Key Engineering Materials Vols. 348-349 (2007) pp. 5-8
online at http://www.scientific.net
© (2007) Trans Tech Publications, Switzerland
Fretting Fatigue Phenomena on an all Aluminium Alloy Conductor
Marco Boniardi1,a, Silvia Cincera1,b, Fabrizio D’Errico1,c, Chiara Tagliabue2,d
1
Dipartimento di Meccanica, Politecnico di Milano, Via La Masa 34, 20156 Milano, ITALY
2
Hammer S.r.l., Via Risorgimento 69/22, 20017 Rho (MI), ITALY
marco.boniardi@polimi.it, silvia.cincera@mecc.polimi.it, fabrizio.derrico@polimi.it,
dir.laboratorio@hammerlabo.com
Keywords: fretting fatigue, all aluminum alloy conductor, aeolian vibrations.
Abstract This paper concerns about a failure analysis of an electric all aluminum alloy conductor
(AAAC) damaged and broken for fretting fatigue phenomena induced by aeolian vibrations.
Life of electric conductors is often reduced by various degradation mechanisms such as repeated
bending, fluctuating tension, distortion, fatigue, wear and corrosion phenomena. However the main
limiting factor of the electrical conductors is related to aeolian vibrations in the high frequency
range (between 5 to 50 Hz). Conductor oscillations may lead to fretting fatigue problems (otherwise
called fretting wear) caused by wind excitation, mainly in the suspension clamp regions, spacers or
other fittings. The induced aluminium wire fracture imply a drastic reduction in the transmission
line service. Vibration dampers are considered the most effective method to extend service life of
electric conductors, as they are the means to reduce fretting damage of aluminium wires.
The aim of the present work is to investigate the failure of an AAAC conductor of a 400kV
overhead transmission line (twin conductors) located in Touggourt Biskra (Algeria); the damaged
and broken conductors were operated in-service only for six months without spacers or dampers.
Three different types of conductors have been taken as experimental samples: the in-service broken
conductor, another in-service damaged conductor and a new conductor from warehouse as terms of
comparison. Samples have been analysed to identify the root cause of the failure and to verify the
conformity of the conductor elements to the international standards. The investigation has outlined
the morphology of the fretting damage: in all cases the fractured wires have shown typical static
deformation marks and dynamic fretting wear tangential marks associated with intense presence of
Al2O3 debris.
Introduction
Overhead electrical conductors consists of many wires twisted together to make a complex structure
combining axial strength and stiffness with flexibility in bending [1]. It is well known in the
electrical power industry that wind vibrations are first damage cause of conductors. Because of
aeolian vibrations, overhead transmission line service life is often drastically reduced. Wind vortex
produce up and down oscillations in the high frequency range (between 5 to 50 Hz) increasing up to
the resonance frequency of cable [2-3]: it causes alternating bending stresses in cable strands and it
leads to fretting fatigue/wear phenomena among cable wires above all in the suspension clamp
regions, spacers or other fittings. This damage leads to partial or complete failure of the overhead
transmission line [2-4].
Fretting fatigue phenomena
Fretting fatigue damage occurs when two contacting wire surfaces are subjected to a normal
clamping force and they undergo a relative movement on the two surfaces due to a cyclic tangential
shearing. Fretting damage causes formation of microscopic elliptical marks on the wire surfaces and
it produces particle debris that originate from chemical reaction between the abraded surface
substances and the external environment. The oxidation products consist of blackish aluminium
oxide - Al2O3. From fretting wear zones sometime originate fretting-fatigue phenomena due to the
applied axial forces [5]. Fretting fatigue phenomena are influenced by many parameters: wideness
of oscillations, value and distribution of contact pressure, type of wire material, condition of contact
All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the
written permission of the publisher: Trans Tech Publications Ltd, Switzerland, www.ttp.net. (ID: 129.132.210.6-11/07/07,09:40:17)
6
Advances in Fracture and Damage Mechanics VI
surfaces, friction forces and stresses nearby surface, cycling frequency, lubrication, temperature and
environment aggressiveness [6]). Three different fretting fatigue/wear systems can be described: (i)
Stick regime - this regime occurs when contact pressure is high and slips are short; wear is very low
and a great number of cycles are necessary to originate cracks. [6]. Considering electrical cables
this case occurs typically inside clamps. (ii) Slip regime - this regime occurs when contact pressure
are low and slips are long; wear is very high but nucleation of cracks is unusual. This system is
typical of cable zone far away from clamps. (iii) Mixed regime - this regime is in the middle of the
other two described. This system is the most critical for the early cracks initiation, as very short
cracks (less than 100µm) may be observed after a very short number of cycles.
