Model calculations for maximum allowable leak rates of hermetic packages

Abstract

Calculations are presented which apply to electronic packages prefilled with helium for leak testing. Depending on the geometry of the leak, the leak rate of the package, while in use in air at ambient pressure, may be equal to or less than that measured while it is tested in vacuum. Two principal flow mechanisms are identified: viscous (including slip), driven by the difference in total pressure between the inside and outside of the package; and diffusive (diffusion and molecular flow), driven by the difference in partial pressure of helium. Air leaks into the package at a rate comparable to the rate at which helium escapes by diffusive flow. An equation is presented which relates leak rate to leak geometry for a capillary leak, and generalization to other leak shapes is made. It is shown that for a 100−cm3 package, with a maximum allowable leak rate of 5×10−6 atm cm3 sec−1, if pressurized to 1 atm (gauge) with helium, most of the pressurization will be lost within a year. If unpressurized, water vapor will accumulate at nearly 3000 ppm/yr. A more stringent standard is proposed of 10−7 atm cm3 sec−1 for a package of this size when pressurized to 1 atm (gauge), or 10−8 atm cm3 sec−1 if unpressurized. The concept is introduced of the ’’marginal gross leaker,’’ defined as a package with the largest leak rate undetectable by a gross−leak test, taken here to be 10−3 atm cm3 sec−1. It is shown that the maximum allowable length of time after filling and sealing a 100−cm3 package with such a leak rate, before it loses enough helium to erroneously pass a helium leak test is 6 days if the standard is 5×10−6 atm cm3 sec−1, and 9 days if the standard is 10−7 atm cm3 sec−1.