Gas Jet Plume Measurement Techniques??

The only true measurement of a system is one taken while the system is in actual, everyday use. That doesn't mean that useful information can't be teased out of a proto-type etc. There are quite a few variables to consider, my approach would be to not consider them at all, rather test the thing with the gas lit.

A thermal imaging system will give you a decent picture of the gas jet, the flame and any other thermal inputs from surrounding parts. This way you can have a look at things without using some visible gas which might have completely different flow characteristics, not to mention the effects that elevated temperatures might be having on your orifice.

Agreed that the only true test is the final test of this assembled item, but it is then extremely costly to have to rework the item to change the jet etc...

The only economic way to manufacture is to make sure the component parts are suitable to be built in to the final item, as with any other complex appliance.

The testing of these jets can not be by using LPG for safety reasons, even if I could recomend using LPG the test results of burning gas from a jet alone will not be indicative of the jet's final performance fitted in a venturi to get the air gas mix correct.

John.

Fair enough. However, thermal imaging will still give you a clear picture of what is happening even if you use a gas that is invisible, unignited (is that a word?)etc.

Yes.... you have a point there, as the gas will be coming from a compressed source just before the jet so the plume should show as a slightly colder area of gas...

Hopefully any swirling and turbulence could be spotted as well as an offset jet...

I was hoping for a simpler solution using a gas that shows up...

Anyone know if those cans used to test smoke alarms have a visible gas or is it just a gas used to trigger the ionisation type of alarm??

John.

It is possible that the holes are not all the same. some may have tiny burrs or other roughness inside or at the exit. At 60 microns such things will make for huge variances in symmetry and flow rate.

Suggest you have a follow uo polish and deburring of the hole to meake them as smooth as possible. This would take a tool made for the task. In any event a detailed examination of the failing and passing holes should be done,

Aurizon, The critical dimensions of the jet do show up nicely on the flow meter we have designed for this company...

So out of tolerance jets are discarded at the point of drilling and cleaning them. We have supplied special tools for this purpose...

I've also recommended them to buy a laser drill for this jet as well as all the other jets for their other products.

John.

This is strange. I wonder what is ruining the jets after they pass approval? Something bends, breaks or somehow interferes with the jet?

When you take the failed jets that have passed the drilling incpection back to the drilling inspection...do they still pass?

Perhaps, the damage seen, after some use, is caused by carbon deposits inside the drilled hole or some other contamination of flying debris. Send it for testing by ESCA or SIMS and they will tell each atom that makes the problem. These instruments you can't buy but can get sample analyzed. They are in 100000$ plus range instruments. They are good for corrosion study and are mass spectrometers sensitive to detect each atom type on the surface or even in depth. A cheap surface plasmon sensor you will find here Texas Instruments makes one low cost sensor for this near infrared wavelength of 840 nm for surface plasmons. It is on their website www.ti.com under sensors list. I am not sure if that is good for your expeiments but is excellent device for sensing many bio-molecules in gas and liquids using surface plasmons. This sensor works on the exponential wavefunction nature after reflection or refraction interface from unmatched surfaces like glass and air or glass and water, glass and vacuum where light changes velocity in the media and also direction on motion is reveresed or altered as happens in fiber optic sensors under total internal refraction. Perhaps tappered fiber optic sensor will also work and will be thin enough (10 micron) to scan the zone. You need mono mode fiber. Remove all top layer, heat it and pull it slightly to make it thin such that wavelength will find hard to get through or will be so called frustrated in the fiber and will have wavefunction extended outside the fiber. It is now ready for interection with external environment. If there is gas then you will get less reflected light within the sensor. It is a quantitative measurement technique.

Aurizon, the jets are only tested for gas flow rate and not for direction of the exit plume or turbulence of the plume... for the venturi to mix the correct amounts of gas and air the jet plume, I'm guessing, should be directed straight down into the venturi... If it isn't and the gas is hitting the side of the venturi then I am assuming that that would give a different mixture ratio and perhaps the failure of the completed appliance?

The problem is with getting hold of a sample of the rejected jets is that they are press fitted into a brass body and so are very likely to be damaged when removed...

Also the company is based a thousand or so miles away, so I've not been able to inspect their assembly techniques..

