You are on page 1of 8

Engineering Failure Analysis 35 (2013) 286–293

Contents lists available at SciVerse ScienceDirect

Engineering Failure Analysis


journal homepage: www.elsevier.com/locate/engfailanal

Failure of a crankshaft of an aeroengine: A contribution


for an accident investigation
V. Infante a,⇑, J.M. Silva b, M.A.R. Silvestre c, R. Baptista d
a
ICEMS, Instituto Superior Técnico, TULisbon, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
b
AeroG/LAETA, University of Beira Interior, 6201-001 Covilhã, Portugal
c
CAST, University of Beira Interior, 6201-001 Covilhã, Portugal
d
ESTSetúbal (Instituto Politécnico de Setúbal), Campus do IPS, Estefanilha, 2941-508 Setúbal, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: The objective of this work is to determine the main cause of failure of a crankshaft from an
Available online 20 February 2013 ULM airplane aeroengine. The core of this paper is focused in the analysis of the damage
mechanisms which were in the base of a catastrophic failure of the crankpin journal. Based
Keywords: on a preliminary observation of the fracture surface, there are clear evidences of a fatigue
Airplane process as both beach and striation marks have been identified. The judicious characteriza-
Crankshaft tion of the failing mechanism, including the identification of the crack initiation site and
Failure
the assessment of the crack propagation rate, is of paramount importance to support the
Fatigue
Fractography analysis
investigation of this aircraft accident. In this context, an exhaustive observation of the frac-
ture surface by means of optical and electronic microscopic techniques, in parallel with
microstructural examinations, were carried out as determinant information for the correct
assessment of the contributive causes to this accident.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction

During the last years, a considerable effort has been addressed towards the application of methodological scientific
approaches regarding the investigation of aircraft accidents resulting from the failure of structural or mechanical compo-
nents. Distinct promoting factors can be in the base of this type of occurrences, like an inadequate design of the component,
an incorrect manufacturing procedure, any eventual intrinsic defect in the material or a misjudged operation procedure of
the aircraft (i.e., beyond the design limits). On the other hand, aircrafts typically have to withstand stringent operational
requirements which force aeronautical components to severe working conditions, such as high loading factors, cyclic and
time dependent damage mechanisms due to creep-fatigue, steep temperature gradients (including elevated temperature
environments), corrosion under stress, vibrations and others. As a consequence, deficiencies always arise which makes it
necessary to assess the significance of the defect with respect to its impact in terms of the airworthiness requirements,
and even on an economic perspective, related with the normal operation of aircrafts [1]. Moreover, a failure investigation
and subsequent analysis should determine the primary cause of failure aiming at determining adequate corrective actions
that will prevent similar failures.
In the particular case of the present paper, efforts have been made to investigate the contributory factors for the failure of
a crankpin journal of an ultralight airplane (ULM) engine. This crankpin belongs to a crankshaft of a four cylinder
reciprocating engine. The crankshaft is one of the most important parts in an engine since it bears large cyclic loads during
the working process. This component has typically a complex geometry, being responsible for the conversion of the

⇑ Corresponding author. Tel.: +351 218417643; fax: +351 218474045.


E-mail address: virginia@dem.ist.utl.pt (V. Infante).

1350-6307/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.engfailanal.2013.02.002
V. Infante et al. / Engineering Failure Analysis 35 (2013) 286–293 287

reciprocating linear displacement of the pistons to a rotary movement of the power shaft. The need for a smooth operation in
the base of the conversion between these two movements implies a very precise alignment of all components which, in turn,
must withstand severe cyclic loads as consequence of the gas combustion in the combustion chamber of the engine. These
strict operation conditions may promote distinct forms of failing mechanisms which are most of the times responsible for
the premature catastrophic failures of crankshafts. In particular, fatigue is the leading ruin process in the most part of failures
in crankshafts, which are typically caused by bending loads and/or torsional loads on the journals fillets [2,3]. On the other
hand, misalignments, even on a small extent, can produce very high stress concentration regions and therefore acting as trig-
gering factor for the onset of a fatigue process.
With increasing of rotational speed and power density, bending fatigue fracture becomes the main failure mode of crank-
shafts [4]. The crankshaft, in a normal situation, is subjected to torsion and bending loads. Generally the loads application
needs to be analyzed considering two main scenarios [5]:

