Proceedings of 36th Joint Propulsion Conference, Huntsville, AL, July 2000.

Aerospace Leak Test Requirements[*]

Barry T. Neyer, Senior Member AIAA
John T. Adams
Terry L Stoutenborough

Contact Address
Barry T. Neyer
PerkinElmer Optoelectronics
1100 Vanguard Blvd
Miamisburg, OH 45342

(937) 865-5586
(937) 865-5170 (Fax)
Barry.Neyer@PerkinElmer.com

Abstract

Most aerospace components are required to have a minimal leak rate to ensure that contamination, such as water vapor, does not enter the explosive cavity. Energetic components and systems are typically required to have a leak rate of less than 510-6 standard cubic centimeters of gas at a pressure of one atmosphere (510-6 STD cm3/sec).

The aerospace community typically uses the leak test requirements specified in MIL-STD-1576, which is based on MIL-STD-202. However, the DOD community uses MIL-STD-331, and the Electronics community uses MIL-STD-883. Although each of these tests is designed to measure the same thing, the test methods differ substantially. Moreover, many of the assumptions used to derive the equations in the test methods may not apply to the typical components used in the aerospace community. This paper discusses leak test requirements in general, details the differences between the test methods, discusses the applicability of leak test requirements to typical aerospace components, and gives suggestions for generalizing the leak test method.

 


Introduction

Energetic components are normally required to be sealed to guard against degradation of the component through contamination. Contaminants such as water or other solvents have the potential to damage the component. The damage may occur to the energetic material by changing the morphology or chemistry of the energetic material. Damage may also occur to the bridgewire through various corrosion mechanisms.

During the manufacturing process of typical energetic components, great care is taken to ensure that there are no significant sources of contamination. The component is also designed to ensure that there is not an easy path for any contaminate to enter the component.

All energetic components have various seals. There are the seals around the various conducting pins or optical ports, and there are one or more joints where the several pieces of the component are welded together. Because it is impossible to have a perfect seal, all energetic components are susceptible to having small leaks.

The typical leak rate for an energetic component used in the aerospace community is a rate of less than 510-6 standard cubic centimeters of gas at a pressure of one atmosphere (510-6 STD cm3/sec). This rate is assumed to be sufficiently low to prevent the introduction of significant contamination during the life of the component.

The aerospace community typically uses the leak test requirements specified in MIL-STD-1576, which is based on MIL-STD-202. However, other leak specifications are also found in the military specifications. For example, the DOD community uses MIL-STD-331 to measure very similar components. The Electronics Community uses MIL-STD-883 to measure electronic components.

Although each of these standards is designed to measure the same type of leak, the test methods and requirements differ substantially.

Moreover, many of the assumptions used to derive the equations in the test methods may not apply to the typical components used in the aerospace community.

This paper discusses leak test requirements in general, details the differences between the test methods, discusses the applicability of leak test requirements to typical aerospace components, and gives suggestions for generalizing the leak test method.

Ideal Leak

The leak rate of a component depends on several factors, some dependent on the component, and some dependent on the environmental conditions. Leaks are caused by holes, cracks, and pores in the outer shell of the component. Thus, the number and sizes of the various defects determine the leak rate. However, the leak rate is also governed by various environmental factors, such as the pressure differential, the temperature, and the gas mixture.

The standard leak rate for a component is defined as the leak rate that would be present at a temperature of 25 C with a pressure differential of one atmosphere. Many times, measurements are made under these experimental conditions. It is also possible to conduct the tests under different conditions, which give a more sensitive indication of leak, and to use well known transformations to give the results that would be obtained under standard conditions.

It is possible to determine the leak rate for a simple system using only two well-known laws from thermodynamics, the Ideal Gas Law and the Equipartition of Energy Law. The Ideal Gas Law,
(1)
equates the pressure, P, times the volume, V, to the number of moles of gas, n, times the ideal gas constant, R, times the absolute temperature, T.

