Test Like You Fly

Although test-like-you-fly directly relates to aerospace projects, the underlying intention is to mitigate risk by running the most appropriate tests. In that sense, the phrase applies more broadly. Skipping or compromising on any part of a test, even one that may seem to be inconsequential, has led to mission failure. So, it is no wonder that test-like-you-fly has gained a special meaning among mil/aero professionals.

It seems obvious that environmental testing should expose equipment to the conditions likely to be experienced in service. However, determining the actual deleterious process occurring in a real environment may not be straightforward. Temperature extremes might cause less damage than a rapid rate of change, and high levels of vibration could be less important than dwelling at a specific resonance.

Many other environmental factors are less well defined than temperature and vibration. Nevertheless, the confidence with which equipment can be put into service directly relates to its environmental test performance, assuming that the tests are truly representative.

In addition to ensuring that the DUT functions properly, testing also must verify its modes of failure. This is especially important when failure results from exposure to fire and has security or safety implications.

Fire

For example, network equipment building system (NEBS) fire tests are designed to ensure the continued operation of a telephone exchange. These tests are distinct from those required to prove operational functionality within the telephone network. Equipment intended for use in an exchange must comply with strict design guidelines, including the use of self-extinguishing materials. A rack of equipment must have sufficient baffles and separate chassis to minimize the spread of flames from any part of the rack to the rest.

There’s clearly a security risk associated with telephone exchange fire damage but typically a small safety risk. In contrast, aircraft or shipboard fires endanger the lives of all on board if not quickly brought under control. Extensive fire testing of fuel lines and hydraulic hoses is performed to confirm their capability to continue functioning while exposed to flames.

It’s especially important that the rate of degradation allows sufficient time to extinguish the fire before the fuel or hydraulic fluid ignites. To verify hydraulic-hose performance, AERO NAV Laboratories designed and constructed the fire test chamber shown in Figure 1. Several propane gas burners can be used, the number depending on the size of the DUT.

Sheldon Levine, the lab’s vice president of marketing, described typical test requirements, “Fire-resistant hydraulic hoses must be subjected to direct flame impingement while flowing fluid under specified flow and pressure conditions. The hose is subjected to the flames for a period of 30 minutes, a time sufficient to allow the fire to be sensed and extinguished. The hose must withstand the flames while maintaining pressure without leaking. In addition,” he concluded, “it must pass a hydrostatic pressure test after the fire test.”

Lloyd’s of London Flexible Hose Fire Test OSG/1000/499 requires that a hose withstand 705°C (1,300°F) for 30 minutes at rated working pressure. Typically, testing to this Lloyd’s specification is performed on hydraulic hose used in marine applications such as offshore drilling rigs as well as ships.

High-pressure hydraulic hoses are made with several layers of steel spirals that provide strength while maintaining flexibility. The innermost tube often is synthetic rubber, and adhesive layers may be applied between each steel layer with an overall protective sheath providing abrasion resistance. For this type of hose, a static pressure fire test is the most stringent because heat conducted from the flame to the central tube is not removed by the flowing fluid. Mr. Levine said that the fire chamber also was used to test fluid-line gaskets in a no-flow condition.

In a 1976 article about brazing the titanium alloy tubing used in Concorde’s hydraulic system, the author described the direct-flame tests applied to materials and components within the engine bays. This environment is quite different to the rest of the airframe, which was expected to operate at slightly elevated temperatures because of the plane’s high speed.

For the engine bays, components “are required to withstand a ‘torching flame’ test that simulates the conditions that can arise should perforation of an engine combustion chamber occur while the engine is delivering thrust, projecting a searing tongue of flame into the engine bay until engine shutdown is achieved…. [These] tests are carried out on typical assemblies that include both mechanical and brazed joints, pressurized to the working pressure (4,000 lb/in.²) with the appropriate hydraulic fluid under no-flow conditions to offer no benefit from cooling flow and to equate to the most severe conditions that can occur in service.”¹

The paper goes on to discuss the unsatisfactory results of initial tests and the subsequent change to a higher melting-point, gold-bearing brazing alloy that could withstand prolonged exposure to such extreme temperature.

