QA advice and best practice for the correct selection of
calibrated test samples for your application of inspection systems according to
BRC, EFSIS and HACCP.
A test piece is just a test piece?
Not exactly - just like inspection systems one size really
doesn't fit all. After all you haven't invested the time and resources into
inspection capital equipment to detect the presence of a test piece. You need
to validate the fact that the inspection system will keep your products free
from foreign body contamination.
Contaminant type and shape
If you look at the test pieces you are using now, you will
notice that the contaminate embedded is a sphere and the diameter is shown in
mm. Most metal detectors work on the principle of a balanced coil, as
contaminates and product passes through the system it unbalances this
arrangement.It is this unbalance that determines whether the contaminate will
be detected. As well as the material type the cross sectional area of the
contaminate plays the most significant part in this. That's why we use spheres
- no matter what orientation the test sample enters the metal detector the
cross sectional area will not vary.
The most common materials used to test inspection systems
are Ferrous (Fe) iron, Non Ferrous (NFe) typically brass, and stainless steel
(SS) typically non magnetic 316 grade. Unfortunately in the real world the most common contaminants
will not be spherical, metal is most likely to be in wire or shaving form.
Remember the metal detector works on cross sectional area? You could in theory
get a long thin "needle" type contaminate - if it was only 1mm in
diameter and your metal detector detection limit is 2.0mm it could be 12"
long and you still may not detect it.
X-ray systems operate differently, they work on density, the
amount of energy absorbed by the contaminate together with the shape determines
detectability. You could have a 1mm thick blade and if your X-ray is set to
detect 2.0mm you may miss the contaminate even if it was 1" in diameter if
it was to lay flat perpendicular to the X-ray fan shaped beam. When selecting
glass contaminates for test samples - care should be taken to understand the
type of glass sphere being used. Density varies greatly from one glass type to
the next. Commonly Soda Lime Glass (SLG) is used as it provides a semi
standard. If you package your product in glass jars, the most likely glass
contaminate is probably going to be your jar glass, following a breakage etc.
it may be worth considering having custom samples made that use your glass
type.
Contaminant Size Used for Testing
Most commercially available test samples use calibrated
spheres of different materials. When you select your size specification to test
with it is important to choose one that is reliably and repeatably detected.
Too small and the inspection system may miss it, too large is not a
representative test. As a rough rule of thumb if you have a wet or salty
product (reactive) and your metal detector will detect 2mm Ferrous, it will be
around 1.5 times this for Non Ferrous so 3mm. Double the ferrous for stainless
steel (as its non magnetic and non conductive) so 4mm approx.
If you have a dry product - such as grains, or deep frozen
product the Ferrous and Non Ferrous will be the same and the stainless steel
approximately 1.5 times this.
X-ray systems are not affected by the frozen/thaw state of a
product, working on density (X-ray absorption) so Ferrous, Non Ferrous and
Stainless Steel will be the same size. Hence most X-ray systems are only tested
with Stainless Steel and Glass. As mentioned earlier different glass has
different density so ensure you know the type of glass in the test sample.
Your product and application will affect the size of
contaminants detected. When you write up your HACCP procedure ensure you are
using the correct size contaminates. Your metal detector or X-ray system
provider will be able to assist with determining the sensitivity of your
system. Thereby establishing a realistic specification. Your inspection sample
provider should also be able to assist you with setting sensitivity and
arranging routine testing procedure and annual calibration.
Calibrated Test Sample Format
As mentioned earlier the purpose of your inspection system
is to detect unwanted foreign bodies in your products - to ensure final product
quality for your consumers. Laying a test piece on the belt of a metal detector
may be a repeatable test, but do you know how that relates to real world metal
detection? Consider your reject type and ensure a contaminated product can be
rejected positively, so consider the position of the test sample.
You should consider a test piece format that allows you test
your system in a repeatable, representative manner. For example if it is a free
flowing product the sample needs to move with your product. If it is a conveyor
based system with a pneumatic or mechanical reject then ensure that the
contaminate can be located anywhere within the packaging and still be detected
and rejected. On the subject of location - product itself can shield the
effects of the contaminant. A 2mm ferrous test piece that is easily detectable
on the outside of a block of cheese may not be detectable at all inside.
Consider inserting the test piece so that the contaminate sphere is embedded in
the product.
Do not use plastic test pieces on X-ray systems - the way
the contaminate spheres are embedded makes it far easier for the X-ray to
"see" them. The image analysis software is looking for density
differences and an area of high absorption (contaminate) surrounded by low
absorption (plastic test piece) makes it highly detachable, even if embedded in
the product. A real world contaminate would most likely be surrounded by higher
density product. Use flat X-ray test cards to test X-ray systems they are
supplied in flat card form to minimize the effect of the contaminate
"holder".
The center of a metal detector
aperture is the least sensitive part, closer to the lining (top, bottom and
sides) is more sensitive. Consider this when you determine where to place the
contaminate for your routine testing. Make sure you can recover the test piece
following the test, you don't want to cause a contaminate problem with testing.
If you make up test packs make sure they are clearly marked and do not get
mixed up with good product.