Fatigue Failures in Industry - Case Studies: Ž. Domazet and T. Piršic
Fatigue Failures in Industry - Case Studies: Ž. Domazet and T. Piršic
Fatigue Failures in Industry - Case Studies: Ž. Domazet and T. Piršic
FATIGUE
STUDIES
FAILURES
IN
INDUSTRY
CASE
1. Introduction
In spite of numerous and expensive researches in the field of fatigue, cracks and failures caused by
fatigue occur every day in all fields of human activity. The paper presents some typical fatigue
damages in industry and transport. Fatigue failure of the main engine lateral support (at bulk carrier),
fatigue cracks at large portal crane, and the fatigue cracks and failures in large gear wheel of cement
mill are described.
ANALYSIS TECHNOLOGIES
1153
Also it was not known whether crack started from the middle of the beam and spread towards edges or
vice versa. After cracking one of the supports, loads were probably redistributed, each time in a
different manner. According to the description of cracks and failures, it was obviously that fatigue of
material took a place again, and its causes should be detected by detailed stress analysis.
2.2 Stress analysis
Due to complicated shape of crack area stress analysis was preformed by means of strain gages. Strain
gages were installed on all four beams (figure 2), in order to obtain operational loads (axial forces and
bending moments).
1154
ANALYSIS TECHNOLOGIES
ANALYSIS TECHNOLOGIES
1155
it is interesting to notice that the stresses were mostly caused by the manoeuvres of the crane
stresses caused by the loads were not significant.
1156
ANALYSIS TECHNOLOGIES
S1
FEM
Not
measured
100
Measured
FEM
100
100
S2
S3
S4
S5
S6
50
150
50-60
300
70
60-80
50-70
100
ANALYSIS TECHNOLOGIES
1157
to determine the stress distribution at the gear rim. Calculated maximum stress amplitudes have been
found at the tooth fillet, approximately 50 - 115 MPa, depending on their positions at the surface.
Stress intensity decreases very fast in the depth of the gear rim. These stress values could not be the
only reason for cracks initiation and propagation. In spite of great number of cracks and one complete
failure of gear ring, repair welding was performed. All necessary steps for the best quality insurance
(best welders, best welding rods, pre-heating, very slow cooling conditions, NDT inspection following
every layer, hammering of all layers, etc.) have been respected and documented. All described
activities took two months and cost approx. $50 000, instead of $300 000 for new gear ring and four
months for its delivery and montage. Three years after repair and frequent controls during the service,
no further cracks have been reported.
5. Conclusion
The case studies presented in this paper illustrate the circumstances of inappropriate design, from
fatigue point of view. It is obvious that in the case of variable loads, special attention should be paid to
fatigue crack avoidance and fatigue crack repairs as well. A lucky circumstance with many fatigue
failures is a relatively long crack propagation period from its origin to the final failure and crack can
be discovered easily. What to do with discovered fatigue cracks is a well known question in such
situations. The usual answer is one of the following actions:
instantaneous unloading of the entire system and replacing the cracked component
reducing the external loads and continuing careful crack growth control, and
retarding, stopping or even eliminating the crack (crack repair) in a very short time.
As the complete replacement can be time consuming and expensive, and reduction of service loads
with existing fatigue crack is very dangerous and mostly unacceptable, fatigue crack repairs seem to
be best solution. The necessary steps for a successful repair of fatigue cracked component should be
[Domazet 1996]:
a) Damage analysis: the first step with any damage and its possible repair should be damage analysis.
It should give answers to some important questions, such as: what is the reason for the fatigue
crack, how dangerous is the existing fatigue crack, how long is the remaining component life, etc.
b) Damage repair: the most frequent fatigue crack repair methods are: repair welding, metal
reinforcements, CFRP patches, arrest holes, etc. The final choice of adequate method and its
parameters depends on all data obtained by damage analysis and knowledge of repair methods.
The role of experience and case studies from literature should not be avoided either.
c) Reliability of repaired component: reliability of repaired components estimation of components
remaining life in accordance with new stress distribution and possible improvements of fatigue
strength. For this reason, the S-N curve of base component should be known. In this stage control
interval and control type (some of non-destructive testing methods) should be defined.
d) Documentation: correct and complete documentation of all undertaken activities, as well as quality
insurance, represents evidence of good work and valuable source for future repair jobs.
References
Domazet, ., Engineering Failure Analysis, Vol. 3, No. 2, 1996., pp. 137-147.
Domazet, ., Lozina, Z., Piric, T., Miina, N., ukic, P., FESB Report No. 01/99, Split, 1999.
Domazet, ., Barl, J., ukic, P., FESB Report No. 14/01, Split, 2001.
Prof. dr. eljko Domazet
Faculty of Electrical Engineering, Mechanical Engineering
and Naval Architecture, University of Split
Mechanical Engineering Department
R. Bokovica bb, 21000 Split, Croatia
Telephone (021) 305-770, Telefax (021) 463-877
Email: [email protected]
1158
ANALYSIS TECHNOLOGIES