Engineering Failure Analysis 109 (2020) 104356

Contents lists available at ScienceDirect

Engineering Failure Analysis

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

T
Analysis of premature failure of aircraft hydraulic pipes
Zahid Mehmooda, , Asad Hameeda, Ali Javeda, Azhar Hussainb
⁎

a
College of Aeronautical Engineering, National University of Sciences and Technology, Islamabad, Pakistan
b
Department of Metallurgy and Materials Engineering, University of Engineering and Technology Taxila, Rawalpindi, Pakistan

A R T IC LE I N F O ABS TRA CT

Keywords: Premature failure of aircraft metallic hydraulic pipes was reported frequently and almost all
Pipeline failures pipes failed in a similar fashion. The failure of pipes during flight depleted hydraulic pressure and
Failure analysis put aircraft at potential risk. Failure analysis of a fractured pipe and a cracked hydraulic pipe was
Fatigue failure carried out. Fatigue striations on the fractured surfaces were observed during scanning electron
Circumferential groove
microscopy (SEM). The fracture initiated from the constrained (inlet) end as crack nucleated
Stress concentration
under the metallic sleeve on the pipe surface. Metallic sleeves are crimped at the pipe ends to
help these connect with other hydraulic system components. On removal of metallic sleeves,
stress concentration sites in the form of circumferential grooves (depth 35 ± 7 µm, radius
76 ± 2 µm) were observed. SEM of cracked pipe also confirmed the crack nucleation from
circumferential groove under the metallic sleeve. Fluctuating hydraulic pressure caused by
random movement of control surfaces produced cyclic loading at the constrained (inlet) end.
Forces generated by hydraulic pressure on pipe bends were obtained computationally and were
validated analytically by solving impulse-momentum equations. Computational stress analysis of
the pipe, based on calculated fluid forces, revealed that maximum Von Mises stress increased by
~ 123% in the presence of circumferential groove at the constrained (inlet) end. Higher stress
levels associated with circumferential grooves at constrained end nucleated early cracks, which
propagated under fluctuating hydraulic pressure till pipes failed prematurely.

1. Introduction

Failure of hydraulic pipes is a critical problem in many industries like nuclear, marine, petroleum, chemical, and aerospace.
Factors associated with pipe failure are material characteristics [1], pipe geometry [2], environmental conditions [3], external/
internal loading conditions [4,5], residual stresses [6] and manufacturing flaws [7]. The interaction of these factors is very complex
to analyse; however, the main contributing factors to failure can be segregated on the basis of operating conditions [8]. For example,
in an aircraft hydraulic system, pressure pulsations and flow transients cause cyclic stresses leading to fatigue failure of pipelines. Due
to the same reason, a limiting factor in the operational life of a hydraulic system is the fatigue life of hydraulic pipes [9,10].
Generally, an aircraft hydraulic system consists of a reservoir, pressure pump, antisurge booster, solenoid valves, filters, pressure
accumulators and displacement actuators [11]. For the aircraft understudy, a schematic of its hydraulic system is shown in Fig. 1,
indicating main accessories interconnected by metallic pipelines. The metallic sleeves are crimped on pipe ends during the manu-
facturing process for connecting pipes with other system components through connectors as seen in Fig. 2. After installation on
aircraft, pipe ends with metallic sleeve get constrained by tightening of connectors to form a leak-proof connection. The maximum
pressure maintained by the hydraulic pump is 21.0 ± 0.5 MPa and its maximum volume flow rate reaches 0.0006937 m3/s when

⁎
Corresponding Author.
E-mail address: [email protected] (Z. Mehmood).

https://doi.org/10.1016/j.engfailanal.2019.104356

Available online 23 December 2019

Received 11 July 2019; Received in revised form 17 December 2019; Accepted 17 December 2019

1350-6307/ © 2019 Elsevier Ltd. All rights reserved.

Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

Pressure line pipe

Flight
which failed frequently
Accumulator

Antisurge
Accumulator

Hydraulic
Pump
Randomly moving flight
control surfaces actuators

Hydraulic
Reservoir

Fig. 1. A simple schematic of the hydraulic system of aircraft under study. Hydraulic pipe in pressure line which fractured prematurely is high-
lighted in red colour. The main components of the hydraulic system are also shown. (For interpretation of the references to colour in this figure
legend, the reader is referred to the web version of this article.)

