Sýdahl riser and umbilicals - fatigue damage assessment rev0

Assessment of Fatigue Damage

Stefan Palm07.05.2008

Application to risers and umbilicals

Version Slide 230 April 2009

Objectives

Give introduction to principles for assessment of fatigue damage with reference to design codes and engineering practice

Give an overview of typical fatigue loads, analysis methodology and fatiguecapacity

Show a few examples for typical riser configurations


Typical

riser fatigue

assessment

procedure

Task Comment

Define fatigue loading. Based on operating limitations including WF, LF and possible VIV

load effects.

Identify locations to be assessed. Structural discontinuities, joints (girth pipe welds, connectors, bolts), anode attachment welds, repairs, etc.

Global riser fatigue analysis. Calculate short-term nominal stress range distribution at each identified location.

Local joint stress analysis. Determination of the hot-spot SCF from parametric equations or detailed finite element analysis.

Identify fatigue strength data. S-N curve depends on environment, construction detail and fabrication among others.

Identify thickness correction factor. Apply thickness correction factor to compute resulting fatigue stresses.

Fatigue analyses. Calculate accumulated fatigue damage from weighted short-term fatigue damage.

Further actions if too short fatigue life. Improve fatigue capacity using:-

more refined stress analysis-

fracture mechanics analysis-

change detail geometry-

change system design-

weld profiling or grinding -

improved inspection /replacement programme


Content

Fatigue loading

Analyses methodologies

Critical hotspots and SN-curves

Damage calculation

Combined damage from two different processes

Fatigue considerations for typical riser configurations-

Steel Catenary

Risers (SCR)

-

Top

Tensioned

Risers (TTR)


Fatigue

load

history

H \ T 1.5 4 6 8 10 12 14 16 18 20 22 24 260.5 6563543 15665333 11341123 4560268 1584350 569519 226896 97258 44501 21500 10893 5760 77221.5 749462 9371771 12490552 6030618 2014398 551039 147376 42996 13543 4293 1370 484 3132.5 38841 1991649 5671979 3978348 1481485 396341 90799 21387 4966 1136 254 62 243.5 2557 326059 2059422 2267390 989181 275053 61169 12542 2308 397 61 10 14.5 220 46297 655471 1161634 639748 183157 40824 7931 1225 175 20 15.5 22 6170 187711 547886 398108 121309 27030 5134 706 81 76.5 2 776 48968 239758 240142 80648 17611 3327 439 43 37.5 90 11805 98656 138934 53372 11415 2153 280 24 18.5 8 2589 37831 77371 35332 7446 1362 182 15 19.5 1 540 13929 40916 22965 4927 861 119 10

10.5 103 4931 20643 14584 3357 561 80 611.5 18 1676 9997 9023 2292 363 51 412.5 3 547 4663 5406 1565 239 34 213.5 167 2095 3137 1065 160 22 214.5 48 903 1760 721 109 14 115.5 13 371 955 479 74 9 116.5 3 145 501 311 50 6 117.5 1 55 252 198 33 418.5 20 121 122 22 319.5 7 56 73 15 120.5 2 24 42 10 121.5 1 11 24 622.5 4 13 423.5 1 7 324.5 3 125.5 2 126.5 127.5

Long-term description of individual waves


Content

Fatigue loading



Damage calculation



Steel Catenary

Risers (SCR)

-

Top

Tensioned

Risers (TTR)-

Umbilical, Bellmouth

area


Global riser response

analysis

-

Fatigue

stress in steel

pipe

Time histories of fatigue stress calculated for a selectednumber of hotspots around the pipe circumference at relevant locations along the riser

where

r is radius out

to the

location where

the

fatigue

is to be checked

(inside, outside

or midwall)

steel pipe thickness used in stress calculation is normallyreduced by half of the corrosion/wear allowance

t=tsteel

-0.5*tcorr

( ) ( ) ( ) ( )AtTr

ItM

rI

tMt yx +⋅+⋅= )cos()sin( φφσ

rt

φ

y

x

( )

