Anda di halaman 1dari 19

International Journal of Fatigue 30 (2008) 22202238

Contents lists available at ScienceDirect

International Journal of Fatigue


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

Axial fatigue design of sheathed spiral strands in deep water applications


Mohammed Raoof *, Timothy J. Davies
Civil and Building Engineering Department, Loughborough University, Loughborough, Leicestershire, LE11 3TU, United Kingdom

a r t i c l e

i n f o

a b s t r a c t
Using the previously reported axial fatigue model of Raoof, and by carrying out extensive theoretical parametric studies on large diameter (i.e. realistic) sheathed spiral strand constructions which very nearly cover the full manufacturing range of lay angles (with the lay angle being the primary geometrical parameter controlling the strand axial fatigue performance), a new set of design SN curves for sheathed spiral strands in deep water applications have been developed. These SN curves provide a very simple means of predicting the axial fatigue life of sheathed spiral strands (under uniform cyclic loading) to rst outermost (or innermost) layer wire fracture, both at (or in the vicinity) of the end terminations, as well as in the free-eld (i.e. away from the detrimental effects of end terminations), with the sealed strands being subjected to simultaneous actions of a wide range of external hydrostatic pressures, such as those in deep-sea conditions with water depths of up to 2000 m. It has been shown that, although the fatigue life to rst innermost layer wire fracture is largely unaffected by the magnitude of the applied external hydrostatic pressure, the fatigue life to rst outermost layer wire fracture may be signicantly reduced (cf. corresponding in-air conditions) under the inuence of sufciently high levels of external hydrostatic pressure. The proposed design SN curves have been compared with others (which are based on tests under in-air conditions) and it has been shown that the application of such previously available in-air design SN curves to sheathed spiral strands in deep water applications may lead to unsafe designs. 2008 Elsevier Ltd. All rights reserved.

Article history: Received 23 February 2007 Received in revised form 18 April 2008 Accepted 9 May 2008 Available online 16 May 2008 Keywords: Spiral strands Hydrostatic pressure Sheathing Cyclic loading Axial fatigue Offshore platforms

1. Introduction The integrity of many major cable supported structures, in both onshore and offshore applications, is strongly dependent upon the cable anchorage systems, which are usually expensive to install, and, if it becomes necessary, very costly and difcult to replace. Since the mid-1970s there has been a signicant increase in the size of helically wound steel cables (spiral strands and/or wire ropes) being used, particularly in the offshore industry, where a spiral strand (or strand) is a collection of a number of helically laid wires wound in a number of concentric layers around a central straight King wire (or an equal lay core, whereby the core is composed of several wires), while a wire rope is composed of a number (typically six) of spiral strands helically wound around a common straight core, with the core either being made of bres or composed of an independent wire rope (IWRC). The cost of conducting large scale tests on steel cables is considerable and, particularly in the case of fatigue tests, very time consuming. Simply scaling up the cable diameters, via extrapolation of the orthodox designs, is (in the absence of a sound theoretical understanding) a risky process. To address this, Raoof [1,2] has used an extension of the
* Corresponding author. Tel.: +44 1509 222610. E-mail address: m.raoof@lboro.ac.uk (M. Raoof). 0142-1123/$ - see front matter 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijfatigue.2008.05.002

previously reported orthotropic sheet theoretical model of Raoof and Hobbs [3] (in connection with the patterns of interwire/interlayer contact forces throughout a spiral strand) to develop a theoretical model for predicting the axial fatigue life of the spiral strands to rst outermost (or inner) layer wire fractures, both at the xed end and/or away from the detrimental effects of end terminations - i.e. in the free-eld. Large scale experimental data, on more than 10 different spiral strands with outside diameters ranging from 25 mm to 127 mm, as tested by a number of independent institutions, have provided ample support for the general validity of Raoofs theoretical predictions [1,2,46]. Based on Raoofs theoretical axial fatigue model, design SN (fatigue stress-number of cycles to failure) curves, which take the construction details of large diameter multi-layered spiral strands into account, have been produced [7], which are applicable to any spiral strand construction, and enable one to design against rst outermost (or innermost) layer wire fracture at either the xed end termination or in the free-eld, where the term innermost layer excludes the central straight King wire (or, indeed, the helical wires forming an equal lay core). These design SN curves were produced by conducting extensive theoretical parametric studies on three different 127 mm outside diameter spiral strands with lay angles of 12, 18 and 24 with the specic construction details for these strands provided to the rst author by the, then,

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

2221

Nomenclature D d Ksi Kbi N n R r S0 S01 Sa wire diameter spiral strand outside diameter stress concentration factor in trellis contact points of layer i in a spiral strand 1/Ksi number of cycles to rst wire fracture number of layers of helical wires in a spiral strand axial load range (expressed as a percentage of the ultimate breaking load (U.B.L.)) helix radius endurance limit for the wire material mean axial strain on the spiral strand axial stress range Sei Sm Smax Smin Sult endurance limit of the helical wires in layer i of a spiral strand mean axial stress maximum axial stress minimum axial stress ultimate tensile strength of the wire material lay angle Poissons ratio for the wire material the nominal axial stress in the helical wires in layer i of a spiral strand maximum effective Von-Mises stress at the trellis contact points of layer i in a spiral strand

a m
 0i r  0maxi r

international wire rope manufacturer Bridon Ropes. Raoofs design SN curves were compared with other (purely empirical) design SN curves which, at the time, were commonly referred to in the literature; namely those of Tilly [8], API [9] and Chaplin [10]. In line with the earlier arguments reported elsewhere [11], it was found that, in certain cases, particularly the API recommendations provided unduly unconservative estimates of axial fatigue life. In addition, it was argued by Raoof [7] that, in the context of strand axial fatigue, the lay angle was the main (rst order) geometrical parameter. Back in the early 1980s, cable manufacturers offered high density and supposedly impermeable thick polythene sheaths (with the polythene jacket typically having a thickness of 15 mm or more) to protect large diameter spiral strands against corrosion in offshore applications. Raoof [12] theoretically showed that a sufciently high external hydrostatic pressure can (in the ever-presence of substantial air-lled voids inside the internally lubricated cables) signicantly inuence the patterns of interwire/interlayer contact forces in sheathed spiral strands. It was also argued that the application of an external hydrostatic pressure on a sealed spiral strand will increase its levels of interwire/interlayer contact stresses with a consequent associated marked effect on a sealed strands fatigue life: in Ref. [2], using a realistic 39 mm outside diameter spiral strand, it was demonstrated (theoretically) that substantial increases in a sealed strands trellis contact patch stresses can lead to signicant reductions in its axial fatigue life in long term applications. The purpose of the present paper is to provide a more complete set of axial fatigue design SN curves than those reported by Raoof [7], which cater for the effects of an external hydrostatic pressure on sheathed spiral strands, and are believed to be of particular importance in deep water offshore platform applications: it is, perhaps, worth mentioning that Raoofs previous work [7] only related to in-air axial fatigue conditions. As a pre-requisite to this, the following section presents a brief account of certain aspects of the axial fatigue performance of helically wound steel cables (spiral strands and/or wire ropes), which are of direct relevance to the subsequent development of design SN curves for such elements. 2. Practical considerations The criterion of fatigue failure for spiral strands and ropes is of necessity more complex than that applied to continuous structures where crack length measurements or a simple observation of loss of integrity may sufce. Discard criteria for helically wound steel cables are based on a mixture of past experience and personal preferences; as well as national or international standards, depending on the type of cable application. The occurrence of an unacceptable

number of wire breaks is by far the most common measure adopted for spiral strand fatigue damage assessment. This, however, is not free from serious pitfalls. Wire breakages can both be internal and/or external, and although some progress has already been made, reliable methods for detecting internal wire failures (under service conditions) in very long helically wound steel cables such as those used in offshore applications (which can be hundreds of meters long) are still to be developed [13]. The use of a standard length for counting the number of wire breakages is essential because multiple breaks can take place along any individual wire due to signicant levels of interwire friction [14]. The works of Stonesifer and Smith [15] and Smith et. al. [16] indicate that wire break density in rope is not a simple indicator of residual tensile strength, and prior knowledge of load history (and associated rope axial compliance at various stages) is necessary for a reasonable estimation of the residual strength. Much emphasis has, over the past 2030 years, been placed on suitable forms of discard criteria based on the remaining fatigue life (or strength) of the cable. It is now well established that depending on the type of cable construction and nature of application, the inuence of broken wires on the strength of a cable is not directly equivalent to a loss of area of steel: the number and distribution of wire breaks around a cable cross-section and also along its length are both important. This, in turn, depends on the type of cable construction and its state of internal lubrication. As mentioned previously, formation of multiple breaks along any individual wire is a relatively common occurrence and the performance of a wire rope or spiral strand is usually not affected by an occasional broken wire in the cable: with sufcient friction, a broken wire will be capable of developing its full share of the load in a relatively short length, often referred to as the recovery length [17,18]. Obviously, multiple wire breaks in a relatively short length of the cable may have a signicant inuence on the cables load carrying capacity [14]. Very often, cables are kept in service until they either break or are rejected by visual inspection. Even under laboratory conditions, external wire breakages have sometimes been assumed to reect the degree of actual fatigue damage. The work of references [15,19,20] conrms that the number of inner wire breaks can be a very signicant proportion of the total wire breakage density. End terminations can often concentrate very signicant numbers of wire fatigue breakages in their vicinity e.g.[21]. Metcalf and Matanzo [22], have performed fatigue tests on various types of rope constructions with different terminations. In particular, they emphasize the importance of the quality of workmanship. Hanzawa et. al. [19] also carried out axial fatigue tests on large diameter ropes and locked coil strands with zinc and epoxy lled sockets. Both of these references illustrate the generally, but not

