Industrial Consulting Work : Failure Analysis and Mechanical Property Testing Services
The School of Engineering at the University of Plymouth is well equipped to perform in-depth failure analysis and to determine mechanical property data. Very significant expertise exists in this area, particularly in the area of fracture, fatigue crack growth and fractography (the analysis of fracture surface features to identify mechanisms and causes of failure). Professor Neil James has some 18 years experience in this field, has published some 60 papers on various aspects of fatigue, fracture and failure analysis, and has performed more than 110 failure analyses for industry. These investigations were often undertaken for insurance purposes, or as part of a litigation process. Reports are comprehensive and give recommendations to avoid future similar failures wherever possible. He has published a number papers dealing with failure analysis, some of which are listed below:
M
N James (1995), Some potential
pitfalls in failure analysis, International Journal of Fatigue, 17
pp.457-462 ISSN 0142-1123.
http://dx.doi.org/10.1016/0142-1123(95)00039-V
M N James (2002), Crashing aircraft, sinking ships - fractographic and SEM support for unusual failure hypotheses, Engineering Failure Analysis 9 No. 3 pp.313-328 ISSN 1350-6307. http://dx.doi.org/10.1016/S1350-6307(01)00016-4
M
N James (2005), Design, manufacture
and materials; their interaction and role in engineering failures,
Engineering Failure Analysis, 12 No. 5 pp.662-678 ISSN 1350-6307.
http://dx.doi.org/doi:10.1016/j.engfailanal.2004.12.012
M N James (2008), Designing against LMAC in Galvanised Steel Structures, Engineering Failure Analysis, 16 pp.1051-1061 ISSN 1350-6307. http://dx.doi.org/10.1016/j.engfailanal.2008.05.019.
Work has been commissioned by such clients as:
Mining and Industrial
Bucyrus Africa (Pty) Ltd; Anglo Platinum Waterval; Pioneer Ready Mixed Concrete, South Africa; Hartebeestfontein Gold Mine, South Africa; Slagment (Pty) Ltd, South Africa; Nestle (South Africa) Pty Ltd; Lloyd Aviation (Pty) Ltd, South Africa; Blue Circle Cement (Pty) Ltd, South Africa; South African Breweries; Rossing Uranium Mine, Namibia; Anglo-Alpha Cement, South Africa; AECI, South Africa; Gencor Ltd, South Africa; Fraser Alexander Bulk Materials Handling (Pty) Ltd, South Africa; Warman Africa (Pty) Ltd; Imerys, Cornwall; Schlumberger Technology Centre, Gloucestershire.
Manufacturing
Gyrus Medical Ltd, South Glamorgan; Galvanizers Association, West Midlands; Eaton Aerospace Ltd, Hampshire; Invensys Controls UK Ltd; ReedHycalog, Gloucestershire.
Lawyers, Loss Adjusters and Insurance Companies
Cubberley & Associates, South Africa; VR Salvage & Associates, South Africa; Theunis Joubert Assessors, South Africa; IGI Insurance Company Ltd, South Africa; Webber Wentzel Bowens, South Africa; Deneys Reitz Attorneys, South Africa; Bell, Dewar Incorporated, South Africa; Foot & Bowden Solicitors, Plymouth; Wolferstans Solicitors, Plymouth; Scott Bailey, Solicitors and Mediators, Hampshire.
Examples of Failure Analyses
The principles of failure analysis are the same irrespective of the scale of the cracked structure or component, and a successful investigation must answer general questions related to the service conditions, material parameters, stress flow and operation history. A multidisciplinary background is therefore useful. Several illustrative examples are given of the type of information which may be obtained during failure analysis.