Today’s electronic assemblies seldom are subjected to direct flame tests, but dust and salt spray are among the many types of environmental tests often applied, especially when qualifying mil/aero and automotive equipment. Each of these tests has a long history beginning with actual dust and salt spray as the names suggest. However, over the years, several test variants have been created to more precisely address a range of conditions.

Dust

Completely sealed equipment is capable of working well regardless of the dust level in the environment. Of course, this is not the case for equipment that relies on external air and fans for cooling. Filters generally are used, and they eventually become clogged unless regularly cleaned.

Air filters also are used on internal combustion engines, and it was in conjunction with AC Spark Plug engine filter design that dust testing began. Arizona road dust, actual dust from the Salt River Valley, originally was used.²

A large percentage of extremely fine, highly abrasive particles was the feature most concerning about Arizona dust. Before a standardized production process was developed, a canvas cloth collected airborne dust from behind or around tractors. A comparison of Salt River Valley and Imperial Valley California dust is shown in Table 1. The large difference in particle size distribution is obvious.

From 1982 onward, Powder Technology (PTI) became progressively more involved in the production of test dust to Society of Automotive Engineers Air Cleaner Test Code Specification J726. In 1992, AC Spark Plug ceased test dust production, hiring PTI to fulfill outstanding orders. And, in 1997, ISO 12103-1 Road Vehicles—Test Dust for Filter Evaluation was published listing four grades: ultrafine, fine, medium, and coarse.

PTI literature describes each of the test dust grades in detail, comparing them to AC Spark Plug dust and listing suitable test applications. For example, ultrafine test dust is nominally smaller than 10 microns in size and used as a test contaminate for fuel-system components and water-filter performance evaluation.

Trace Laboratories recently conducted a test that simulates dust infiltration occurring in the working environment. Chris Finch, the lab’s technical sales manager, described the situation: “A customer that makes joystick controllers for earth-moving equipment required that the product be life-tested for 5,000,000 cycles while subjected to a dust-filled environment. During normal dust testing, you place the item in the chamber for a set period of time without cycling the unit under test but while agitating the dust and maintaining a set concentration level. Trace’s dust durability test setup includes a LabVIEW-based program to mechanically and electrically cycle the joysticks under a simulated vehicle load. The program provides continuous monitoring and storage of all test data.

“The dust is constantly cycled for 45 seconds and allowed to settle for four minutes,” he explained, “and the software is designed to monitor and control the test rig and stop the testing if there is either a mechanical or electrical product failure. Throughout the test cycle, the units also are observed visually for other anomalies.”

Several key points summarize the joystick dust test:
• Simulates dust infiltration over a 23°C to 40°C temperature range.
• Uses ISO A12103-1 A4 coarse dust.
• Continuously monitors electrical function.
• Simulates in-vehicle electrical load conditions.
• Verifies the operation in real-world conditions found during the normal operation of the system.
• Detects both electrical and mechanical wear or malfunction that would not be detected in separate test sequences.

While researching dust testing, several articles were found relating to the U.S. Army M4 carbine dust test conducted in late 2007. In this case, a loaded carbine was exposed to extreme dust conditions for 30 minutes and then test fired using 120 rounds. This cycle was repeated with the carbine wiped down and lubricated after every 600 rounds and cleaned after every 1,200 rounds until 6,000 rounds had been fired.³