)(4

6422

44

IDODA

IDODI

−=

−=

π

π


Fatigue

stress in component

of

flexible

riser or umbilical

Simplified method often used where one assume that e.g. pipe in umbilical cross section is located at the center of the pipe having the same curvature as theglobal model:

-

where

κ

is curvature

and r is radius to hotspot

e.g. (OD-t)/2 for midwall

stress end E is module

of

elasticity

Calculation of stress in each component in cross section-

Need

purpose made

software to find

relation

between

the

global responses

and stress in each

component

(i.e. cross section

analyses)-

Important

to consider

friction

stress due to contact

pressure

Testing of components and complete cross-sections required for designs outside previous experience

SCR flexible

riserumbilical

)cos()()sin()()( φκφκσ ⋅⋅⋅+⋅⋅⋅= rEtrEtt yx


Content

Fatigue loading

Global Load Effect Analyses methodologies

Fatigue analysis and SN-curves

Damage calculation



Steel Catenary

Risers (SCR)

-

Top

Tensioned

Risers (TTR)-


area


Method

for fatigue

analysis

Fatigue analysis based on SN-data-

SN-data

determined

by fatigue

testing of

considered

weld

detail

-

based

on

linear cumulative

damage-

most commonly

used for risers

Fatigue analysis based on Fracture Mechanics-

used as supplement to SN data

-

document

sufficient

time interval

from crack

detection

during inspection

and time of

unstable

fracture

-

document

that

fatigue

cracks

occuring

during operation

will

not exceed

the crack

size

corresponding

to unstable

fracture


Fatigue capacity for constant stress range

The basic fatigue capacity is given in terms of S-N curves expressing the number of stress cycles to failure, N, for a given constant stress range, S:

mSaN −=

)Slog(m)alog()Nlog( −=

where a and m are empirical constants established by experiments.

Equivalently:

1

10

100

1000

1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08 1.00E+09 1.00E+10

Number of cycles, N

Stre

ss ra

nge,

S

log(ā)=intercept of log N-axism= negative inverse slope)


Corrosion

fatigue

test set-up

Testing setup

with

4 segment specimens

linked together

Test specimen

with corrosion

chamber


Pipieline

girth

weld

test specimen






10

100

1000

1,E+04 1,E+05 1,E+06 1,E+07 1,E+08

Number of load cycles

Stre

ss ra

nge

(MPa

)

RP C203, original NFailureDesign curve

10

100

1000

1,E+04 1,E+05 1,E+06 1,E+07 1,E+08

Number of load cycles

Stre

ss ra

nge

(MPa

)


Fatigue

cracking

failure

modesfatigue cracking from weld toes/roots into the base material

-

frequent

failure

mode-

most common

weld

in risers

is symmetric, single sided

with

welding

from outside-

more difficult

to inspect/have control

of

the

root-

weld

toe discontinuities

generally

present and behave

like pre-excisting

crack-

crack

initiation

time short

fatigue cracking from a surface irregularity or notch into the base material (e.gcorrosion)

-

concern

for components

with

stress cycles

of

high

magnitude-

crack

initiation

time is long, crack

propagation

time is short


Fatigue

crack

growth

Base material

Large defect/Unstable

fracture

Ni

Crack

size

Initiation

period Propagation

period


Weld Base material

Ni (weld) Ni

Crack

size


fracture


Fatigue

crack

growth

–

Paris law

mKCdNda )(Δ⋅=

agK ⋅⋅⋅= πσ

Paris law:

σ stress

MPa

K stress intensity

factor

MPam-1/2

a crack

length/size

m

g function

dependent on

crack

size

and geometry

(e.g. presence

of

stress concentrations)C

dimensionless

constant

m dimensionless

constant

(typically

in the

range 3-5)