Connector secured by
metallic sleeve

Metallic
sleeve
(black ring)
crimped on
pipe end

Antisurge booster

Fig. 2. A hydraulic component (antisurge booster) is shown next to a pipe end with a metallic sleeve and a connector nut. The metallic sleeve helps
in holding pipe connectors when they are attached to other hydraulic system components.

system pressure drops to 18 MPa and below. The pressure produced by pump acts on control surfaces of aircraft through a network of
pipes and components. The random movement of aircraft control surfaces during flight induces fluctuations in hydraulic pressure.
The fluctuations are sensed by a pressure transducer and are recorded by the onboard data acquisition system (DAS) of aircraft. An
example of a hydraulic pressure profile of a typical flight produced by DAS is shown in Fig. 3.
According to the aircraft history, pressure line pipe installed with antisurge booster failed frequently which is highlighted in
Fig. 1. No limit on the service life of hydraulic pipes has been specified by the original equipment manufacturer (OEM), however,

2
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

Hydraulic pressure
fluctuations in one sortie

Fig. 3. A typical example of hydraulic pressure fluctuations recorded by aircraft DAS system during one flight. Hydraulic pressure fluctuates with
the movement of control surfaces of aircraft.

Cracked pipe
(C-pipe)

4.2 mm crack

Fractured pipe Fractured portion with metallic sleeve

(F-pipe)

Fig. 4. A fractured hydraulic pipe (F-pipe) and a cracked hydraulic pipe (C-pipe) are shown. Visual inspection revealed that pipe fractures from its
constrained (inlet) end near the metallic sleeve.

pipes are replaced after 1200 flying hours based on the physical appearance (e.g. chaffing, dents, nick marks). In our study, two
failure cases of the same pipe were selected for the analysis, one case of fractured pipe (named F-pipe) and the other was cracked pipe
(named C-pipe). The F-pipe fractured after flying almost 225 hours and C-pipe fractured after flying 185 hours. The outer diameter of
the pipe is 10 mm and its wall thickness is 1 mm. It is pertinent to mention that midair fracture/cracking of the aircraft hydraulic
pipes caused loss of hydraulic pressure, impaired the movement of control surfaces and endangered aircraft. Therefore, root cause
analysis of the failure of hydraulic pipes was deemed essential.

2. Experimental procedure

The failure of hydraulic pipes was investigated by visual examination, optical light microscopy (OLM) using Nikon SMZ 18
microscope and scanning electron microscopy (SEM). SEM model was VEGA3, used in secondary electron (SE) imaging mode, at
20 kV and a working distance of 15~20 mm. Energy Dispersive X-Ray Spectroscopy (EDS) analysis was carried out using Oxford
Instrument X-max detector and Inca software. Metallographic samples of hydraulic pipes were prepared by the conventional
mounting, grinding and polishing techniques for microstructural analysis. A solution of 5 g Ferric Nitrate (Fe(NO3)3 + 70 ml

3
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

(a) Metallic sleeve

Hydraulic pipe

(b)

A region of fatigue
crack propagation

Fig. 5. Optical light microscopy (OLM) images of fractured surface of F-pipe are shown. (a) Fractured face of F-pipe with metallic sleeve at 20 X, and
(b) A region of fatigue crack propagation on the fractured surface at 50 X.

water + 25 ml Hydrochloric acid (HCl) was used to reveal the grain structure [12]. Zeiss Axio microscope (A1) was used for
microstructural analysis. The hardness test of pipes was carried out using a microhardness tester; Model No HM221-AT400, Wolpert
Wilson ® Instruments. Hydraulic flow simulation through the pipe and its stress analysis was carried out using ANSYS ® Workbench.

