Ferrous metal detector

[sachindra]

Starting with The Fe3O4 material, which in its natural form is magnetite and on which PI metal detectors will give no substantial response. Any weak response will be due to magnetic lag, or viscosity as it is often called. This weak response has no relation to the quantity of magnetite present and would not be useful in your project. Induction balance continuous wave metal detectors do respond to magnetite but are not suitable for other reasons. What is needed is a magnetic susceptibility meter which is designed for measuring even very small percentages of such minerals within a non-magnetic matrix.
Eric.

Ture. PI did not give any response at all. Magnetic susceptibility meter sounds like a better choice for Fe3O4 than eddy current approach.

The steel fibre is a different problem as much depends on the diameter of the fibres. Steel wool can be measured with the MS2 and the readings will generally be higher than for the magnetite, but will depend on the percentage of course. The fact that you can detect it with a PI metal detector is indicative of magnetic lag (viscosity) rather than an eddy current response. Bartington make probes for the MS2/MS3 sensors which may be good for your application. It all depend whether it is a small sample you intend to analyse, or a large one.

There is also a Magnetic Viscosity meter based on PI technology that will detect viscous iron minerals and also ferrous metal. This something I designed the accompany the MS2B for a fuller analysis of rock samples as it does not respond to susceptibility. Using a 1 inch ferrite cored probe it also makes a good rebar pinpointer in concrete structures.

Eric.

Right now, I intend to do some testing in lab (I am a graduate student, so in lab) and have smaller samples. However, I intend it to apply in larger scale. It is longer rather than lager ? because I plan to test it on concrete pavements (roadways as they say in UK ? not sidewalks).
I used device with probe, like that from Bartington. I used eddy current probe from ZETEC ?MIZ-21B (link) to find the difference in response at various frequencies (from 200Hz to 200kHz) in these samples. I will post some of the impedance plane graph in response to another post. That would give more clear indication of what material I am trying to measure exactly.
Rebar pointer . I wish there were some way to map rebar in concrete pavements without having to go down and touch the surface. Last year I used a Proceq device (similar to this link) to map steel in a highway near Chicago. It is not fun to use these devices to locate rebar then note the data on a paper while you are working with lane closure in one lane and vehicles driving at 45 mph (72 kmph) just 12 feet from you . But, you got to do what you got to do.

I missed the above, but the susceptibility meter approach may still be valid as an 8in coil is available from Bartington. Best discuss it with them.

Eric.

Thank you Eric for the response. I will email them and see if they can give me some idea.

[sachindra]

Just a thought:
Would the Iron filling respond like 'hot ground'?
If so then maybe the reverse of GEB may work.

I am not sure what exactly ?hot ground? is or GEB (sorry, I am too new to this field I guess).
I used an eddy current probe from ZETEC ?MIZ-21B (link) and tested the iron filling sample.
I did not have the sample inside concrete ? I can make one and test it. However, I tested it in epoxy (ie. I added iron filling in epoxy resin and let it sit for 24 hours to harden). I have attached images to show what it responds like.
The images are impedance curve.
Image 1 ? 60kHz test (Samples: Iron powder in epoxy; Steel; Brass; Aluminum; Copper)
Image 2 ? 600kHz test (Samples: Iron powder in epoxy; Steel; Brass; Aluminum; Copper)
Image 3 ? 200 Hz (frequency of PI) and 200kHz test (Samples: Iron powder in epoxy; Steel; Concrete prism that I tested with PI; Copper)

[sachindra]

we can't get closer than 10"

What environment are we trying to test the concrete in. Chernobyl ?

I think we need to know the exact situation.

Rebar, services and so on will affect result. Normal fibre reinforcement is for floor slabs. But doesn't fit with can't get within 10" ?

Right now I am doing some experiment in lab. However, I plan to apply it in larger scale such as concret pavements (ie. concrete roadways in UK english). So, I want something that I can ideally mount on vehicle and drive slowly (20mph or 32 kmph) and record continuous data.
I specifically intend to achieve these things:

1. Develop a way to detect concrete samples with steel fibers or iron powder filling.
   • There are some efficient way to do those using touch method. I have tried it in past in projects in real world. They are very dangerous when we have partial lane closure with vehicles driving fast in other lanes. Fast = 45 mph or 72 kmph ? which is standard work zone speed limit.
   
   
2. Mount those sensors on the bumper of the car or some other mounting device.
3. Instead of traditional audio signal or single point display, hook it up to Arduino and use SD card shield to continuously record the data. I could hook up some GPS sensors to know the location and from that sensor reading I would know if that location has iron fillings or powders or dowels.
   • Dowels are used in concrete pavement with joints. It is spaced every 15 ft. and sometimes, believe it or not, contractors ?forget? to put dowels. How do they check it? Drive every 15 ft. get down, take these contact method to scan for it, then get in the car and repeat it for miles and miles and miles. ? Although this is not what I am trying to solve right now. My focus is in detecting concrete with steel fiber or iron filled section. This can be another application, that I learnt hard way, of this device.

