Dear All,
Looks like this thread is similar to previous, ie plenty of confusion. Part of the problem is the number of possible variables, eg package contents-shape-presentation / test piece shape / magnetic array shape / other geometrical aspects / false positives. Plus the inability to draw pictures in a thread post.
Objective : To ensure that the occurrence of a defined( eg 3D shape / sized metallic object) contaminant “in” a defined product unit (eg 3D s/s package) is “adequately” detected when occurring at a
position within / at the surface of forementioned package
such that the likelihood of detection is a minimum. ( presumably it may be required to establish whether placement on the surface gives a significantly different result as compared to being “in” the surface.) From memory the theory ideally prefers to (initially) nominate the position to be a point location (as per usual ideal).
So the question is apparently something like – assumptions - (a) that the geometry of the detector/conveyor configuration is fixed and in the conventional array style, (b) a uniform rectangular box is involved with its long line parallel to the belt flow and placed symmetrically about the centre of belt. Then - where is the point (see 1st para.) of minimum likelihood of contaminant detection, ie minimum sensitivity of detector. Answer from memory is at some point on a hypothetical line within the product in box/in the surface of product within box / parallel to conveyor belt and which achieves closest approach to analogous line through geometrical center of coil array. Don’t remember if any significance as to where the point is along this line, eg near end which enters field first or last or in the middle (I recall this aspect was discussed in published (Tesco?) test procedure in previous thread) ?
And therefore, as GMO initially postulated, for the “low” box scenario, the desired line is presumably “in” the top surface of box, ie at the intersection of this surface and a vertical plane which passes through the box’s geometrical center and is parallel to long box axis. Or maybe (to be validated) can be “on” the box top surface also.
And the easiest way to validate is presumably to simply test, ideally perhaps initially with a spherical standard size object of required diameter, eg a ball bearing, but most commercial test pieces seem to be strips, eg wands. Yet another variable to take into account.
And as already mentioned, the available detector strength may be limited by possibility of false positives.
Now even more confused ?
Rgds / Charles.C
