Detecting coils inside steel cylinders

[discover66]

TL;DR Summary

If I have a coil inside a piece of steel pipe, can I detect it from outside of the pipe using a tuned transmitter / receiver array?

Just some idle Sunday musings. I have a steel pipe that is 2" ID, 2.5" OD, 10' long. I want to place a coil inside the pipe midpoint along the pipe and detect it's presence with a transmitter / receiver array. The array can consist of two coils wrapped around the outside of the pipe and can be adjacent to the inner coil.

Can I detect the inner coil?

 

[anorlunda]

Is the inner coil just sitting there, or is it connected to anything?

 

[Baluncore]

I think you must say what you mean by a "coil".
Is it an LC circuit, or just a short coil of insulated wire with open ends?

 

[DaveE]

I would look at something like TDR. Maybe an ultrasonic wave launched from the end of the pipe. Or maybe something like tapping along the length and analyzing the vibrations (this is how I used to find studs underneath drywall).

 

[Twigg]

If the inner coil is open-ended, I think you are SOL for detection with an outside coil.

If the inner coil is connected to some circuit, the simplest method would only involve one outer coil. With one outer coil, a pipe, and one inner coil connected to some circuit, you have a stray-field transformer. You can measure the position of the inner coil (in one dimension only) by measuring the impedance across the outer coil. When the inner coil is infinitely far away, you would measure only the base reactance of the outer coil; when the inner coil is close, you would measure additional impedance from the inductively coupled inner circuit. I can't comment of the feasibility of this method because there are too many free parameters here.

 

[DaveE]

OK how about a single coil that you resonate with a capacitor (probably best at low frequency). Using a signal generator/analyzer you could look at the Q as the coil is scanned down the pipe. Like a metal detector, sort of.

 

[discover66]

Is the inner coil just sitting there, or is it connected to anything?

The inner coil is not connected to anything except itself.

 

[discover66]

I think you must say what you mean by a "coil".
Is it an LC circuit, or just a short coil of insulated wire with open ends?

Coil can have as many turns as necessary and the ends of the wire connected, i.e. LC circuit.

 

[discover66]

Is the inner coil just sitting there, or is it connected to anything?

Just sitting there. Once installed there is no physical access. What I really want to know is if the coil loses integrity.

 

[discover66]

OK how about a single coil that you resonate with a capacitor (probably best at low frequency). Using a signal generator/analyzer you could look at the Q as the coil is scanned down the pipe. Like a metal detector, sort of.

I cannot connect any electronics to the inner coil. I was thinking of inducing a current in the coil and detecting the field with the second coil. But as this is inside a steel pipe, can I do that?

 

[discover66]

If the inner coil is open-ended, I think you are SOL for detection with an outside coil.

If the inner coil is connected to some circuit, the simplest method would only involve one outer coil. With one outer coil, a pipe, and one inner coil connected to some circuit, you have a stray-field transformer. You can measure the position of the inner coil (in one dimension only) by measuring the impedance across the outer coil. When the inner coil is infinitely far away, you would measure only the base reactance of the outer coil; when the inner coil is close, you would measure additional impedance from the inductively coupled inner circuit. I can't comment of the feasibility of this method because there are too many free parameters here.

I considered stray field transformer, but I don't have access to the inner coil. The position of the inner coil isn't important, it's integrity is. I need to know if the inner coil fails.

 

[Twigg]

For my earlier transformer suggestion, you only need access to the outer coil to measure its impedance. The inner coil would in theory add to the outer coil's impedance through the inductive coupling. However, if the inner coil is just shorted together I doubt the change in the outer coil's impedance will exceed your measurement uncertainty. Just a hunch, could be wrong. Also, if the pipe is a ferromagnetic steel (i.e. anything you'd usually encounter that isn't 300 series stainless), you're probably SOL because that will butcher any kind of inductive coupling with huge stray fields.

 

[Baluncore]

Also, if the pipe is a ferromagnetic steel (i.e. anything you'd usually encounter that isn't 300 series stainless), you're probably SOL because that will butcher any kind of inductive coupling with huge stray fields.

What does "SOL" mean ?

 

[DaveE]

I cannot connect any electronics to the inner coil. I was thinking of inducing a current in the coil and detecting the field with the second coil. But as this is inside a steel pipe, can I do that?

The idea is that the external excited coil will detect the metals in the environment, mostly by eddy currents or a change in the magnetic properties. Highly resonant circuits (high Q) can be quite sensitive to parameter changes. A change in the magnetic field distribution could change the inductance and shift the resonance. A change in losses from eddy currents could change the width of the resonance (Q). However, I'm not sure how well this will work in your case. You would have to experiment a bit to figure out feasibility and the best set up. You'll probably need someone to help with the electronics if you aren't already familiar with LC circuits.