2222

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

infallibly, superior fatigue performance of epoxy resin compared with zinc. The occurrence of premature failures such as socket eye breakage and sudden pull-out is reported by Hobbs and Ghavami [21]. In published work, useful information such as cable construction details and grade of wire are often not given. Moreover, each specic test result reported in the available literature, refers to a different cable construction with different wire materials, conditions of testing, specimen specication (i.e. length and end termination) and fatigue failure criteria. In view of this rather unsatisfactory state of affairs, it is not, then, surprising to nd a very large degree of variation among the experimental SN curves reported in the literature, with the attempts, in recent years, at codifying the SN curves having invariably taken the form of suggesting lower-bound curves to published experimental SN data. In view of the growing test database on steel-cable axial fatigue performance, the statistical treatment of the test data has, fairly recently, enjoyed some attention [23,24]; the subject area is, however, still in its infancy and much remains to be done. Central to the validity of these statistical treatments is an appropriate choice of the underlying distribution function. Various distributions such as normal, log-normal and Weibull have, in the past, been assumed. In a publication by Raoof and Hobbs [24], by examining a statistically acceptable population of axial fatigue data for specimens tested under nominally identical conditions, it was shown that the Gumble distribution (as opposed to the Weibull or normal types) was the preferred one for the statistical analysis of axial fatigue test data on helically wound steel cables. Raoof and Hobbs [24] argue that more attention should be paid to the design of cable fatigue tests. Because there are termination effects, obviously when wires fail at the terminations and arguably in the rope lengths adjacent to the terminations affected by disturbances during the termination process, there is a place for termination tests and for tests on longer specimens, to determine the free-eld cable behaviour. The latter might be facilitated by high-performance terminations: the objective is to combine termination data and cable data extrapolated to service lengths with condence, condence that can only come with statistically signicant numbers of test results and an understanding of their distribution. Raoof and Hobbs [24] also address the question of the minimum acceptable length of test specimens to be used for axial fatigue tests on wire ropes and spiral strands when resources are limited. Based on theoretical estimates of the recovery length of broken wires in spiral strands as reported by Raoof and Kraincanic [17], a minimum length of test specimens of about 10 lay lengths is recommended for axial fatigue tests on spiral strands, with the length of specimens used for obtaining the axial fatigue test data reported in the available literature being (in the vast majority of cases) considerably less than 10 lay lengths. 3. Theory Using the orthotropic sheet model of Raoof and Hobbs [3], Raoof [12] presents a detailed account of the appropriate steps for nding the pattern of interwire/interlayer normal contact stresses in axially pre-loaded sealed multi-layered spiral strands experiencing external hydrostatic pressures. In particular, Raoof [12] has shown that high external pressures, such as those associated with deep water applications, can signicantly increase the magnitudes of interwire/interlayer normal contact forces. Later work by Raoof [2] has, on the other hand, demonstrated that such increased levels of interlayer (trellis) contact forces can signicantly affect (i.e. reduce) the sealed spiral strands axial fatigue life. Space limitations do not allow a full reporting of the rather lengthy formulations here, and instead, in the following, the salient features of the theoretical axial fatigue model of Raoof, as applied to sheathed spiral

strands, will be briey discussed for completeness: this will, then, enable the reader to better understand (and appreciate) the developments as reported in the subsequent sections for the full details of the presently used theoretical model for predicting the axial fatigue life of axially pre-loaded multi-layered spiral strands, the interested reader may refer to Refs. [1,2]. Very briey, using the orthotropic sheet theoretical model, reliable estimates of the interlayer (i.e. trellis) normal contact forces (and stresses) throughout an axially pre-loaded multi-layered sheathed spiral strand experiencing a given level of external hydrostatic pressure, can be obtained [12]. In-air axial fatigue tests on multi-layered spiral strands suggest that individual wire failures are largely located over the trellis points of interlayer contact and, assuming that this is as a result of high-stress concentration factors in these locations, the magnitudes of maximum effective  0max , over all the trellis points of contact Von-Mises stresses, r throughout the multi-layered sheathed strand, for a given mean axial load and level of external hydrostatic pressure, are calculated with the stress concentration factor, Ksi, for the trellis contact points in layer i, then dened as [1,2]

K si

 0maxi r  0i r

 0i , is the nominal axial stress in the helical wires associated where, r with the trellis contact points in layer i, as calculated using the method developed by Raoof [12], with the proviso that, due to typographical errors, Eq. (15) in Ref. [12] should be replaced by

X Ri X MSi 2PMSi cos bi

where Ref. [12] gives the denition of the various terms in Eq. (2), with all the results reported by Raoof and his associates to date being based on the above equation rather than Eq. (15) in Ref. [12]. Raoof [1,2] deals with the topic of strand axial fatigue at some length. Using the values of Ksi in conjunction with axial fatigue data on single wires, a theory has been developed which predicts the axial fatigue of sheathed strands (under constant amplitude cyclic loading) from rst principles, with the direct (axial) stress ranges in the individual helical wires in conjunction with the stress concentrations over the trellis points of interlayer contact causing fatigue fractures in the individual helical wires. For carbon steel wires, the SN curve (in the absence of any interwire contact(s)) possesses an endurance limit, S0 , below which no damage occurs. Traditionally, the magnitude of S0 is compared to the ultimate wire tensile strength, Sult: tests on single galvanised wires suggest an approximate value of S0 = 0.27Sult [25]. The reduced magnitude of the endurance limit, Sei, for the trellis contact points in layer i (i = 1, n) of a multi-layered spiral strand, the individual wires of which not only experience cyclic variations in axial tension but are also subjected to severe states of interlayer contact stresses, may be dened as

Sei K a K bi S0

where Sei takes interwire contact and fretting plus surface conditions and size effects, etc., into account, with K bi K1si , and Ka being a constant, the wire surface constraint factor (or, as previously called in our earlier publications (e.g. [1,2]), surface nish factor). The values of the parameters Sei, are then used to produce the SN curves for fatigue life to rst outermost (with i = 1) (or inner) wire fractures in sheathed spiral strands using the SN curves available in the literature for axial fatigue life of individual wires of a given grade [1,2], with the proviso that i = n relates to the innermost layer, with n = total number of layers of wires in the strand (excluding the core). Fig. 1 compares the theoretical predictions with experimental data for a 51 mm outside diameter spiral strand [1]. The criterion for fatigue initiation has been the occurrence of rst wire failure

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

2223

40
Theory

Load range/maximum breaking load, %

Experimental

30

20

10

0 1.0E+05 1.0E+06 1.0E+07 Life to first outer layer wire fracture, cycles 1.0E+08

Fig. 1. Axial fatigue of 51 mm outside diameter spiral strand comparison of the theory and test data.