The Equipartition of Energy Law,
, (2)
relates the energy per degree of freedom to the temperature. Here m represents the mass of a mole of the gas, and represents the mean square velocity in one direction. A similar equation can be expressed for the two other directions also. Use of equation (2) allows the calculation of the average velocity of gas molecules. The component of velocity in the x direction is given by
, (3)
with a similar equation for the other 2 dimensions. Since the atmosphere is composed of approximately 80% nitrogen and 20% oxygen, ordinary air can be assumed to have a molecular weight of 29 grams per mole. Thus, the average velocity of air molecules at room temperature is approximately 500 m/s, which is significantly faster than the speed of sound.

Consider an ideal case of a wall with a circular hole of area A. Figure 1 shows a schematic diagram. At any given moment, some gas molecules flow from side 1 to side 2, while other molecules move from side 2 to side 1. The number of molecules in the small cylinder with area A and depth vx Dt is
, (4)
where Pi represents the pressure on side i (i = 1 or 2) as shown in Figure 1. One half of the molecules in the box on side 2 will enter side 1, while one half of the molecules on side 1 will enter side 2 because, on the average, one half of the molecules are traveling towards the wall and one half are traveling away. (The number of molecules that move out of the box without entering the hole is exactly balanced by the number of molecules that move in.)

Figure 1: Ideal Leak

Taking the limit as Dt approaches zero and subtracting to get the net flow yields the instantaneous net change in number of moles of gas on each side of the wall. It is generally more convenient to use the Ideal Gas Law to convert the number of moles into volume of gas. This yields the equation

. (5)
Thus, the leak rate (dV/dt) is proportional to the pressure difference times the area and inversely proportional to the square root of the mass of the gas molecule. The standard leak rate, L, is the rate when the pressure differential is one atmosphere, the gas is air, and the temperature is 25 C. The standard leak rate is given by the constant

. (6)

It is instructive to consider what size of a hole corresponds to a given leak. For example, consider a hole of diameter 0.2 m. Such a hole is almost impossible to drill, even with a laser system, and is difficult to see except under extreme magnification. However, an improperly sealed joint can easily have a hole of this size. The standard leak rate for this hole is
(7)

Equation (7) shows that a relatively large single hole will still result in a device that meets standard leak rate requirements. Because the hole size is very large compared to the size of an air or water molecule (<410-10 m), all types of molecules will leak through. Even a hole with a diameter a factor of 100 smaller, which results in extremely small leak rate of 510-10 STD cm3/s is several times larger than an air or water molecule. However, it is not true that liquid water will leak through. The surface tension of water is large enough to prevent liquid from entering the cavity. A 510-6 STD cm3/sec is generally considered sufficient to guarantee that liquid water will not flow into a cavity.

One important aspect of leaks that is often overlooked is that the leak rate is given separately for each species of gas molecule. The pressures in equation (5) are the partial pressures of the individual gas types. For example, if one side of a wall contains 100% argon, and the other side contains 100% nitrogen, both gasses will flow to the other side, completely ignoring the flow of the other types of gas. The argon will flow out of the chamber with 100% argon into a chamber with a complete vacuum at the same rate that it will flow into one that has several atmospheres of pure nitrogen.

Because Helium is the lightest inert gas molecule, it is often used to measure the leak rate because it will leak by a factor of 2.7 more than standard air. As mentioned previously, the environmental effects on the leak rate are usually removed by performing the leak test at a standard temperature of 25 C and a pressure differential of 1 ATM using air. Equation (5) gives the correction factor if different environmental conditions are used.

Real Leaks

The previous section discussed a simple leak, which allows a simple formulation of leak rate as a function of various physical parameters. Components manufactured for use in the aerospace community may be represented by such an ideal case, or by quite a different case.

If the total leak is created by a number of leaks, each of which is large compared to a typical air molecule, then the total leak rate will be the sum of the individual leak rates. Equation (5) still holds because the total area is also the sum of the individual areas.

However, another possibility is that there is an large number of small leaks, such as in an O-ring seal. The leaks in an O-ring seal are caused by pores in the O-ring material, in addition to any leaks caused by gaps or defects in the seal. Although the leak rate will still follow the form of equation (5), the area factor, A, will no longer be the geometrical area of the hole. In addition, the effect of the gas type might be more complicated than the simple inverse square root form of equation (5). If the leak path of a device is composed only of such small leak paths, then it is possible that some of the larger species of gas molecules might not be able to leak through.