I have been sent many thousands of components but its only now with the new flowmeter we've been able to weed out about half the rejects and now we're looking for other possible causes...

I'm afraid the company has put me on a strict budget, so visits to this country / company would put me over budget...

I have a sneaky suspicion, from past events, that the comapny has perhaps had operatives who have rejected appliances for reasons not related to the jets... but hopefully a simple test like this to find the secondary effects might focus their minds a bit into putting their production line onto a more understanding footing...

John.

PS apologies for the secrecy of the company and product but I had to sign a nondisclosure agreement with them...

Try converting image color different for each temperature zone. This gives much better picture and each zone gets separated for the eye due to high contrast.

Perhaps, you can also try with inert gas first without fire. You will get dip in temperature on the exit points and good thermal image can be acquired. If you want a cheap solution then try scanner with single detector or micro thermocouple. You can also use colored gas like they use in rockets for weather monitoring which leaves long tail like trail of gas plume. This can be seen by ordinary camera. If this all fits well then it may work for combustible gases also.

Good morning Shyam,

At the moment we are testing these jets with compressed air of high purity, through the special flowmeter we have designed and supplied them... It is catching the majority of defective jets, but some are still passing and getting through to be assembled into a complete appliance...

It is the jet at fault as when one defective product is re-worked and a new jet fitted the product performs well...

The solution doesn't need to be cheap, this production line is suffering an extremely costly 10 to 15 % re-work of finished products which is costing them many 100's of thousands of Euros per year...

This is why I'm trying to persuade them to invest in a laser drilling machine...

Your idea of coloured gas is possibly a good starting point to experiment with...

I dont want to experiment with expensive thermal cameras if we find that the problem isn't related to the gas plume at all...

The main problem I have is that this company has been manufacturing gas powered heating appliances of various sorts for 25 years and yet when I ask them basic questions they haven't a clue as to how some of them work!! Only that if they make them 'this way' then they do work!!

John.

"The main problem I have is that this company has been manufacturing gas powered heating appliances of various sorts for 25 years and yet when I ask them basic questions they haven't a clue as to how some of them work!! Only that if they make them 'this way' then they do work!!" Interestingly, there's a large software company based in Redmond, Washington that behaves exactly in this same way. Evidently the problem is not unique to a particular industry. -- Europium

Dear Europium,

Well, Taguchi method has been tried and worked well. Nature is that way and we all are just because of that. We live if it works else we become fossils and some one have to dig for our bones and genes.

If something works then you never deviate the path easily. Look at how all those food stuffs are made. No one easily like changes. and you follow then 1000s of years and not one man's life time. I can't blame some one. It is also not easy to prove that their idea is poor and your idea is better. You prove and they will change.

While Taguchi was making alcohol sensor using 1 cent torch light bulb filament as heater (glass cover removed) coated with Tin Oxide, there were many working with electron microscope and other great equipments to make similar sensors. Taguchi actually made a small car that moved if you drink alcohol and say to it "move". It sensed alcohol and moved away. His sensor was no bad than any one developed with million dollar research. I think Taguchi was wise as he used one 1 cent bulb.

Taguchi also used to make gold electrodes by bangging a gold coin and make a thin foil in his kitchen.

I have seen pictures of his home made sensors and I was highly impressed. I can esily give him all five +++++ for his work.

Have you thought about using a polarized light source viewed through a second polarizing filter at right angles to the first. You place the object between them and the movement of the air causes the light to diffract and hence forms patterns when viewed through the second filter.

You may have seen this technique used to display the miniscule air currents that are formed by the heat of a human body. It should certainly be able to display the jet of air coming from the port you described and give you some idea of symmetry and alignment. Its certainly worth a try and would only cost a few dollars to test.

Ahhhhh masu, that's an interesting idea...

So you're saying that if light is projected through the flow polarised in one direction... then if viewed through another polarised filter the gas flow will show up...

If that's the case I suppose the gas movement is causing the light to alter its polarisation?

Hmmmmmmm that's a nice cheap and simple experiment I can try right now!

I will give it a try thanks.... John.

I have never tried this exactly but have seen the results reportedly produced by others. I have however used the same technique to examine the crystalline structure of rock samples under a microscope, The results can be quite spectacular. Regardless though its easy to try and if it works bingo.