in the outermost layer. A fairly signicant degree of scatter was found in the experimental data which may be covered by an (empirical) surface nish factor Ka in the range 0.5 6 Ka 6 1.0. The ultimate tensile strength of the wire material is Sult = 1640 N/ mm2. With the strands having epoxy resin end terminations, all the initial wire failures in this strand occurred away from the ends. However, as discussed elsewhere [24], for the end terminations to have no effect on the wire fractures remote from the ends, the minimum length of test specimens must be around 10 lay lengths with the wire fractures occurring within the central region, which extends by 2.5 lay lengths on either side of the middle of the test specimen (i.e. within the central portion with a length of 5 lay lengths). It then follows that due to the total length of the tested 51 mm outside diameter strands being signicantly less than 10 lay lengths, even for the wire fractures away from the ends, certain test data points in Fig. 1 have been inuenced by the detrimental end effects, with the correlation between the theory and such test data suggesting Ka = 0.5 as an appropriate factor in the presence of end effects. Otherwise, for wire fractures which happen far enough away from the ends not to be inuenced by end effects, one may assume Ka = 1.0. This, then, provides the reader with an insight into
5000 Theory (Ka= 0.5) Experiment (N = Number of events) (N) 4000

the role of the parameter Ka in the proposed theoretical model, with the appropriate range of values for Ka (i.e. 0.51.0) having been established by tting the theoretical SN curves to the test data in Fig. 1. The theoretical predictions have been supported by a very extensive set of large scale in-air test data relating to spiral strands of more than 10 different diameters, d, covering the wide range 25 mm 6 d 6 127 mm and tested by Bridon Ropes (private communication), Imperial College, London [21], Transport Research Laboratories (private communication), and National Engineering Laboratories (private communication) in the UK, and the University of Alberta in Canada [26], with lay angles, a, within the wide range 11 6 a 6 21, and wire diameters, D, covering the range 3 mm 6 D 6 7.10 mm. These results very nearly encompass the presently adopted geometrical manufacturing limits. In all cases, the correlations between the theoretical predictions and such an extensive set of large scale test data has been very encouraging [1,2,46]. For the present purposes, Figs. 2 and 3 present the encouraging correlations between some new (i.e. not previously reported) large scale axial fatigue test data and the theoretical predictions based

Axial Load Range (kN)

3000 (2) (1) Outermost Layer

2000

1000 Innermost Layer 0 1.E+05

1.E+06 Fatigue Life, Cycles

1.E+07

Fig. 2. Axial fatigue of 164 mm outside diameter spiral strand comparison of the theory and test data.

2224

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

3000 Location of wire failures Free Field Socket

2500 Axial Load Range (kN)

Test No . 1 2

2000

1500 Number of visible broken wires 1000 (2) 500 Theory (Ka = 0.5): Innermost layer Outermost layer 0 1.E+05 1.E+06 Fatigue Life, Cycles 1.E+07 (2) (1) (2)

Fig. 3. Axial fatigue of 102 mm outside diameter spiral strand comparison of the theory and test data.

on Raoofs model (with an assumed Ka = 0.5), which relate to 164 and 102 mm outside diameter (i.e. very large) spiral strands, respectively. Tables 1 and 2 give the construction details for these strands, with calculated breaking loads of 22,370 kN and 8770 kN for the 164 mm and 102 mm strands, respectively, and both spiral strands having wires with a tensile grade of 1520 N/mm2. The 164 mm strand was subjected to one axial fatigue test with a mean axial load of 8045 kN and maximum and minimum cyclic axial loads of 9230 kN and 6860 kN, respectively (i.e. an axial load range

of 2370 kN), while two tests with an axial load range of 940 kN were carried out on the 102 mm strand, both of which employed maximum and minimum cyclic axial loads of 3320 kN and 2380 kN, respectively (i.e. a mean axial load of 2850 kN). As regards the test on the 164 mm strand, all the wire failures happened at the socket, with no free-eld failures taking place. Importantly, unlike our previously used axial fatigue test data in Refs. [1,2,46] which (in the vast majority of cases) had adopted either the total collapse or the number of cycles to initial wire fracture(s) in the

Table 1 Construction Details for the 164 mm outside diameter spiral strand with an equal lay core Layer 1 2 3 4 5 6 7 8 9 10 Core Number of wires, n 72 66 60 54 48 42 36 30 24 18 14 7 7 7 1 Lay direction RH LH RH LH RH LH RH LH RH LH Wire diameter, D (mm) 6.50 6.50 6.50 6.50 6.50 6.50 6.60 6.60 6.60 6.60 6.00 3.70 4.85 4.95 6.70 Lay angle, a (deg) 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 17.99 12.98 12.15 7.53 Helix radius (theo.), r (mm) 78.34 71.81 65.29 58.77 52.24 45.72 39.80 33.18 26.56 19.95 Net steel area, Ani(mm2) 2512.28 2302.92 2093.56 1884.21 1674.85 1465.50 1295.09 1079.24 863.39 647.54 340.58 69.64 120.82 131.25 35.25

Table 2 Construction details for the 102 mm outside diameter spiral strand with an equal lay core Layer 1 2 3 4 5 6 Core Number of wires, n 48 42 36 30 24 18 12 6 6 6 1 Lay direction RH LH RH LH RH LH Wire diameter, D (mm) 6.00 6.00 6.00 6.05 6.05 6.05 6.15 3.60 4.60 4.35 4.45 Lay angle, a (deg) 17.00 17.00 17.00 17.00 17.00 17.00 15.92 10.85 10.02 5.72 Helix radius (theo.), r (mm) 48.06 41.97 35.98 30.00 24.02 18.04 Net steel area, Ani (mm2) 1419.18 1241.78 1082.20 901.83 721.47 541.10 317.00 57.86 95.22 87.84 15.55

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

2225

outermost layer (i.e. visible wire failures) as the criteria for fatigue failure, in the course of the tests on the 164 mm strand, use was made of certain non-destructive (acoustic) monitoring devices which could reasonably detect the occurrence of even internal wire fractures (events). Indeed, Fig. 4 presents a plot of the experimentally recorded number of events against the number of cycles, for the 164 mm diameter strand; clearly demonstrating the progressive nature of the fatigue failure for such a highly redundant element. Included in Fig. 4 are also two test data points which indicate occurrence of two visible (i.e. external) wire failures. It is noteworthy that, according to the plots in Fig. 4, a number of

internal wire fractures have occurred before any visible wire fractures are found in the outermost layer, with the number of visible wire fractures over the fatigue life span of the strand being only a small fraction of the total number of wire failures. In other words, a mere visual inspection of the strand can result in grossly unreliable estimates of the true remaining (residual) strength of the cable. On the other hand, in view of the highly redundant nature of the strand, the use of the initial wire fracture as fatigue failure, will in practice, leave a highly desirable margin of safety against total collapse. Bearing this in mind, in Fig. 2, only the test data for the rst three acoustic events are given with these occurring at

10 Acoustic Visible

Events

0 0 5 10 Cycles (10 )
Fig. 4. Progression of wire failures in the 164 mm outside diameter spiral strand with increasing axial fatigue cycles.
5

15

20

Table 3 Construction details for the 127 (a = 12) mm outside diameter spiral strand with an equal lay core Layer 1 2 3 4 5 6 7 8 Core Number of wires, n 56 50 44 38 32 26 20 14 7+7 7 1 Lay direction RH LH LH RH LH RH LH RH Wire diameter, D (mm) 6.60 6.60 6.60 6.60 6.50 6.50 6.50 6.60 4.00 and 5.20 5.20 7.10 Lay angle, a (deg) 12.00 12.00 12.00 12.00 12.00 12.00 12.00 12.00 8.57 and 8.03 4.92 Helix radius (theo.), r (mm) 60.17 53.73 47.29 40.85 33.89 27.56 21.23 15.15 Net steel area, Ani(mm2) 1958.67 1748.81 1538.96 1329.10 1085.58 882.03 678.49 489.67 85.05 and 144.33 147.02 39.59

Table 4 Construction Details for the 127 (a = 18) mm outside diameter spiral strand with an equal lay core Layer 1 2 3 4 5 6 7 8 Core Number of wires, n 54 48 42 36 31 25 19 14 7+7 7 1 Lay Direction RH LH LH RH LH RH LH RH Wire diameter, D (mm) 6.55 6.55 6.55 6.55 6.55 6.55 6.55 6.30 3.90 and 5.10 5.25 7.00 Lay angle, a (deg) 18.00 18.00 18.00 18.00 18.00 18.00 18.00 18.00 13.07 and 12.20 7.62 Helix radius (theo.), r (mm) 59.22 52.64 46.07 39.50 34.02 27.46 20.90 14.85 Net steel area, Ani (mm2) 1913.31 1700.72 1488.13 1275.54 1098.38 885.79 673.20 458.90 77.29 and 133.53 142.55 38.49