Consider a typical aerospace energetic component, shown schematically in Figure 2. There are two places in which such a component commonly leaks: in the glass to metal seals around the pins, and in the seal joining the two halves together. Usually, any leak is due to an improper joining of the two halves together, or due to cracks in the glass. In such a case, there is generally only a relatively small number of leaks, most of which have a physical size very large compared to air molecules. Thus, such a device will be governed by equation (5) and will allow any gas to leak through the cracks.

Figure 2: Leak Points in Typical Component

In the case of a device that has the potential for both types of leaks, the leak rate alone does not determine the ability to prevent various species from leaking into the device. Rather, the leak rate determines only the rate at which molecules leak in to or out of the component.

Leaks in Cavities

All energetic components have an internal free volume. The volume may be just the space between the particles of the energetic material, or it may also include bigger voids, spaces etc. If there is no void, then there can be no leaks, because there is no place for the gas to go. Thus, all energetic components may be approximated as a cavity with a leak path into the cavity.

The derivation in the previous section is valid only if the pressure difference remains constant. This is the situation for short amounts of time or for measuring the leak rate through a wall where the pressure on each side is kept constant. However, for sealed components, the partial pressure will change as the molecules leak into and out of the device. Eventually the device will reach a steady state where the number of molecules that leak in is exactly balanced by the number that leak out. After equilibrium is established, the partial pressures of each gas in the cavity will equal the partial pressures of each gas externally.

Since the leak rate, and thus the change in pressure, is proportional to the pressure difference, the pressure is governed by an exponential decay law. For example, if the initial pressure inside the cavity is Pi, and the pressure outside is Po, then the pressure will obey the equation
. (8)
where L is the leak rate, t is the time, and V is the volume of the internal cavity.

The cavity leaks, even after the cavity has the same equilibrium pressure as the outside medium. The net change in gas is zero, but the cavity continues to leak at the same rate, L.

It is instructive to consider how long it would take to make a complete air exchange in a typical cavity of an energetic component used in the aerospace community. For many of these components, the internal free volume is a small fraction of a cubic centimeter. Assuming an internal free volume of 0.1 cubic centimeters, and a leak rate of 10-7 STD cm3/sec means that it would take 106 seconds for all of the air inside to be completely replaced by outside air. Since there are 3.15107 seconds per year, this means that the air is replaced 31.5 times each year, or approximately once every 12 days. Thus, the air inside of a component that meets the typical hermetic sealing requirements is identical to the outside air within two weeks.

The preceding paragraph assumes that all leaks are due to holes and cracks, and not to the pores in an o-ring type of seal. If the component has only an o-ring type of seal, and is guaranteed to have no cracks or holes, then it may turn out that numerous species of molecules are prevented from leaking. However, it is not easy to prove that such a device has only a porous type of leak, and no cracks.

Equation (8) is valid only if there is one surface from which leaks occur. If the cavity contains a double wall or contains explosive powder that absorbs gas, then both the leak and bombardment processes are more complicated, and equation (8) must be generalized. In most energetic devices the cavity is filled with a very porous explosive mixture. Gas will diffuse relatively slowly throughout the powder. In most of these cases, the diffusion rate will be slower than the leak rate for a hermetic component.

Leak Test Methods

There are four main methods of measuring the leak rate of a sealed component: bubble, dye penetrant, helium bombardment, and radioactive gas. The through leak rate can also be measured on a header using some of these techniques.

Helium Bombardment

The helium bombardment method consists of bombing the components with helium for a time T, and then reading the helium with a mass spectrometer. Although any gas could be used in principle, helium is almost always the gas that is used for several reasons. Helium is the lightest inert gas. Since the mass is much smaller than air, an atomic mass of 4 versus the 29 for air, it leaks 2.7 times faster. It is present in such small amounts in the natural atmosphere that there is likely to be little present except for that introduced in the test. Finally, it is easy to detect in a mass spectrometer.

Unfortunately, the reading on the mass spectrometer does not directly correspond to the standard leak rate of the component. The factor of 2.7 for different gas or the temperature correction factor is corrected by multiplication. The measured leak rate, l, is a more complicated function of the intrinsic leak rate, L., the bombardment pressure, P, the bombardment time, T, and the time from removal of the chamber to measurement, t. The equation takes the form

(9)

Figure 3 shows the indicated leak rate versus the standard leak rate as a function of when the leak rate of the component is performed. The leak rates shown in this figure are calculated directly from equation (9). The internal volume of the cavity is 0.1 cm3, typical of energetic devices used in aerospace applications. The bombardment time of one hour and two atmosphere are also typical. The figure illustrates several important points.

