5 replies to this topic

[AtomicDancer]

An auditor recently asked how we challenge our metal detectors. 

Our response: we calibrate via 3rd party once per year, we validate/verify by the production team running test cards during the run / QA observing this check daily. 

They responded that wasn't sufficient. We needed to have a challenge to make sure the equipment was working at its optimal settings. 

 

How would you challenge your metal detector? 

 

*Note: We have some finished product bag metal detectors and some in-line gravity fed metal detectors.  

[Tony-C]

Hi AtomicDancer,

 

In an ideal world the metal detectors should be challenged using a test pack of the food with a test piece inserted into the pack. It doesn’t seem that you do although that doesn’t seem to be unusual. Maybe the auditor was looking for challenge tests with test packs?

 

Gravity fed is a different ball game as using test packs isn't feasible.

 

Do you have the metal detector commissioning data which should have included setting up the metal detector with test pieces in the product (test packs)?

 

If you aren't doing so, I would normally be running an annual validation using test packs and also when there are changes to the product, product size, packaging material or equipment including conveyor speeds. It would make sense to do that during your annual calibration.

 

Kind regards,

 

Tony

[GMO]

This is one of the best sources I've ever found for explaining metal detector technology including how to test them etc.

 

Guide to Metal Detection Technology

 

It's good as a reference for HACCP plans etc and free to download. 

 

In general you normally want to make up test packs with the test piece in the lowest sensitivity portion of your metal detector (normally the middle) which might or might not be inside the test pack depending on the size of the aperture and the depth of your pack. You'd normally use these test packs to test the machine at the start of run, end of run and an appropriate frequency throughout production where the product would still be in your control in the event of a test failure. Also sites often choose this test frequency by how much product will need to be rechecked if a test fails. It depends a lot on throughput and how much on site storage you have.

 

The manual you can download from the link explains it all though and is very useful.

[Baohan88]

Hi AtomicDancer,

 

In an ideal world the metal detectors should be challenged using a test pack of the food with a test piece inserted into the pack. It doesn’t seem that you do although that doesn’t seem to be unusual. Maybe the auditor was looking for challenge tests with test packs?

 

Gravity fed is a different ball game as using test packs isn't feasible.

 

Do you have the metal detector commissioning data which should have included setting up the metal detector with test pieces in the product (test packs)?

 

If you aren't doing so, I would normally be running an annual validation using test packs and also when there are changes to the product, product size, packaging material or equipment including conveyor speeds. It would make sense to do that during your annual calibration.

 

Kind regards,

 

Tony

Hello, I'm in a similar situation, but the metal detector I'm using is a gravity drop type. If that's the case, are there any other ways to conduct a challenge test it?

[Tony-C]

Hello, I'm in a similar situation, but the metal detector I'm using is a gravity drop type. If that's the case, are there any other ways to conduct a challenge test it?

 

Hi Baohan88,

 

If money is no object then I believe auto test system are available. See the image below from Mettler Toledo Guidance ‘sophisticated automatic test systems are available to guarantee center-line performance.’

 

 MT How to Correctly Test Your Industrial Metal Detector.png   1.25MB   0 downloads

 

Mettler Toledo Tips & Tricks:

‘It is highly unlikely that the above test procedures (or any other manual test procedure undertaken) will result in the test piece actually passing through the center of the metal detector’s aperture. As such the test will not actually confirm worst case performance.

In such instances, a more controlled test should be done at agreed intervals where the test sample is guaranteed to pass through the center of the aperture i.e. worst case testing. To facilitate this, sophisticated automatic test systems are available to guarantee center-line performance. These can be used to supplement the manual test method described above or used periodically to confirm center line performance. Other benefits of using auto test devices are covered later in this document - turn to page 10 for further details.’

 

Kind regards,

 

Tony

[GMO]

Hello, I'm in a similar situation, but the metal detector I'm using is a gravity drop type. If that's the case, are there any other ways to conduct a challenge test it?

 

A "throat" metal detector? Is this linked to a bagger or is the product packed further down the line?

 

Throat metal detectors work but they are a compromise over an in pack detector. You cannot test them fully in the same way that you would a conveyor. Most people use a "ball" like test piece and put that into the flow. For example, if you have a multihead weigher, into the product feed. If there is a bagger, normally what happens is the outfeed makes a double bag and stops.

 

These are the kind of test pieces. It's just the one which came up first. Other people make them.

 

Test Samples, Tablets & Cards - Overview - METTLER TOLEDO

 

The question is whether you should restart the flow and run product before doing your other test pieces. Ideally you should. But then that means you can't test them consecutively. If you don't, you're not testing them with product flow. So either way is a compromise over "best practice" metal detection. Also you can't realistically do a memory test etc.

 

Note that timing issues can cause jaw damage if you do have a flow wrapper on the end of your throat metal detection so timing is important. Also, let's be honest, losing your test piece is very possible too. You would want someone at the base of the machine checking for when it comes through so that doesn't happen. This makes testing your detector a 2 person job.

 

So for all of the above, unless you have some packaging you're using which is impossible to metal detect in, a better situation is to detect once the product is already wrapped later in the flow (which may also have the benefit of testing the outputs from more than one machine and also it's better practice to detect once wrapped). Then if you retain your throat detection as well you can get away with testing less frequently. 

 

I've never seen the automatic test systems Tony is suggesting but I can't see how some retailer standards in the UK would accept them with the way they're written. Obviously not a problem for you but I only mention that as it's something you might find heavily questioned at any audit but also you need to be sure it's genuinely working as you think for food safety reasons.  You'd need to make sure you validated the efficacy of that auto test in some way in my mind and were really sure you were confident and could prove to others that it worked. It's particularly reject timing that I'd worry about.