2226

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

Table 5 Construction Details for the 127 (a = 24) mm outside diameter spiral strand with an equal lay core Layer 1 2 3 4 5 6 7 8 Core Number of wires, n 54 48 42 36 30 24 18 14 7+7 7 1 Lay direction RH LH LH RH LH RH LH RH Wire diameter, D (mm) 6.40 6.40 6.50 6.50 6.60 6.60 6.80 6.10 3.90 and 5.10 5.25 7.00 Lay angle, a (deg) 24.00 24.00 24.00 24.00 24.00 24.00 24.00 24.00 17.89 and 16.75 10.58 Helix radius (theo.), r (mm) 60.23 53.54 47.58 40.79 34.53 27.64 21.38 14.94 Net steel area, Ani (mm2) 1901.58 1690.29 1525.58 1307.64 1123.49 898.79 715.57 447.87 72.07 and 125.56 143.94 38.49

50 45 40 Load Range (% U.B.L.) 35 30 25 20 15 10 5


= 12, 18 & 24 degrees Water Depth = 0 m S1 = 0.002867 U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

50 45 40 Load Range (% U.B.L.) 35 30 25 20 15 10


= 12, 18 & 24 degrees Water Depth = 1500 m S1 = 0.002867 U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(a)

(d)

0 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

0 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

Fatigue life, Number of Cycles (log scale)

Fatigue Life, Number of Cycles (log scale)

50 45 40 Load Range (% U.B.L.) 35 30 25 20 15 10 5


= 12, 18 & 24 degrees Water Depth = 500 m S1 = 0.002867 U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

50 45 40 Load Range (% U.B.L.) 35 30 25 20 15 10


= 12, 18 & 24 degrees Water Depth = 2000 m S1 = 0.002867 U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(b)

(e)

0 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

0 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

Fatigue life, Number of Cycles (log scale)

Fatigue life, Number of Cycles (log scale)

50 45 40 Load Range (% U.B.L.) 35 30 25 20 15 10 5


= 12, 18 & 24 degrees Water Depth = 1000 m S1 = 0.002867 U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(c)

0 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

Fatigue Life, Number of Cycles (log scale)

Fig. 5. Composite plots of the theoretical predictions of strand fatigue life to rst wire fracture for all the layers at a given level of mean axial strain, S01 0:002867, varying magnitudes of lay angle, a, and at varying water depths; (a) 0 m; (b) 500 m; (c) 1000 m; (d) 1500 m; (e) 2000 m Ka = 1.0.

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

2227

838,497, 842,276 and 904,648 cycles. It should also be noted that Raoofs theoretical model can predict the rst wire fractures in not just the outermost layer, but also in any internal layer (with the exception of the core wire(s)) of a multi-layered spiral strand, with the innermost and outermost layers showing the lowest and highest axial fatigue lives, respectively, and the SN curves for the intermediate internal layers invariably lying in-between these two limits. With this borne in mind, in Fig. 2, it is very encouraging that the initial three test events are found to lie in-between the two theoretical SN curves relating to the outermost and innermost layers, with the assumed Ka = 0.5 for these theoretical plots correctly predicting the actual location of the initial failures which, in the tests were found to be at the socket.

Turning now to the two tests (i.e. 1 and 2) on the 102 mm diameter strand, Fig. 3, in test 1, by 2 106 cycles, two visible free-eld failures had occurred, while in test 2, the rst two failures happened at the socket, with these followed by two more in the freeeld, and a fth wire failure occurring at the socket. In Fig. 3, the differences between the test data for nominally identical axial fatigue tests is noteworthy, and it is encouraging to nd the test data lying in-between the theoretical SN curves (with Ka = 0.5) for the outermost and innermost layers of the 102 mm diameter strand. Bearing the above in mind, there is, therefore, strong support for the general applicability of Raoofs proposed theoretical model for in-air conditions, which can predict both initial outer and/or inner wire breakages with the initial inner wire fractures generally hav-

50 45 40 Load Range (% U.B.L.) 35 30 25 20 15 10 5


= 12, 18 & 24 degrees Water Depth = 0 m S1 = 0.002867 U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

50 45 40 Load Range (% U.B.L.) 35 30 25 20 15 10


= 12, 18 & 24 degrees Water Depth = 1500 m S1 = 0.002867 U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(a)

(d)

0 1.0E+04

1.0E+05

1.0E+06

1.0E+07

0 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue life, Number of Cycles (log scale)

Fatigue life, Number of Cycles (log scale)

50 45 40 Load Range (% U.B.L.) 35 30 25 20 15 10 5


= 12, 18 & 24 degrees Water Depth = 500 m S1 = 0.002867 U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

50 45 40 Load Range (% U.B.L.) 35 30 25 20 15 10


= 12, 18 & 24 degrees Water Depth = 2000 m S1 = 0.002867 U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(b)

(e)

0 1.0E+04

1.0E+05

1.0E+06

1.0E+07

0 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue life, Number of Cycles (log scale)

Fatigue Life, Number of Cycles (log scale)

50 45 40 Load Range (% U.B.L.) 35 30 25 20 15 10 5


= 12, 18 & 24 degrees Water Depth =1000 m S1 = 0.002867 U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(c)

0 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue Life, Number of Cycles (log scale)

Fig. 6. Composite plots of the theoretical predictions of strand fatigue life to rst wire fracture for all the layers at a given level of mean axial strain, S01 0:002867, varying magnitudes of lay angle, a, and at varying water depths; (a) 0 m; (b) 500 m; (c) 1000 m; (d) 1500 m; (e) 2000 m Ka = 0.5.

2228

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

ing (for in-air conditions) a lower fatigue life than the outer wires [1,2]. As regards the axial fatigue performance of sheathed spiral strands experiencing external hydrostatic pressures, it should be noted that axial fatigue tests on a, say, 127 mm outside diameter and axially pre-loaded sheathed spiral strand (subjected to a mean axial load of, say, 0.500MN) with a length of, for example, 10 m under the simultaneous action of, say, 1000 m of water pressure is next to an impossible exercise, and the value of a reliable theoretical model capable of predicting the reductions (cf. in-air performance) in the axial fatigue lives of sheathed spiral strands in deep water applications, is then obvious. In what follows, based on an extensive series of theoretical parametric studies, simple to use SN curves will be developed for axial fatigue design of large

diameter sealed strands in deep-sea applications, which are aimed at the busy practicing engineers. 4. Theoretical parametric studies 4.1. Background As discussed elsewhere [7,27], following extensive theoretical and experimental work, the lay angle has been found to be the primary factor which controls a number of overall structural strand characteristics (including the axial fatigue performance). Bearing this in mind, in what follows, results are reported which are based on extensive theoretical parametric studies using Raoofs axial fatigue model [1,2] as applied to the constructional details for three

100
0.001 = S1 0.002 0.002867 0.004

100
0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

Load Range (% U.B.L.)

10

10

= 18.00 degrees Water Depth = 0 m U. B. L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(a)
1.0E+06 1.0E+07

= 18.00 degrees Water Depth = 1500 m U. B. L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(d)
1.0E+06 1.0E+07

1 1.0E+04

1.0E+05

1 1.0E+04

1.0E+05

Fatigue life to first outer layer fracture, Cycles

Fatigue life to first outer layer fracture, Cycles

100
0.001 = S1 0.002 0.002867 0.004

100
0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

10

Load Range (% U.B.L.)

10

= 18.00 degrees Water Depth = 500 m U. B. L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(b)
1.0E+06 1.0E+07

= 18.00 degrees Water Depth = 2000 m U. B. L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(e)

1 1.0E+04

1.0E+05

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue life to first outer layer fracture, Cycles

Fatigue life to first outer layer fracture, Cycles

100
0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

10

= 18.00 degrees Water Depth = 1000 m U. B. L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(c)
1.0E+06 1.0E+07

1 1.0E+04

1.0E+05

Fatigue life to first outer layer fracture, Cycles

Fig. 7(ae). Lower bound SN curves for the 127 mm outside diameter sheathed spiral strand (a = 18) based on the fatigue life to rst outermost layer wire fracture and subjected to a wide range of mean axial strains, S01 , and varying levels of water depth; (a) 0 m; (b) 500 m; (c) 1000 m; (d) 1500 m; (e) 2000 m Ka = 1.0.