 

[Vanadium 50]

SOL = So Out of Luck.

 

[Vanadium 50]

You have a coil with inductance L capcitatively coupled to the tube. Can't imagine that this will be very high Q, which means that it can't be sensitive to the exact configuration, which is what we want, isn't it?

 

[Baluncore]

Coil can have as many turns as necessary and the ends of the wire connected, i.e. LC circuit.

I think you have overgeneralised the problem statement, to the point of impossibility.
Why do you need to place a coil there in the first place ?
Why do you need to know where it is ?
What flows through the pipe ?
Why not put a permanent magnet there instead, then you can find it with a magnetic compass from outside the iron pipe.

 

[DaveE]

You have a coil with inductance L capcitatively coupled to the tube. Can't imagine that this will be very high Q, which means that it can't be sensitive to the exact configuration, which is what we want, isn't it?

No, I'd have a coil and a capacitor configured as a high Q tank. Then as the coil is brought up close to the tube I would evaluate how the inductance and/or losses vary along the pipe. Still, it may not be sensitive enough to detect extra metal bits inside the pipe. Similar to a metal detector. I'm not sure if it would work or not. You'd need some good instrumentation, something like a network analyzer or a lock-in amplifier; at least until you figured out the signature.

However, I still prefer the TDR idea. Like launch and EM-wave down the tube and look for reflections. Or the frequency domain version where you send CW uwaves into the pipe and look at the frequency response of the impedance. If the coil is big enough it may look like a stub, or such.

Or, probably my favorite: fasten a microphone to the end of the pipe, connected it to an oscilloscope. Then you hit the end of the pipe with tiny hammer (i.e. delta function excitation) and look for sound reflections from discontinuities. It's cheaper than waveguide uwave stuff and sound moves slower.

Then there's the idea of putting a long stick down the tube and measuring it when it hits something. I suspect that has already been ruled out though.

 

[alan123hk]

Summary:: If I have a coil inside a piece of steel pipe, can I detect it from outside of the pipe using a tuned transmitter / receiver array?
Can I detect the inner coil?

As far as I know, stud finder uses capacitive coupling, while metal detection uses inductive coupling. Therefore, it is possible to detect the coil in the pipe by placing another coil array at a very close distance outside, but this may not be an easy task. You may need to use quite complex electronic circuit to increase the sensitivity of the outer coil array.

 

[Baluncore]

The problem with a resonant circuit inside the pipe is that the skin effect on the outer wall of the pipe will prevent communication. You need very low frequencies to get through a thick metal screen. To have a sufficiently low frequency the coil will need a huge number of turns in parallel with a huge capacitor.

At the same time, if the coil was coaxial within the pipe, the inner wall of the pipe will have an induced counter-current that magnetically cancels the current in the coil.

 

[Rive]

Once installed there is no physical access. What I really want to know is if the coil loses integrity.

That coil... Is it your own making? Does it have some other function than being there?
More details about the actual goal would be useful.

If that coil is your own making then you might consider making some mechanical energy to penetrate the tube instead of magnetic field. With good orientation and an additional magnet you can set the tube vibrate from the outside with that ' tuned transmitter '. With carefully considering the orientation of the coil and with a small magnet you can have AC voltage there. Add a diode bridge, and you have AC with doubled frequency: if you transmit that outside (mechanically) it might be picked up.

 

[alan123hk]

the skin effect on the outer wall of the pipe will prevent communication. You need very low frequencies to get through a thick metal screen

Agree, the use of low frequencies is the key. However, low frequency means high inductance, but I think it may not be allowed to insert the magnetic core into the coil inside the pipe because I believe that for some purposes, something must pass through the pipe.

 

[DaveE]

Agree, the use of low frequencies is the key. However, low frequency means high inductance, but I think it may not be allowed to insert the magnetic core into the coil inside the pipe because I believe that for some purposes, something must pass through the pipe.

Yes, I think putting things in the pipe isn't allowed, or else it would be a tape measure you would put in there.

I was thinking you could have a coil with a core outside on a radial axis.

However, I just reread the OP and he said two coils wrapped around the outside as a transmitter and receiver. That seems hopeless to me given the close magnetic coupling to the pipe, in the transformer business we would call this a "shorted turn". Too bad the pipe isn't plastic, then it sounds similar to an LVDT.

 

[discover66]

I think you have overgeneralised the problem statement, to the point of impossibility.
Why do you need to place a coil there in the first place ?
Why do you need to know where it is ?
What flows through the pipe ?
Why not put a permanent magnet there instead, then you can find it with a magnetic compass from outside the iron pipe.

