Detecting Metal on Metal


Thread Starter

Lalo Mora

What type of sensor can detect a little piece of metal when it is placed in a bigger piece of metal?? Consider that they are the same color.

If you try to sense it from one side..
the piece to be sensed is set just 1/4" above the surface of the bigger piece and there are variations in position.
If you try to sense it from above
The big part has a hole then the little part(to be sensed) goes in the hole but it also has a hole in the center..

right now I´m using photoelectric sensor (difusse system) but I´m not getting good results...

Any of you can help?

Thak you

Steve Myres, PE

If the reflectivity is right, you might be able to use a PE with a focused beam. This is pretty good at only sensing objects at some preselected range. Otherwise, you might try a pair of see through sensors in a horseshoe configuration, and move the workpiece relative to the sensor, till you find the protrusion, then rotate 90 degrees and repeat.

Kind of kludgey, I know, but when you dont know where the protrusion is supposed to be, what can you do? You might try a vision system, if all else fails.

Depending on the degree of position uncertainty, you might be able to use a proximity sensor. I once used the short detection range of small diameter proxes to my advantage, by pressing a dowel pin into the tool surface, left projecting about 0.060, but I knew the sensor would cross over the pin with good accuracy. The drawbacks with this will be the restricted sensing area (lateral) with the small dia prox, and some uncertainty about lateral position, when you do find the insert.


A triple-beam sensor can be used from above to sense the height difference. They can be set to detect within a very tight range 'window'. A fiber optic 'thru-beam- sensor with the emitter on one side, and the receiver on the other, could also see the height difference looking at the side. Diffuse is about the worst method of photo sensing that you could use....

--Joe Jansen
Does the small piece fit into the hole in the big piece?

Can you just look for hole diameter? (small hole - part , large hole - no part)


Christian Dow

If you use a fiber and a lens that is focussed to the height of the top peice you can get a lot of value difference while using a digital fiber optic amplifier.

ricardo araujo

Good afternoon

Well, I think you can try a sensor system based in Laser difusion. The Laser is more thin and the process can be more easy to detect, and the cost is not so higher...

Other solution could be a couple of inductive sensors, comparing the input. The metal peace above the bigger can be identiffied with two pulses, and the bigger with one pulse ...
I've used a Sick optics laser for very similar applications with great results. It's a pricey option, but if it's cost-effective for ya, you can set these things with a relay output that you can set without ever having to worry about taking in the analog signal. Very tight resolution, very expensive.

Good luck to ya,
Jack Benner
Caster Automation
Nucor Steel Hertford

Gurdas Singh

Can you avail the service of a PC? If yes, try using a camera based system. With pattern matching the system is able to find the presence of the part anywhere in the focus area. I know of cameras from Basler. You'll need to use a videop capture card also (like from "": ). This will work but is expensive.
try to use a ccd camera ,first capture the picture
and then do some kind of image processing
just a guess try it

Nathaniel H. Ericson

Your best bet would probably be an SIE capacitive sensor. They can be adjusted to sense very minor differences in parts. I have been able to sense the presence or absense of support wires inside a lamp (easy to do with a photo sensor on a clear lamp; very difficult on a frosted or colored lamp).

Berns J.Buenaobra

My suggestion is all together rid off proximity sensing techniques and employ machine vision system which could use a monochrome CCD camera and a frame grabber. There will be some image processing programming using "segmentation" techniques. I've used NI IMAQ and LabVIEW6i when I designed custom algorithms for QA inspection of timepieces before.
I presume that your requirement comes in the "Assembly Line", to check whether the small metallic component is placed on the larger metallic piece or not.

The solution I recommend is based on "Doppler Effect". You need to place two sensors in the assembly line, in such a manner that one sensor is placed before the "Smaller" Metallic piece is introduced and the next sensor is to be placed after.

The difference in frequency between the "Incident and the Reflected" beam in both the cases (Travel Time is the Actual Measurement) will help you to check the presence of the "Smaller Metallic" piece.

"If your line permits the placement of Inductive proximity switches, then this will be cost effective" than the above one.

Derek Appleton

There are devices called 'Line Arrays' that are a optical device that are equivilant to many, small photoswitches in one sensor. These devices can give a analog output based upon the % of blacked-out region they may be just what you need. I used them years ago but can't remember a manufacturer. Ask around some of the suppliers. They will probably test one on site for you.

Hope this helps.
I recommend a CCD Laser Displacement Sensor from Keyence It's a good choose for your application.

Luis Cardona
Control Systems Eng.
Ethicon Endo-Surgery Inc.

D Ruggles, Power/mation

One thing that must be determined first is what is the sensing environment? Second; What is your budget?

Photo electrics (PE) are not always the best for dirty environments, but must be used in certain cases, therefore the PE type, i.e. diffuse, convergent, etc. is crucial.

Inductive proximity switches are very rugged and can be introduced in very harsh environments. They work well in applications like sensing metals in punching and stamping operations. Capacitive sensors should be used in very specific applications. They work well for sensing wire, liquids, etc. items with high dielectric constants (Dk). Air has a Dk of 1.0 while water has a Dk of 48-88. Add humidity to air and you may have a problem if you adjusted your capacitive sensor to a carbon steel piece that could very well have a Dk of 22.

Not knowing too many specifics about your application I would first try PE's for cost sake. Test for reliablity and determine if you have a fit for PE's in you appliction. Banner Engineering has some of the most robust and widest selection of PE's on the market. They have some very low cost laser convergent type sensors that I believe would fit your app.

If your budget allows try vision.

Good luck.

Vineet Kumar Goyal

Did you try Retro-reflective sensor; or just an inductive sensor at a fixed distance from the top!

Mike Mitchell

Can of corn, use a Cutler-Hammer E58 2" Perfect Prox. This sensor can be used in harsh environments and has the cut-off (background supression) that a standard diffuse can not offer. The spot size of visible light is .250 and you should be able to detect from either direction. Another option is to use the iProx from Cutler-Hammer, which is a fully programmable inductive sensor. I would suggest the band sensing mode. The mode of operation would allow you to select a band as small as .040 and that is the only time the sensor will activate. You would mount the sensor looking down on the two pieces, select a distance to the top of the ridge of the center piece, then add about .100 to that distance to set up your band and problem solved. Contact CH at 1-800-426-9184 for a local distributor.

Mike Mitchell


Try using a magnetic sensor, looking for changes in polarity. this app was used by the DOT for rail brake analysis in realtime. Key in a absolute location device, set up a memory map of the N/S poles. Changes in N/S will indicate the change(s) in material, even when the same material is used.