3. Results

3.1. Visual examination

In two different occurrences, the failure of the aircraft hydraulic system was reported during flight and post-flight inspection

4
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

(a) (b) Parallel lines of

fatigue striations
A

Hydraulic pipe

Metallic sleeve
A

(c) B (d) B

Fatigue Failure

Shear dimples
Fast Rupture

Fig. 6. SEM images of the fractured surface of F-pipe. (a) An overall SEM image of the fractured surface with zones marked for in-depth analysis, (b)
A higher magnification image of zone A shows fatigue striations indicating crack propagation, (c) Fatigue marks and fast fracture zone are clearly
visible at zone B, and (d) and a magnified view of shear dimples of zone B is presented.

revealed failed pipes as the cause of system failure. F-pipe was fractured from the constrained (inlet) end in two pieces, whereas a
large through-thickness crack (length 4.2 mm) was present on C-pipe at the same end. The pipes were fractured/cracked at the
constrained (inlet) end near the crimped metallic sleeve as shown in Fig. 4.

3.2. Optical light microscopy (OLM)

The OLM was used at 5 ~ 100 X magnification to analyse the fractured surface of the hydraulic pipe. Overall OLM image of
fracture end of F-pipe with sleeve is shown in Fig. 5(a). At higher magnification, regions that could be associated with fatigue crack
propagation were observed and are shown in Fig. 5(b). The width of the crack propagation zone varied from 0.2 ~ 0.6 mm along the
pipe thickness of 1 mm as highlighted in the figure.

3.3. Scanning electron microscopy (SEM)

3.3.1. SEM of F-Pipe

The fractured surface of F-pipe was analysed using SEM and Fig. 6(a) shows an overall SEM view of the fractured end of F-pipe
with metallic sleeve. Two zones, A and B are marked on the surface which are analysed in detail. At zone A, fatigue striations were
clearly visible at higher magnification which are shown in Fig. 6(b). A relatively flatter fatigue region, as well as a fast fracture region,
was observed at zone B, as shown in Fig. 6(c). The fatigue striations of zone B are not clear possibly due to rubbing during service.
However, shear dimples of zone B are shown in Fig. 6(d), which indicate the fast fracture of the pipe.

3.3.2. SEM of C-Pipe

During the crimping process, a metallic sleeve grips pipe surface with its internal threads as shown in Fig. 7(a, b). The threads
produce circumferential grooves as observed on the cracked end of C-pipe as shown in Fig. 8(a, b). The crack nucleation from the

5
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

(b)

(a)

Threads on
internal side of
metallic sleeve

Fig. 7. The metallic sleeve which is crimped on pipe ends for securing connections is shown. (a) An overall view of the metallic sleeve is presented,
(b) A cutaway section of the metallic sleeve revealed its internal threads.

(b)
(a)

Circumferential groove

Crack nucleated
Crack propagation
from groove
on pipe surface

Fig. 8. A cracked end of C-pipe is displayed. (a) Cracked end with metallic sleeve, (b) After removal of sleeve, circumferential grooves formed by
internal threads of the sleeve are observed on the pipe surface. Crack nucleates from grooves and propagates on the pipe surface.

circumferential groove was observed using SEM as shown in Fig. 9. The cracked end of C-pipe was opened and then cut carefully
avoiding rubbing of the fractured surface. The overall SEM view of the fractured surface of the C-pipe is shown in Fig. 10(a) and two
zones C, and D are marked on it. The root of the circumferential groove is focused at zone C as shown in Fig. 10(b). Multiple cracks
have nucleated from the root of the groove as indicated by the arrowheads. The steps are also observed on the fractured surface,
indicating the formation of ratchet marks by multiple cracks. A magnified view (Fig. 10(c)) of zone D showed a flatter region
containing fatigue striations, however, striation pattern has lost its clarity possibly due to rubbing during service and/or post fracture
damage.

6
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

Crack nucleation from

circumferential groove

Circumferential Crack growth

grooves made by
metallic sleeve

Fig. 9. SEM image of the cracked end of C-pipe. Circumferential grooves made by the metallic sleeve and crack propagation on pipe surface are
magnified.

(a) C (b)

Cutting marks

Root of circumferential
groove

(c)

Fig. 10. SEM analysis of fractured surface of C-pipe. (a) Overall view of fractured surface of C-pipe is revealed after opening cracked end and
different zones are marked on it, (b) Nucleation of multiple cracks from root of the groove and ratchet marks are observed at zone C, and (c) At
higher magnification, fatigue striations at zone D are observed.