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

2229

different sheathed 127 mm outside diameter (i.e. realistic) spiral strands with lay angles of 12, 18 and 24 (as designed and made available to the rst author, some years ago, by the then International wire rope manufacturerBridon Ropes). Tables 35 give the full construction details for these three strands which, amongst them, cover the full range of manufacturing limits for the lay angle, with each strand having an equal lay core and the same lay angle in all its layers while the other geometrical factors (such as number and diameters of the helical wires) for the three different strand constructions are kept very nearly the same. The assumed Youngs modulus for galvanised steel wires E = 200 kN mm2, and the Poissons ratio for wire material m = 0.28. The estimated ultimate breaking loads (U.B.L.) for all three strand constructions are assumed to

be equal to 13,510 kN, with the assumed ultimate tensile strength for the wire material Sult = 1520 N/mm2. It should, however, be noted that, although all the following numerical results are based on Sult = 1520 N/mm2 and U.B.L. = 13,510 kN, the nal recommended design SN curves are, based on Raoofs [7] arguments, of general applicability irrespective of the grade of wire material and the magnitude of U.B.L., because all the axial load (or stress) ranges in the proposed design SN curves have been non-dimensionalized with respect to the strand ultimate breaking load (or stress). As regards the effect of the mean axial load on the SN curves, Alani and Raoof [28] have shown that the previous argument by certain researchers, who have advocated the use of a Goodman

100 0.001 = S1 0.002 0.002867 0.004

100
0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

Load Range (% U.B.L.)

10

10

= 18.00 degrees Water Depth = 0 m U. B. L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(f)

= 18.00 degrees Water Depth = 1500 m U. B. L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(i)
1.0E+06 1.0E+07

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1 1.0E+04

1.0E+05

Fatigue life to first inner layer fracture, Cycles

Fatigue life to first inner layer fracture, Cycles

100
0.001 = S1 0.002 0.002867 0.004

100
0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

10

Load Range (% U.B.L.)

10

= 18.00 degrees Water Depth = 500 m U. B. L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(g)
1.0E+06 1.0E+07

= 18.00 degrees Water Depth = 2000 m U. B. L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(j)
1.0E+06 1.0E+07

1 1.0E+04

1.0E+05

1 1.0E+04

1.0E+05

Fatigue life to first inner wire fracture, Cycles

Fatigue life to first inner layer fracture, Cycles

100
0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

10

= 18.00 degrees Water Depth = 1000 m U. B. L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(h)
1.0E+06 1.0E+07

1 1.0E+04

1.0E+05

Fatigue life to first inner layer fracture, Cycles

Fig. 7(fj). Lower bound SN curves for the 127 mm outside diameter sheathed spiral strand (a = 18) based on the fatigue life to rst innermost layer wire fracture and subjected to a wide range of mean axial strains, S01 , and varying levels of water depth; (f) 0 m; (g) 500 m; (h) 1000 m; (i) 1500 m; (j) 2000 m Ka = 1.0.

2230

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

or Gerber approach for spiral strands in order to cater for the effect of mean axial load, when presenting axial fatigue results, does not lead to any advantages when compared to the use of axial stress range: it is shown that the traditional axial stress range (as opposed to an equivalent axial load range based on the Goodman or Gerber approaches) leads to less scatter of the axial fatigue data about the tted mean curve in the axial fatigue SN plots, and the alternative approach of using an equivalent axial load range for plotting the SN curves does not (despite the extra efforts involved) lead to any advantages. For the present purposes, therefore, all the following proposed SN curves are based on the strand axial stress range, Sa, with the mean axial stress, Sm, dened as

Sm

Smax Smin 2

where Smax and Smin are the maximum and minimum values of the externally applied axial stress, with

Sa Smax Smin

Based on extensive theoretical parametric studies reported elsewhere [7,29], it is suggested that for a given axial stress range, the larger the outer diameter of a spiral strand, the lower its axial fatigue life to rst outermost (or innermost) layer wire fracture. In other words, there is a size effect and for generally applicable design SN curves, one needs to carry out experimental and/or theoretical parametric studies on large diameter (preferably full-scale)

100
0.001 = S1 0.002 0.002867 0.004

100
0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

10

Load Range (% U.B.L.)

10

= 18.00 degrees Water Depth = 0 m U. B. L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(a)

= 18.00 degrees Water Depth = 1500 m U. B. L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(d)
1.0E+06 1.0E+07

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1 1.0E+04

1.0E+05

Fatigue life to first outer layer fracture, Cycles

Fatigue life to first outer layer fracture, Cycles

100
0.001 = S1 0.002 0.002867 0.004

100
0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

Load Range (% U.B.L.)

10

10

= 18.00 degrees Water Depth = 500 m U. B. L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(b)

= 18.00 degrees Water Depth = 2000 m U. B. L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(e)

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue life to first outer layer fracture, Cycles 100

Fatigue life to first outer layer fracture, Cycles

0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

10

= 18.00 degrees Water Depth = 1000 m U. B. L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(c)

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue life to first outer layer fracture, Cycles

Fig. 8(ae). Lower bound SN curves for the 127 mm outside diameter sheathed spiral strand (a = 18) based on the fatigue life to rst outermost layer wire fracture and subjected to a wide range of mean axial strains, S01 , and varying levels of water depth; (a) 0 m; (b) 500 m; (c) 1000 m; (d) 1500 m; (e) 2000 m Ka = 0.5.

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

2231

strands. Based on the work reported elsewhere [7,29], obtaining results based on a scaling factor of, say, 2 or 3 is, however, not unreasonable and (considering the usual scatter problem associated with axial fatigue data for spiral strands) may be adopted in practice for experimentation as an alternative approach with considerable nancial savings provided only that the lay angles are kept similar during the scaling process. For our present purposes, it should, therefore, be noted that although, over recent years, larger (e.g. 164 mm outside diameter) spiral strands have, indeed, been manufactured, the use of 127 mm diameter strand construction details, as given in Tables 35, is quite acceptable for developing the presently reported design SN curves, especially bearing in mind that strands with diameters signicantly larger than 127 mm are (at least for the present) rarely used in practice.
100
0.001 = S1 0.002 0.002867 0.004

Figs. 5(ae) and 6(ae) show the individual data points for all the layers of the three different 127 mm outside diameter spiral strands with lay angles of 12, 18 and 24, assuming a constant mean axial strain S01 0:002867, with Ka values of 1.0 and 0.5, respectively, and at various water depths ranging from 0 m to 2000 m. It is, perhaps, worth mentioning that, as explained previously, in the present terminology, the term innermost layer wire excludes the helical wires in the equal lay core of the three different strand constructions in Tables 35, with the layer 8 in Tables 35 being considered as the innermost layer throughout the presently reported parametric studies. A comparison of the plots in Figs. 5(ae) and 6(ae) suggests that the fatigue life, dened as the number of cycles to rst wire fracture for all the layers, ranging from the innermost to the outermost ones, is, for any given load
100
0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

10

Load Range (% U.B.L.)

10

= 18.00 degrees Water Depth = 0 m U. B. L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(f)

= 18.00 degrees Water Depth = 1500 m U. B. L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(i)

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue life to first inner layer fracture, Cycles

Fatigue life to first inner layer fracture, Cycles

100
0.001 = S1 0.002 0.002867 0.004

100
0.001 = S1 0.002867 0.002867 0.004

Load Range (% U.B.L.)

10

Load Range (% U.B.L.)

10

= 18.00 degrees Water Depth = 500 m U. B. L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(g)

= 18.00 degrees Water Depth = 2000 m U. B. L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(j)

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue life to first inner layer fracture, Cycles

Fatigue life to first inner layer fracture, Cycles

100 0.001 = S1 0.002 0.002867 0.004

Load Range (% U.B.L.)

10

= 18.00 degrees Water Depth = 1000 m U. B. L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

(h)

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue life to first inner layer fracture, Cycles

Fig. 8(fj). Lower bound SN curves for the 127 mm outside diameter sheathed spiral strand (a = 18) based on the fatigue life to rst innermost layer wire fracture and subjected to a wide range of mean axial strains, S01 , and varying levels of water depth; (f) 0 m; (g) 500 m; (h) 1000 m; (i) 1500 m; (j) 2000 m Ka = 0.5.

2232

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

range (expressed as a percentage of the U.B.L.), lower for an assumed value of Ka = 0.5 than for Ka = 1.0. It can be seen that for both sets of plots, particularly for smaller water depths, a signicant degree of scatter is exhibited. It should be noted that these plots all assume a constant S01 0:002867, and changing S01 from, for example, 0.00050 to, say, 0.004 will lead to an even larger degree of scatter. One may decide to draw a lower-bound curve to all such data for each given water depth (as shown in the gures) and dene these to be design SN curves for various depths of water: indeed, for the case of in-air conditions, this has been the way Chaplin [10] and Tilly [8] have produced their recommended SN curves based on purely experimental data which, incidentally, exhibited a somewhat similar large degree of scatter to those in, say, Figs. 5a and 6a, although these references adopted different failure criteria to be discussed later. As explained next, using Raoofs theoretical model, one is able to come up with more satisfactory recommendations.