Figure 3: Indicated Leak versus Standard Leak for a 0.1 cm3 component.

There are two different standard leak rates that yield the same indicated leak rate. There is no way to tell from a single measurement if the indicated leak rate is low because the actual leak rate is low or because the actual leak rate is so high that most of the gas has leaked out in the time between the bombardment and the measurement. Because of this ambiguity, a method of determining gross leaks, such as a bubble leak or a visual inspection, is required. It is also possible in principle to perform a second leak check and find the intersection point of the two corresponding curves on a figure like Figure 3, but this method will not distinguish between components with no leak and those with substantial leaks.

Furthermore, for any given leak rate, the indicated leak rate is a function of when the leak measurement is performed. As Figure 3 shows, waiting for long times before measuring the leak will always produce a lower indicated leak rate.

Military Specifications

All of the military specifications studied allow the operator to use the exact formula given by equation (9) to convert the indicated leak to the standard leak rate. However, because most operators who perform these leak checks do not want to have to produce a curve for each measurement time, general rules have been established that allow the operator to compare the indicated leak rate with a cutoff value. If the indicated leak rate is less than the cutoff value, then the actual standard leak rate is less than the acceptance value. Each military leak standard has a different table of values. Even though they are designed to give the same results, they are quite different.

Moreover, there is also a widely different set of requirements for determining the presence of gross leaks. MIL-STD-1576 requires a visual inspection, while MIL-STD-202, MIL-STD-331, and MIL-STD-883 require test such as a bubble leak test. Typically it is assumed that a fine leak method will find all leaks finer than 10-4, and a gross leak will find all greater leaks. Because the size of a hole that can produce a 10-4 leak rate is less than 1 m, extreme caution must be used when performing visual inspection to look for such small leaks.

Tables 1 through 4 give the bombardment parameters for each of the military standards. Below each table are also listed the standard leak failure criteria. It is implied by these tables that performing the bombardment leak test according to these parameters will yield the equivalent standard leak. A quick inspection of the tables shows that the bombardment parameters and specification limits are quite different.

Table 1: MIL-STD-202 Bombardment Parameters

Internal Free Volume (cm3)

Bomb Pressure (ATM Absolute)

Bomb Time (min)

Measure Time (min)

Reject Limit (cm3/s He)

< 0.4

5

120

60

510-8

>= 0.4

5

120

60

210-7

>= 0.4

3

240

30

110-7

*The standard leak failure criteria are 510-8 STD cm3/s air for internal cavities of less than 0.01 cm3, 110-7 for cavities between 0.01 and 0.4 cm3, and 110-6 for cavities greater than 0.4 cm3.

 

Table 2: MIL-STD-331 Bombardment Parameters

Internal Free Volume (cm3)

Bomb Pressure (ATM Absolute)

Bomb Time (min)

Measure Time (min)

Reject Limit (cm3/s He)

< 1

2

240

15

110-8

1 5

2

720

15

110-8

5 10

2

1440

15

110-8

The standard leak failure criterion is 110-6 STD cm3/s air for all cavity sizes

Table 3: MIL-STD-883 Bombardment Parameters

Internal Free Volume (cm3)

Bomb Pressure (ATM Absolute)

Bomb Time (min)

Measure Time (min)

Reject Limit (cm3/s He)

< 0.05

5

120

60

510-8

.05 .50

5

240

60

510-8

.50 1.0

3

120

60

110-7

1.0 10

3

300

60

510-8

10 20

3

600

60

510-8

The standard leak failure criteria are 510-8 STD cm3/s air for internal cavities of less than 0.01 cm3, 110-7 for cavities between 0.01 and 0.4 cm3, and 110-6 for cavities greater than 0.4 cm3.