Failure may occur at the position of maximum bending; this may be at the center of the crankshaft or at either end. In
such a condition the failure is due to bending and the pressure in the cylinder is maximal;
The crankshaft may fail due to torsion, so the crankshaft needs to be checked for torsion loads at the position of maximal
torsion. The pressure at this position is a combination of a fraction of maximal bending pressure and the maximal engine
torque.

In order for fatigue to occur, a cyclic tensile stress and a crack initiation site are necessary. The crankshafts run with har-
monic torsion combined with cyclic bending stress due to the loads of the cylinder pressure transmitted from the pistons and
connecting rods – to which inertia loads have to be added. Although crankshafts are generally designed with high safety
margins in order not to exceed the fatigue strength of the material, high cyclic loading and local stress concentration may
lead to the formation and growth of cracks even when the fatigue strength is not exceeded in terms of average values
[2]. One critical spot in crankshafts is the interface region between the crankpin journals and the webs, where stress concen-
tration issues are normally mitigated through a fillet with an adequate geometry.
For this purpose, a fillet rolling technology has been widely used for many years [4]. The radius itself reduces the stress in
these critical areas, but since the fillets in most cases are rolled, this also has a beneficial effect due to the induced compres-
sive residual stress in the surface, which prevents cracks from forming. It is difficult though to account for the exact stress
distribution in the fillet region as this depends on several factors.
The main objective of this work is to undertake the analysis of the damage mechanisms which promoted the catastrophic
failure of the aforementioned crankpin journal aiming at identifying the crack initiation site and characterizing the main fea-
tures of the fracture surface. This qualitative information is crucial for assessing the loading conditions that promoted the
fatigue process in order to support the conclusions in the report emanated by the air safety authority having as main premise
the avoidance of a similar hazardous situation in the future.

2. Case study

The failed crankshaft presented in this work belongs to a four stroke, four cylinder liquid air cooled engine of an ultralight
airplane that had an accident. The nominal power output of this engine is 73.5 kW at 5800 rpm. At the time of the accident,
the engine had a total of 996 h of reported service. According to the technical specifications provided by the manufacturer,
this engine’s TBO (Time Between Overhauls) is 2000 h. The failed component is depicted in Fig. 1 where it is possible to visu-
alize the fracture zone in the interface between the crankpin journal #2 and the counterweight. Each of the crankpin journals
has been press-fit into the web sections in the opposite end to the counterweight.
Fig. 2 shows each side of the failed crankpin journal. These images are clear about a perpendicular fracture surface pattern
regarding the longitudinal axis of the crankpin journal. The journal was found to have an extensive crack running from the

web main journal crankpin journal #2

lubricating holes counterweight

Fig. 1. Image of the failed crankshaft. Fracture has occurred in crankpin #2.
288 V. Infante et al. / Engineering Failure Analysis 35 (2013) 286–293

(a) (b)
Fig. 2. General view of both sides of the fractured crankshaft. (a) Side A; (b) Side B.

Fig. 3. Fracture surface: (a) General view of Side A; (b) Crack initiation zone in Side B (dash line indicates the crack initiation zone).