7
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

Groove radius

Groove depth

Fig. 11. A magnified SEM image of a side view of the circumferential groove is shown. Dimensions of the groove were measured and subsequently
used in stress analysis.

Table 1
The average material composition of hydraulic pipe and precipitates.
Element Cu Zn Al Ni Si

Overall alloy Composition Wt % 82.0 ± 2.0 15.0 ± 2.0 1.5 ± 0.3 1.3 ± 0.2 1.1 ± 0.2
Composition at precipitate Wt% 73.5 ± 2.0 8.0 ± 2.0 0.5 ± 0.3 10.5 ± 0.2 7.5 ± 0.2

3.3.3. SEM of circumferential groove

The circumferential groove is a stress concentration site and the corresponding stress concentration factor is dependent on its
dimensions/geometry [13]. In order to measure the circumferential grooves, the pipes were longitudinally cut using EDM wire
cutting technique. The depths and radii of grooves were measured using high magnification images from SEM as shown in Fig. 11(a,
b). The average values of the groove’s depth and radius are 35 ± 7 µm and 76 ± 2 µm respectively.

3.4. Chemical composition and metallography

EDS analysis of a sample from C-pipe was performed at multiple locations and the average material composition of the pipe is
provided in Table 1. From material composition results, it is found that hydraulic pipe was made of Ni-Al-Si alpha-brass alloy, UNS
C69100 and widely known as Tungum [14,15]. Longitudinal sections of C-pipe were cut in the proximity of crack initiation site and
metallographic samples were prepared by conventional techniques. The OLM image of the microstructure is displayed in Fig. 12(a).
Equiaxed grains of alpha phase (Cu-Zn) can be observed in a light colour and dark dots (submicron size) were dispersed on the grains.
For detailed analysis of dark dots, SEM and EDS analysis was carried out. It was found that the dark dots are Ni-Si rich precipitates
as shown in Fig. 12(b and c). The chemical composition of precipitates obtained from EDS is tabulated in Table 1 alongside the overall
composition. The concentration of Ni and Si in the chemical composition of precipitate (10.5% and 7.0%) was found to be higher
than the main matrix (1.3% and 1.1%). The presence of Ni-Si precipitates in similar copper alloys such as C64700 has also been found
in literature during microstructural studies [15]. The Tungum hydraulic pipes were manufactured as per ASTM-B706 of seamless
Copper alloy (C69100) pipes and tubes. The pipes bear the markings “TF00” by the manufacturer which means that they are pre-
cipitation hardened. The average microhardness value of pipes was also measured as 132 ± 3 HV which is in the range specified by
the ASTM B-706 [16].

3.5. Stress analysis

DAS data of hydraulic fluctuations (Fig. 3) was sorted using level crossing technique and most recurring hydraulic pressure values
were obtained. Hydraulic pressure fluctuated in a range of 8.5 to 21.5 MPa and the maximum number of values occurred in between
19.0 and 21.5 MPa pressure as shown in Fig. 13. The fluctuations are resulted by the movement of aircraft control surfaces which
create a favourable pressure differential in the hydraulic system. By sensing lower pressure, the hydraulic pump drives fluid through
the complete system. The mass flow rate of the pump increases with a drop in hydraulic pressure; however, it attains a constant value
(0.0006937 m3/s) when the corresponding pressure becomes 18 MPa or less. Subsequently, fluid velocity through the pipe also varies
with hydraulic pressure as shown in Fig. 14.

8
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

(a)

Ni-Si Precipitates (b) (c)

Fig. 12. The OLM and SEM results of Tungum alloy are shown. (a) OLM image of microstructure of Tungum alloy (UNS C69100) shows light colour
grains of alpha phase (Cu-Zn) and dark dots of Ni-Si precipitates, (b) SEM micrograph of same microstructure indicating Ni-Si precipitates, and (c) A
representative EDS spectra of the Tungum alloy pipe is shown with a table containing wt % composition of overall alloy and the precipitates.