4.2. Design recommendations Figs. 7(ae) and 7(fj) present plots (in loglog scale) of the load range (expressed as a percentage of the U. B. L.) against axial fatigue life, for the 127 mm outside diameter spiral strand (a = 18), to rst outermost and innermost wire fractures, respectively, with the assumed value of Ka in these plots equal to 1.0 i.e. the wire fractures associated with these plots are assumed to happen in the free-eld, away from the detrimental inuence of end terminations. These plots cover a wide range of cable mean axial strains, 0:001 6 S01 6 0:004. Figs. 8(ae) and 8(fj), show similar plots, corresponding to the 127 mm outside diameter spiral strand (a = 18), but for an assumed Ka value of 0.5 i.e. the wire fractures associated with these plots are assumed to be inuenced by the detrimental end effects. Each plot in Figs. 7(aj) and 8(aj) includes a lower-bound straight line to all the individual theoretical data

100
The theoretical lower bound to first outer layer fracture at 0 m water depth

100
The theoretical lower bound to first inner layer fracture at 0 m water depth

Chaplin API Tilly

Tilly

Load Range (% U.B.L.)

Load Range (% U.B.L.)

API

10

= 12 18

10

Chaplin

= 12 18 24

24

for the theoretical curves: U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(a)

for the theoretical curves: U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(b)

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

Fatigue Life, Cycles 100


The theoretical lower bound to first outer layer wire fracture at 0 m water depth

Fatigue Life, Cycles 100


The theoretical lower bound to first inner layer wire fracture at 0 m water depth

Load Range (% U.B.L.)

10
API

Load Range (% U.B.L.)

Tilly

10

API

= 12

Chaplin

Tilly

Chaplin

for the theoretical curves: U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

18 24

(c) 1.0E+07 1.0E+08

for the theoretical curves: U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

= 12 18

24

(d) 1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue Life, Cycles

Fatigue Life, Cycles

Fig. 9(ad). Comparison of alternative design SN curves for in-air conditions, based on various theoretical criteria for axial fatigue failure and different values of Ka: (a) and (c) Ka = 1.0 and 0.5, respectively, for the fatigue life to rst outermost layer wire fracture; (b) and (d) Ka = 1.0 and 0.5, respectively, for the fatigue life to rst wire fracture in the innermost layer.

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

2233

points which are (cf. Figs. 5(ae) and 6(ae)) found to exhibit less degree of scatter this, therefore, strongly suggests the merit in separating the results for each individual value of lay angle: in this way, more sensible lower-bound SN curves may be obtained with the individual data points relating to each lower-bound SN curve exhibiting reasonable scatter. The above exercise has been repeated for the other two 127 mm diameter spiral strands with lay angles of 12 and 24: space limitations do not allow presentation of the plots for all such cases here these have been fully reported elsewhere[29]. Sufce it to say that in all cases, the scatter of the individual theoretical data points (to which a lower-bound SN curve is added) is similar to those presented in Figs. 7(aj) and 8(aj). An examination of all the plots for the three different strand constructions with lay angles of 12, 18 and 24, suggests that increasing the lay angle tends to give a lower SN curve for all levels of hydrostatic pressure, for both Ka = 1.0 and 0.5, with

the axial fatigue life dened as the number of cycles to rst outermost or innermost layer wire fracture. Figs. 9(ad), 9(eh), 9(il), 9(mp) and 9(qt) present all the lower-bound SN curves, based on the present extensive theoretical parametric studies, assuming Ka values of 1.0 and 0.5, with fatigue failure dened as the number of cycles to rst outermost (or innermost) layer wire fracture, and at various levels of external hydrostatic pressures, equivalent to water depths of 0 m, 500 m, 1000 m, 1500 m and 2000 m, respectively. In each of all these gures, three theoretical lower-bound SN curves, corresponding to lay angles of 12, 18 and 24 are presented. Included in Fig. 9(ad) are the purely empirical lower-bound in-air SN curves, as recommended by API [9], Chaplin [10] and Tilly [8], for comparison purposes. Most importantly, the level of external hydrostatic pressure, within the range 0 m 2000 m, appears to have very little (if any) effect on the fatigue life to rst innermost

100
The theoretical lower bound to first outer layer fracture at 0 m water depth at 500 m water depth

100
The theoretical lower bound to first inner layer fracture at 0 m water depth at 500 m water depth

Load Range (% U.B.L.)

10

= 12

Load Range (% U.B.L.)

10

18 12 24 18

= 12 18 for the theoretical curves: U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm 24

for the theoretical curves: U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

24

(e)

(f)

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

Fatigue Life, Cycles 100


The theoretical lower bound to first outer layer wire fracture at 0 m water depth at 500 m water depth

Fatigue Life, Cycles 100


The theoretical lower bound to first inner layer wire fracture at 0 m water depth at 500 m water depth

Load Range (% U.B.L.)

10

Load Range (% U.B.L.)

10

= 12

12

for the theoretical curves: U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

18 24

(g) 1.0E+08

for the theoretical curves: U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

= 12 18 24

(h) 1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue Life, Cycles

Fatigue Life, Cycles

Fig. 9(eh). Comparison of alternative design SN curves at 0 m and 500 m water depths, based on various theoretical criteria for axial fatigue failure and different values of Ka: (e) and (g) Ka = 1.0 and 0.5, respectively, for the fatigue life to rst outermost layer wire fracture; (f) and (h) Ka = 1.0 and 0.5, respectively, for the fatigue life to rst wire fracture in the innermost layer.

2234

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

100
The theoretical lower bound to first outer layer fracture at 0 m water depth at 1000 m water depth

100
The theoretical lower bound to first inner layer fracture at 0 m water depth at 1000 m water depth

Load Range (% U.B.L.)

10

= 12

Load Range (% U.B.L.)

10
= 12 18 24

18 12 18 24 24

for the theoretical curves: U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(i)

for the theoretical curves: U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(j)

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

Fatigue Life, Cycles

Fatigue Life, Cycles

100
The theoretical lower bound to first outer layer wire fracture at 0 m water depth at 1000 m water depth

100
The theoretical lower bound to first inner layer wire fracture at 0 m water depth at 1000 m water depth

Load Range (% U.B.L.)

10

Load Range (% U.B.L.)

10

= 12

12

for the theoretical curves: U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

18 24

(k) 1.0E+08

for the theoretical curves: U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

= 12 18 24

(l) 1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue Life, Cycles

Fatigue Life, Cycles

Fig. 9(il). Comparison of alternative design SN curves at 0 m and 1000 m water depths, based on various theoretical criteria for axial fatigue failure and different values of Ka: (i) and (k) Ka = 1.0 and 0.5, respectively, for the fatigue life to rst outermost layer wire fracture; (j) and (l) Ka = 1.0 and 0.5, respectively, for the fatigue life to rst wire fracture in the innermost layer.

layer wire fracture (Figs. 9(b, d), (f, h), (j, l), (n, p) and (r, t)) regardless of the assumed value for Ka. As regards the lowerbound SN curves with fatigue life dened as the number of cycles to rst outermost layer wire fracture (Figs. 9(a, c), (e, g), (i, k), (m, o) and (q, s)), increasing the level of external hydrostatic pressure is shown to reduce the fatigue life, but only slightly, once the level of external hydrostatic pressure exceeds 500 m, although, by increasing the water depth from 0 m to 500 m, there is found to be a practically signicant reduction in the fatigue life (compared to the in-air conditions). 5. Discussion As already discussed by Raoof [1], a spiral strand undergoing constant amplitude cyclic axial loading in air, does, indeed, possess an endurance limit, below which no fatigue damage is assumed to

occur and the plots based on theoretical parametric studies using the same 127mm diameter sheathed spiral strand constructions with lay angles of 12, 18 and 24, as reported elsewhere [29] provide a simple theoretical means of determining the magnitudes of these endurance limits, not only for in-air conditions but also when the sheathed strands are subjected to external hydrostatic pressures, equivalent to water depths of between 500 m and 2000 m. However, in offshore platform applications, a sheathed spiral strand may become exposed to the potentially detrimental effects of seawater in the form of corrosion, which is of particular concern in the regions of end terminations within the water splash zone where the sheaths are in their most vulnerable state, and may, over the design life of the structure, become damaged. For this reason, it was decided to assume a non-existence of an endurance limit for the proposed design SN curves. In other words, the straight line SN curves (in loglog scale) for sheathed strands are assumed

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

2235

not to have any cut off point below which no fatigue damage occurs, and any level of axial stress range is assumed to cause some (although, perhaps, small) level of damage. This point may have signicant practical implications in offshore applications: in such cases, the small amplitude forces (from, for example, waves and/ or wind) are the ones with the highest number of occurrences, and a knowledge of small amplitude/long life behaviour is of particular importance. Figs. 9(ad), 9(eh), 9(il), 9(mp) and 9(qt) present all the presently proposed lower-bound design SN curves, based on extensive theoretical parametric studies, for all levels of hydrostatic pressure, and for both values of Ka = 1.0 and 0.5. Considering that these plots are presented in loglog scale, the signicant inuence of the lay angle on the axial fatigue life of sheathed spiral strands in deep water (and, indeed, in-air) applications, is obvious.