So, why 10?? Is it hard and fast distance?
So that it is above the ground while driving slowly. The distance can vary but I want good clearance with ground, just in case. I could get PI go up-to 6 inches above and detect steel fiber sample. Would be great to add more height for two reasons: (i) good clearance, (ii) for same height, it can scan deeper == more sample within the influence zone == better signal ? maybe can drive faster.

Can 1 or 2 inches work?
Too low clearance.

Why ferrous only?
PI would also respond to Aluminum or other metals in the car. I can mount a longggg PVC and mount it far from my car. And I don?t want people staring at me and say: ?What is that wierdo doing????

[sachindra]

try go from simplest to difficult on your way of the learning. begin from Barracuda, then build Delta Pulse or Clone or Mirage PI.
you will get first experience and will be know where to run to.

I have tried Clone PI, and other eddy current touch probe devices. I have more details in above posts.

[waltr]

I am not sure what exactly ?hot ground? is or GEB (sorry, I am too new to this field I guess).

Here is a good paper that describes hot ground (“magnetic soils” or “mineralised soils”) and GEB methods.
https://www.minelab.com/__files/f/11...S_&_THEORY.pdf

[Koala]

there's a fair few vehicle based sensing systems. Nearly all of them involve pumping a signal out and measuring how much is reflected back.


I would have thought someting microwaves

[sachindra]

Here is a good paper that describes hot ground (“magnetic soils” or “mineralised soils”) and GEB methods.
https://www.minelab.com/__files/f/11...S_&_THEORY.pdf

Thanks. I will look into this.

[sachindra]

there's a fair few vehicle based sensing systems. Nearly all of them involve pumping a signal out and measuring how much is reflected back.


I would have thought someting microwaves

Actually, I will use GPR as well for the testing. However, one main limitation of GPR is weather (water, snow, ice, etc.). Water will significantly attenuate signal that means we have to wait for dry condition. With snow for many months plus rain few more months, it is good option with limitation of using condition. Eddy current/other magnetic based system is not affected by these adverse conditions.

[Ferric Toes]

Can you tell us more about the iron powder filling (Fe3O4) such as it's particle or mesh size? Is it pure Fe3O4, or does it include other oxides, Fe2O3, or impurities? Also the steel fibers, the cross section size?

What horizontal resolution do you require? i.e. for this type of application a noise cancelling rectangular search coil is often used. The long dimension determines the lane width of the search.

Provided the road surface is fairly smooth, a towed non-metallic trolley or sled is preferable as the coils can be closer to the road and further from the vehicle.

Eric.

[sachindra]

Can you tell us more about the iron powder filling (Fe3O4) such as it's particle or mesh size? Is it pure Fe3O4, or does it include other oxides, Fe2O3, or impurities? Also the steel fibers, the cross section size?

What horizontal resolution do you require? i.e. for this type of application a noise cancelling rectangular search coil is often used. The long dimension determines the lane width of the search.

Provided the road surface is fairly smooth, a towed non-metallic trolley or sled is preferable as the coils can be closer to the road and further from the vehicle.

Eric.

I used Fe3O4 of size 30 microns (500 mesh) for the sample I was testing (Link of actual product). It is a pure Fe3O4. Steel fiber is 0.75 mm in diameter (Datasheet link) with 30 mm length.

Whole lane scan is not required. I was thinking of multiple point scan instead of one big loop. i.e. few (5 or 6) smaller loops 1 ft apart from each other. That way if there is any anomaly in some section it could be picked up. This is assuming that getting that many readings and saving is not significantly difficult than doing one loop. Typical lane width is 12 to 14 ft, so we have enough spacing between the sensors (that is after figuring out what sensor which is the main thing) to not interfere potentially. I hope this makes sense.

Towing non-metallic trolley is good option. I was just thinking of PVC pipe extended 4-5 ft from vehicle (front or back whatever position... or both ) with wires for each loop inside the pipe. Loops then being mounted on concrete form cardboard (picture) to prevent from bending and vibration.

[Ferric Toes]

I used Fe3O4 of size 30 microns (500 mesh) for the sample I was testing (Link of actual product). It is a pure Fe3O4. Steel fiber is 0.75 mm in diameter (Datasheet link) with 30 mm length.

Whole lane scan is not required. I was thinking of multiple point scan instead of one big loop. i.e. few (5 or 6) smaller loops 1 ft apart from each other. That way if there is any anomaly in some section it could be picked up. This is assuming that getting that many readings and saving is not significantly difficult than doing one loop. Typical lane width is 12 to 14 ft, so we have enough spacing between the sensors (that is after figuring out what sensor which is the main thing) to not interfere potentially. I hope this makes sense.

Towing non-metallic trolley is good option. I was just thinking of PVC pipe extended 4-5 ft from vehicle (front or back whatever position... or both ) with wires for each loop inside the pipe. Loops then being mounted on concrete form cardboard (picture) to prevent from bending and vibration.