undercut fillet (Fig. 2a) near the lubrication hole. The fillet was machined by a conventional turning operation, with a 0.8 mm
nominal radius.
A close examination of the fracture surface revealed fatigue marks in the two separate components. In fact, the presence
of beach marks propagating in a flat plateau, as visible in Fig. 3a, is one of the main characteristics of a fatigue process, and in
this case these can be found in almost 2/3 of the fractured area, the remaining part having a rougher pattern consistent with
a sudden brittle fracture preceding the catastrophic failure of the component. Another interesting remark about the beach
marks is the absence of a clear spiralling movement during crack propagation which is indicative of a non-prevailing tor-
sional loading condition. This means that the crack propagated mainly due to a cyclic bending state condition as explained
in Section 1.
Fig. 3b (dash line zone) shows the crack initiation site which is situated in the interface of the crankpin with the coun-
terweight. As expected, this is a critical region in terms of stress concentrations even considering the relieving effect of the
undercut fillet. Moreover, the existence of an adjacent lubrication hole contributes to a steepest stress gradient in this zone
and a lesser bearing area. Another interesting observation in this image is the existence of very subtle ratchet marks in the
crack initiation site. Ratcheting is a common phenomenon in very high stress concentration regions, indicating that the fa-
tigue initiation phase was set off from a multiple plane crack formation which typically produces small tear steps or shear
walls [6].

3. Results and discussion

As mentioned before, the main objective of this work was to undertake an exhaustive analysis of the failed component
with a particular emphasis in the fracture surface of crankpin #2. Nevertheless, a general inspection of the crankshaft
was made in order to check out for the eventual existence of degrading marks in the main journals or any evidences of mis-
alignments, which was not the case.
Having stated the above, most observation efforts were directed towards a fully understanding about the fatigue process
in the origin of the failure of the component. In addition to the information in the previous section, it is important to try to
identify the root causes for the crack initiation stage. Fig. 4 is a magnified view of the crack initiation region presented in
Fig. 3a. Careful examination revealed the presence of some form of damage mark extending from the fillet region close to
the crack initiation site (circular dashed line in Fig. 4), which could be considered as a contributing factor to set off the fatigue
process. Yet, if this was the case, then a mirrored corresponding mark would have to be detected in the other side of the
fractured part (Fig. 2b), which in fact did not occur. This means that the probable cause in the origin of this defect mark
was the abrupt failure of the component during the engine’s operation yielding likely rubbing between the two sides during
the separation phase. Other possible explanation for the presence of this mark is a probable mishandling procedure at some
point throughout the investigation of the accident (note that the component was received disassembled from the engine).
V. Infante et al. / Engineering Failure Analysis 35 (2013) 286–293 289

Fig. 4. Magnified view of the crack initiation region.

Detail “d” (Figure 9)

Detail “e” (Figure 10)

Detail “c” (Figure 8)

Detail “b” (Figure 7) Detail “a” (Figure 6)

Fig. 5. Mapping of the different locations of the fracture surface submitted to SEM.

After discarding this hypothesis, efforts were focused towards the observation of the crack initiation site in the neighbor-
ing of the fillet region by means of a scanning electron microscope (SEM). Fig. 5 shows the different locations of the fracture
surface observed using SEM, and which will be discussed subsequently. In particular, detail ‘‘a’’ refers to the observed ratchet
mark (Fig. 6), whereas the undercut fillet corresponds to detail ‘‘b’’ (Fig. 7) and the fatigue striations to detail ‘‘c’’ (Fig. 8). The
transition between the crack propagation and final rupture zones are represented by detail ‘‘d’’ (Fig. 9). and secondary crack-
ing in detail ’’e’’ (Fig. 10).
The aforementioned ratchet mark was clearly identified using the SEM technique, as depicted in Fig. 6. As explained,
ratchet marks are typically related to the presence of high concentration features in the component, such as the undercut
fillet in the present case, and they follow from the confluence of the boundary between two adjacent crack initiation planes.
In this case, the localized ratcheting visible in the figure means that the crack origin was very close to this position. A close
examination of the surface condition of the undercut fillet (Fig. 7), which was made using a conventional turning operation,
allowed discarding any possibility of an inadequate finishing or any defect introduced during the machining process.
The information provided by the ratchet mark combined with the overload zone, which is roughly limited to 1/3 of the
total fracture surface (the remaining part being characterized by well defined beach marks) allows concluding that the fa-
tigue process was driven by high cyclic loading with moderate amplitude [7]. It is clear that the major contributing factor to
the onset of the fatigue process is the high stress concentration in the interface between the crankpin journal and the coun-
terweight, which is not conveniently mitigated (or at least minimized) by the existing undercut fillet in this region. More-
over, it should be stressed out that the stress gradient in the vicinity of the crack initiation site is upheld by the presence of a
lubrication hole, as visible in Fig. 4.
Despite the fact of the high nominal hardness of the steel 15CrNi6 alloy used in this component, striation marks have
been detected during SEM observations, as visible in Fig. 8. Notwithstanding the blurring of this image, which comes from
290 V. Infante et al. / Engineering Failure Analysis 35 (2013) 286–293

