find just to repay his generous information exchange about RC system.
Please accept my apologies for that.
If one positive thing will surface from this is that it could be Ed’s best solution since two separate people thought the same solution which is an “engineering nonsense”. You will always have n+ 1 solutions from n designers. Each one has a different solution and the +1 is the one that should be used.

Again, it was just my superficial attention to this topic and I’m sorry for it.

Best regards,

Nick

[Ramius]

Thanks Nick!
If it is 1" is that the body of the paddle wheel or the pipe inlet size? The reason I am asking and trying to keep everything as small as possible is that the bow are is starting to become very crowded as this is where the torpedoes and missles are plus a mechanism to deploy and retract the bow planes!

Ed

[NickMu]

Hi Ed,

The sensor I used does not have your traditional cylindrical shape.
Just to have an idea of the way it looks go here:
http://www.jets24.com/bombardier-brp-seadoo-jet-ski/electric/speed-sensor-9198.html?___store=english&___from_store=german
Also if you search Sea-Doo part 278001998 you might find more pictures to get a better idea. A view from the oposite side will give you more information.
The paddle wheel is about 1 ¼” diameter. The mounting plate is about 2”x5” but you can cut most of it off. With a little creative engineering you can hide most of the body and expose only the tips of the paddles.
It is used but in working order. I will send you pictures in a few days when I will get back to work.
My offer still stands but it will take few more days until I will be able o mail it to you.

Regards,

Nick

[AvionicsMaster1]

Sorry. I was just ecstatic someone agreed with me. You're right I should have done a better job of that.

Anyway, the downside is I finally got hold of an engineer I know who has a maritime background. He said paddle wheel thing is about your best bet as you wouldn't have to compensate for density: an inch of water is an inch of water. You could have inlet in forward facing torpedo tube and outlet in aft facing tube. You could also shift it to a non watertight compartment, creating room in bow, if you can find one that is submersible.

He also suggested a way I havent' seen posted and that is to use and accelerometer. True it would only work in still water but there are several phones that have internal accels built into them. Just a thought.

[NickMu]

The paddle wheel I used is completely sealed and, after finding the right spot for it, all you have to do is run the three wire cable inside. There will be only one time calibration required (determine the “active diameter” of the wheel) and after that just use your favorite method (frequency or period of the pulses) to calculate the speed.

Regards,

Nick

[languer]

I agree the paddle wheel is a good, simple idea. But I don't think the paddle wheel by itself can tell you the speed of the boat. If I remember my dynamics class, the paddle wheel will tell you the flow in the wheel. Which will be a function of two forces; (1) thrust of the sub, and (2) thrust of the water. In other words if the sub is travelling against a strong stream, the paddle wheel (I think) will incorrectly state the sub travelling at higher speed. If the sub is travelling in the same direction of the stream, the paddle wheel (again I think) will incorrectly state the sub travelling at slower slower speed. If the sub is travelling in still water, the paddle wheel will correctly indicate the speed. I believe you also would need to know the thrust of the sub (i.e. propeller movement) to properly calculate the actual speed of the sub (or maybe I am all screwey - which happens quite often).

[NickMu]

Hi Languer,

40 – 45 years ago these things were much clearer in my mind.
I can safely say that you are correct. And to add to the injury when checking the speed with paddle wheel in a stream not only that your readings are shifted with ± stream’s speed but they will be different depending on your position within the stream’s body: the closer to the walls and bottom the slower the water speed.

So the only accurate way to check the speed will be counting the time spent between two telegraph posts (LOL) or GPS based. Any other of the suggested methods will require some sort of calibration before each use if the environment changed.

Let’s look at the bright side: I recall some posts ago that Ed’s answers to various questions suggest that he will be using his sub only in fresh steady water.

Regards,

Nick

[AvionicsMaster1]

Ok, here's what you do. Instead of all this rigamoro (sp?), get yourself a Defense Department contract to come up with a sonar useable on an inlet waterway. Sequester, schmester the DOD always has money somewhere floating (no pun intended) around.

Then you develop this totally awesome system but forgetting to tell them it can be sensed by a North Korean satellite orbiting near earth. Then, when they ask you to protect it from emissions, you tell them you can't and they will suddenly lose interest. Presto, you get to keep the hardware cause the government can't figure out where to store it. You don't care if the North Koreans can track your submarine. They get to spend lots of money 'cause that's they do and we're all happy. Well, we'll be happy if you send pix and code. That'll solve your problem and you might make enough money so you won't have to work again. Only in America could this scheme work.

