In this two
part blog series we will shade some light into a very interesting and important
topic: plastic part failure.
Parallel to
this two part blog post series I made a presentation which can
be watched here on YouTube:
The field of
plastic part failure analysis is wide and we will focus first on the “why
plastic parts fail”, together with showing the main reasons. After this we
focus on the antidote – what can we do to prevent failure?
Why
Plastic Parts Fail
In
general, product failure is a costly business. There may be several consequences of part failure such as product liability which can result in significant settlements and penalties
[3]. For example a manufacturer may be held liable if the product is defective. Furthermore, if the product is manufactured in a
defective way and proper testing as well as inspection was not conducted, then
product liability may be enforced too. There are several more reasons (missing
of adequate labeling; missing instructions and warnings of the product).
“Nobody
Wants to Air Their Dirty Laundry in Public”
In
the past it was difficult to guide designers on plastic part failure. Plastic
part failure was kept secret, since nobody wanted to air their dirty laundry in
the public.
However,
there was a study published by David Wright [1] which classified the causes of
failure for over 5,000 failed plastics parts. One of the amazing key findings was
that the vast majority of failures were avoidable.
What
was the big problem? The know-how on how to prevent failure was publicly
known , however it was inadequately communicated along the plastic part manufacturing
chain. This chain usually consists of specifiers, designers, processing
companies, purchasing, and material suppliers. Designers might be aware of
certain material differences and their impact on the overall part performance.
Contrary, material purchasers might choose a cheaper material without knowing
the aforementioned impact on the part performance.
Causes
of Failure
Mr. Wright
shows in his study two viewpoints on the causes of failure [1]:
1. Phenomenological
causes of failure: in this viewpoint failures are attributed to a physical
mechanism (Figure 1).
2. Human
viewpoint: in this viewpoint failures are attributed to human related decision
making and execution (Figure 2).
 |
| Figure 1: Overview phenomenological causes of failure [1]. |
 |
| Figure 2: Overview causes of failure from the human viewpoint [1]. |
Figure 1 shows that environmental stress cracking (ESC) is the biggest cause of failure in plastic parts (30%), followed by static notch fracture (20%), and dynamic fatigue (19%).
Interesting to see are the human caused failures in Figure 2. Here, material misselection and poor specification are with 45% by far the biggest reason for plastic part failure. The other reasons are fairly equally distributed.
There are several known cases where misselection and poor
specification lead to catastrophes.
One of them was the space shuttle Challenger disaster from
1986 (Figure 3) [2].
 |
| Figure 3: Space Shuttle Challenger Disaster [2]. |
The space shuttle broke apart 73 seconds into its flight.
All seven crew members were killed. The so-called “Roger Commission” was
initiated, where Dr. Richard Feynman was part of the investigation. They found
that the accident was caused by a failure of the O-ring sealing joint on the
right solid rocket booster. The selected O-rings showed less resilience at
10°C. On the flight day it had 2°C and the seals were never tested at 10°C and
below temperatures. Dr. Feynman presented the low resilience by putting the
O-ring in ice water. He took it out and stretched them. The rings did not
return to their original position.
This example highlights that material specification, selection, and testing are crucial points of having a proper function plastic part.
Polymer material selection as the antidote of plastic part failure will be discussed in the second part of this blog series: Plastic Part Failure - Part 2: The Antidote
Thank you for reading and #findoutaboutplasticsGreetings,
Herwig Juster
No comments:
Post a Comment