FATIGUE ANALYSIS METHODOLOGY ON MOORING SYSTEM

By Murdjito
Dept of ocean Engineering FMT ITS
1. General
Mooring chain fatigue performance is an important design consideration for mooring terminal
systems. Overly conservative design approaches can be costly, whilst less sophisticated
analysis methods may not be truly representative, with the possibility of non-conservatism.
Connectivity calculating fatigue life of mooring system involving catenaries is a complex
problem. First order loading and vessel motions, low frequency second order vessel motions,
vessel springing and vibration due to vortex shedding can all contribute towards the long-term
fatigue damage of catenaries in a mooring system. The interaction of all these loads must be
assessed thoroughly to avoid unrealistic results.
First order fatigue is evaluated in time domain since the non-linearities of the mooring
system to be analysed was envisaged to be large. This approach is numerically demanding
but considered necessary to achieve an adequate level of confidence in the results. The first
order motions are described by vessel Response Amplitude Operator (RAO's).
Low frequency motions are very much dependent on the mooring system configuration. Low
frequency vessel motions are commonly considered insignificant in terms of fatigue due to
their very long period in comparison to wave loading. However this assumption does not hold
for many cases" ln fact, for most cases involving non- symmetrical spread mooring,
especially in shallow water, the low frequency motion can cause the highest quantities of
damage.

2. Statutory Requirements
ln accordance with the API RP 2SK [2]: 'Fatigue design is required for permanent moorings
only. A predicted mooring component fatigue life of three (3) times the design service life is
recommended. The factor of three accounts for uncertainties in lifetime load predictions."
ln accordance with the DNV-OS-E301 [3]: "The fatigue limit state is intended to ensure that
each type of component in an individual mooring line has a suitable resistance to fatigue
failure."
The design equation for fatigue limit state in a permanent mooring is as follows. lf the
mooring line is regularly inspected ashore, then the safety factor of three (3) should be
applicable

1d c F 0

Where :

d c = the characteristics fatigue damage accumulated during design lifetime.

F

= the single safety factor for fatigue limit state values of which is as follows.

Criteria 1 :

F =5 when

dF

< 0.8

( d 0.20.8 )

when

dF

Criteria 2 :

F =5+3

> 0.8

Where :

dF

= the adjacent fatigue damage ratio, which is the characteristic

fatigue damage ratio, which is the characteristic fatigue damage dc in two

adjacent lines taken as lesser damage divided by greater damage.
Fatigue damage from first order loads and motions and low frequency motions maybe
determined using either time or frequency domain analysis method; or the combination of
both. Three (3) methods can be considered for combining fatigue damages due to the low
frequency and wave frequency tensions, as follows.
o
o
o

Simple summation method. ln this approach, low frequency and wave frequency
fatigue damages are calculated independently. The total damage is assumed to be the
sum of the two.
Combined spectrum. ln this approach, the combined low frequency and wave
frequency spectrum is first calculated. Fatigue damage is estimated using the
combined root mean square (RMS) tension range.
Time domain cycle counting. ln this approach, the combined low frequency and wave
frequency tension spectrum is transformed into tension time history. A special cycle
counting method such as RAINFLOW method is used to estimate the number of tension
cycles and the expected value of tension range from which fatigue damage is estimated.

Among the three (3) methods above, time domain cycle counting method is considered to be
more precise.

3. Analysis Method

The characteristic fatigue damage, accumulated in a mooring line component as result of

cyclic loading, is summed up from the fatigue damage arising in a set of environment states
chosen to discretise the long term environment that the mooring system is subjected to,
expression of which is as follows
i=n

d c = di
i=1

Where :
di

= fatigue damage to the component arising in state I and

the discretisation into state i=1,n state

The fatigue analysis of the mooring chains was carried out utilising the computer program such
as Visual Orcaflex, etc. Fatigue analysis was performed following the procedures of
deterministic irregular wave fatigue analysis using Rainflow cycle counting method, as
recommended in API RP 2SK [2], section 7.5.
The long term environment was represented by eight (8) environmental direction and five (5)
reference wave classes, amounting to a total of forg (a0) discrete design conditions. Each
design condition consist of a reference direction and reference sea state characterised by a
significant wave height, peak period, current velocity and wind velocity. The probabilities of
occurrence of these conditions are detailed in section 4.6.
Each design condition was analysed analogously. Time domain simulations, for each design
condition, were simulated for 400 seconds in time step of 0.001 second. The simulation time
includes an additional 100 seconds ramp to minimise the starting transient effects. For each
design condition, the time history wave train where defined such that the maximum wave height
falls within the selected 400 simulation period.
The computed tension time history component obtained from each time domain simulations
were converted to stress time history component by dividing the corresponding tension by the
nominal cross-sectional area of the chain, where the cross-sectional area of the chains were
computed based on the expression as follows.