Fatigue

crack

growth

testing

Testing arrangement showing corrosion

chambers

2 off compact

tension crack

growth

test

specimens

instrumented with

strain

gauges


Compact

tension

specimen

–

fatigue

crack growth


Measurement

of

crack

growth

rate

Base Material - Sea Water

1,E-06

1,E-05

1,E-04

1,E-03

1,E-02

1 10 100

dK MPam1/2

da/d

N m

m/c

ycle

Base Material 5+6 Regres BS-B BS-B +2 sd BS-A BS-A +2sd


Fatigue

crack

growth

0

2

4

6

8

10

12

14

16

18

20

0,E+00 5,E+05 1,E+06 2,E+06 2,E+06Number of cycles, N

Cra

ck h

eigh

t, a [m

m]

Crack growth, Ds=40MPa, a0=2mmCrack growth, Ds=50MPa, a0=2mmCrack growth, Ds=60MPa, a0=2mm

mKCdNda )(Δ⋅=


Fatigue

crack

growth

–

intiation

period

Ni

Crack

size

Initiation

periodPropagation

period

Macroscopic

defect


Examples

of

Riser fatigue

critical

hotspots

Threaded connectors-

example

of

use: coupling

between

riser joints in C/WO and drilling risers

-

critical

location: hotspot

with

SCF>1 at transition

between

pipe and connection

Bolted flanges-

example

of

use: coupling

between

riser joints in permanent TTR

-

critical

location: weld

between

flange and pipe, flange w/bolts

Welds-

example

of

use: SCR

-

critical

location: weld

root

and cap

Base material in the pipe-

critical

location: areas with

large

responses


Selection

of

SN curves

construction details;

fabrication process – welded, clad, forged, machined, etc;

base material or weld;

welds - hotspots on the inner surface and outer surface

weld details and tolerances, weld type (welding with or without backing, double sided weld);

stress concentration factors from concentricity, thickness variations, out of roundness and eccentricity; angularity;

environment - air, free corrosion or cathodic protection in sea water.


Weld

classes

–

DNV RP C203


Weld

classes

–

DNV RP C203


SN-curves


SN-curves

(DNV RP-C203)Non-welded

sections:

B1 SN-curve

Longitudinal seam

weld:

B2 SN-curve

Cast

nodes:

C SN-curve

Forged

nodes:

B1 SN-curve

if

DFF=10

C SN-curve

if

DFF < 10

An SCF is used that accounts for the actual fabrication tolerances.

Eq. (2.9.1)

Eq. (2.9.1)

The

nominal stress on

the

outside

of

the

pipe to be used for fatigue

assessment

of

outside

hotspots

The

nominal stress on

the

inside

of

the

pipe to be used for fatigue

assessment

of

the

inside

hotspots


Stress concentration

factor

due to fabrication

tolerance:

δ is total eccentricity (δthickness + δovality)

δ0 is eccentricity inherent in SN data (=0.1t)

t is pipe thickness

D is pipe outer diameter

Fabrication

tolerances

Dt

et

SCF−−

+= )0(31

δδEq. (2.9.1)

Total eccentricity

is sum of

fabrication

tolerance

of

thickness

and ovality:

4/)(

2/)(

2/)(

minmax

minmax

minmax

minmax

DD

DD

DDtt

ovality

ovality

ovality

thickness

−=

−=

−=−=

δ

δ

δδ

(no

pipe centralisation)

(pipe centralisation

during contruction)

(pipe centralisation

during contruction and rotated

until

good

fit)


Eccentricity


Thickness

effect

Fatigue

strength

of

welded

joints to some

extent

dependent on thickness

Reduced capacity due to increased local stress in toe for increased thickness

Thickness effect accounted for by modification of the stress

Reference thickness tref=25mm

k is thickness exponent

(recommended

k=0.15 for pipes)

0.80

0.85

0.90

0.95

1.00

1.05

1.10

1.15

0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 0.040 0.045 0.050 0.055 0.060pipe thickness t (m)

(t/t re

f)k

tref=0.025, k=0tref=0.025, k=0.15


S-N curves for different environment (media)

10

100

1000

1.00E+04 1.00E+05 1.00E+06 1.00E+07 1.00E+08 1.00E+09

Number of cycles

Stre

ss ra

nge

(MPa

)

