Challenge test of an Automatic Machine for the Non-Destructive Detection of Integrity Defects in Blister Containers

Case study

The use of this CCIT method in the pharmaceutical field, asks for the several requirements to be met, in particular:

1. the non-destructive nature: the container must not be altered in any way by the testing. This aspect, means that a limitation must be placed on the vacuum levels to be applied in the testing chamber. The generated vacuum, must determine neither alterations in the quality of the lid, nor integrity defects. A non-destructive nature also allows testing to be repeated (if necessary) and applying to 100% of production, with the ability to subsequently place on the market all of the tested containers.
2. the qualification of measurement sensitivity: the system must be able to reliably and repeatedly detect containers affected by defects of particular size. The use of artificially created defects, makes the determination of sensitivity level possible. Among the various options available on the market, it has been decided to use containers micro-drilled by laser (Set P_5, P_10, P_15, P_20). Further sets of containers, with larger holes (macro-drilled), have been created internally using calibrated metal bits (Set P_55, P_100, P_250).

The case study, took place at the company manufacturing the LF-BLI machine. Two sets of items to be subjected to tests have been taken into account, in particular containers:

with no integrity defect (negative controls)

with defects artificially created by using laser (positive controls) (Figure 7)

Figure 7. Positive controls – containers micro-drilled by laser (5µm, 10 µm, 15 µm, 20 µm)

Table A specifies, respectively, the name and number of various sets of containers and the sizes of the holes in positive controls.

Table A. Specification of the containers used

This case study has the following objectives to be achieved:

1. model calculation and characterization of recipes: set the operating parameters of the CCIT by using the containers of Set N
2. sensitivity verification: check the detection of each container relating to Set P
3. statistical evaluation: determine quantitatively the statistical indicators, respectively, in terms of reliability, repeatability and performance measurement of the CCIT:

false positive rate

false negative rate

Sigma level.

Negative controls

Set N, was implemented by using the LF-BLI and by assessing the lid deflection of containers displayed on the HMI graphic matrix (Figure 5). This selection process, was carried out starting from a large set of containers, visually inspected earlier by qualified operators and, as a result, deemed to conform to closure integrity specifications.

The following actions were performed for each of these containers:

1. loading into the testing chamber
2. bringing the testing chamber to the hermetic closure condition
3. activating the vacuum by means of a dedicated HMI pushbutton: generating the vacuum level (V₀) and keeping this level for a preset time period (T₅=90 sec)
4. checking the behavior of the lid b the graphic matrix
5. approving and inserting in Set N if all Δ_ij ≥ THR-, or all Δ_ij ≤ THR+
6. exhausting the vacuum
7. starting a new operation.

Positive controls

Each of the containers belonging to Set P, has been drilled at the top and in the centre of one of its reference cells, in accordance with the aforementioned stated methods. This configuration, therefore, puts the hole at the point of maximum contact with S_ACQ in the period (T₁+T₂), in which the container is subjected to the differential pressure deforming the lid. This may potentially prevent air inside the cell from escaping; the lid of the cell in question, may not deflect fully, giving a false positive result. This choice was made in order to examine the worst case scenario.