Introduction to Accelerated Testing Types

[Please note that the following article — while it has been updated from our newsletter archives — may not reflect the latest software interface and plot graphics, but the original methodology and analysis steps remain applicable.]

In typical life data analysis, the practitioner analyzes life data (times-to-failure or times-to-suspension) for a sample of units operating under normal conditions in order to quantify the life characteristics of the product and make predictions about all of the units in the population. For a variety of reasons, manufacturers may wish to obtain reliability results more quickly than they can when the data comes from products operating under normal conditions. Instead, they may use accelerated life tests to capture life data for the product under accelerated stress conditions, which cause the units to fail more quickly. Performed correctly, accelerated testing can significantly reduce test times, resulting in reduced time to market, lower product development costs and lower warranty costs, as well as other benefits.

There are a variety of types of accelerated testing approaches and the accelerated testing strategy must be carefully designed to fit the product under consideration. This article presents a brief introduction to accelerated testing types and a discussion of some test design issues that impact the analysis of the data obtained from these tests. ReliaSoft ALTA PRO software provides a complete array of analysis tools for data from quantitative accelerated life tests.

Qualitative Accelerated Tests
Accelerated testing methods can be either qualitative or quantitative. Qualitative accelerated tests (such as HALT, HAST, “torture tests” or “shake and bake” tests) are used primarily to reveal probable failure modes for the product so that engineers can improve the product design. These tests are performed on small samples with the test units subjected to a single severe level of stress (e.g., stress cycling, cold to hot, etc.). If the specimen survives, it passes the test. Otherwise, appropriate actions will be taken to improve the product’s design in order to eliminate the cause(s) of failure that were identified during the test. A good qualitative accelerated test quickly reveals the failure modes that will occur under normal use conditions but does not introduce failure modes that will never be encountered in real life situations. These tests can provide valuable information about the types of stresses and the stress levels that should be applied in subsequent quantitative accelerated life testing. However, in general, they do not provide information that can be used to quantify the life characteristics of the product under normal use conditions.

Quantitative Accelerated Life Tests

Quantitative accelerated life tests (QALT), on the other hand, are designed to quantify the life of the product and to produce the data required for accelerated life data analysis. This type of test involves the controlled application of accelerated stress conditions in order to stimulate product failure and provide life data more quickly. The life data obtained from these tests can be used to estimate a probability density function (pdf) for the product under normal use conditions and to calculate reliability, probability of failure, mean life, failure rate, B(10) life and other important reliability metrics for the product.

QALT tests can employ usage rate acceleration or overstress acceleration to speed up the times-to-failure for the products under test. With usage rate acceleration, which is appropriate for products that do not operate continuously under normal conditions, the analyst operates the products under test at a greater rate than normal to simulate longer periods of operation under normal conditions. For example, if an appliance manufacturer assumes that the average washing machine will be used about six hours per week, then the manufacturer can test a sample of products continuously to reduce the test time by a factor of 28. With this testing approach, one week of continuous testing can be used to simulate 28 weeks of operation under normal use conditions! The data from this type of test can be analyzed with standard life data analysis techniques.

However, this testing method is not effective for products that have a very high or continuous usage rate under normal conditions. Some electronic devices, for example, are expected to operate continuously under normal use conditions. Usage rate acceleration is not an option for this type of product and a different type of accelerated life test must be used in order to obtain data for these products quickly. In these cases, overstress acceleration is used instead. With overstress acceleration, one or more environmental factors that are known to cause the product to fail under normal conditions (such as temperature, voltage, humidity, etc.) are increased in order to stimulate the product to fail more quickly during the test. The stress types and stress levels used in overstress acceleration tests must be carefully chosen so that they accelerate the failure modes for the product but do not introduce failure modes that would never occur under normal use conditions. Normally, these stress levels will fall outside the product specification limits but inside the design limits. The data sets from this type of test require special accelerated life data analysis techniques, which include a mathematical model to “translate” the overstress probability density functions to normal use conditions.

Stress Types and Stress Levels

In an effective overstress acceleration life test, the analyst chooses one or more stress types that cause the product to fail under normal use conditions. This can include temperature, voltage, humidity, vibration or any other stress type that directly affects the life of the product under normal use conditions. The stress(es) are then applied at various accelerated levels and the times-to-failure and times-to-suspension for the units under accelerated test conditions are measured. For example, if a product normally operates at 290 Kelvin and high temperatures cause the product to fail more quickly, then the accelerated life test for the product may involve testing the product at 310K, 320K and 330K in order to stimulate the units under test to fail more quickly. In this example, the stress type is temperature and the accelerated stress levels are 310K, 320K and 330K. The use stress level is 290K.

The application of the stress (under test conditions and/or during normal use) can be constant (time-independent) or time-dependent. When the stress is constant, the stress level applied to a sample of units does not vary with time. Each unit is tested under the same accelerated temperature for the duration of the test. For example, ten units are tested at 310K for 100 hours, ten different units are tested at 320K for 100 hours and ten different units are tested at 330K for 100 hours.

When the stress is time-dependent, the stress applied to a sample of units varies with time. Time-dependent stresses can be applied in a variety of ways. For example, if temperature is the stress type, each unit may be tested at 310K for 10 hours then increased to 320K for 10 hours then increased to 330K for 10 hours over the duration of the test. Alternatively, the units may be placed in a test chamber where the temperature starts at 310K and increases by five degrees every ten minutes until the chamber reaches 330K. Some common types of time-dependent stress profiles include step-stress, ramp-stress and various profiles in which the application of the stress is a continuous function of time. Figure 1 and Figure 2 display two examples of the many time-dependent stress profiles that can be used in an accelerated life test design.

Analyzing Data From Accelerated Life Tests

Using the life data obtained at each accelerated stress level, standard life data analysis techniques can be used to estimate the parameters for the life distribution (e.g. Weibull, exponential or lognormal) that best fits the data at each stress level. This results in an overstress probability density function (pdf) for each accelerated stress level. Another mathematical model, the life-stress relationship, is then required to estimate the probability density function (pdf) at the normal use stress level based on the characteristics of the pdfs at each accelerated stress level. The plot in Figure 3 demonstrates the relationship between life and stress for a particular product.

When analyzing accelerated life data, the analyst must select a life-stress relationship that fits the data for the particular product. Available life-stress relationships include the Arrhenius, Eyring and inverse power law models. These models are designed to analyze data with one stress type (e.g. temperature, humidity or voltage). The temperature-humidity and temperature-nonthermal relationships are combination models that can be used to analyze data with two stress types (e.g., temperature and humidity or temperature and voltage). The general log-linear and proportional hazards models can be used to analyze product life data where up to eight stress types need to be considered. Finally, the cumulative damage (or cumulative exposure) model has been developed to analyze data where the application of the stress (either at the accelerated stress levels or at the use stress level) varies with time. The formulation of these life-stress relationship models for use in a software package was the result of extensive research and development. ReliaSoft ALTA PRO is the only commercially available software package capable of analyzing data with time-dependent stresses.