As mentioned previously, the assumed values of the U.B.L. and grade of wire for producing the proposed theoretical lower-bound SN curves are 13,510 kN and 1520 N/mm2, respectively. However, following the arguments in Ref. [7], because the axial load range in these plots is non-dimensionalized with respect to the U.B.L., all these theoretical lower-bound SN curves are of general applicability, irrespective of the magnitude of the U.B.L. and grade of wire. Also presented in Fig. 9(ad) are the lower-bound SN curves (for in-air conditions) as recommended by API [9], Chaplin [10] and Tilly [8]. In producing these purely empirical SN curves, none of these references differentiate between the various types of spiral strand (and/or wire rope) constructions. Moreover, different types of failure criteria were adopted by these researchers. Chaplin used the failure criteria as being the one to total collapse, while Tilly chose the number of cycles to 5% wire failure (i.e. life to fatigue ini-

100
The theoretical lower bound to first outer layer fracture at 0 m water depth at 1500 m water depth

100
The theoretical lower bound to first inner layer fracture at 0 m water depth at 1500 m water depth

Load Range (% U.B.L.)

10

= 12

Load Range (% U.B.L.)

10

18 12 18 24 24

= 12 18 24

for the theoretical curves: U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(m)

for the theoretical curves: U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

(n) 1.0E+07 1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

Fatigue Life, Cycles 100


The theoretical lower bound to first outer layer wire fracture at 0 m water depth at 1500 m water depth

Fatigue Life, Cycles 100


The theoretical lower bound to first inner layer wire fracture at 0 m water depth at 1500 m water depth

Load Range (% U.B.L.)

10

Load Range (% U.B.L.)

10

= 12

12

= 12

for the theoretical curves: U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

18

24

(o) 1.0E+08

for the theoretical curves: U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

18 24

(p) 1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue Life, Cycles

Fatigue Life, Cycles

Fig. 9(mp). Comparison of alternative design SN curves at 0 m and 1500 m water depths, based on various theoretical criteria for axial fatigue failure and different values of Ka: (m) and (o) Ka = 1.0 and 0.5, respectively, for the fatigue life to rst outermost layer wire fracture; (n) and (p) Ka = 1.0 and 0.5, respectively, for the fatigue life to rst wire fracture in the innermost layer.

2236

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238

tiation), whereas the failure criteria adopted by API is not dened in the code. The lay angles of the spiral strands used for producing Tillys design SN curve were equal to 14, 18 and 21, while Chaplins test strands had lay angles of 18. The strand construction details used by API are not given in the publicly available literature. The potentially unsafe nature of the previously reported lowerbound SN curves of API, Tilly and Chaplin for certain (smaller) levels of axial load range (depending upon the magnitude of the lay angle), and the potentially unsafe use of in-air SN curves as a guide to the fatigue life estimation of sheathed spiral strands experiencing high levels of external hydrostatic pressure is particularly noteworthy. Even for the in-air conditions, the API recommended SN curve can be unconservative for certain practical cases. As regards the use of Tillys or Chaplins recommended SN curves, for in-air conditions, the situation depends on the failure criteria adopted in practice, and the magnitude of the lay angles of the

wires in the strands, which are to be used in a given structure. It is also noteworthy that none of these purley empirical SN curves differentiate between the axial fatigue failures which happen at, or in the vicinity of, the end terminations, or in the free-eld (i.e. away from the detrimental effects of end terminations). 6. Design equations All the lower-bound design SN curves developed in this paper (as presented in Figs. 9(ad), 9(eh), 9(il), 9(mp) and 9(qt)) may be dened by the following simple equation

R aN b

where R = the axial load range (expressed as a percentage of the ultimate breaking load (U.B.L.)), and N = the axial fatigue life to rst

100
The theoretical lower bound to first outer layer fracture at 0 m water depth at 2000 m water depth

100
The theoretical lower bound to first inner layer fracture at 0 m water depth at 2000 m water depth

Load Range (% U.B.L.)

10

= 12 18 12 18 24 24 for the theoretical curves: U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm

Load Range (% U.B.L.)

10
= 12 18 for the theoretical curves: U.B.L. = 13510 kN Ka = 1.0 Sult = 1520 N/mm 24

(q)

(r) 1.0E+07 1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

Fatigue Life, Cycles 100


The theoretical lower bound to first outer layer wire fracture at 0 m water depth at 2000 m water depth

Fatigue Life, Cycles 100


The theoretical lower bound to first inner layer wire fracture at 0 m water depth at 2000 m water depth

Load Range (% U.B.L.)

Load Range (% U.B.L.)

10

10

= 12

12

for the theoretical curves: U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

18

24

(s) 1.0E+08

for the theoretical curves: U.B.L. = 13510 kN Ka = 0.5 Sult = 1520 N/mm

= 12 18 24

(t) 1.0E+08

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

1 1.0E+04

1.0E+05

1.0E+06

1.0E+07

Fatigue Life, Cycles

Fatigue Life, Cycles

Fig. 9(qt). Comparison of alternative design SN curves at 0 m and 2000 m water depths, based on various theoretical criteria for axial fatigue failure and different values of Ka: (q) and (s) Ka = 1.0 and 0.5, respectively, for the fatigue life to rst outermost layer wire fracture; (r) and (t) Ka = 1.0 and 0.5, respectively, for the fatigue life to rst wire fracture in the innermost layer.

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238 Table 6 Values of the constant parameters a and b for sheathed spiral strands, as dened in Eq. (6) Location of rst wire fracture Ka Water depth (m)

2237

a (deg)
12 a b 644 2999 3843 3967 3719 2599 11829 15546 15828 14633 3352 3787 3953 3974 3914 14391 16291 17020 17107 16848 0.2577 0.3953 0.4184 0.4205 0.4141 0.3873 0.5233 0.5483 0.5495 0.5422 0.4099 0.4206 0.4244 0.4249 0.4236 0.5436 0.5546 0.5585 0.5590 0.5576 18 a 1336 3838 5131 6012 6584 12574 16716 22681 26807 29462 5944 6262 6440 6557 6622 26807 28198 29034 29590 29893 b 0.3348 0.4304 0.4569 0.4713 0.4796 0.5420 0.5648 0.5925 0.6082 0.6163 0.4711 0.4756 0.4781 0.4798 0.4807 0.6082 0.6127 0.6154 0.6171 0.6180 24 a 3098 7471 12121 10253 10052 26807 32591 41212 45083 44108 12137 12137 12137 12137 12137 53746 53746 53746 53746 53746 b 0.4143 0.4941 0.5382 0.5226 0.5205 0.6082 0.6283 0.6495 0.6575 0.6554 0.5387 0.5387 0.5387 0.5387 0.5387 0.6742 0.6742 0.6742 0.6742 0.6742

Outermost layer

1.0

0 500 1000 1500 2000 0 500 1000 1500 2000 0 500 1000 1500 2000 0 500 1000 1500 2000

