By Eugene le Roux, FSAIRAC, and Eamonn Ryan
How do you prove that a vehicle will reliably last 150 000km without driving test vehicles 150 000km?
There is an important qualification: accelerated testing does not simply mean testing harder.
Senivpetro | Magnific.com
That is one of the practical challenges facing reliability engineers. Long service-life requirements are essential, but simply reproducing an entire product’s life under normal operating conditions can take too long and cost too much.
This is where accelerated testing comes into its own. The principle is relatively straightforward: identify the stresses responsible for failure and reproduce them in a controlled test environment at an accelerated rate.
Fatigue provides a good example. A component may experience relatively low forces for most of its operating life, with occasional high-force events. If engineers can measure those forces and establish how frequently they occur, they can develop a representative load spectrum.
The component can then be subjected to those loads on a fatigue-testing machine, concentrating the test on the stresses that make the greatest contribution to fatigue damage. The result can be a much shorter test programme than simply operating the complete vehicle until it reaches its required mileage.
But there is an important qualification: accelerated testing does not simply mean testing harder. If the test conditions are made excessively severe, the component may fail through a mechanism that would not occur during normal operation. Such a test could produce a failure, but tell engineers very little about the product’s actual service life.
The acceleration therefore needs to be based on an understanding of the relationship between stress, cycles and damage. It also explains why simply dividing a required service life between several test vehicles does not prove that service life. Twenty vehicles travelling 7 500km each have collectively covered 150 000km, but none has demonstrated that it can survive 150 000km.
The same thinking can be applied beyond mechanical fatigue. Corrosion, ultraviolet exposure, temperature ageing, humidity and thermal cycling can potentially be accelerated in controlled laboratory conditions. The objective is again to reproduce the relevant degradation mechanism rather than simply expose the product to an unrealistically harsh environment.
This requires engineers to understand how the product will actually be used, which stresses it will encounter and which failure mechanisms are most likely to limit its life.
There is consequently no shortcut around good engineering. Accelerated testing still requires representative samples, suitable test equipment, instrumentation, analysis and a properly defined acceptance criterion. It is not a way of avoiding reliability testing; it is a way of making long-term reliability testing practical.
And there is a strong commercial argument for doing it properly. A failed qualification test may delay production or require a design change. A field failure, however, can result in warranty claims, recalls, reputational damage and potentially serious safety consequences.
For organisations under pressure to protect cash flow and meet production schedules, reliability testing can look like an avoidable cost. In reality, the cost of proving reliability is usually easier to control than the cost of discovering unreliability after production has begun.
Accelerated testing does not eliminate the need for endurance testing. It makes it possible to demonstrate long-term performance within a realistic development timeframe.