Whether or not this test regimen adequately represented the actual conditions in which the weapon would be used can be argued. However, possibly more important in view of the detail with which PTI compared the composition of its test dust product to that of the earlier AC Spark Plug dust is a comment in an October 2004 Desert Research Institute report:
“…current chamber test methodology misrepresents real-world conditions. The character of the soils and dust collected from areas of military activity in Iraq is greatly different from the material used in current weapons testing procedures. Current procedures employ laboratory-generated dust that is 99.7% silicon dioxide (i.e., quartz), contains no salt or reactive chemicals, and contains coarser particle sizes than most of the Iraq samples. Use of this material cannot simulate conditions in Iraq that have contributed to weapons failures.” ⁴

Not only is particle size and size distribution important, but so too might be the chemical composition. It’s not clear from reference 3 whether any changes in the test dust composition were made following the 2004 report. For the type of dust used in the 2007 test, the results show that some brands of carbine jammed less often than others. However, unless the test dust represented the actual service environment, this result might be misleading.

Salt Spray

Salt spray or salt fog testing, like dust testing, seems deceptively simple. The objective is to determine a material’s susceptibility to corrosion. But, what if the DUT is a piece of heavy industrial equipment rather than a railing at a seaside resort? Sulfur dioxide (SO₂) often is produced when coal or petroleum is burned, so the SO₂ injected salt fog test might better represent this environment than a plain salt fog test.

Corrosion takes many forms, and salt-spray testing isn’t appropriate for all of them. For example, some grades of stainless steel are subject to stress-corrosion cracking at elevated temperatures. Cracking only at lower temperatures can be studied using the salt spray test ASTM B-117 Standard Practice for Operating Salt Spray (Fog) Testing Apparatus at 95°F. B-117 also can simulate road salt environments to investigate corrosion to vehicles.

For many years, trucks have used galvanized steel panels to reduce corrosion, but a large number of articles question the correlation of B-117 testing with actual vehicle experience. Delving into this aspect of salt spray testing helps to explain the importance of choosing the most relevant test procedure.

The salt spray attacks bare zinc. This is expected, but not to the degree many tests produced. In normal use, trucks are exposed to salt spray but at some point dry out. When the salt spray testing is applied cyclically rather than continuously, a protective zinc carbonate layer that greatly reduces subsequent corrosion is allowed to form on the surface. Continuous spraying prevents this layer from forming.⁵

ASTM G-85 Standard Practice for Modified Salt Spray (Fog) Testing specifies one hour wet-one hour dry cycling to simulate actual environmental conditions. This test further adds ammonium sulfate (NH₄)₂SO₄ to the basic sodium chloride (NaCl) and is run at 95°F for at least 16 hours.

This test is more representative of actual use conditions, but Dr. Galv suggests that none of the accelerated corrosion tests available in 1997 were adequate. He recommended comparing a new part to existing parts in the field that have been used under the same or similar conditions.⁵

Other sources claim that the 5% salt concentration is much higher than found in real-world use. Also, salts other than NaCl are commonly encountered. The result is that salt spray test methods may produce different corrosion mechanisms and corrosion products than actual outdoor testing.

The SAE J2334 Laboratory Cyclic Corrosion Test specification may help to resolve these problems. This was developed over a period of several years during which the real-world corrosion performance of a range of materials was evaluated.

As shown in Figure 2, the test involves separate humid, salt spray, and dry periods within a 24-hour cycle. Also, a much lower concentration of salt mixed with calcium chloride and sodium bicarbonate is used rather than a 5% salt spray. Analysis of test results confirmed that 80 cycles produced corrosion corresponding to five years of outdoor vehicle testing.⁶

The General Motors North America GM9540P cycle B test uses a similar mix of chemicals. A 0.9% salt solution is used rather than 0.5%, the 24-hour cycle is split evenly into three eight-hour periods, and the dry stage is at less than 30% humidity rather than 50%. A paper from the Swedish Corrosion Institute reported on magnesium panel corrosion accelerated by 40 days of GM9540P testing.⁷

If you need to prove compliance to a specification that calls for the use of B-117, you have little choice but to run that test. However, when good correlation between test results and performance in the actual environment is important, an appropriate targeted corrosion test is likely to produce more relevant results than B-117.

