What's a good way to measure liquid's pressure and/or velocity?

[cave_cat]

suppose I have a long thin pipe. There is some liquid inside (probably water, but can add stuff to it if needed). On a certain segment of the pipe I want to measure the velocity and/or pressure of the liquid at various points along its length and feed the results into the computer. Flow of liquid is quite erratic, sometimes changing direction every 30 seconds, experiencing big pressure changes and so forth.

I do not really care about high accuracy (i.e. +- 10% is fine). Time resolution is a bit more important. As long as it sort of gives the general picture and, most importantly, is cheap and easy to build, I am happy.

Questions:

1. how small is a miniature pressure sensor? how much does it cost? can you give me a link to the manufacturer's website? If there are miniature pressure sensors working based on different physical principles, what should I know about them in terms of their advantages/disadvantages?

2. what are the various ways to measure velocity of the liquid using miniature and inexpensive sensors? All I can think up front now is some clever laser shining through the stream onto a sensor on the other side and a magnetic device tracking the motion of iron dust in the liquid. I suspect that both of these ideas are NOT the simplest way to do it and probably are not something I could order from the manufacturer by overnight mail. Not being a trained mechanical engineer, I am basically at a loss here.

Could the experts here please give me some smarter ideas? Is there a miniature liquid velocity sensor that can be bought inexpensively?

Also, in terms of "inexpensively", ideally I want to build it from the start in such a way that the system could eventually be manufactured and sold for $400 or something. I could spend lots of money now (hypothetically, in reality I would need to find it first ) but ultimately I want to use sensors that can be bought in bulk.

3. suppose, through discussion on this thread or elsewhere in private, we figure out which sensors to buy and they prove to be straightforward to use. Does anybody want to try build something like that for me, working under NDA, noncompete etc? How much would such a project cost? How long would it take? Note, for the hypothetical future, if you end up working on this project seriously part-time, I would arrange that you would have a lab separate from your university's, to avoid their IP grabbing policies.

Thanks a lot in advance to all who respond. If you wish to respond privately, please send me an email through the site.

 

[mgb_phys]

Can you access the liquid? ie can you drill a hole in the pipe to allow a sensor to sense the fluid?
That's the simplest - in which case you can use any small off the shelf pressure sensor. They are all basically the same principle, a flexible membrane and a strain gauge to measure how much the membrane bends. You can even buy them with builtin ADC and uProc which will send serial data direct to your PC.

If you can't access the liquid and the pipe is rigid it's a bit trickier - there might be optical or sonic doppler flow meters you can use.

 

[cave_cat]

yes, I can access the liquid and modify it in various ways. The pipe is probably 3 millimeters in diameter and is made of flexible material (like rubber).

How big is the off-the-shelf pressure sensor you speak of? How much does it cost? Can I look at its spec page? Can I have 10 such sensors in different places in a 4 meter pipe? Can I route the output from these 10 sensors into a single physical "bus" cable or do I have to connect each one separately directly to the analog-to-digital converter and/or computer itself?

Also, is there a way to measure velocity? Or nobody measures velocity in thin pipes?

 

[mgb_phys]

they vary from about 3mm square upward re,eber you could run a small tube from your pipe to the sensor - the pressure would be the same at both ends (assuming it doesn't expand).
How much depends on accuracy, range, temperature calibration etc - I can't recommend any particular maker. http://www.sensorland.com/HowPage004.html

 

[brewnog]

Also, is there a way to measure velocity? Or nobody measures velocity in thin pipes?

How thin is thin? Do you have control over routing of the pipe? What range of values are you expecting? What is the fluid?

 

[coreyh]

Hi Cave Cat,
A few things you need to define about your problem before selecting your pressure sensors:
- what pressure range are you talking about?
- what frequency are the fluctuations in the pressure?
- what are their magnitude above the average pressure?
These criteria will help you in determining the right type of pressure sensor for you application. If you need very high time resolution then this will mean your sensors and DAQ will be expensive, $400 might not even get you one of them.
Also, if you just have a rubber hose, then it might be a bit tricky adding sealed ports for your sensors along the length, but i guess you could just drill a small hole and seal it with glue or something.
As for velocity, since water is pretty much incompressible, i would just recommend measuring the flowrate at one end of the pipe (once again, depending on the time resolution needed this will effect the selection and cost of your flow meter) and then calculating the mean velocity: Q=V*A, where Q is flowrate in m3/s, V is velocity in m/s and A is pipe cross sectional area in m2.
As for acquiring the pressure signals, you would need a 10channel data acquisition card, which adapts to your computer via PCI or USB and then some software to log and store the data. As above, the logging frequency that you require will determine the DAQ card you need. If it is more than say 10 or 100Hz, then you are getting quite expensive!

 

[Dadface]

1.By measuring the outflow volume per second and the internal diameter of the pipe you can calculate the velocity of the liquid.Provided that all parts of the pipe are filled with liquid the average velocity should be the same at all parts along the length of the pipe although it will vary across each cross sectional area due to viscous forces and turbulence
2.The pressure at the inlet can be measured by feeding the pipe from a constant pressure head which can be easily knocked together.The pressure difference along the pipe will be h*liquid density*g,where h is the pressure head.Knowing the pressure difference along the whole length of pipe you can get a reasonable estimate of the pressure at any point along the pipe.

 

[fisherj]

You might find these links to products and flow calculations useful:

Minitature pressure transducers
http://www.sensorsone.co.uk/products/42/Miniature.html

Digital differential pressure sensors
http://www.sensorsone.co.uk/products/169/Digital-Differential-Pressure-Sensors.html

Pressure sensors for connecting to a PC
http://www.sensorsone.co.uk/products/303/USB-Cable-Connection-Pressure-Sensors.html

Converting differential pressure into flow
http://www.flowmeterdirectory.com/flowmeter_flow_calc.html