Many factors may influence fretting fatigue phenomena in cables: (i) if bending wide increases,
cable life decreases because the slips number increase causes a faster damaging of the contact
surfaces; (ii) clamp type may influence contact condition among strands and between clamp and
cable inducing different type of loads (and in the same way different fretting-fatigue system)[6];
(iii) cable tension seems to have a scarce effect on fretting fatigue phenomena if it is included
between 15%-40% of tensile strength of the cable [7]; (iv) friction coefficient among strands may
influence life of cables. In fact presence of lubricant leads to reduction of friction coefficient and in
the same way to reduction in oxidation of particles produced by slips. Lubrication is useful above
all in fretting fatigue slip and mixed regime and where slips are the main cause of early cracks
initiation [8].
The studied case
The analysis concerns an overhead transmission line of 400kV (twin conductors) installed in
Touggourt Biskra (Algeria); The damaged conductors were operating in-service only for six months
but without any spacers or dampers. The investigated conductors after the failure are shown in
figure 1.
Three different types of conductors have been taken as experimental samples: the in-service broken
conductor, another in-service damaged conductor and a new conductor from warehouse as terms of
comparison. The aim of the study is to verify the conformity of the elements to the standards, to
identify the root cause of failure and to propose suggestions to avoid future problems in the
transmission line.
Cables are made of aluminium alloy (AAAC type), they are formed by 61 wires (3,45 mm in
diameter) rolled around a central wire in four layer with the following sequence: 1 central wire, 6
wires first layer, 12 wires second layer, 18 wires third layer, 24 wires fourth layer. Nominal cable
diameter is 31,05 mm with a specific weight of 1629 kg/km.
Visual analysis
Visual examination of the in-service broken cables highlights the presence of various wire rupture
in the third layer and partly in the fourth one: the fretting damage is evident nearby and at the point
of the hang clamp. Moreover in the damaged zone is evident the presence of blackish debris due to
the presence of aluminum oxide (Al2O3). (see Figure 1).
Figure 1 - The examined damaged cables: presence of aluminium oxides is evident in the dark grey zones
near the clamping region.
Key Engineering Materials Vols. 348-349
7
In order to fully analyse the third layer, some wires of the fourth layer were removed: the observed
damage show continuous elliptical marks at the same distance as the fourth layer pattern (see Figure
2). The damage is due to the stress induced by the clamping force and the wear caused by rubbing
among wires. The elliptical marks are present on the wires starting from clamp to 1,5-2m.
Figure 2 - Examples of damage on cables: white arrows show marks, white circle shows broken wires
In-service damaged cable does not have broken wires, but there is systematic presence of elliptical
marks (even if not so deep), and blackish debris (aluminum oxides). New cable does not show any
type of damage.
Observation of fracture surface
All samples were taken by manual cutting hacksaw to avoid thermal changes nearby cutting point.
All fracture surfaces have been cleaned in an acetone solution [9]. Macroscopic examination shows
two type of damage: longitudinal marks (from now type A damage) and elliptical marks sloped 30°
to longitudinal axis of wire (from now type B damage). This two kinds of damage depend on the
contact modes between wires and the loading condition: A type damage is related to line contact
between wires of the same layer and B type damage is the typical inter-layer point contact between
wires from different layers (see Figure 3).
Type A damage
Type B damage
Figure 3- Examples of the type A damage (started from longitudinal marks) and of the type B damage
(started from marks with 30° slope)
Microscopic examination of the fracture surfaces shows a typical fatigue fracture, mainly started
from A type damage, nevertheless wires have both type of damage. Only in few cases fatigue cracks
started from B type damage.
