Fitness-for-Use Testing
How do we ensure that engineered materials will meet all of the customer needs
including the ones that the customer is not aware of? What is the optimal balance
of price and performance? How do we quantify the value of our product in different
The most important word in the vocabulary of advertising is TEST. If
you pretest your product with consumers, and pretest your advertising,
you will do well in the marketplace.
David Ogilvy
One large American company has been a major brand name in home appli-
ances for decades. However, its market position was squeezed between pre-
mium product lines appealing to afuent consumers and low-cost imported
In the 1980s, the company updated its line of refrigerators. After study-
ing the options, it chose to implement turbine compressor technology.
Kitchen refrigerators commonly employed piston compressors to pres-
surize and circulate the refrigerant. This older technology was being
replaced by turbines in larger cooling applications, such as commercial
air conditioners. Turbines offer fundamental advantages in noise, vibra-
tion, maintenance, and cost of manufacturing, because they have fewer
moving parts. Unfortunately, the design team focused too intently on cost
Once the design was nalized, 600 turbine units were built and subjected
to an accelerated life-testing regime. This protocol was believed to simulate
5years of typical service and had been validated during an evaluation of
earlier models. The company was condent in the turbine approach because
every unit passed the 2-month torture test. The new product line was subse-
quently introduced with a 5-year warranty and started to appear in homes
across North America.
Within 8months, the company began to receive reports of premature com-
pressor failures. The rst units to fail were located in poorly ventilated
spaces. An investigation was launched to determine the root cause of a situ-
ation that was quickly becoming a crisis. Attention quickly focused on com-
pressor bearing failures. Critical wear components were produced with an
economical metal sintering process instead of the dependable and expensive
24 Process Techniques for Engineering High-Performance Materials
casting technique. Ironically, the powdered metal technology had previously
been rejected by the company’s air-conditioning division due to concerns
over wear problems.
The 2-month test cycle had rst hinted of the problem. The prototypes
exhibited the onset of lubricant breakdown and heat damage. Under the
test protocol, these distressing observations were not recorded as failures,
because the systems were functional throughout the prescribed evaluation
The testing regime could produce failures in piston-style compressors
that correlated well with the actual eld performance. However, the tur-
bine failure rate was nonlinear. They could easily survive the test period,
but would experience lubricant breakdown if the ambient conditions were
too warm. The low ow rate of the refrigerant in a home refrigerator was
insufcient to properly cool the small compressors. This was not an issue
on large equipment, where the bearings had superior lubrication and better
wear resistance.
The primary testing error was in not determining that the ambient tem-
perature was the controlling factor in the failure of the sintered bearing com-
ponents. The secondary mistake was in not running the trial for a longer
period, which would have revealed the inection point where the failure
rate increased. Rather, the units were tested only to an arbitrary time limit
dictated by previous experience with a different technology. This test regime
is similar to driving a car at night in heavy fog. No danger is visible, but you
have no means to judge the safety margin between a good performance and
a disaster. An appropriate tness-for-use (FFU) test program will reveal the
safe window of operation.
25Fitness-for-Use Testing
Performing an unsuitable suite of tness tests will create a false
sense of security, masking potential problems. Dramatic material
failures often have their origins in tests that seem to validate perfor-
mance, while overlooking serious aws. Test methods that are ideal
for one application may be unsuitable for another, despite apparent

Get Process Techniques for Engineering High-Performance Materials now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.