DNV F1-curve CP

DNV F1-curve in airDNV F1-curve free corrosion

factor 1.2

factor 4.5

factor 3


Bi-linear S-N curves

1

10

100

1000

1.E+03 1.E+04 1.E+05 1.E+06 1.E+07 1.E+08 1.E+09 1.E+10

No of cycles, N

Stre

ss R

ange

, S

NSW

SSW

(a1;m1)

(a2;m2)⎟⎟⎠

⎞⎜⎜⎝

⎛ −

= 1

sw1m

)Nlog()alog(

sw 10S

⎩⎨⎧

≤⋅>⋅

=−

−

swm

2

swm

1

SSSaSSSa

N2

1

Log(Nsw) is typically 6-7


SN-curves

–

Umbilicals/flexible

risers

Project specific data based on testing applied for:-

armour

wires (flexible

risers, umbilical)

-

copper

conductors

(umbilicals)-

super duplex

pipes (umbilicals)

-

DNV-RP-C203 => SN-curve

for small diameter super duplex

steel

pipe (pipe OD=10-100 mm)

Sn-curve

applicable

for umbilicals

that

have been

reeled:

number of cycles under reeling < 10

strain range during reeling < 2%


When

to use

SN-curves

and da/dN

?

SN-curves:

The detail has to be specified and possible to be represented by one ofthe classes.

Alterantively, component specific design curve can be established by testing.

Fatigue crack growth caclulcations (da/dN):

The initial and final crack sizes have to be known.

Crack growth parameters in Paris law, m and C, has to be known. Somestandardised m/C values given in BS 7910. Otherwise, have to be determined by testing.

Detailed stress distribution has to be known


Content

Fatigue loading



Damage calculation



Steel Catenary

Risers (SCR)

-

Top

Tensioned

Risers (TTR)-


area


Fatigue capacity for variable stress range

n(Si) : Number of stress cycles with range Si N(Si) : Number of stress cycles to failure given by S-N curve D : Fatigue damage η : Usage factor (0.1-0.3)

The Miner-Palmgren

rule is adopted for accumulation of fatigue damage from stress cycles with variable range:

η )()(∑ ≤=

i i

i

SNSnD

mi

ii SSn

aD )(1 ∑= Single slope S-N curve

Equivalently:


Fatigue

analysis

-

Short term fatigue

damage

Long

term stress range distribution:

Number of stress blocks (Nb) and each block stress range (Δσ) calculated from the analysis. Number ofstress cycles (ni) with range Δσ is counted

Number of stress blocks (Nb) should not be less than 20

Total fatigue damage for the short term sea statefound by summation (Palmgren-Miner):

( )∑=

Δ⋅⋅=bN

i

miitermshort SCFn

aD

1_

1 σ

Block

no. Stress range (Δσ)

Number

of

cycles

(ni

)

1 0-10 1928372

2 10-20 2342732

3 20-30 1338753

4 30-40 453132

5 40-50 34321

6 50-60 4332

7 60-70 433

8 70-80 223

:

:

:

Nb 120-130 3

Example

of

stress histogram for one

seastate


”Damage

accumulation”

–

fatigue

crack

growth calculation

0

2

4

6

8

10

12

14

16

18

20

0,E+00 5,E+05 1,E+06 2,E+06 2,E+06Number of cycles, N

Cra

ck h

eigh

t, a [m

m]

Crack growth, Ds=40MPa, a0=2mmCrack growth, Ds=50MPa, a0=2mmCrack growth, Ds=60MPa, a0=2mm

mKCdNda )(Δ⋅=

Unstable

fracture

Nf

D = (Number

of

load

cycles)/Nf


”Damage

accumulation”

Crack

size


fracture


Detailed

fatigue

analysis

necessary?

