In process particle size inspection

[TIMMAAYY]

Any suggestions or ideas on a sensing device that can detect particle size within a stream of DI water? Flow rate is 6 l/min, ambient temperature, 3/8" ID tube.

Looking for a way to detect any type of contamination in the stream ( rinse water after pressure testing ) and have specifications that finding any particle over 200 microns in the stream will require rinsing again, repeat until no particles of that size are detected.

Needs to have real time sensing, capable of 6l/min flow rate, and repeat every 90 seconds. I have found many sensors that are capable of this task, except for the flow rate. The most I can find is 60ml per minute. I would need ten of these on a manifold to do this, and that would introduce more possibilities for errors and false readings.

Any suggestions? Thanks.

 

[MintJulep]

Could you get away with ensuring (or assuming) that your flow is well mixed and divert a tenth of it for measurement with the 60ml/min sensor?

 

[btrueblood]

You mean 1/100th of the flow, Mint. OR - buy 100 sensors.

Can you not catch and subsequently filter the water thru a millipore filter? How often is this process going to run, do you really need a real-time indication?

 

[MintJulep]

Damn metric system. How can anyone be expected to do factors of ten in their head

I think the intent of the measurement is an indication of system cleanliness, not that the fluid itself needs to be clean.

 

[TIMMAAYY]

For clarification, it would require 100 sensors, my bad. 60ml x 100 = 6,000 ml - 6 l. Damn.

And I'm not looking to clean the water. I'm looking to inspect what's IN the water. This is a final rinse on the part prior to packaging. The part is pressure tested, and any burrs or shavings that would still be entrained are flushed out. If anything over 200 microns is seen, the flushing operation is to be repeated until nothing of that size is seen in the rinse stream.

I was hoping I could find a sensor/meter capable of handling 6l/min and detection of particle size in the range, whether it be by direct measurement or a relative measurement (condcutivity, etc. ). I know conductivity is out. 8^)

 

[TIMMAAYY]

Tacky double post.

Process needs to be repeatable every 90 seconds. This is 100% inspection, as it's a critical part of the fuel delivery system. Real time is needed.

 

[cloa]

Are these particles magnetic ie. ferritic steel?

 

[TIMMAAYY]

Some are, some aren't. Could be tooling, could be cast iron, could be aluminum.

 

[NickE]

Try looking at mining/ore processing companies. There were several methods discussed in the one class I had, that was 15 years ago though.

nick

Try Autocompu

Nick
I love materials science!

 

[arunmrao]

http://adsabs.harvard.edu/abs/1985crrm.reptS....S

Single particle sizing by measurement of Brownian motion
Stanton, A. C.; Cheng, W. K.
In AFOSR The 1985 AFOSR/AFRPL Chem. Rocket Res. Meeting 3 p (SEE N86-13507 04-28)
A new optical technique for the measurement of submicron particles, using a laser interferometric system to measure the Brownian motion of isolated particles in a gas stream is investigated. By measurement of the motion of many such particles, the size distribution may be obtained. This technique is unique in that it requires no a priori assumption on the particle size distribution, nor knowledge of the particle optical properties. The experimental approach utilizes a laser interferometer system for measurement of a time dependent signal sensitive to particle displacement. To understand this approach, an appreciation of the characteristic aspects of the Brownian motion of an isolated particle is required. A representation of the motion of a 0.1 micrometer particle is shown. Random changes in particle velocity occur in characteristic times of 10 to the minus 14th power seconds due to collisions with gas molecules. With sufficient time resolution to measure instantaneous particle velocities, one would observe a Maxwellian velocity distribution. The characteristic time for the Maxwellian distribution to evolve is the relaxation time, which is a function of particle size and fluid properties. To understand the properties of the signal, the Brownian motion sensor was simulated using a Monte Carlo technique.
Keywords: BROWNIAN MOVEMENTS, COMBUSTION, LASER INTERFEROMETRY, PARTICLE SIZE DISTRIBUTION, SIZE DETERMINATION, MONTE CARLO METHOD, OXIDATION, RELAXATION TIME, SOOT, TIME DEPENDENCE, VELOCITY DISTRIBUTION


I have not failed. I've just found 10,000 ways that won't work." — Thomas Edison
_____________________________________

 

[NickE]

Ooopss.. That was Outotec...

Any way this machine can sample at 170L/min, and size ranges from 1-500um (0.001mm - 0.500mm)...

http://www.outotec.com/pages/Page.aspx?id=36490&epslanguage=EN

Nick
I love materials science!

 

[TIMMAAYY]

I'll check it out. Thanks. I'm looking for particles in fluid, not a gas, but maybe the same principles apply.

 

[Dissymmetry]

Sounds like you are testing for high pressure fuel or hydraulic lines for aircraft. I'm not sure if they are competitors of yours or not, but Aerofit Products, Acra Aerospace and Airdrome Parts Company all make these kinds of fittings and perform this test in-house. They may be willing to either perform the testing for you or give you information on how they perform the test.

 

[TIMMAAYY]

That's not the case at all. It's a tube leading to the injectors on a diesel engine.

So far I think I found the correct technology, laser diffraction, but I have not found the equipment that will handle the range of flow and minute amount of particles that are to be sensed in the stream.

One one hand, there are a couple dozen small units capable of 60mL/min, on the other end of the spectrum there is equipment made for the mining industry that costs about $300k and is the size of a large refrigerator.

I need something in between.