Polarized light is good for rotational molecules like sugar solution which will rotate the plane polarized beam. That is neither quantity dependent nor direction dependent. So what do you get out of it? Water molecules also have rotational motion but you need proper energy or the wavelength close to that of the vibrations of the molecules which can contribute to the angular motion of the molecules and interact. I am not sure what is that you are planning to detect by this experiment? This experiment is done in this way. You first have a polarized light source and have nother polarized plane disk with angle maked. Alight the two polarizers to see light beam. Now insert the media which will twist the plane and you will see nothing. Now swist the marked disk until you see the light once again at maximum and that is the twsist given by the media molecular motion. Now this is fixed amount for a type of molecule say Glucose, fructose, sugar etc when they are desolved in water and are in ionic state. I am bit confused here. What is that we want to see in the LPG jet? Gas molecule polarization ? CO2 polarization? or what? CO2 absorption is about at 1200nm in MID IR Zone. Hence, will not be detected by the normal IR detectors of Silicone that goes only 1100nm.

Are you talking about surface Plasmon sensor technology?

Hmmmmm I've not had any immediate success with observing the gas flow using polarising filters...

I've got a feeling that, as Shyam has said the gas plume to be examined must contain some compound that induces a change in light polarisation.... either that or the light may have to be monochromatic?

The pictures you see of heat rising from a human body I suspect are taken in the infrared region and show the temperature of the air flow around a heated body??

Still its worth another experiment of two...

Afterall using a cloured gas may still be the best and cheapest solution.... just as long as the colouring agent is not a particulate nature, that may interfere with the jet's performance?

I did think of smoke filled gas, but the particles would contaminate the jet.

If its of any interest these jets are ultrasonically cleaned in clean room conditions, so any testing gas must be clean or have to have expensive extraction system to evacuate the gas to ambient.

John.

The polarizing filters is probably to simple a solution to be true, knowing the luck I have. I havn't tried it myself and am going on what I have been told but I will try and dig up some bits and give it a go myself. I have seen video clips of the currents of air around a person and a candle flame and with both of these there may be considerable mixing of various gasses like CO₂, H₂0 etc. I am having some surgery done next week (nothing too serious) so will be out of circulation for a few days but I will defiantly have a try as soon as possible and will post the results here. Anyway I hope you find a simple solution.

Dear John, This reference may be of some use. It is on flame or combustion gas imaging technique using a laser beam. http://cat.inist.fr/?aModele=afficheN&cpsidt=3156186 While you do ultrasonic cleaning, some solution that may remove corrosion may be used. If nozel is made from SS316 or SS316L stainless steel then it can be acid media and ultrasound together to etch out ant deposit in the nozel. Otherwise ultrasound alone may not work. At least use agitated ultrasound pulses for better results. One of my friend Mr. T.S. Neelakantan manufacturers large ultrasound cleaners for industries and he may perhaps provide more information. I will ask him. I was with him with Air India to clean Boing 747 Engines. Hence, some special agents must be available with him. Yes, most of the optical processes need tunned laser wavelength that is resonant frequency absorption band of the Gas contents (any one of the molecule as they are to molecular vibrations of low energies in IR and MID IR zones and not like atomic orbital energies of Bohr Model as in atomic spectroscopy). There is another very interesting sensor possible and that is Open Photoaccoustics Spectroscopy. Here, when light hits the gas or liquid or solid, it absorbs energy and then gives back as sound vibrations or accoustics signal that you can hear in a microphone. You deliver energy in pulses such that material is not continuously heated to high temperatures. In this method even a opaque material can be tested which was not possible by transmission measurement method. I used PAS in 1990 for my Ph.D. work. I used almost every possible method. including Atomic, Molecular, XRF, ESR, PAS, ACP, PES etc. I was too much involved to get to the truth for a clue for my developed gamma radiation sensing phosphors that were 300% more sensitive than any one ever made in the world. I have this technology now and is known only to me. I think PAS has given me results that were not possible by any other means.

Masu, I hope your surgery is minor and successful.

Shyam, Photoacoustics spectroscopy? interesting... I have little knowledge of this technique, but I did conside using a sensitive microphone to listen to the adio as well as the ultrasonic noise of the gas being vented from the jet.

