9 replies to this topic

[Silviapa]

Hello Everybody,
Right now, I am updating a "metal detector checking of accuracy SOP" for a meat processing plant. The previous procedure that this plant had only consisted on running 3 test samples (Ferrous, Non Ferrous and Inox) every hour. I have my doubts with this procedure.
First, I think that the every test sample should be inserted into a product. In that way, we will know if the metal detector can reject that product. Is it right? What do you think?
Second, I think that every hour is overkilling. I've read online that the verification tests should include start up and at least every two hours or scheduled at break times, if appropriate. What do you suggest?
Please, I appreciate if someone can help me with this doubts or can provide me an example of "metal detector checking of accuracy procedure".
Kindest regards,
Silvia

[Charles.C]

Dear Silviapa,

Welcome to the forum!

Regarding the first question in yr post, there are 2 long discussions here on this aspect. The later one is here -

http://www.ifsqn.com...dpost__p__42547

and contains a link to the earlier one (similar length) inside it.

The second question/answer rather depends on yr actual situation (eg product, failure rate?) and/or how much of a (corrective action) gambler you are ?

Rgds / Charles.C

[Silviapa]

Thanks so much for your soon response Charles. Now, I have a much better idea.

Another quick question. The metal detector that I have is a THS model (manufacturer: eia). Does anyone know who can provide a calibration or maintenance service on this equipment? We are in Illinois.


Thanks,

Silvia

[KTD]

Is the man'f 'eia' or 'ceia'? Ceia is an Italian brand with US offices. www.ceia-usa.com will get you a corp address and phone number. They should be able to help you find a local contact...

[Charles Chew]

First, I think that the every test sample should be inserted into a product. In that way, we will know if the metal detector can reject that product. Is it right? What do you think?

Not necessarily so but I would like to share with you that for a "ferrous contaminant the orientation effects to trigger the "magnetic field" is very good when on its horizontal axis BUT is very poor when on vertical axis position. (it works on the reverse for "Non-Ferrous"). Therefore, the uncertainty in metal detection is a very common problem and cannot be relied upon. You may want to consider these issues when conducting metal detector re-calibration.

[Dr Ajay Shah]

It is also important to get your metal detector calibrated by the supplier who commisioned the machine at leat every year as the phase and sensitivity will change over time.

What Charles Chew has mentioned is a valid point.

I have implemented procedures where one has to check the metal detector at the start of the operation for Ferrous, Non Ferrous and Stainless Steel pieces supplied by the manufacturer and every two hours to ensure that the metal detector is performing as it should do.

i hope this helps.

Cheers

Ajay Shah

[MQA]

Hi silviapa,

In one plant, testing of the metal pieces was conducted every thirty minutes.

Why so often? Production volume was over a thousand units per hour. It was decided that if the pieces failed (which occasionally occurred), it was easier to re-test 500 units rather than 1,000 or 2,000.

So I guess you could say, your production volume and the quality of your metal detector will determine how often you test the metal pieces.

[MQA]

Also, a good read: http://haccpfood.blo...pliance+Blog%29

[Dr Ajay Shah]

Also, a good read: http://haccpfood.blo...pliance+Blog%29

Good Article JAKMQA.

Cheers

[GMO]

It is also important to get your metal detector calibrated by the supplier who commisioned the machine at leat every year as the phase and sensitivity will change over time.

I agree, however, I disagree that it has to be the supplier who commissioned the machine; as long as its serviced by a competent company that should suffice.