Fig. 6. Ratchet mark.

Fig. 7. Surface finish of the undercut fillet in the interface of the crankpin journal and the web.

Fig. 8. Fatigue striation marks.

the very high magnification (20,000), this information, more than the confirmation of the cyclic induced deformation pro-
cess, is crucial for the estimation of the crack propagation rates, which will be a topic of a forthcoming work.
Fig. 9 shows the transition between the crack propagation region and final overload zone. In this latter region, higher
roughness is indicative of a sudden failure caused by the insufficient extension of the load bearing section. Secondary
cracking is also typically prevalent in the final fracture zone,as shown in Fig. 10. In this case, transverse cracks are visible
V. Infante et al. / Engineering Failure Analysis 35 (2013) 286–293 291

Fig. 9. Transition between the fatigue crack propagation region and final overload zone.

Fig. 10. Secondary cracking.

perpendicular to the plane of the primary fracture, which are associated with crack propagation mechanism of the primary
fracture. In fact, transverse cracks occur when the material is unable to accommodate the strains induced by the plastic
deformation of the primary fracture. Fractures of this type are invariably associated with planes of weakness or brittleness
induced by inclusion stringers or grain boundaries of highly worked structures [8], and they have been reported in the con-
text of other types of fatigue failures where high nominal stress level conditions are present [9].
In addition to microscopic observations, hardness measurements were carried out on a polished surface of the failed com-
ponent using a Mitutoyo AVK-C2 Vickers hardness tester considering a load of 1 kg. The objective of these tests was to con-
firm the existence of a hardening surface treatment in the crack initiation region.
Fig. 11 shows the hardness test results (HV1) considering a straight measuring path extending from the surface towards
the center of the crankpin journal. As we can see, there is a notorious difference between the hardness values in the surface
and core material, respectively 880HV1 and 472HV1, which means that a case-hardening treatment was applied to this com-
ponent. Considering this, a metallographic analysis was undertaken considering the aforementioned cross section in the
etched condition (using a 4% NITAL solution), which is visible in Fig. 12. The presence of a white layer on the surface of
the section is consistent with that which is typically formed during a case-hardening process. The nominal case depth
was found to be around 0.2 mm, although the dimension of the surface layer was not perfectly uniform along all the perim-
eter of the section.
From the observations in the crack initiation region, an interesting surface feature was found in the vicinity of detail ‘‘a’’ of
Figs. 5 and 6. In fact, Fig. 13 shows an uncharacteristic defect which is similar to a cavity with an average dimension of
around 150 lm. According to the information provided by the manufacturer of the engine, the crankshaft is fabricated using
a conventional forging process. This type of fabrication process may imply some forms of defects within the material whose
topology is comparable to the characteristics of the cavity shown in the image. A possible explanation for the presence of this
defect can be attributed to the high stresses produced during the forging operations, which can result in the formation of
bursts or hot tears. In general, these type of defects are promoted due to the existence of weaken points in the material, such
as porosity, segregation or inclusions. High forging temperatures can also contribute to this type of imperfections.
292 V. Infante et al. / Engineering Failure Analysis 35 (2013) 286–293

Fig. 11. Hardness vickers profile obtained in the crankpin journal.

Fig. 12. Metallographic analysis of the crankpin journal, with a clear visualization of the hardened surface layer. Magnification: 50.