In other air flow work sometimes listening to the escaping sound of the air and varying the flowrate by adjusting the inlet pressure can show up forms of turbulence which can cause all sorts of anomalies when measuring gas flow accurately.

I have to go out this evening so I will not be back until tommorrow, if anyone has any further thoughts on this subject I would be pleased to hear them.

John.

Dear John PAS is totally different from micro hole exit noises. This is generated after laser energy is absobed by the material under test and released in the surround air as sound. Unless lest material has the resonance absorption band for a particular wavelength of the light, it can not take energy from the beam and hence can not generate that sound. This technique is so great that signature of almost all materials in coumpond form are now available as library. Look for IR spectra library and perhaps it may be a paid service as data is very valuable. One problem with this method is that there is some background absoption may be there all over the wavelength due to scattering so there will be this noise also everywhere as some smooth line trace. You get riding peak over the background when special signature of the molecule is present. Hence, scanned and recorded data is essential. Just hearing some sound is no good. I think these people were the first to develop commercial PAS model nearly 20 years ago. MTEC can analyze your few samples for free and also have small paid service. I have sent many samples from India and they always helped me with free data. They are great people and you can discuss with them as they have good idea of IR spectra and access to IR data library. www.mtecpas.com http://www.mtecpas.com/Special_deals.html/Samples.html Here you may get some IR spectra data www.springerlink.com/index/Q1776P492518U352.pdf

I will prefer to try first water jet experiment with high pressure and that can be seen. You know those water jets used for PCB drilling? Then use a capacitor sensor to sense the water jet. I am sure this will give interesting results. Laser drilling is not a bad idea but not goint to be any cheap. If 1550nm Sapphire laser is used with 100ns pulses then slowly you can evaporate the material and have a nice hole. Perhaps look at www.photonicsspectra.com or www.laurin.com and ask the experts for laser drilling instruments. I think electrostatic discharge etching may also work. It is done spark erosion in kerosene filled tank. I am not expert of that. Wire pulling also used to drill microholes. You burry the hard wire in the metal initially and then pull out. Perhaps your company may have CNC fabrication facility, hence drilling may be ideal. Laser can be used to drill micron size up to 10cm deep.

A useful lab method is Schlieren Photography. While the details are to great to go into here, look into the technique, it is described in G. S. Settles book "Schlieren and Shadograph Techniques". Suitable simple parabolic mirrors can be bought from many sources, or the physics department of a nearby educational institution might be able to help. See also http://en.wikipedia.org/wiki/Schlieren_photography

If you think about it Schlieren photography is similar to the process with the polarizing filters. I hadn't thought about it before but it probably requires a parallel light source to work. An easy way to test this would be to use sunlight as the source then pass it through the polarizing filters.

Ahhhh interesting 4CD.... and Masu...

I must try that tomorrow, I've been on the phone to the company today and talked about turbulence and the production engineer casually mentioned that sometimes the jets do make a 'whistling' noise, which he had put down to turbulence and he had told the operators to reject those jets.

So maybe they are catching the majority of turbulent jets? They wouldn't be catching the ones where the plume is coming out skewed or off centre though, so this idea of having a visible jet of air to be seen and checked is still important I think...

I will let you know how some further tests proceed but this Scier photo article looks very interesting...

Many thanks - John.

Dear John,

I have some experience with similar (but a bit larger) gas jet orifices. One type of problem I encountered was that the issuing jet was affected by the shape of the exit chamfer or radius caused by deburring after drilling.

Have you inspected the orifice production line rejects and the orifices removed from the 10-15% repaired products to ensure that no such regular orifice exit anomoly is occurring?

Steve

Good advise, you cannot have a large radius or the fluid will follow the curve to a degree. There is a name for this, coanda effect. google it

The gas mus exit to a squared off edge and that way the radius will not bother it.

If you have a nick in the radius, that will also cause a problem.

Oh yes..... of course this is occuring and that is what i want to be able to detect quickly and accurately and for a production operative as well...

With the size of hole being only 60 microns or 2½ thou" diameter the only way to look for problems is with a high powered microscope and even then a tiny score mark inside the jet would not be noticed...