Based on fluid velocity and hydraulic pressure, a detailed stress analysis was conducted, using ANSYS ® Workbench to work out
the complete stress picture on the pipe under different operating conditions. The values of lowest pressure (8.5 MPa), most recurring
pressure values (19.0, 19.5, 20.0, 20.5, and 21.0 MPa) and the highest value (21.5 MPa) were used for the analysis. Quadratic
elements mesh was used to achieve better results and the flow problem was solved using ANSYS ® Fluent. Velocity boundary con-
dition was applied at inlet while the pressure boundary condition was applied at the outlet. The standard k-ω turbulence model was
used for solving pipe flow [17]. The contours of static pressure developed with inlet fluid velocity of 13.8 m/s along with outlet
pressure of 8.5 MPa are shown in Fig. 15. Prior to computational stress analysis based on Fluent results, analytical verification of fluid
forces was carried out. It is evident from Fig. 15 that pressure head loss occurs as hydraulic fluid flows through the pipe. For the
analytical calculation of fluid forces, head loss coefficient for pipe flow was required. The head loss is due to surface friction, pipe
bends and flow separation and is calculated by following analytical expressions [18]:-
∆P = P2 − P1 = Khl V 2ρ /2 (1)
where ‘P1’ is pressure at the inlet to a pipe section and ‘P2’ at the outlet. ‘V’ is the velocity of hydraulic fluid through a constant area
pipe, ‘ρ’ its density and ‘Khl’ is head loss coefficient, given by
L
Khl = f ; for straigth section
D (2)
‘L’ is the length of the straight section of pipe, ‘D’ is the internal diameter of the pipe and ‘f’ is a frictional factor given by Eq. (3)
and Eq. (4).

9
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

Fig. 13. Hydraulic fluctuations data recorded during a flight is sorted using the level crossing technique. Maximum data points occurred in a
pressure range of 19.0 to 21.5 MPa.

Fig. 14. The graph of fluid velocity through pipe against hydraulic pressure is shown. Fluid velocity is zero at the maximum rated pressure of the
pump and its highest value (13.8 m/s) occurs at 18 MPa pressure and below.

For fully developed laminar flow, i.e. Reynold number, Re < 2100;
f = 64/Re (3)
4
For fully developed turbulent flow with smooth pipe walls and Reynold number, Re < 2 × 10 [19];
f = 0.316Re−1/4 (4)
‘Khl’ for pipe bend is

r α 6.6[ sin(α /2) + sin(α /2)]

Khl = fα + (0.10 + 2.4f ) sin ⎛ ⎞ + ; for pipe bend
D ⎝2⎠ (r / d )( π )
4α
(5)
where ‘r’ is bend radius and ‘α’ is bend angle.
As highlighted in Fig. 15, hydraulic pipe contains four bends and fluid flowing through pipe bends experiences change in mo-
mentum and produces forces on pipe bends. The forces on a pipe bend with angle ‘θ’ are given by the impulse-momentum equation as
follows [20]:-
Fi = ρQ (V1 − V2 cosθ) + P1 A1 − P2 A2 cosθ (6)

Fj = ρQ (−V2 sinθ) + P1 A1 − P2 A2 sinθ (7)

10
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

Inlet

Outlet

Bend 1 Bend 4

Bend 3

Bend 2

Fig. 15. Contours of static pressure developed in the pipe as a result of ANSYS ® Fluent simulation. Velocity inlet condition of 13.8 m/s and a
pressure outlet condition of 8.5 MPa was used.

Fig. 16. Schematic of forces generated on pipe bend by the change in fluid momentum through the bend.

In the above equations, ‘Fi’ and ‘Fj’ are the forces along the i and j axes respectively as shown in Fig. 16. ‘V1’ and ‘P1’ are the
velocity and pressure at the inlet to the pipe bend, ‘V2’ and ‘P2’ velocity and pressure at the outlet. ‘ρ’ is the hydraulic fluid density and
‘Q’ is volume flow rate. By solving Eqs. (1) to (5), pressure values across pipe bends were calculated and inserted in Eqs. (6) and (7)
along with other prescribed parameters. The resultant of fluid forces obtained analytically (by solving Eqs. (6) and (7)) and from
computational approach (Fluent) are plotted against pressure values in Fig. 17. The forces calculated from the computational ap-
proach were found in a close approximation of analytically calculated forces with a maximum difference of 5.3%.
After verification of fluid forces, computational stress analysis of hydraulic pipe was performed based on Fluent results. In this
step, fluid forces were imported and applied internally on pipe walls. The two ends of the pipe were constrained; similar to the case of
physically installed pipe on aircraft. The structural problem was solved using ANSYS ® Static Structural module. High quality
quadrilateral elements (average orthogonal quality ~ 0.99 and skewness ~ 4 × 10−2) were used for mesh generation. The elements
were concentrated at the high stress region by using biased edge sizing, for capturing the stress gradients accurately. The stresses
produced as a result of fluid forces corresponding to hydraulic pressure of 8.5 MPa are shown in Fig. 18(a). A groove was modelled at
the constrained (inlet) end for simulation of stress concentration caused by a circumferential groove. The dimensions of the groove