A c=

2 d
4

Where :
d = nominal diameter of chain
The stress component time history is analysed using the rainflow cycle counting technique,
based on Palmgren -Miner [15] & [16] law. This gives a number of stress ranges for half cycles.
The associated single-occurrence load case damage value is as illustrated by the following
expression.

i=n

1
S= D ( kS ( i ) )
2 i=1
Where :
K = stress concentration factor
S(i) = stress range

4. S-N Curves
The most common approach for the assessment of mooring chain fatigue is the stress-cycle (SN) approach. The equation used to determined chain fatigue life is based on the following
expression.
Component capacity against tension fatigue:
m

nc ( s )=a D S

The linearised logarithmatic form:

log ( n c ( S ) )=log ( a D ) m. log ( s )

Where :

nc ( S ) = number of stress ranges (number of cycles)

S = stress range (double amplitude)

a D = intercept parameter of the S-N curve

m = slope of the S-N curve

5. Environmental Condition and Loads

Waves. The design significant wave heights for the long term environmental events were
represented by eight (8) discrete design conditions; each condition consists of a reference
direction and a reference sea-state characterized by five (5) wave height distribution
associated with a recommended wave period. The irregular random waves were represented in
the analysis by the ISSC spectrum [7], expression of which is as follows.
ISSC spectrum

H 2
S ( )= s
4 Tz

( )

5 exp

[ ( ) ]
4

1 2
4
Tz

and

Tz=Tp

5+
11+

1
2

Where:
Hs = significant wave height
Tz = average zero up-crossing wave period
Tp = spectral peak period
= encounter freq
= peakness parameter

Table 4.4a represents the summary of the wave data for the various environmental directions
used for the fatigue analysis

Wind. Dynamic wind loads consist of a steady component and a time varying component known
as the gust, which generates low frequency motion. The steady component of the wind speed
is represented by a 1-hour average mean wind 10 m above sea level. The time varying wind is
described by a wind gust spectrum. The gust spectrum was represented in the analysis by the
API wind spectrum formulation as given in API RP 2A [5], expression of which is as follows
API wind spectrum:

fS ( f )
=
2
(z)

f
fp

[ ( )]
1.5 f
1+
fp

5
3

And

( z )=I ( z ) V ( 1 hr , z )
Where:
S(f) = spectral energy density
Z = reference elevation
f = freq.

( z )2 = standard deviation of wind speed

Table 4.4b represents the summary of the wind data for the various environmental directions
used for the analysis.

Current. Current were modelled as varying velocity component at different water level but
remaining constant with respect to time. Table 4.4c represents the summary of the current data
at various water levels and the various environmental directions used for the analysis.

Mean Sea level (MSL). The water depth, referenced to mean sea level (MSL) was considered.
The seabed was assumed to be flat in the analysis.
Tide. The highest water level (HHWL) was considered above MSL while the lowest water
level (LLWL) was considered below MSL in the analysis.

6. Corrosion Allowance
Corrosion allowance including wear and tear of 0.2 mm/year (50% of the chain corrosion
allowance), as recommended in DNV-OS-E3O1, was considered for the entire length of the
chain. The characteristic capacity of the chains, which forms the basis for the mooring
calculations, were adjusted for reduction in capacity due to corrosion, wear and tear accordingly.
The reduction in capacity was calculated based on the following expression.

BS cor=CBS

D nom2 T cor
Dnom

Where :

BS cor = corroded breaking strength

CBS = catalog breaking strength
T cor

= corrosion thickness for total design life

D nom = nominal diameter of chain

7. Time Domain (Coupled) Simulation

Time domain simulation allows non-linearities for all terms in the equation motion. Both high
frequency and low frequency wave excitations are applied simultaneously in the time domain. In
a simplified form, the equation of motions for a mooring system with six degrees of freedom in
waves can be expressed as follows.
6

{( M ij+a ij ) x+ bx+ c x }=Fi

j=1

And:

F=Fm + F w + F l
Where:

for

i=1 6

x = displacement matrix
M = mass matrix

a = added mass matrix

c = hydrostatic stiffness matrix, exist only for heave roll and pitch component
F = force matrix incl. wind, current, first & second order wave forces

Fm

= steady (mean) component of force

F w = wave freq component of force

Fl

= low freq component of force

Appendix: Cx, Cy, Cyx for Wind (OCIMF)

Appendix: Cx, Cy, Cyx for Current (OCIMF)