Outermost layer

0.5

Innermost layer

1.0

Innermost layer

0.5

wire fracture (i.e. fatigue initiation) in cycles, with a and b being constant parameters depending on the assumed value of Ka, water depth, and lay angle, a, plus the position of the rst wire fracture within the sheathed spiral strand (i.e. as to whether it happens in the outermost or innermost layer). For a given spiral strand (irrespective of the imposed level of mean axial load), values of the parameters a and b may simply be obtained from Table 6, which cover external water depths ranging from 0 m to 2000 m, with the intermediate values to be obtained by interpolation. 7. Conclusions Based on an extensive series of theoretical parametric studies, conducted on some very substantial (127 mm outside diameter) multi-layered spiral strands with realistic construction details, covering a wide range of lay angles 12 6 a 6 24, a new set of design SN curves for predicting the axial fatigue life of sheathed spiral strands to rst outermost (or innermost) layer wire fracture, both at, or in the vicinity of, the end terminations, as well as in the free-eld, experiencing a wide range of external hydrostatic pressures (equivalent to water depths of 0 m to 2000 m) have been developed. As mentioned previously, the theoretical model of Raoof (presently used to produce the design SN curves) has already been veried against a large number of carefully obtained large scale experimental data (under in-air conditions), using specimens with diameters ranging from 25 mm to 127 mm, with the tests carried out by a number of independent institutions, using test specimens from different cable manufacturers hence, providing ample support for its general validity (at least for in-air conditions). By using extensive theoretical parametric studies, it has been shown that, at 0 m water depth (corresponding to in-air conditions), modest increases in the lay angle, within current manufacturing limits, can lead to practically signicant reductions in the strand axial fatigue life. It has also been demonstrated that applying an external hydrostatic pressure to sheathed spiral strands can result in a practically signicant reduction in their axial fatigue life, and that, over a wide range of sheathed spiral strand constructions, the presence of an external hydrostatic pressure only affects (cf. in-

air conditions) the number of cycles to rst outermost (and, not innermost) layer wire fracture. The nal numerical results have also demonstrated the important practical implications of taking the detrimental effects of the end terminations into account, with the fatigue life to rst outermost (or innermost) layer wire fractures being substantially lower than that for cases when fatigue failure (dened as the number of axial load cycles to either outermost (or innermost) layer wire fracture) occurs in the free-eld i.e. away from the end termination. The proposed SN curves have been compared with others (for in-air conditions) recommended by API, Chaplin and Tilly, which are the ones most commonly referred to in the literature. It has been shown that, even for in-air conditions, in certain cases, these purely empirical SN curves, particularly the one presented by API, may provide unconservative results for practical applications. The implication by API that their design SN curve, although originally produced at 0m water depth (corresponding to in-air conditions), can be used as a guide to the fatigue behaviour of sheathed spiral strands experiencing substantial levels of external hydrostatic pressure, has been shown to be misleading. Unlike the presently proposed design SN curves, all of the other ones (i.e. those proposed by API, Chaplin and Tilly) have been produced, based on purely empirical approaches, using test data relating to specimens which are unlikely to have covered the full range of rst order design parameters (particularly the lay angle) hence, the practical signicance of the presently proposed design SN curves which very nearly cover the full range of current manufacturing limits, as far as the most important geometrical parameter (i.e. the lay angle) is concerned. Finally, with the often experimentally observed signicant remaining fatigue life after the rst wire fracture(s), the use of the number of cycles to rst wire fracture in the presently proposed design SN curves, although by no means a universally accepted criteria for fatigue failure, is believed to be a practically sensible (conservative) approach, leaving (in the absence of a full understanding of the fatigue performance of helically wound steel cables) a highly desirable margin of safety against total collapse in practice.

2238

M. Raoof, T.J. Davies / International Journal of Fatigue 30 (2008) 22202238 [15] Stonesifer FR, Smith HL. Tensile fatigue in wire rope. In: Proceedings of the 11th annual offshore technology conference, Houston, Texas; 1979. p. 53945. [16] Smith HL, Stonesifer FR, Seibert ER. Increased fatigue life of wire rope through periodic overloads. In: Proceedings of the 10th annual offshore technology conference, Houston, Texas; 1978. p. 17718. [17] Raoof M, Kraincanic I. Determination of wire recovery length in steel cables and its practical applications. Comput Struct 1998;68(5):44559. [18] Wiek L. The inuence of broken wires on wire rope strength and discarding. In: Organisation Internationale Pour LEtude de LEndurance des Cables (OIPEEC), round table conference, Luxembourg; October, 1977. [19] Hanzawa M, Yokota H, Toda Y, Yokoyama K. Fatigue behaviour of large diameter wire ropes. In: Proceedings of the 8th annual offshore technology conference, Houston, Texas; 1981. p. 43542. [20] Chaplin CR, Tantrum NRH. The inuence of wire break distribution on strength. In: Organisation Internationale Pour LEtude de LEndurance des Cables (OIPEEC), round table conference, National Engineering Laboratories, Glasgow; June, 1985. p. 3.3.119. [21] Hobbs RE, Ghavami K. The fatigue of structural wire strands. Int J Fatigue 1982;4(April):6972. [22] Metcalf JT, Matanzo F. Wire rope terminations, selection, and replacement criteria. In: Proceedings of the 12th annual offshore technology conference, Houston, Texas; 1980. p. 51724. [23] Castillo E, Canteli AF. Statistical models for fatigue analysis of long elements. In: Proceedings of the IABSE workshop on length effect on fatigue of wires and strands, El Paular, Madrid, Spain, report 66. 1992. p. 1532. [24] Raoof M, Hobbs RE. Analysis of axial fatigue data for wire ropes. Int J Fatigue 1994;16(October):493501. [25] Thorpe, TW, Rance A, Silvester DRV. The fatigue of electrogalvanised wire used in the manufacture of wire ropes. Report by the Materials Development Division, Atomic Energy Research Establishment (AERE), Harwell, UK; January, 1985. [26] Papanikolas PK. Axial fatigue of multi-layered wire strands. PhD thesis submitted to the University of Alberta, Edmonton, Canada; 1995. [27] Raoof M. Effect of lay angle on various characteristics of spiral strands. Int J Offshore Polar Eng 1997;7(1):5462. [28] Alani A, Raoof M. Effect of mean axial load on axial fatigue life of spiral strands. Int J Fatigue 1997;19(1):111. [29] Davies TJ. Static, dynamic, and fatigue characteristics of helical cables. PhD thesis submitted to Loughborough University; 2000.

References
[1] Raoof M. Axial fatigue of multi-layered strands. J Eng Mech, ASCE 1990;116(10):208399. [2] Raoof M. Axial fatigue life prediction of structural cables from rst principles. In: Proceedings of the institution of civil engineers, Part II, vol. 91; March, 1991. p. 1938. [3] Raoof M, Hobbs RE. Analysis of multi-layered structural strands. J Eng Mech, ASCE 1988;114(7):116682. [4] Raoof M. Axial fatigue life prediction of bridge cables. In: Proceedings of the 3rd international conference on bridge management, inspection, assessment, and repair, E. & F.N. Spon, University of Surrey, Guildford, UK; 1996. p. 52331. [5] Alani A, Raoof M. Axial fatigue characteristics of large diameter spiral strands. In: Proceedings of 5th international offshore and polar engineering conference, The Hague, The Netherlands, vol. II; 1995. p. 2605. [6] Raoof M, Alani M. Axial fatigue of spiral strands in offshore platform applications. In: Proceedings of 7th international offshore and polar engineering conference, Honolulu, Hawaii, vol. II; 1997. p. 16976. [7] Raoof M. Design of spiral strands against axial fatigue. Int J Offshore Polar Eng 1999;9(2):14957. [8] Tilly GP. Performance of bridge cables. In: Proceedings of the 1st Oleg Kerensky memorial conference, Institution of Structural Engineers, London, session 4; June, 1988. p. 22/48/4. [9] API 2FP1 (RP2FP1). API recommended practice for design, analysis, and maintenance of moorings for oating production systems; 1992. [10] Chaplin CR. Prediction of offshore mooring ropes. In: Proceedings of the round table conference on applications of wire rope endurance research, Organisation Internationale Pour LEtude de LEndurance des Cables (OIPEEC), round table conference, Delft, The Netherlands; 1993. p. 5075. [11] Raoof M. A critical review of draft API recommended practice 2FP1 regarding fatigue life estimation of moorings. In: Proceedings of the 11th international conference on offshore mechanics and arctic engineering, ASME, Calgary, Canada, vol. III, Part B; June, 1992. p. 52132. [12] Raoof M. Effect of hydrostatic pressure on strand behaviour. J Strain Anal 1990;25(2):7584. [13] Raoof M, Davies TJ. Possible shortcomings of the calibration methods for certain non-destructive monitoring devices for helically wound steel cables. J Strain Anal 2006;41(3):22138. [14] Chaplin CR, Potts AC. Wire rope in offshore applications. The Marine Technology Directorate, London; 1988. Publication 88/100.

Anda mungkin juga menyukai