So the only way to check for this economically and with the speed needed is to flow air through while checking the flow rate also doing a visual check on the outlet plume shape...

John.

Would it make sense to blow the plume through a water column and see which direction the trail of bubbles travels? Or alternatively set up a fluidic chamber as a go no go test, if the plume does not go directly out the exhaust drilling then it will impinge on the walls and give a reduced and turbulent flow to the exhaust port and a partial flow to the sensing port.

Hey..........!!!!!!!!

That's an interesting idea..... let's see, if I put a tube a set distance away from the jet outlet, then for a normal straight flow of air most of the aire flow would be into the tube and the tiny pressure could be measured? or maybe the tiny air flow could be measured?

Then if a jet with a plume off to one side was tested the air jet would mostly miss the tube and reduce the flow / pressure....??

As the air flow is so small at this tiny jet size, its only about 60 cc/min I don't think the bubbles in water would show much....

Last night I was watching the smoke machine used in a disco and wondered if that gas / vapour could be used... But I don't know how those things work...?

Thanks for the idea though - John.

Possibly a microphone very close to the exit of the nozzel will show a difference on a scope, or even amplified in a speaker.

This would need you to test 50 or so to see if there is any sound variance. Cost should be just a few bucks for a microphone and some cord into an input jack of a cassette recorder.

Aurizon, Yes, when I mentioned to the customer that any turbulence etc... could be detected by sound he said that the test operators had been throwing away some jets which were 'whistling' as he put it....

It did make me think that a more sensitive microphone system could be used... But that would only detect turbulent jets which create a sound pattern, what if the jet of air at the outlet was just off to one side??

Good idea though, as the jets are so small I'm sure any sound generated would be in the ultrasonic range and only audible when the turbulence is really bad? So maybe something like a combination of microphone detection and a tube air flow collector might be worth investigating...?

John.

just listen to them and than see which ones later on show up as flawed.

It would be easy to tape 100 of them with a number spoken just as you test one.

number 1 hiss, number 2 hiss.

Then after the bad ones show up, listen to that number and see if there is a difference. Even if you do not know the cause just reject those with the noise.

Later analysis might show the actual flaw

Hi

What you can do is to use another visible gas or very fine particle / power spray at 90 degree to the plume and then plume will become visible.

All organic material become blue green when illuminated by 250nm UV light. This will also make the plume visible to sensitive CCD camera.

I think solution to your requirement is rather simple.

A couple more good ideas, thanks!

Shyam, I think the solution is probably simple too, trouble is that nobody has designed anything to do this job before!!

Everything sounds simple until, like me, you are asked to design a reliable piece of equipment to be used by none technical personnel to find and eliminate this problem which even the manufacturer's don't know what is causing the problem!

Then things aren't so simple...

John.

Another thing that can be done is an ultrasonic deburring operation, which uses a small ultrasonic bath with a very fine abrasive suspended in the water. 15-20 micron might do it and use a few minutes immersion.

This is an idea that needs to be tested to determine the efficacy of the approach.

The theory being that the problem may be caused by some very small surface features that the ultrasound bath would remove.

Thanks Aurizon, they are already using ultrasonic cleaning and deburring in a clean room environment.

Each jet is immediately sealed to keep it free of contaminents, after testing that is...

The jets are only opened in another clean room to be assembled to the bodies and then they are ultrasonically cleaned again and sealed until they reach final assembly ...

At each point they are flow tested, its only after being finally assembled in to the finished product that testing problems occur... But then the test is different... it involves inspection of the flame from the jet for stability and colour / air mixture etc...

This is why I'm positive just plain flow testing is not enough to test for.

John.

In the assenbly operation...you need a larger hammer :)

From the looks of it the problem occurs in assembly with the stress causing a distortion in the orifice?

For a flow sensor, what I had in mind is precisely excited RTDs. A ring containing 4 or 8 RTDs as a bridge instead of fluidics may also work. The disco smoke machines I've seen heat up an approved oil.

Emjay, the flow sensor is no problem, there are many off the shelf ones which are perfectly adequate for this measurement.

As for the smoke machines, it sounds like they produce more of a particulate or vapour cloud which may cause problems if the vapour were to condense in or on the test jet?

