PPT3000 M-Rule Shelf-life Field Test Comparison ® PPT3000 M-Rule Shelf-life Field Test Comparison This document contains information on the PPT3000 M-Rule Shelf-life field test comparison study conducted at a major PET converter with the assistance of an independent test lab Information on the procedure and the materials used to obtain the data are as follows Test Objective: The objective of this series of tests was to validate the test results of the M-Rule Shelf-life Prediction model using the Agr PPT3000™ and document the efficacy of this method in comparison to traditional methods for measuring CO2 loss in a given PET container Purpose of Tests: Verify that M-Rule CO2 predictive shelf-life test results obtained using the PPT3000 test system are comparable to results obtained through traditional real-time test methods Tests Performed: A detailed test plan was developed to adequately test the M-Rule Shelf-life prediction capabilities of the PPT3000 A common set of bottles was produced and used to test for CO2 loss Tests were performed and results compared according to the following test methods: ● M-Rule Web Model ● PPT3000 M-Rule Feature ● Headspace measurements Zahm-Nagle CO2 loss test ● Headspace measurements Septum CO2 test Conditions for all tests were established in advance and controlled Equipment used for tests: PPT3000: Agr Proto unit #001 Software v1.0.0.5 Test Samples: Mono-layer PET containers, of three different styles and capacities, were used for this series of tests Styles included: ● 20 oz (591 ml) ● liter ● liter Samples of each style were selected from the same production run, cavity sorted and randomized Strict attention was paid to the preparation and selection of samples to reduce the chance of variable test results due to sample defects or production anomalies Sample Set Preparation Samples were produced on SIPA blowmolder Three production sets of bottles were gathered for each style/capacity in the following manner: (# cavities + 100 additional samples representing all cavities) To ensure that at least three containers from each mold of the blow molder were contained in the set, the SIPA machine was sampled from one side All containers were aged (set aside in ambient conditions) for a period of at least 24 hours All containers were measured for thickness in the shoulder, mid-sidewall and heel area One measurement per location per container Containers were sorted for testing as follows: a) One set (one bottle from each cavity) was randomly pulled from the population for routine pressure and volume expansion tests on the PPT3000 pressure tester The purpose of this was to determine if the population is part of a statistically normal distribution b) A second set (one bottle from each cavity) was randomly pulled from the population This sample set was tested in the PPT3000 for CO2 shelf-life using the M-Rule profile c) The remaining samples were divided into two groups, boxed and shipped to two different test labs for long-term testing using the Zahm Nagel and Septum CO2 loss testing protocols Methods of testing: ● AGR PPT3000 w/M-Rule calculation ● Long-term shelf-life using Zahm-Nagel CO2 loss test ● Long-term shelf-life using Septum CO2 loss test ● M-Rule Prediction using M-Rule Web Model Actual Testing for Study Long-term tests: Bottles were cold-filled with carbonated water and tested at the normal intervals, per Pepsi qualifications, using the Zahm-Nagle and Septum test methods PPT3000 M-Rule tests: Test profiles were developed for each of the three bottle styles and loaded into the PPT3000 Sample sets were tested on the PPT3000 based on these profiles and results documented M-Rule Web Model: M-Rule calculations were run using the “empty” containers model, surface area was supplied by the Design group, and weights of the bottles prepared for this test Test Results The following graphs show consolidated data on each of the sample sets Each chart plots test results based on the Zahm-Nagel and Septum real-time tests as well as predictive data from the M-Rule Web Model and the PPT3000 M-Rule Shelf-life tests Notes: 1) PPT3000 test results are shown on the chart as a cluster of points rather than a trend This is because the PPT3000 is an instantaneous test and offers only one value for the predicted end result shelf life 2) The Septum tests were performed with a higher initial carbonation level than the other tests This accounts for the position on the graphs Study results Shelf Life - CO2 loss (days) Test run 20 oz liter liter (591 ml) M-Rule Model PPT3000 M-Rule Pressure-based test Internal CO2 testZahm-Nagle Internal CO2 testSeptum Test Test Duration 55 55 58 62 83 89 20 minutes 60 seconds 56±6* 56±7* 88±8* 84 days 54±6* 67±7* 89±8* 84 days *Note: Standard deviation for the Septum and Zahm-Nagel tests is between 0.02 and 0.1 volumes of CO2 Executive Summary of Results Real-Time Tests: Long-term shelf life tests conducted during this study performed as expected Both the ZahmNagle and Septum produced similar CO2 loss patterns This is evidenced in each of the charts for the representative bottle type Note that during this study, the Septum test was performed using a higher carbonation level than the other tests This accounts for the off-set of the graph However, when a correction is factored in for the higher initial carbonation levels, one can see a direct correlation of test results Predictive Methods: M-Rule Web-based Model, when appropriate bottle parameters such as surface area, brimful capacity, material characteristics, sidewall thickness, and test conditions were input into the program, produced similar results to both of the real-time tests The PPT3000 using the M-Rule Model likewise produced results that corresponded favorably to the long term tests Note, however, that the results from the PPT3000 represent shelf-life at a specific point in time In this case it was when the CO2 loss reached the 20% value Conclusions This series of tests show that the M-Rule prediction methods for shelf-life determination provide comparable results to long-term tests using the same group of bottles under similar conditions The results of these tests also demonstrate that it is possible to provide accurate shelf-life prediction of a container design using the M-Rule Model in the laboratory based on the analysis of bottle design characteristics Tests with the PPT3000/M-Rule Model further validate the prediction capabilities of the M-Rule Model Tests involving the PPT3000 also show that shelf-life prediction capabilities can be taken to the production floor to provide real-time shelf-life data on production bottles, at the time of production, and at an accuracy level that corresponds with long-term test results Rev These test results confirm that the PPT3000 with M-Rule can provide shelf-life prediction data with reasonable accuracy in a fraction of time it takes to traditional tests This combination of predictive analysis along with actual bottle performance data using the PPT3000 pressure/ expansion tests stands to offer a number of benefits to bottle manufacturers and fillers including: ● Capability to monitor shelf-life on a ● ● ● ● ● real-time basis Improved shelf-life management of production containers Faster response in the event of processrelated issues Up-front data and documentation of predicted shelf-life performance prior to container filling and shipment Potential to reduce or eliminate long-term testing Ability to monitor material utilization as it pertains to package performance, at the time of production The predictive capabilities offered by the M-Rule in the PPT3000 system make it possible to perform on-the-floor evaluation of production bottles in a fast and economical way, with accurate results that correspond to traditional long-term testing methods When used in conjunction with a traditional long-term testing program, the PPT3000 with M-Rule can be a valuable tool for managing and maintaining the expected shelf-life of production bottles on an on-going basis Copyright 2007 Agr International, Inc ● Phone: +1-724-482-2163 ● Fax: +1-724-482-2767 ● Web: www.agrintl,com ... ● AGR PPT3000 w /M- Rule calculation ● Long-term shelf-life using Zahm-Nagel CO2 loss test ● Long-term shelf-life using Septum CO2 loss test ● M- Rule Prediction using M- Rule Web Model Actual Testing... PPT3000 A common set of bottles was produced and used to test for CO2 loss Tests were performed and results compared according to the following test methods: ● M- Rule Web Model ● PPT3000 M- Rule Feature... each of the sample sets Each chart plots test results based on the Zahm-Nagel and Septum real-time tests as well as predictive data from the M- Rule Web Model and the PPT3000 M- Rule Shelf-life