Page 198 - Failure Analysis Case Studies II

183

The failure cracks started from cold slugs or voids, because they occurred near the base of an
adjacent buttress, which possessed a sharp corner with the sidewall, and represented a stress
raising factor of about 4.2. The cold slugs and/or voids were more serious in their stress
magnifying effect, possibly having a stress raiser of about 6, so the effective stress concentration
in this zone was about 25 times the nominal applied stress of only I .55 MN mP2.
3. The final factor which made the situation critical was the presence of a substantial level of
frozen-in strain, produced by cold moulding the material. This effectively added some 20 MN
m-’ extra stress to the surface of the cold slug, producing a total stress in the region of about
59 MN m-’, a stress comparable with the mean strength of the material. Brittle cracks were
thus initiated at the surface of cold slugs or voids when the system was first pressurised, and
grew progressively at each use of the cooling system. The cracks had penetrated through to the
interior of the radiator, but only small quantities of water leaked out under pressure, judging
by the several traces of contaminant found near the crack. When the cracks reached a critical
size, they propagated catastrophically, releasing pressure from the system, and hence resulted
in loss of the cooling facility.
4. The failed specimen appears to be a maverick which was probably accidentally included in the
batch sent to the car manufacturers. Normal QC procedures usually prevent such mouldings
being used, but when this batch of tanks was moulded, they failed to catch the rogue product.
Careful visual inspection at the moulding machine would probably have caught the rogue,
provided the operator was aware of what to look for in terms of defects such as weld lines and
cold slugs. It is a difficult product to examine quickly for such defects owing to its black
colouration, which were only revealed by dusting with whiting and by very close visual inspec-
tion.

Acknowledgements

Naomi Williams (OU) for SEM, Jim Moffatt (OU) for mechanical testing, Gordon Imlach (OU)
for etching experiments, and the manufacturer for permission to publish this edited account of a
more substantial report.

References

[I] Lewis PR (Course Chair), Design and manufacture with polymers, T838. Post-graduate OU Course in the Manu-
facturing Programme, 1998, Block 5.2, section 4:47.
[2] Lee SM, editor. International encyclopaedia of composites. New York: VCH, 1990: for entry on Processing, void
formation in melt flow thermoplastic composites, p. 302.
[3] International encyclopaedia of composites. Op cit., for entry on Characterisation, a general review covering X-
radiography, OM and SEM.
[4] Folkes MJ, Russell DAM. Orientation effects during the flow of short-fibre reinforced thermoplastics. Polymer
1980;21: 1252-1258.
[5] Peterson RE. Stress concentration factors. John Wiley & Sons, Inc., 1974, Fig 39; also in Pilkey WD, Peterson’s
stress concentration factors 2nd ed. John Wiley & Sons, 1997, Chart 2.29.
[6] Peterson, op cit., Figure 150; Pilkey, op cit., Chart 4.71.