A detailed fatigue analyses can be omitted if the largest local stress range is less than the stress range at 1.107 cycles (i.e. fatigue limit)

Guidance applicable for air and seawater with cathodic protection (i.e. two sloped curves)

In case of DFF > 1.0, the allowable fatigue limit needs to be reduced by a factor (DFF)-1/3

If one cycle is above the fatigue limit, fatigue damage from all stress cycles has to be includedDetailed

fatigue

assessment

can

be omitted Detailed

fatigue

assessment

required


Fatigue

analysis

-

Short term fatigue

damage

Rainflow

counting:

Number of cycles (Nc) in stress time series and stress ranges (Δσ) calculated by Rainflowcounting

Fatigue damage calculated for each cycle and total fatigue damage for the short term sea statefound by summation (Palmgren-Miner):

( )∑=

Δ⋅=cN

i

mitermshort SCF

aD

1_

1 σ


Fatigue

analysis

-

Long

term fatigue

damage

Long term fatigue damage as a weighted sum of short term fatigue damages:

where-

DL

–

accumulated

long-term

fatigue

damage

at given location-

Dij

–

Short term fatigue

damage

for seastate

i in direction

j-

Pij

–

Probability

of

occurrence

for seastate

i in direction

j-

Nd –

number

of

wave

directions-

Ns

–

number

of

sea-states

in the

wave

scatter

diagram

∑∑= =

=d sN

jij

N

iijL PDD

1 1


Design Fatigue

factors

Design fatigue factors (DFF) versus Safety Class (DNV OS F201)

Low (API-RP-2RD) Normal High(API-RP-2RD) 3.0 6.0 10.0

1≤⋅ DFFDL

Fatigue

criterion:

A risk based

fatigue

criterion

benchmarked

against

reliability

analyses is outlined in DNV RP-F204 Riser Fatigue. Relevant for novel

concepts

to evaluate

the

standard DFF and relative importance

of

each

parameter.

Steel risers:

Flexible risers and umbilicals => DFF=10


Reflection : Desired properties of integration scheme

Need good physical understanding of the system to select proper analysis methodology

Simplified analysis methods need validation

Three important contributions to fatigue damage are wave-induced, low-frequency and vortex-induced stress cycles

Recommended SN-curves and SCF’s for relevant riser/pipeline geometries is given in DNV-RP-C203

Methods for improving fatigue capacity.


Improving

fatigue

performance

Reduce stress concentrationsChange geometry: tapering, increase fillet radius,Grinding

Remove defectsGrinding NDT - repair

Reduce stress levelReduce global responseReduce stress concentrationsIncrease dimensions

Reduce number of load cyclesUse a bend stiffener instead of a bellmouth


References

“Dynamic Risers”. Offshore Standard DNV-OS-F201. October 2003

“Submarine Pipeline Systems”. Offshore Standard DNV-OS-F101. October 2007

“Riser Fatigue”. Recommended Practice DNV-RP-F204. July 2005

“Fatigue Design of Offshore Steel Structures”. Recommended Practice DNV-RP-C203. October 2006

“Environmental Conditions and Environmental Loads”. Recommended Practice DNV-RP-C205. October 2007


References

Faltinsen, O.M. “Sea Loads on Offshore Structures”. Cambridge University Press

Tucker, M.J. & Pitt, E.G. (2001) “Waves in Ocean Engineering”. Elsevier Ocean Engineering Book Series. Vol. 5

Ochi, M. (1998) “Ocean Waves – The stochastic approach”. Cambridge Ocean Technology Series – 6. Cambridge University Press.

Sarpkaya, T. and Isaacson, M. (1981) “Mechanics of wave forces on offshore structures”. Van Nostrand Reinhold Co.


References

Sparks, C.S. “The Influence of Tension, Pressure and Weight on Pipe and Riser Deformations and Stresses”. Transactions of the ASME. Vol. 106. March 1984. pp.46-54

Newland, D.E. “An Introductin to Random Vibrations and Spectral Analysis”. Longman Scientific and Technical

Blevins, R.D. “Flow-Induced Vibration”. Krieger Publishing Company


http://www.dnv.com/

Sýdahl riser and umbilicals - fatigue damage assessment rev0

Documents

a0 2mm crack growth

dshort term

processes fatigue considerations

typical riser configurations

top tensioned risers

large defect unstable fracture

steel catenary risers

recommended practice dnv