OIL & GAS Subsea Materials Testing and Assessment for HPHT Applications DNV GL Technology Week October 31 st, 2016 9am 12pm Colum Holtam, Rajil Saraswat, Ramgo Thodla 1 SAFER, SMARTER, GREENER
Agenda Welcome and introductions (9am) Safety moment Introduction and background to API 17TR8: HPHT Design Guidelines for Subsea Applications Introduction to fracture mechanics assessment Standard/Handbook approach FEA-based approach Fracture mechanics assessment case study Fully clad subsea HPHT component Break Materials characterization and testing Open forum Close (Midday) Lunch! 2
API 17TR8: HPHT Design Guidelines for Subsea Applications 3
API 17TR8: HPHT Design for Subsea Applications HPHT design is a significant new challenge facing the subsea sector, particularly in the Gulf of Mexico API 17TR8 provides HPHT Design Guidelines, specifically for subsea applications First Edition issued February 2015 Second Edition under preparation Due for release in 2016 (tentative) Could be several future Editions 4
API 17TR8: HPHT Design for Subsea Applications Subsea HPHT design challenges T > 350 F P > 15ksi Harsh environmental conditions Sour production High H 2 S/Elemental S High Cl- Seawater with CP Low T (40 F) Also elevated T? Design approach Stress based vs. Fracture mechanics Failure modes Fracture Fatigue 5
Materials Focus Subsea materials typically high strength steels But elevated T and P generally requires use of high strength nickel based alloys and/or clad material Modification of design philosophy (fracture mechanics vs. stress based) Environmentally assisted fracture & fatigue become critical in design Testing required to characterize environmentally assisted cracking behavior SSR Fracture toughness Fatigue (FCGR / S-N) Operating/Production conditions HPHT Shut in conditions Seawater + CP at low T 6
API 17TR8: HPHT Design for Subsea Applications Annex D: Material Characterization Protocols New in Second Edition Guidelines for use of metallic materials (low alloy steels and CRAs) for HPHT applications Generating material properties suitable for the application of fracture mechanics based approaches to the design of subsea equipment 7
Introduction to Fracture Mechanics 8
Principals of Fracture Mechanics Equilibrium evaluation between: The Crack Driving Force (CDF) (load) The fracture toughness (capacity) Crack Driving Force (CDF) Fracture toughness (materials resistance) Who is the strongest? Analyses/Calculations Testing CDF < CTOD mat : Crack is stable CDF CTOD mat : Crack is unstable (fracture or crack growth) 9
Fracture Mechanics Engineering Critical Assessment (ECA) A complex interaction between many input parameters (3 main types) Design data Geometry Size Location Orientation Installation & Operation Primary (Static and Fatigue) Secondary (residual stress) Tensile properties Fracture toughness Fatigue crack growth data
Fracture Mechanics Engineering Critical Assessment (ECA) Distinguishing between what is safe and unsafe Failure Assessment Diagram used to model failure by fracture / plastic collapse
Pipeline ECA Crack growth modelled to calculate critical initial flaw sizes 12
Flaw Types a 2c B B a B p 2a 2c 2c r m r m W is the mean circumference (2 r m ) r m External flaw Internal flaw Embedded flaw 13
Tensile Properties Stress strain curve Yield UTS Location of flaw Weld metal Parent metal 14
Fracture Toughness Fracture toughness parameter CTOD K J Specimen type Compact Tension (CT) Single Edge Notched Bend (SENB) Single Edge Notched Tension (SENT) 15
Fatigue Crack Growth Analysis Fracture mechanics used to model fatigue crack growth through life Fatigue crack growth rate law (C, m) 1 16
Loads/Stresses 17
Different Approaches to Fracture Mechanics Analysis K r =K I /K IC K I σ σ, Stress (Analytical method, Finite element analysis) L σ σ applied load collapse load Standard/Handbook (BS 7910, API/ASME 579) K I, Crack driving force FEA model with a crack, cracked model Standard/Handbook (BS 7910, API/ASME 579) L r, Plastic collapse ratio FEA model with a crack, model 18
Fracture Mechanics Software CRACKWISE (TWI commercial software) BS 7910 Signal (Quest Integrity commercial software) API/ASME 579 BS 7910 FlawSizer (DNV GL internal software) BS 7910 DNV-OS-F101 App. A DNV-RP-F108 FEA ABAQUS Spreadsheets/MathCAD
Fracture mechanics Assessment Case Study 20