A little more detail of the problem.
The pipe has an inner pipe inside it that is carrying high pressure fluid. The high pressure fluid will corrode / erode the inner pipe and eventually break through to the outer pipe. What I want to do is wrap a coil around the inner pipe (ferrous steel) so that when breakthrough occurs the coil is cut and the system suffers loss of signal, which is detected from outside the outer pipe. The coil doesn't carry any information except that the coil has integrity. The coil failure is the signal.
Because of the nature of the system, I don't have any method of direct contact with the inner coil after initial assembly. An LP circuit could work and low frequency would be the only way through the skin effect. The capacitor is going to be an interesting challenge, as the pipe will be running at temperatures up to 250 deg f.

 

[Rive]

In such system the breakdown of the inner tube would not necessarily cause the coil to fail. I think you should go for detecting the pressure change between the tubes.
Maybe some airtight capsule with set (low) pressure inside. That would be safely crushed if the external pressure rises.
I would opt for some ultrasonic solution. Magnetic and steel is just not some simple thing.

 

[discover66]

In such system the breakdown of the inner tube would not necessarily cause the coil to fail. I think you should go for detecting the pressure change between the tubes.
Maybe some airtight capsule with set (low) pressure inside. That would be safely crushed if the external pressure rises.
I would opt for some ultrasonic solution. Magnetic and steel is just not some simple thing.

Unfortunately the tubes are at the same pressure. The inner tube has abrasive multiphase fluid and is pressure balanced with the outer pipe.

 

[discover66]

On failure there is no pressure change. they are not in pressure isolation. The coil will get cut. :)

 

[DaveE]

I agree you are probably better off with a more mechanical (acoustic?) approach. It's just really, really hard to use an EM approach when your trying to look inside a pretty good Faraday cage.

 

[Averagesupernova]

On failure there is no pressure change. they are not in pressure isolation. The coil will get cut. :)

If there is no pressure differential and the inner pipe fails then what is it that causes the coil to get cut? The equal pressure will not cause the inner pipe to expand and cut the coil.

 

[Rive]

Unfortunately the tubes are at the same pressure. The inner tube has abrasive multiphase fluid and is pressure balanced with the outer pipe.

Guess the 'stuff' in the inner tube is flowing, while the external tube is used only as a pressure vessel? Then there is already an inevitable pressure difference (since the pressure drop along the flow). The inner tube should be able to handle it, with a decent reserve too.

It's just a small step to add (or: take) some more pressure and use it for leakage detection.

 

[alan123hk]

Maybe you can try to do an experiment.

Just try to apply a frequency of about several kilohertz to several tens of kilohertz, and also try to adjust the current of the outer coil, you can study the data and evaluate the feasibility.

By the way, what is the thickness of the inner steel pipe?

 

[Tom.G]

What is the purpose of the outer pipe, does it transport anything, is it just a redundant container when the inner pipe fails, something else?

Details please! In engineering data/information is golden. So far we seem to be playing a game of "20-Questions".

 

[Fredbonyea]

OK how about a single coil that you resonate with a capacitor (probably best at low frequency). Using a signal generator/analyzer you could look at the Q as the coil is scanned down the pipe. Like a metal detector, sort of.

A steel pipe will conduct a local field via eddy currents. The conduction will be week, but at the right frequency, and with a differential detector, in most cases it is possible to detect a coil inside a steel barrier at very short distances.
Your source current must be very close to the pipe, preferably within a millimeter. The frequency of the source should be the same or a multiple of frequency of the resonant circuit you have constructed.
A differential detector could have two field coils, both butting the steel pipe at the same distance. What you would look for is small current variations between your two bridged fields as you pass the paired coils along the pipe, and it will be very small. The thicker the pipe, the more current you will need to couple the eddy currents.

 

[DaveE]

A steel pipe will conduct a local field via eddy currents. The conduction will be week, but at the right frequency, and with a differential detector, in most cases it is possible to detect a coil inside a steel barrier at very short distances.
Your source current must be very close to the pipe, preferably within a millimeter. The frequency of the source should be the same or a multiple of frequency of the resonant circuit you have constructed.
A differential detector could have two field coils, both butting the steel pipe at the same distance. What you would look for is small current variations between your two bridged fields as you pass the paired coils along the pipe, and it will be very small. The thicker the pipe, the more current you will need to couple the eddy currents.

This idea, like most in the thread doesn't work. We all thought he was looking for the presence or position of an internal coil. Later, we found out he is looking for physical damage (or not) to said coil, which means we have to evaluate structure, not just existence.

My latest idea: a portable x-ray machine.

 

[Averagesupernova]

My latest idea: Forget about this thread. Too much info is coming too late. No good description from the get-go. My question has still not been answered, so I say what's the point?