Evaluation of gas flow models and simulation of package integrity testing

Vivek Gnanasekharan, Purdue University

Abstract

Ensuring the integrity of food packages has become critical because of the proliferation of flexible and semi-rigid packaging. The goal of this work was to develop models for predicting gas flow through capillaries and microorifices and use the models to numerically simulate two types of pressure difference package integrity tests. Published models for predicting gas flow through capillaries and microorifices were evaluated using experimental data for molecular, transition, laminar and turbulent flow. Models for capillary flow were modified to improve their predictive power, particularly under turbulent flow conditions. An empirical function was developed to predict orifice coefficients as a function of Reynolds number. The final capillary and orifice flow models adequately represented the experimental data. The validated models were used to numerically simulate pressure difference testing. The effects of seal defects and pinhole leaks with diameters 0.1 to 100 $\mu$m were examined using the simulated tests. Lid deflection, test chamber pressure, package characteristics and other parameters were incorporated into the simulation. Based on commercial sensor sensitivity and a test time of 2 s, simulation results showed that 50 $\mu$m seal leaks and 20 $\mu$m pinhole leaks could be detected by both pressure and vacuum lid deflection tests. Leaks of a given size could be detected at lower pressure differentials by the vacuum test as compared to the pressure test. Detection based on test chamber pressure allowed detection of 30 $\mu$m seal leaks and 5 $\mu$m pinhole leaks. The use of increased test time (40 s) and pressure differentials (91 kPa) improved leak detection sensitivity. Under such conditions, the detection of 10 $\mu$m seal leaks and less than 1 $\mu$m pinhole leaks was possible. Test sensitivity was not affected by a $\pm$5 $\sp\circ$C change in temperature. Sensitivity increased with decreasing package lid thickness, leak length, headspace volume, and test chamber volume. Sensitivity was also affected by lid material and test gas.

Degree

Ph.D.

Advisors

Floros, Purdue University.

Subject Area

Packaging|Food science|Mechanical engineering

COinS