X-Ray Detection of Air Pockets in Oil

Have you considered ultrasonic testing? It is used by medical technologists to detect and measure the flow of blood, so perhaps it might work for you.

I'm not all that familiar with industrial x-ray photography and not too knowlagable on your device either, but can you add a simple contrast agent in the oil? Perhaps one of the iodine varieties used in people?

http://en.wikipedia.org/wiki/Radiocontrast

FWIW, here's a site which at least calls out a few viscosities, along with some other info. Not sure it's helpful.

http://www.scribd.com/doc/3970872/Contrast-Media

First it would be helpful if you could tell us how much air you believe is still being injected. I mean that it will certainly be difficult to find 100 molecules of air in 1 liter of oil. But it might be a radical project to identify the sole molecule expected to not be N2 or O2.

For realistic amounts of gas, I would consider doing a compression deflection test since the oil would not be compressed but the gas would be.

We're not looking for anything that small really - we suspect that some drillings may be half filled, in which case it would be bubbles 1 -2 mm. Or we may have an oil air emulsion in some places. A higher resolution, i.e. down to microns would be desirable though. The gas will be atmospheric air that was present upon assembly.

A compression test is a good idea, and will be useful in quantifying the amount of trapped air, but it is important that we know where the air is within the Injector.

Yes, you certainly need a much higher atomic number element present in the oil to get any contrast. The difference in the absorbtion of x-rays for carbon/hydrogen (oil) and oxygen/nitrogen (air) is too small to be of use in x-ray radiography. If you can find some additive for the oil that contains heavier elements - the heavier the better - you should get a much better result. Molybdenum disulphide (Mo = At. No 42) has been used in oil to reduce friction. You should be able to purchase some already dispersed in oil, making addition quite simple. Perhaps you could even use neat MoS2 solution. Tungsten disulphide is also a possibilty.

Good luck!

I'm not familiar with your process so here comes my (stupid?) question:

Can you pull a vacuum on the parts and backfill (draw-in) the oil?

Seems like no air would result in no air pockets and no need for x-rays.

Not a stupid question at all - we do vac fill them and thought it was an effective process until some test results led us to question this.

I don't know anything about this sort of thing, but, you may find that applying the vacuum for a lot longer (hours) before filling with oil will help.

I've used oil impregnation for dielectric insulation and cooling in electrical assemblies.

The right combination of pre-heating, drying, and sufficiently good vacuum (with time to allow for material outgassing) prior to the oil back-fill was critical. Just pulling the oil in with a "partial" vacuum left many air pockets/bubbles and did not work well for my application.

Surely adding chemicals to your experiment will affect the results anyway.

Maybe air plus maybe a barium compound for instance.

I would maybe go down the route of light detection methods or ultrasound as suggested.

http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6THG-4WGF10M-1&_user=10&_coverDate=08/31/2009&_rdoc=1&_fmt=high&_orig=search&_sort=d&_docanchor=&view=c&_searchStrId=1408968041&_rerunOrigin=google&_acct=C000050221&_version=1&_urlVersion=0&_userid=10&md5=d98846df5cf699acb5f90f4d54fe54c2, this may be of help.

Hi all. I just stumbled upon this forum. As an ASNT Level III in radiography, I may have some info that could help. With radiography (x-ray) it is all about densities. The image you see is a difference in the densities of the materials being tested. The more of a difference, the easier it is to radiograph. To get the clearest picture, you would need to use a low power x-ray machine and slow film. If you can increase the difference in densities by using something other than oil, that would help. Adding dye to the oil would be useless. The film is loaded in cassettes that have lead filters. You can load multiple pieces of film into one cassette to get different pictures. I would not use more than 3 in one cassette. Use 3 different speeds of film.

That being said, the ultrasonic suggestion is a viable one. If you have a good NDT company in your area, I would contact them.

I didn't realize my post was a come on to get money from you. Hopefully, it will help point you in the right direction.

K,

its not a 'come on'. The entire science publishing industry on the web does exactly that sort of 'excerpt' thing, and it is intended to attract buyers. It is valid to point out research doc of that type, and if the user needs it enough, they will pay for it.

don't feel bad.

Chris

You might weigh the injector prior to, and after filling, and measure the fluid volume used to fill it. Knowing the densities of air and oil will allow you to calculate the amount of air in the injector.