Fig. 13. SEM image of an internal imperfection close to the crack initiation region.

Whichever is the root cause of this defect, it seems clear that its effect in the crack initiation stage should be regarded as a
contributing factor for the onset of fatigue damage in a cumulative context considering the other aforementioned stress
V. Infante et al. / Engineering Failure Analysis 35 (2013) 286–293 293

raising features, namely the undercut fillet and the lubrication hole, which are all contiguous to the crack initiation region
show in Fig. 3.

4. Conclusions

The analysis of this failed crankshaft envisaged the determination of the contributing factors for the occurrence of the
resulting accident. A throughout visual and microscopic inspection of the fracture surface of the crankpin journal allowed
concluding that the crankshaft failed as a result of a high cycle fatigue process with a crack initiation region in the interface
of the crankpin journal and the counterweight.
From the analysis of the component, some important secondary conclusions can be highlighted:

 Ratchet marks have been identified in the crack initiation site adjacent to the existing undercut fillet in the interface of the
crankpin journal and the counterweight. This is a clear indication of a high stress concentration condition in this region
which therefore promoted the onset of initial micro-cracks and a subsequent crack front progression due to fatigue. On
the other hand, there were no visible indications that suggested any material or manufacturing defects in the fillet region.
 Beach marks consistent with a fatigue process have been identified along almost two thirds of the fracture surface,
whereas the remaining part has a brittle fracture pattern as a consequence of a final overload condition. The evolution
of the crack front geometry seems to indicate that a predominant cyclic bending state condition has been the promoting
factor in the base of the failing process.
 Hardness measurements in the crankpin cross section revealed that there is a significant increase of the nominal hardness
near the surface of this component indicating the existence of a case-hardened layer extending along its perimeter, which
is consistent with the information provided by the engine’s manufacturer.

As a final conclusion, the relevance of the stress concentration in the neighboring of the fillet region (as a result of the
combined contribution of the undercut fillet, lubrication hole and a possible forging defect) should be stressed out has
the major likely root cause for the failure of this component. Having in mind the importance of this matter, a following com-
putational analysis will be undertaken to assess the effect of the stress gradient in the interface of the crankpin journal and
the counterweight aiming at better understanding its influence in the crack initiation process due to fatigue.

References

[1] Reddy AV. Investigation of aeronautical and engineering component failures. Florida (USA): CRC Press; 2004.
[2] Becerra JA, Jimenez FJ, Torres M, Sanchez DT, Carvajal E. Failure analysis of reciprocating compressor crankshafts. Eng Fail Anal 2011;18(2):735–46.
http://dx.doi.org/10.1016/j.engfailanal.2010.12.00.
[3] Fonte M, de Freitas M. Marine main engine crankshaft failure analysis: a case study. Eng Fail Anal 2009;16:1940–7. http://dx.doi.org/10.1016/
j.engfailanal.2008.10.01.
[4] Bao K, Liao RD, Zuo ZX. Stress calculation of fillet rolled crankshaft in bending fatigue tests. Appl Mech Mater 44–47:2798–2804. doi: http://dx.doi.org/
10.4028/www.scientific.net/AMM.44-47.2798.
[5] Failures of shafts. Metals handbook, vol. 10: failure analysis and prevention, 8th ed. American Society for Metals; 1975. p. 373-96.
[6] Wulpi DJ. Understanding how components fail. ASM Int 1999:126–7.
[7] Fatigue and fracture. ASM handbook, vol. 19. Metals Park (OH): ASM International; 1996. p. 665–79.
[8] Secondary cracks. Metals handbook, vol. 9: fractography and atlas of fractographs, 8th ed. American Society for Metals; 1975. p. 32.
[9] Silva JM, Infante V, Freitas M. de Reis L. Microscopy analysis of damaged aeronautical components. In: Méndez-Vilas A, Díaz J, editors. Microscopy:
science, technology, applications and education – microscopy book series nr. 4. Badajoz (Spain): Formatex Research Center; December 2010. p.
1838–45.