11
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

Fig. 17. Comparison of analytically and computationally (Fluent) calculated fluid forces produced by flowing fluid on the pipe bends. The forces are
plotted against the system pressure which drives the hydraulic fluid.

were set the same as measured using SEM (Fig. 11). A refined mesh was made near the groove to ensure accurate calculations of the
stress field. Fluid pressure was applied again on pipe inner walls and results obtained from simulation are shown in Fig. 18(b). From
stress analysis, it was found that the value of maximum Von Mises stress at constrained (inlet) end is 53 MPa without groove and
118 MPa with groove (hydraulic pressure ~ 8.5 MPa). Similarly, computational stress analysis for other pressure values (19 to 21.5
MPa) was carried out for both (with and without groove) conditions of constrained (inlet) end. The comparison of stresses with and
without circumferential groove, for most recurring pressure values, is shown in a plot of Fig. 19.

4. Discussion

4.1. Role of pipe bends and end constraints

The pipe is designed with multiple bends and fluid flowing through the bends experienced a change in momentum and resultantly
produced forces on the bends. Additionally, it was also noted during the study of the system that the connector nut is tightened over
the sleeve and the pipe at the sleeve became mechanically constrained. The forces present at the pipe bends had a bending effect and
produced flexural stresses at the constrained end with sleeve. The results of computational stress analysis also indicated that the
location of maximum Von Mises stresses is at the constrained end which is indeed the fractured region in both the failed pipes. In fact,
it is the constrained end underneath the sleeve/nut from which crack nucleation takes place on the pipes (Figs. 4 and 10). For straight
pipes or pipes with no bend angle, the resultant of forces calculated by impulse-momentum equations [Eqs. (6) and (7)] is zero;
subsequently, the flexural stress are absent at the constrained end. Therefore, pipe bends coupled with the constrained end played an
important role in increasing the stress level at the constrained end which corresponds with the crack initiation/fractured region.

4.2. Role of hydraulic pressure fluctuations

The hydraulic pressure, produced by the pump, fluctuated following the movement of aircraft surfaces during flight.
Subsequently, the fluid forces acting on pipe bends (discussed in the previous section) also varied (Fig. 17). With the increased
hydraulic pressure, the fluid forces also increased as predicted by impulse-momentum equations and shown by the computational
analysis. The corresponding stresses at the constrained (inlet) end also increased and the same can be observed in Fig. 19 where the
Von Mises stresses calculated from computational stress analysis are plotted against the pressure values. From DAS data it is evident
that the maximum number of pressure values occurred in the range of 19.0 to 21.5 MPa, and calculation of corresponding stresses
varied from 253 MPa to 283 MPa, respectively. By definition, stress fluctuation causes fatigue. In the present study, the optical and
electron microscopy undoubtedly revealed the presence of fatigue at the fractured/cracked surfaces of both the pipes.

12
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

Fig. 18. Computational stress analysis of hydraulic pipe showing high stress areas corresponding to 8.5 MPa system pressure. (a) Maximum Von
Mises stress at constrained (inlet) end is 53 MPa without circumferential groove (b) Maximum Von Mises stress increases (from 53 MPa) to118 MPa
in the presence of circumferential groove. The location of maximum Von Mises stress at the constrained (inlet) end matches the fractured/crack
nucleation region of failed pipes (Figs. 4 and 9).