John.

Yes, I believe particulates would be a problem.

Excimer Laser is the best tool for small hole corrections with ns pulse. However, Safire or CO2 laser can give kW of power.

That UV imaging thing is like excitation by UV 250nm Mercury wavelength and emission at 400nm which is almost common to all organic material with slight shift. Use high power UV lamp 100W pen type. Make sure you do not inhale alots of Ozone generated by this lamp. This simple looking lamp can make ozone to eat away you lots of neurons in few hours as it will generate as high as 100ppm level of O3 in air. Permitted levels are only 0.1ppm for an hour exposure. I experienced this in my research hence keep safe or use fume hood for experimenting. It may be fun to watch the plume in darkness so make the room dark. Also force any organic gas that has this emission in visible zone to see the plume. You can even watch this under a microscope or eyepiece x10. Edmund scientific may have a telescope whith UV injector and eyepiece. You can see it in magnified form and from several meters away. It may be costly item to about US$10000 range. I do not know excat price now but have used those for remote sensing applications. You can also use 100$ UV land and Bishop Graphics eyepiece for another few dollars.

Higher energies always excite the electrons and you get good light emission after that in visible reason. You will have poor light output if flow is less and then you will need image intensifier to see. I am 100% sure, you will get good image if you have right tools.

A year ago I discussed this problem with a couple of laser manufacturer's , laser drilling is the way to go with this problem...

The cost is minimal compared to the £500,000 a year this company is having to spend on reworking its final products... Also as the company is producing several millions of jets a year the laser drilling in-house is the best solution...

Laser drilled jets will be extremely reproducable and will need minimal testing...

I've presented this solution to the company, at first to have the jets drilled for them and then to accept that for cost reasons as well as for Research and development the laser drill should be bought in-house...

Unfortunatly they, like so many other companies are run by accountants who only look at the balance sheet for the short term and so it was thought better to continue as they have been going but with improved testing equipment...

Of all these ideas perhaps the most important factor is that the equipment has to be used by a semi-skilled operator and not rely on visual and objective views... i.e. the equipment must provide a green lamp for pass and a red lamp for fail!! As well as needing the bare minimum of setting up and to test each jet within one second...

Not a simple problem at all... John.

In that case I can make cheap device to monitor the defective part. What is the amount that can be paid or what will be their budget plan. I need to do research for them and can get it ready in about 6 months and perhaps may cost about US$100,000. Will that be affordable to them? This will give go-no go immediate results. I know how to do it and planning to go for Patent for it. I will set up a test system very soon and I am sure I will get good results.

Your discussion has made me to take it seriously and I will design the monitoring sensor soon. I will work on having a test set up ready.

For cleaning, if nozzle can be raized to about 500C then all carbon deposits will become CO2 and will leave the hole clean. You sure have to go above 400C. If there is sone other dust in the nozzle. Pressurized acid flow may also clean.

No no no.... $100,000 is much to much..

They are very begrudgingly only offering $10,000 for a solution....

However, I am charging them for consultation and research and development on top of that...

Its the only way they can get approval past their accountant's boss's nose!!

John.

Shyam,

I admire your percieved ability to solve this problem so quickly and at such a cheap price.

But I have to ask myself how you can achieve your stated aim both in price and time?

After all, the problem hasn't been identified yet.

I am still researching possibilities, including plume direction.

I wonder, does your 'price' include multiple visits to the Republic of Ireland to investigate what they are making, how and why?

Do you realise that they don't know why they are doing what they are doing?

I certainly wouldn't and haven't committed myself to any fixed price for this job, as I'm still trying to find out what the job involves!!

John.

John,

Once again in the interests of keeping it simple, how about a light source from the HP side of the jet shining through the hole onto a collimator.

Hmmmm that's an interesting idea...

Light wouldn't give an exact equivalent idea of gas flow but maybe it would give an indication of problems with the hole shape and such like...

I have been pre-occupied with using a gas as its more representative of the actual use as well as we can measure the flow rate at the same time as watching for unusual gas flow patterns. This is mainly because the test time for each jet is only about one second...

Its worth a thought and a play on the work bench though..?

Thanks - John.

Did this idea move any further? Perhaps I can help on this.