 

Table 4: MIL-STD-1576 Bombardment Parameters

Internal Free Volume (cm3)

Bomb Pressure (ATM Absolute)

Bomb Time (min)

Measure Time (min)

Reject Limit (cm3/s He)

< 0.05

3

20

10

110-6

.05 .10

3

30

10

110-6

.10 .20

3

60

10

110-6

.20 .30

3

90

10

110-6

.30 .40

3

120

10

110-6

.40 .50

3

150

10

110-6

> .50

3

180

10

110-6

The standard leak failure criterion is 510-6 STD cm3/s air for all cavity sizes

Comparison of tables with Equation

It is possible to compare the tables with equation (9). The tables in military standards 202, 883, and 1576 are fairly consistent with the times and pressures that can be calculated from equation (9). At some volumes they would reject components with a factor of two less leakage, and at some volumes they would accept components with a factor of two more leakage.

The tables in MIL-STD-331 are quite different. Using the tables would result in the reject of components that had more than a factor of 10 less leak than the cut off limit. The main reason for this is that the smallest free volume in the table is 1 cm3, which is much larger than the typical free volume in an energetic component. However, even at the larger internal volumes, use of the tables would result in excessive rejection of components that met the specifications.

The other difference between the various tables and equation (9) is that the tables specify a measurement time that is constant with respect to internal free volume; it is also too small in general. Inspection of equation (9) shows that the exponential decay of pressure is proportional to the time divided by the volume. Changing the volume by an order of magnitude should result in a change in the amount of time needed to perform the measurement by the same amount.

A limit on the amount of time between removal from the pressure chamber and the measurement is needed to ensure that there is no confusion between the slightly leaking component and the one that has leaked most of the helium out of the cavity. Generally, it is assumed that a leak of the order of 10-4 can be seen by other methods. The measuring times listed in the table are, in almost all cases, much shorter than what is needed to reliably distinguish between the two sides of the leak curve shown in Figure 3.

In almost all cases, it is far more preferable to use formula (9) to derive the specific test conditions, than to use the table. Many fewer components could be improperly rejected, and a longer measurement time could be utilized.

Radioactive Decay

Radioactive decay is different than a bombardment technique in that a radioactive gas is placed inside the cavity of the component during device manufacture. Radioactive gas is used because it is possible to measure the quantity of gas remaining in the device, not just the gas that is leaking out. If the ratio of radioactivity detected at a later time is the same as when first measured after correcting for the change due to lifetime, then there is no leak. If no radioactivity is detected, then all of the gas has leaked out. The ambiguity of two different leak rates giving the same result of equation (9) is removed.

If a non-radioactive gas is used, it is still possible to measure the leak rate using a mass spectrometer. However, the ambiguity of very small versus very large leak remains.

Bubble Leak

The bubble leak method is used to find large leaks. It also has the advantage of showing directly the locations of the leak. The component to be checked is placed in a bath of a clear liquid such as water or alcohol. A partial vacuum is drawn, and the component is observed for leaks.

The aerospace standard, MIL-STD-1576, which is based on MIL-STD-202, specifies an absolute pressure of 63.5 Torr (mm of Hg). However, MIL-STD-331 method C8 specifies a total pressure of 600 Torr. A larger pressure difference (lower absolute pressure) will show leaks more readily and should be used when ever possible.

Dye Penetrant

The dye penetrant method is similar to the bubble leak method in that it is used to determine gross leaks and will indicate where the leak is. The devices are typically placed in a pressure chamber. They are completely submerged in a dye such as Dye-check, FL-50, Fluorescein, or Rhodamine B, Zyglo. A pressure of several atmospheres is applied for several hours. After the components are washed, they are inspected for any signs of the dye. Any sign of dye penetration is cause for reject. All of the military standards that mention these methods use similar techniques.

Summary

A comparison of the various leak test methods has shown that there are wide variations in the test parameters for measuring the same type of leak. In general, use of the tables in these standards is not advised. Furthermore, the rational for performing leak testing should be evaluated. The typical components used in the aerospace community will undergo many complete air exchanges with the outside environment many times during the life of the component. It is not at all clear that there is any rational basis, other than historical precedent, for requiring sophisticated leak testing for a typical component.

Acknowledgements

The authors wish to thank Julie A Thomes and Alan C Munger for reading the manuscript and Jeremy R Lovely for technical assistance.



[*]Copyright 2000 by PerkinElmer. Published by the American Institute of Aeronautics and Astronautics, Inc. with permission.