4.3. Role of circumferential grooves

The other important observation in the study was that the crack growth always initiated from the circumferential groove imparted
underneath the metallic sleeve during its crimping process on the pipe. The circumferential grooves act as stress raisers in the present
scenario as the pipe stresses (created by the hydraulic pressure) intensified due to the presence of a circumferential groove as found
from the computational stress analysis (Fig. 18). The comparison of stresses with and without the presence of circumferential groove
at different hydraulic pressures is also provided in Fig. 19. The presence of circumferential groove at high stress region of pipes
increased the Von Mises stresses almost by ~123%.

4.4. A design flaw!

In the presented scenario, the hydraulic pipes have failed by the combined effect of forces on pipe bends, mechanical constraint,
pressure fluctuations, and stress raiser (groove). The pressure fluctuations are inherent in the design of an aircraft hydraulic system as
these are produced due to the movement of control surfaces. Whenever a control surface moves, pressure drops in the system which is
replenished by the hydraulic pump. The fluctuations are a source of fatigue loading which produced varying fluid forces at pipe bends

13
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

Fig. 19. Graph of Von Mises stress against the hydraulic pressure value is shown. Von Mises stress varies with hydraulic pressure resulting in fatigue
at the failed pipe end. Stresses were also amplified (~123%) by the circumferential groove (stress concentration site) imparted by the metallic
sleeves.

Flared pipe end

Nozzle shaped
component end

Fig. 20. Flared end fitting of another type of aircraft is shown which can be a potential replacement of crimped sleeve end fittings.

and subsequent varying amplitude stresses at the constrained end. Anti-surge accumulator helps to smoothen small pressure fluc-
tuations from the pump; however it cannot mitigate those completely. For straight pipes, the forces and subsequent stresses due to
bends are obviously not present as discussed in Section 4.1. However, aircraft accommodates multiple onboard components and
accesories from its electrical, hydraulic, avionics, engine, and fuel systems besides its structure, and its hydraulic pipes are designed
with multiple bends for installation in confined space. Therefore, hydraulic fluctuations and pipe design with bends cannot be taken
out of the scenario for failure prevention of the pipes. On the other hand, the role of pressure fluctuations and pipe bends in
premature fatigue failure of pipes was exacerbated by the stress concentration of the circumferential groove. In fact, one of the

14
Z. Mehmood, et al. Engineering Failure Analysis 109 (2020) 104356

important considerations for hydraulic pipes in design against fatigue is the elimination of stress raisers. These stress risers are likely
to initiate crack and subsequently reduce the fatigue life of pipes [21–23]. Thus crimping of the sleeve on the pipe ends is producing a
region with high stress concentration similar to notches and grooves [24,25].
The crimped metallic sleeve on pipe ends is a peculiar design flaw which is generally not a common practice in modern aviation.
As an example, end fitting on the hydraulic pipe of another type of aircraft is presented in Fig. 20. The hydraulic pipe has a flared end
and it is connected to a ‘nozzle-shaped’ end of a component. In the peculiar flared end fittings, stress concentration sites like grooves
and notches are not present. Therefore, crimped sleeve end fittings are not suitable for hydraulic pipes of aircraft and it is the primary
cause of premature failure of pipes.

5. Conclusions

Hydraulic pressure fluctuations, pipe geometry (bends and constraints), and the stress concentration of circumferential grooves
produced due to sleeve crimping are the three contributory factors resulting in fatigue failures of the pipes. However, pressure
fluctuations and pipe geometry variations are always present in aircraft hydraulic systems whereas the circumferential groove in the
present case aggravated the situation (increased the stress by ~123%) to cause premature failure. Therefore, the crimping of sleeve
end fitting on the pipe is the primary cause of the pipe failures and is unsuitable for aircraft hydraulic pipes.

6. Recommendations

It is recommended that, if deemed economically viable, the crimped sleeve end fittings of aircraft hydraulic pipes may be replaced
with the flare-end fittings or any other suitable type which do not produce stress concentration sites.