Thanks for the details on the materials added to the concrete. I have a pot of 100 mesh and also a pot of 1 - 5 micron Fe3O4; both are 99.9% pure. Both pots give small viscous signals on a PI test unit when placed on a 1" diameter coil. Not enough to be useful when dispersed in concrete. Magnetic viscosity is called Viscous Remanent Magnetism (VRM) if you read literature on rock magnetism. The finer material gives the greater signal due to the smaller particle size, of which both samples would have a percentage. For Fe2O3 the particle size has to be below 0.03 microns to exhibit magnetic viscosity, or short term VRM. This is nicely explained in 'Hitchhiker's Guide to Magnetism' by Bruce Moskowitz. Your 500 mesh material may have more of these smaller particles but I doubt it will be enough to be useful for detection by standard PI. A system that responds to magnetic susceptibility is necessary and a PI system could be made to do this. Multi-channel is also possible and I have personally have been involved in the design of such PI detectors for industrial use. Coils are generally rectangular and there is no need for a space between them.

Just downloaded this paper as I was puzzled as to how iron oxide would help to strengthen concrete. Very interesting. In your test samples, what is the percentage of iron oxide in the total volume. The paper suggests 15% as being about optimum. I have only done a speed read, so there may be more info.

Experimental study of concrete made with granite and iron powders as partial replacement of sand.

A question about the steel fibers. Is their orientation in the concrete random, or are they all in more or less in the same direction?

Eric.

[Skippy]

I interpreted the steel 'fibres' as being randomly mixed in with the concrete, so orientation, which is certainly important to detection, should be random, unless the fibres 'flatten out' near to the surface.
They appear to be ferrous. There is no mention of them being stainless steel, and they are galvanised, which also points to regular steel. They are very high strength, I notice, 1200 MPa upwards, so clearly some fancy alloy, though the wire-drawing process is good for increasing tensile strength too.
( stainless steel bicycle spokes have UTS of 1200 MPa, even though the bulk material they are made from is more like 400 MPa )

[Koala]

https://www.geophysical.com/whatisem ??

[sachindra]

Thanks for the details on the materials added to the concrete. I have a pot of 100 mesh and also a pot of 1 - 5 micron Fe3O4; both are 99.9% pure. Both pots give small viscous signals on a PI test unit when placed on a 1" diameter coil. Not enough to be useful when dispersed in concrete. Magnetic viscosity is called Viscous Remanent Magnetism (VRM) if you read literature on rock magnetism. The finer material gives the greater signal due to the smaller particle size, of which both samples would have a percentage. For Fe2O3 the particle size has to be below 0.03 microns to exhibit magnetic viscosity, or short term VRM. This is nicely explained in 'Hitchhiker's Guide to Magnetism' by Bruce Moskowitz. Your 500 mesh material may have more of these smaller particles but I doubt it will be enough to be useful for detection by standard PI. A system that responds to magnetic susceptibility is necessary and a PI system could be made to do this. Multi-channel is also possible and I have personally have been involved in the design of such PI detectors for industrial use. Coils are generally rectangular and there is no need for a space between them.

I will look in detail about magnetic susceptibility. A quick read online shows that magnetometer might be able to detect it. I have ordered one Fluxgate magnetometer system (from https://www.fgsensors.com/) to test the samples. Till the time I receive it, I tried making a simple version of it as discussed in other threads of this forum and other site (this site) but could not make it properly.

You are right about PI detectors not detecting dispersed particles. In fact, I used a 10 inch coil PI detector and could not detect even about 2 lb of powder.

Just downloaded this paper as I was puzzled as to how iron oxide would help to strengthen concrete. Very interesting. In your test samples, what is the percentage of iron oxide in the total volume. The paper suggests 15% as being about optimum. I have only done a speed read, so there may be more info.

Experimental study of concrete made with granite and iron powders as partial replacement of sand.

A question about the steel fibers. Is their orientation in the concrete random, or are they all in more or less in the same direction?

Eric.

That paper has good information. I had around 20% by volume in epoxy sample.

The steel fibers are randomly oriented in the concrete. Usually steel fibers are used in concrete pavements to prevent cracks from opening up rather than for structural purpose.

Thanks Eric.

[sachindra]

I interpreted the steel 'fibres' as being randomly mixed in with the concrete, so orientation, which is certainly important to detection, should be random, unless the fibres 'flatten out' near to the surface.
They appear to be ferrous. There is no mention of them being stainless steel, and they are galvanised, which also points to regular steel. They are very high strength, I notice, 1200 MPa upwards, so clearly some fancy alloy, though the wire-drawing process is good for increasing tensile strength too.
( stainless steel bicycle spokes have UTS of 1200 MPa, even though the bulk material they are made from is more like 400 MPa )

True. Steel fibers are mixed randomly in concrete -- specially to enhance post cracking behavior (ie. hold the cracks tight).