Elite Electronic Engineering provides a wide range of environmental tests beyond the usual temperature and vibration, including dust, salt fog, rain, corrosion, and power wash. Steve Laya, sales and marketing manager, said that in addition to having the capability to run more common Ingress Protection, National Electrical Manufacturers Association, and MIL-STD water exposure tests, Elite had custom built the water spray apparatus required to perform the JIS DO203-S2 specifications. In this test, the UUT completes a 360° rotation at 17 rpm while the nozzle wands articulate in a counter motion spraying 40 l/min.

Similarly, Sypris Test and Measurement’s Michael Loyd, an application engineer, reported that the company has built a custom 8′ x 12′ cabinet with the capability to perform both salt and SO₂ spray (fog) testing. Mr. Loyd said that the usual range of test types was being expanded by customers seeking to supply the same equipment for both DoD and commercial applications.

And, CSZ Testing Services Operations Manager William Jackson discussed the combined temperature/humidity testing his lab’s been doing, “Recently, we performed the damp heat exposure and humidity freeze cycle tests called out in IEC 61215, 61646, and 62108 and ASTM E1171 for solar panels and related components. These tests require chambers capable of controlling both temperature and humidity for parts up to 4′ x 2′ in size.”

Summary

A comment from Gary Delserro, president of Delserro Engineering Solutions, helps to relate testing to reliability: “Beyond the normal specification testing we perform, a growing commercial trend is to help customers develop a plan to test the reliability of a product. We help them design a test plan with quantifiable numbers to estimate a field life.”

Herman Held at Quest Engineering Solutions described noise testing that his company performs, also with an emphasis on reliability, “The combination of audio testing, airflow testing, and thermal imaging with more standard temperature testing gives us the capability to help companies design their products to simultaneously optimize product reliability and minimize audio noise.”

And, Trace Labs’ Mr. Finch contributed his experience to the discussion of test relevance and product reliability: “Recent trends in environmental testing attempt to incorporate all product qualification requirements into a single test. The engineering community is looking for the quick and dirty silver bullet of testing as evidenced by the increased popularity of HALT and kurtosis vibration.

“Modern testing programs utilize HALT, vibration, and temperature testing to locate gross design flaws. However, no single test can account for all possible failure modes,” he continued. “It is imperative to conduct product-specific environmental simulations, which ensure that failures are found due to real-world conditions. The type and severity of the tests depend on the product. When developing a test plan, we consider the product cost and volume, consumer expectations, competition, and potential liability.”

References

1. Rosser, J. D., “Gold Brazing on the Concorde Airframe Hydraulic System,” Gold Bulletin, September 1976, pp. 119-122.
2. “The History of Test Dust,” Powder Technology, http://www.powdertechnologyinc.com/history-of-test-dust/history-of-test-dust.php
3. “The USA’s M4 Carbine Controversy,” Defense Industry Daily, Feb. 2, 2009, http://www.defenseindustrydaily.com/the-usas-m4-carbine-controversy-03289/
4. McDonald, E. and Caldwell, T., “Geochemical and Physical Char-
   acteristics of Iraqi Dust and Soil Samples,” Desert Research Institute Division of Earth and Ecosystem Sciences and Center for Arid Lands Environmental Management, 2004, http://www.defenseindustrydaily.com/files/Iraq_Dust_2004-10_DRI_Report.pdf
5. “Ask Dr. Galv,” January/February 1997, http://www.galvanizeit.org/images/uploads/drGalv/saltspray2.pdf
6. “Is the Salt Spray Test Any Good?,” http://www.corrosion-doctors.org/TestingBasics/B117-Bogus-1.htm
7. Jönsson, M., “Manufacturing of Magnesium Components,”
   http://www.gjuteriforeningen.se/vamp32/Corrosion_test.pdf

August 2009