References

[1] Η. Τawancy, L.M. Al-Hadhrami, Failure analysis of a welded outlet manifold pipe in a primary steam reformer by improper selection of materials, Eng. Fail. Anal.
16 (3) (2009) 816–824.
[2] E.S. Firoozabad, B.-G. Jeon, H.-S. Choi, N.-S. Kim, Failure criterion for steel pipe elbows under cyclic loading, Eng. Fail. Anal. 66 (2016) 515–525.
[3] W. Liu, T. Shi, Q. Lu, Z. Zhang, C. Ming, J. Gong, J. Ren, Failure analysis on fracture of S13Cr-110 tubing, Eng. Fail. Anal. 90 (2018) 215–230.
[4] W. Edjeou, G. Pluvinage, J. Capelle, Z. Azari, Effect of pressure and defects on the pipe flattening factor, Eng. Fail. Anal. 94 (2018) 469–479.
[5] J. Zheng, B. Zhang, P. Liu, L. Wu, Failure analysis and safety evaluation of buried pipeline due to deflection of landslide process, Eng. Fail. Anal. 25 (2012)
156–168.
[6] K. Mankari, S.G. Acharyya, Failure analysis of AISI 321 stainless steel welded pipes in solar thermal power plants, Eng. Fail. Anal. 86 (2018) 33–43.
[7] A. Karani, S. Koley, M. Shome, Failure of electric resistance welded API pipes–effect of centre line segregation, Eng. Fail. Anal. 96 (2019) 289–297.
[8] H. Rezaei, B. Ryan, I. Stoianov, Pipe failure analysis and impact of dynamic hydraulic conditions in water supply networks, Procedia Eng. 119 (2015) 253–262.
[9] D. Olson, Pipe Vibration Testing and Analysis. Companion Guide to the ASME Boiler & Pressure Vessel Code. Chapter 37, ASME PRESS, New York, 2002.
[10] G.A. Antaki, Piping and Pipeline Engineering: Design, Construction, Maintenance, Integrity, and Repair, CRC Press, 2003.
[11] W.L. Green, Aircraft Hydraulic Systems: An Introduction to the Analysis of Systems and Components, John Wiley & Sons, 1985.
[12] G.F. Vander Voort, Metallography, Principles and Practice, ASM International, 1999.
[13] W.D. Pilkey, D.F. Pilkey, Peterson's Stress Concentration Factors, John Wiley & Sons, 2008.
[14] C. Pitt, T. Pitt, Variations in mechanical properties and microstructure of CZ127 (Tungum or Ni–Al–Si α-brass) during aging after solution treatments at 800 and
900° C, Metals Technology 11 (1) (1984) 145–148.
[15] J.R. Davis, Copper and Copper Alloys, ASM international, 2001.
[16] ASTM-B706, ASTM B706-18, Standard Specification for Seamless Copper Alloy (UNS No. C69100) Pipe and Tube, ASTM International, West Conshohocken, PA,
2018, www.astm.org, 2018.
[17] D.C. Wilcox, Turbulence modeling for CFD, DCW industries La Canada, CA, 1998.
[18] D.C. Rennels, H.M. Hudson, Pipe flow: A practical and comprehensive guide, John Wiley & Sons, 2012.
[19] F.P. Incropera, A.S. Lavine, T.L. Bergman, D.P. DeWitt, Fundamentals of Heat and Mass Transfer, Wiley, 2007.
[20] R. Bansal, A textbook of fluid mechanics and hydraulic machines, Laxmi Publications (2004).
[21] B.D.C. Pinheiro, I.P. Pasqualino, Fatigue analysis of damaged steel pipelines under cyclic internal pressure, Int. J. Fatigue 31 (5) (2009) 962–973.
[22] R.S. Motta, H.L. Cabral, S.M. Afonso, R.B. Willmersdorf, N. Bouchonneau, P.R. Lyra, E.Q. de Andrade, Comparative studies for failure pressure prediction of
corroded pipelines, Eng. Fail. Anal. 81 (2017) 178–192.
[23] U. Zerbst, M. Madia, C. Klinger, D. Bettge, Y. Murakami, Defects as a root cause of fatigue failure of metallic components. I: basic aspects, Eng. Fail. Anal. (2019).
[24] H. Yu, X.-L. Xu, Fatigue failure of high-pressure oil-pipes of truck diesel engine, Eng. Fail. Anal. 97 (2019) 145–160.
[25] L. Witek, P. Zelek, Stress and failure analysis of the connecting rod of diesel engine, Eng. Fail. Anal. 97 (2019) 374–382.

15