A walking dragline removes the overburden during open cast mining of coal. A typical large dragline is illustrated below. Walking is accomplished by cam motion and the pontoons seen in the photograph. Such large welded structures readily initiate fatigue cracks and, as downtime is extremely expensive, reasons for any cracking must be quickly identified and remedial steps put in place with the minimum of disruption to mine operation. The drag bucket is filled by a scraping action and is pulled towards the dragline by chain links. These links are integrally cast into chain using high strength, abrasion resistance steel which is suitable for ground engagement operations. Occasionally, such links may suffer fracture during service and the question arises as to culpability for replacement cost and loss of production. To assign culpability requires a complete investigation of chain metallurgy, fracture surface appearance and casting practice. An unusual case of failure is shown below. This shows the existence of internal brittle 'bursts' in the as-cast chain links with a cross-sectional size of about 120 mm by 120 mm. Typically, such defects occur under poorly controlled casting conditions, from hydrogen embrittlement, or as a result of large forging reductions and hot shortness in the metal. These internal defects quickly extended in service via fatigue crack growth and led to failure. In this case, toughness of the cast material had been influenced by relatively high levels of sulphur, phosphorous, aluminium and nitrogen. This led to problems with interdendritic and intergranular separation, due to embrittlement during heat treatment of the links. Control of melt and casting practice had not met the required standards, and the foundry accepted responsibility.
Stress Corrosion Cracking of a Grade 304 Stainless Steel Pipe
Cracking occurred in an austenitic Grade 304 stainless steel pipe transporting glucose solution. The pipe was seam welded, with a thickness of 2 mm and a diameter of about 120 mm. The pipe was lagged and one of the questions to be answered was whether the cracks initiated from the inner or outer surfaces. The figure below shows the inner surface appearance of two typical cracks. The cracks straddle the weld seam, and branching is much more extensive on the inner surface, than on the outer. This indicates that the problem is stress corrosion cracking (SCC), driven by a static residual stress at the weld and a suitable environment inside the pipe. Grade 304 suffers from stress corrosion cracking in the presence of trace chlorides, particularly at elevated temperatures around 60EC (the circulation temperature of the glucose). Investigation of the fracture surface in a scanning electron microscope showed the feathery cleavage patterns often associated with stress corrosion cracking in steels. Recommendations to avoid the problem, included checking the chloride level of the glucose and ascertaining whether this could be controlled below the threshold level for SCC. If this was not feasible, a recommendation was made to change to a ferritic stainless steel, such as type 430. In this case, however, care would have to be taken during welding to avoid so-called weld decay problems.
A number of
failures of zinc plated carburettor bolts were experienced by a local manufacturer.
failure was via longitudinal cracking, as shown in the figure below. These cracks extended
through the bolt into the fuel hole up the centre, leading to spillage in service. These
bolts were required to be made from En8M steel. This is a free cutting steel containing
dispersed lead particles, with the M grades manufactured to close limits on chemical
composition and mechanical properties. The bolts were, in fact, supplied as grade A steel,
indicating no check of mechanical properties. The bolts were manufactured from steel
as-received and then zinc plated. Several of the bolts were cracked open in the laboratory
and examined using low power stereo microscopy. This showed that the outer parts of the
fracture surface in the bolt head region were zinc plated, indicating that the cracks
existed prior to the plating process (see the figure below).
The
bolts were transferred to the scanning electron microscope to inspect the mechanisms of
fracture at the interface between the unplated and plated parts of the fracture surface.
This revealed that intergranular cracking was present along parts of the interface. This
is a brittle cracking mechanism indicative of an embrittlement problem. Microstructural
investigation showed that the steel had a normalised ferrite-pearlite structure with
strong dendritic texture present. This implied that the cast steel had not been
homogenised during manufacture of the bar. This investigation showed that the steel bar as
supplied, contained a pre-existing seam, or surface defect running longitudinally along
it. however, although this is a serious defect, it was not immediately apparent why the
crack should then extend right through the cross-section of the bolt. Ferritic steels are
susceptible to liquid metal embrittlement by liquid zinc at around 400EC, with a sharp
crack-like defect (the seam) providing an excellent initiation site. The textured
structure would probably have a lower fracture toughness, than a correctly homogenised
structure. The embrittlement is shown by the presence of intergranular fracture. Equally,
during tightening of the bolt on assembly, the applied tensile load would cause further
fracture of the sharp crack. If the supplied material met the En8M specification and was
seam-free, this problem would not occur.
In addition, over the last few years a number of failure analyses have been performed on cases of LMAC in galvanised steel. structures.