[Ramius]

Hi All!
Again you guys are very correct. I cannot imagine running the sub in a stream, only a lake, pond, or swimming pool with still waters. In aircraft there are two measurements of speed, IAS (indicated air speed) and TAS (true air speed). IAS is the speed weather into the wind or with the wind while TAS is the speed over the ground as if you were a car. In this case we can realistically only work with IAS as far as I know.

Best, Ed

[AvionicsMaster1]

How much money do you have? I found this http://www.airmartechnology.com/uploads/brochures/CS4500.pdf by accident and wonder if it might work. I'm also wondering if you couldn't find one used at a dock or salvager.

The brochure is light on how to hook it up but it would fit your needs almost perfectly. After a quick search I couldn't find anything on the "Airmar paddlewheel format" this is supposed to put out.

[MITechMike]

Hello Ramius
It looks like folks here were very helpful and I'm curious about what you came up with for a solution. I have a project that I'm working on and need speed through the water, using as small a sensor as I can.
Thanks
Mike

[Ramius]

Hi Mike!
Actually there are now a lot of answers to basic questions. First save yourself a lot of research by knowing that speed = PSI! As for the PIC code you can find it here at:http://www.picbasic.co.uk/forum/showthread.php?t=19572. For me I needed a "differential" pressure sensor from Seroflo ([email protected]). The reason for a differential is that as my sub dives pressure will increase so you have to subtract out "depth" pressure (Static Pressure) from the pressure on the bow (velocity Pressure). One technique I have used with my compass was to mount the sensor on top of the PIC. If you are using a surface mount PIC you will see the PIC fits comfortably under just about any sensor. You can use a CAD program to design your PC board and then convert the CAD drawing to Gerber files. It helps to create "Layers" in Cad that you can turn on and off so you can create you solder mask and "ground plane". You will probably end up with a double sided board so you can make one side say in red traces and the other side in blue traces. Do not forget to specify plated thru holes! Your board size will most likely be .625 x .625". About the only other design consideration will be to isolate the sensor from the water. Hope this helps.

Best, Ed

[Demon]

How much money do you have? I found this http://www.airmartechnology.com/uploads/brochures/CS4500.pdf by accident and wonder if it might work. I'm also wondering if you couldn't find one used at a dock or salvager.

The brochure is light on how to hook it up but it would fit your needs almost perfectly. After a quick search I couldn't find anything on the "Airmar paddlewheel format" this is supposed to put out.

That's very interesting.

I wonder what would be involved in rolling your own version. Ultrasonic implies the use of sound, are there new devices out for robotics that could be used for this?

I'm looking at the speeds involved in the specs at it's not NASA-type stuff. Except for the sensitivity of the sensor device, the rest seems within our grasp.

Robert

[Acetronics2]

Weight:
—1.4 kg (3 lb)—Plastic
—1.8 kg (4 lb)—Bronze


usable aboard a model ... ???

Alain

[Ramius]

Hi Guys!
The Airmar sensor requires a hole almost 3" in diameter! In models this is huge in comparison to a 1/8" pitot tube or just a single pressure tube used with a complete pressure sensor circuit that only takes up .625" (3/8") square! Ultrasonics is great and again the transmitter and receiver are about 1". Mike asked for something as small as possible. 3/8" is about as small as I can think of.

Best, Ed

[Amoque]

What about a radiometric hall effect type sensor/s placed on the inside hull and a magnetic steel wire (guitar string?) running perpendicular on the outside - equipped with a small hydro-resistive plane on the loose end? As the speed of the ship increases, current moves the wire and the hall sensor/s record the reduced/ increase field strength. I think that adjusting the sensitivity is as easy as adjusting the length and wire diameter of the sensor and the placement of the sensor along the length. It might take two sensors - one forward, one reverse.

I also think a small tab hanging below the hull on a hinged fitting... an angle sensor measures the deflection as the sub moves through the water. May be oriented horizontally or vertically, using a spring. Forward is positive angles, reverse is negative...

The ultrasonic holds promise - if the resistance of the external mechanism is not an issue. I imagine either one transmitter mounted flush amidships - a triangular reflector splitting the ping to receivers mounted fore and aft, or a more traditional setup with transmitter and receiver on both ends. Could temperature be eliminated as a factor by calibrating an offset at zero speed? Pressure remains a factor I think, but didn't I read that is currently measured?