I have been following this for a while and have problems getting my head around it. I just have to hear or read the word bubbles and I'll jump on it. Due to a language barrier (I suppose) in the original post I don't seem to fully understand the problem.

I think the x-ray diagnosis of such minute and unstable phenomenon would be hard to interpret or even detect. If I understand the problem correctly here is what I would do.

1. Freeze a sample in a silicon tube (take sample and freeze instantly)

2. Take very thin shavings (like it is done in tumor research). If the bubble size is anticipated to be in the lower end of micron size use electron microscope.

3. View under microscope and measure. Any research, team worth its bubble, would jump on this, I think.

4. Consider shrinkage which may occur.

5. Publish results here in CR4. I'd love to see what these frozen bubbles look like.

This should allow one to be able to distinguish between a bubble and the oil or any other suspended particles for that matter. Medical research centers should be able to help.

I have re-read the OP and can only think of this solution at the moment. It would be good to know more about the "real problem", something seems to be disguised here, which the OP is either not telling us or has the problem of conveying it properly.

If this helps, send me a bottle of bubblies, but I am not holding my breath, Ky.

Hi. I think that my description may have not been adequate. What I am trying to do find bubbles inside an assembled Fuel injector - where the bubbles are and how many/how much volume do they occupy. If you look at the "Detailed Function" section of the Wikipedia link below it will give you an idea of what I am talking about.

http://en.wikipedia.org/wiki/Fuel_injector

The difficulty is that the air bubbles are inside a drilling inside a machined steel part, which is inside a casing. Taking the Injector apart will allow the air bubbles to escape - hence the use of X-ray.

The freezing idea is not a bad one though - it would allow the component to be disassembled without loosing the bubbles. The casing is screwed on however and it is likely that the threads will be seized when it is cooled.

I didnt go into too much detail about what I am trying to do as I was after a fairly specific answer regarding x-raying. I didn't think that I needed to go into too much detail about the actual components, but having said that the alternative techniques I have been recommended have been interesting and may well prove useful.

Cheers for your help, by the way.

There are two fundamental things I don't understand. First, why the trapped air bubbles could be a problem? If your injector consistently delivers the metered amount of fluid, then regardless of any volume or location of trapped bubbles inside the injector the injector is properly doing its job. Second, what is the source of these mystery bubbles? If the bubbles come from the initial assembly then any injection error quantity will be a one time occurrence that can be remedied by a few initial purging cycles. If instead the air(?) is dissolved or suspended pockets in the fluid then the location and magnitude of these bubbles will change with conditions.

Thank you for clearing this up.

This gives me a better idea of what your "problem" is. Like redfred said, I can't see what the need is to quantify the amount, volume, position of these air bubbles. I think that measuring them or making them visible, will be impossible IMHO. X-ray or what ever diagnostic tools would be introduced. Schroedingers cat?

Schrödinger's cat - Wikipedia, the free encyclopedia

Depending on the pressure inside the injector any gas will hide or better be forced into solution. Imagine a hydraulic accumulator with out a skin, container. Even at pressures 4-5Kpsi any gas will dissolve in the ambient fuel. Read up the gas solution tables which consider the temperature of the fluid as well.

Because I am still not sure how you identified this problem as being one, you might want to have a look at this thread What Happens to the Balloon? In there I just use one balloon, although there could be n'th of them. It was the principle I was after and it might assist in what you are doing.

It is not the question I asked back then but the various approaches, thoughts that were suggested that gave me a better understanding, of a 'somewhat' similar situation. This is now a while back and I can't thank the participants in that thread enough for helping me get my head around that hypothetical question.

Bottom line: Even if you were able to picture, detect some bubbles they will always be of a transient nature. Either in solution or just coming out of solution. At point of injection it will hardly make a difference, or so I think. Fascinating subject, thanks for posting this.

Good luck and keep us informed, Ky.

You may want to contact http://www.sonoscan.com/. They are a leader in ultrasonic imaging and may be willing to provide some testing for you

As far as I can see here- a portable doppler machine would be the answer- these use ultrasound as another poster said- but also the doppler principle to identify spaces in a fluid etc- there are different types & frequencies- generally 3-10 MHz send -receive sound waves-can be print-out or audible according to model- many available on market.

I don't agree that x-rays are a good method for locating bubbles in oil, but to achieve better contrast you might use ferrofluid in place of the oil.

Surely air rises in your fuel oil so why not take the feed from the lowest part of the tank and then you can be sure it has the lowest amount of air in it ?