Updated: Jun 9, 2020
A wide mouth food jar, such as a pickle or salsa jar, began to have failures on the filling line and it was reported that a competitor’s bottle was performing well on the same filling line with no failures.
This issue involved a standard stock container that could be sourced from multiple sources. Similar enough to other stock containers that it could easily be sourced elsewhere without label, closure, or filling line changes. This is a competitive situation for the glass manufacturer and a position of strength for the glass container user.
Various glass defect failures had occurred in recent history on this same container and same filling line. There were instances of glass defects in the finish area (threads and closure area) that had caused production losses and efficiency issues. Customer had warned the glass container supplier on quality and wanted to see improvement.
Business expectations were not that of a large filling location, where disruptions and chargebacks later would be an acceptable solution. Customer expected material to run as it always had, with no failures. The customer expected to be given a timely resolution, without having to make any changes to their line, their process, or their quality expectations.
Only a few samples and unclear information on their location and mode of failure were available early in the investigation, so an onsite meeting during a fill day was arranged with the customer. This would allow a face to face discussion on the failures and circumstances, and the ability to witness the breakage and the fracture analyst to collect samples directly from the filling line.
Upon arriving, it was found that the customer had switched to using the container on a different filling line with a different process. Non-technical staff at the broker who were previously notified of this process change did not pass it on to technical staff at the broker or the manufacturer. The customer went from a hot fill, to a hot fill retort process with hotter product and cooling water that was excessively cold resulting in a large delta T at the point of failure just prior to retort exit. This cooling water was below 70 degrees F. which resulted in a delta T of 120 degrees F in excess of the 75 degrees F. recommended in ASTM procedure C149 due to the filling temp of over 190 degrees F entering the retort process.
This change subjected the containers to a higher stress load and different mode of thermal shock. While the containers performed admirably during the hot fill process with little or no breakage, when they were subjected to the hotter product on the other filling line, and then cooling water below recommend temperatures per delta T standards, they began to fail at a high rate.
Unrelated quality issues had strained relationships and used up much of the normal flexibility expected during troubled times. Potential resolutions of inventory management to a process at the customer with no failures until complementary production that would survive the current process was produced was not able to be negotiated.
Containers that failed in the process were retrieved, cleaned, and analyzed. Unprocessed and intact samples from the same lot were also reviewed. The failed containers showed origins of failure at the ring and blank match, the seam of two mold tooling parts, underneath the bead of the finish. This was consistent across the failed sample lot, with some cavity numbers more prevalent than others suggesting a manufacturing condition may be contributing to breakage.
With thermal shock the identified load, and the resulting failures in an abnormal location under the finish bead at the ring and blank match, the unprocessed sample containers were reviewed. They showed an undercut or mismatch condition where the two tooling parts meet, showing an improper match from tooling ware and repair. Since this is a difficult to stress area in standard thermal shock tests, it was not surprising that this condition was not detected during routine quality checks. Visual checks were not performed, as this is normally not needed as this condition is related to tooling wear and repair, and tools in proper order and repair will not yield containers exhibiting this condition.
Intermediate non-technical staff were not able to properly communicate between the glass manufacturer and the filling location, on how to manage current inventory and future expectations of performance. It is certainly unlikely that a review of this change in filling process would have been enough to identify the underlying problem, as a review of thermal shock testing and other records would have given a clean bill of health. Possibly, had the information been given, it may have improved the reaction of the bottle manufacturer as it would have put the situation in the context of a process change and compatibility issue, rather than a simple review of standard quality data and confirmation. The glass manufacturer, based on the first report which did not contain the process change notification, may have reviewed records and believed there was not a glass problem and reacted accordingly.
The glass manufacturer slipped on good practices, which had not manifested as a problem before. By the time they could react the account was gone. The tolerances allowed on the mold tooling were too large, resulting in a mismatch condition that went undetected for an unknown amount of time. Standard test methods used, such as a thermal shock tank with the proper delta T, were not enough to detect the failure of the ring and blank match. Generally, this would not be an issue, as proper repair and change of the mold tooling makes this failure unlikely. The poor repair guidelines combined with the lack of available test methods and larger tolerances on the mold equipment, resulted in the glass containers being produced in a manner that was not compatible with the customer’s change in filling process.
Glass Guy LLC – works on the problem
The problem needs to be corrected by the glass manufacturer. Work can be done on the filling line to bring process parameters under better control for performance reasons, and to show technical staff during future problems what process controls are to eliminate erroneous avenues of investigations early on in conflict resolution scenarios. However, mold repair processes and tolerances need to be improved. When measured on a 10X comparator, the offset of the ring to the blank as measured on the glass container, in some cases exceeded .015” at various points when checked visually through a 360 degree rotation. Visual checks for these conditions can be done as required to verify mold repair.
The filling line escaped responsibility in this case due to the nature of the glass container as found during the fracture analysis. However, it can be said that having a product with an internal temperature exceeding 190 degree F and passing through a retort process, needs to have adequate cooling water temperature. While 70 degree F cooling water was detected, there were points were cooling water was found to be 54 degrees F when measurements were taken! The delta T of this process was a whopping 120 degrees F or higher, well exceeding and sometimes doubling the ASTM standard Delta T of 75 degrees F. It would be recommended that a more controlled process be implemented to reduce the risk of failure in this process.