Chemical analysis
From damaged cable a piece of wire of 1 m length was taken and vacuum melted in order to obtain
a surface wide enough to carry out a quantometric analysis. Results are shown in table 1. Chemical
analysis respects requirements of standard EN AW 6101-A.
1
Si
0,629
Fe
0,238
Cu
0,005
Mn
<0,001
Mg
0,558
Cr
0,005
B
-
Ti
-
Zn
0,05
Table 1 - Quantometric chemical analysis of the damaged cable
V
0,005
Al
Bal.
8
Advances in Fracture and Damage Mechanics VI
Tensile tests
Samples of wires were taken from three type of cables (in-service broken cable, in-service damaged
cable, new cable) in a random way, unrolled manually. Results are indicated in the table 5.
In-service broken cable
Layer 1st
Layer 2nd
Layer 3rd
Layer 4th
Test
Rm
A%
Rm
A%
Rm
A%
Rm
A%
[MPa]
[MPa]
[MPa]
[MPa]
Mean.
318
7,4
325
5,9
325
7,4
329
6,9
Std. dev.
13,78
0,5
6,98
0,66
6,93
0,64
11,97
0,76
In-service damaged cable
Mean
336
10,13
341
7,95 339,25
6,84
339,5
7,05
327,3
6,84
Std. dev.
--6,38
1,03
13,25
0,86
8,39
0,34
12,69
0,96
New cable
Av.
338
9,32
318,8
7,69
330,7
6,82
319,7
7,17
330,2
8,88
Std. dev.
--10,05
0,74
5,85
1,52
7,97
0,62
7,27
0,79
Table 5 - Mechanical characteristic of wires taken from the in-service broken cable, from the
in-service damaged cable and from the new cable.
Central wire
Rm
A%
[MPa]
327
8,1
---
Mechanical properties of wires satisfy all requirements of CEI/IEC 1089/91 and CEI/IEC 104
standards about wires of diameter less than 3,5mm type B. Moreover three cables are identical,
from a mechanical point of view. These results have excluded the presence of aluminium wire
defects.
Conclusions
The studied electrical conductors have failed due to fretting fatigue phenomena. Static characteristic
of the studied cable scarcely influence fretting fatigue phenomena; instead operating conditions
have great influence (i.e. the elastic deformation caused by vibration of the cable).
The observation of the fracture surface confirms fretting fatigue failure: starting form superficial
longitudinal marks, fatigue cracks take origin and lead to various wire rupture thus drastic reduction
in the transmission line service life can be expected. Fracture surfaces are perpendicular to wire axis
and fretting fatigue system of failure is mainly governed by relative movements along line contact
between wires of the same layer (type A contact). Tensile test and chemical analysis show that
conductors satisfy standard prescription for 6101 alloy. Fretting fatigue failure can not be attribute
to alloy defects or cable anomalies; this is evident also considering that the dynamic damage is
limited to the clamp region. Failure was caused by aeolian vibration: the use of lubricant is not so
effective (as it is a fretting fatigue stick system failure); instead the application of dampers may be
useful.
Bibliography
[1] Internet site www.deoragroup.com/dwnm.htm .
[2] R.K.AGGARWAL, A.T.JOHNS, J.A.S.B.JAYASINGHE, W.SU, Electric Power System Research 53
(2000) pp.15-22.
[3] F.E.KRUEGER, in Failure Analysis and Prevention, Metals Handbook, Vol.10, 8th Ed., ASM
International, pp.154-160
[4] Z.R.ZHOU, A.CARDOUT, S.GOUDREAU, M.FISET, Tribology International Vol.29 No.3 (1996)
pp.221-232.
[5] T.C.LINDLEY, Int. J. Fatigue Vol.19, Supp. No.1 (1997) pp.S39-S49.
[6] C.R.F.AZEVEDO, T.CESCON, Engineering Failure Analysis 9 (2002) pp.645-664.
[8] Z.R.ZHOU, M.FISET, A.CARDOU, L.CLOUTIER, S.GOUDREAU Wear 189 (1995) pp.51-57.
[9] S.NISHIDA, Failure analysis in engineering applications, Butterworth-Heinemann, Oxford (1992) p.8.