Case Study Background The latest HPHT design guidelines in API 17TR8 require that a fatigue and fracture assessment be conducted assuming there is a flaw located at the highest stress location The initial size of the assumed flaw is conventionally based on the reliability of the NDT methods used during manufacture of the component Test programs required to quantify the behavior of high strength nickel based alloys exposed to HPHT production conditions including sour service as well as low alloy steels exposed to seawater with cathodic protection Material properties data, including fracture toughness and fatigue crack growth rates, are used as inputs to fatigue and fracture assessments FEA required to determine static and cyclic stresses based on design information 21
Objectives Conduct a fracture mechanics assessment case study for a fully clad subsea HPHT component using standard FAD-based assessment methods, such as those found in BS 7910 and API/ASME 579, and FE-based methods Consider internal circumferential flaw Evaluate the differences between the approaches 22
Assessment Approach Design Basis Geometry, Material, Loads Finite Element Analysis Stress, Strains Fracture mechanics analysis Fatigue Life 23
Component Details Material: F22 internally clad with alloy 625 (weld overlay) 6mm thick clad layer Simplified representative geometry: Low cycle fatigue dominated by pressure loads Cycle between low temperature/low pressure and high temperature/high pressure Operating pressure = 20ksi (137.8MPa) Operating temperature = 400 F (204 C) Hydrotest pressure 25ksi (172.25MPa) 1.25 x working pressure (ASME VIII Div. 3) Sour environment 24
Steps in a Finite Element Analysis Create structure geometry and develop mesh Loadings and boundary conditions Pressure, supplemental loads, temperature distributions, residual stresses Initial boundary conditions Material properties Perform analysis Extract displacement/strains/stress, stress intensity factor, collapse loads 320 45 152.5 20 20 45 150 298.5 450 1520 Units: mm 25
Simplified Component Geometry F22 steel Inconel 625 clad (6mm) 2D axi-symmetric model 26
FE Model Thermal Analysis Sequential coupling of thermal and stress analysis Steady state thermal analysis Loc1 Sea water temperature 4 C Heat transfer coefficient, 50 W/m 2 K Loc2 Fixed temperature 204 C 27
FE Model Stress Analysis, Boundary Conditions Axial load and pressure end load Refined mesh Internal pressure Fixed Second order 8 noded axi-symmetric elements 28
Material Properties High temperature stress-strain, CTE, specific heat, thermal conductivity F22 material properties were obtained from ASME VIII Div. 2 Inconel 625 material properties were obtained from Special Metals F22 steel Inconel 625 29
Loading steps Hydro-static test pressure of 172.25MPa (25 ksi) Remove hydro-static pressure Apply axial load of 1.5 million lbf Apply operating pressure of 137.8MPa (20ksi) and corresponding pressure end load Import thermal strains from the thermal analysis model 30
Results Axial Stress Hydrotest Axial load Axial load + Internal pressure Axial load + Internal pressure + Temperature 31
Axial Stress Location 1 Location 1 Location 2 Location 2 32
Uncracked Model vs. Cracked Model Uncracked FEA model Simpler as crack is not modelled Assessment is easier as commercial software is available Cracked mesh FEA model Complicated as crack is explicitly modelled Accurate as the actual geometry is modelled instead of using simplified handbook solutions 33
Development of Finite Element Mesh in Cracked Model Flaw location, orientation and size Focused mesh around crack tip along crack front 3D FE model: surface cracks and other cracks that are not well reflected by 2D models 2D FE model: significantly reduces time and resources needed, provides same or good approximate results when problems can be described in 2D. Collapsed elements (Spider web) 34
Model with Cracks under Axial Load and Internal Pressure 35
Crack Driving Force (CDF) estimated using 7 models Location 1 Location 2 K I 856.4MPa mm 0.5 for 1mm crack 36
Collapse Load, 1mm Crack, Location 1 L r = P/P c P, Applied load P c, Collapse load (Load at which ligament goes beyond yield) Elastic fully plastic analysis P c =245.284MPa L r =137.8/245.284=0.561 Mises Stress Plastic strain 37
Fracture Mechanics Assessment Inputs for Internal Flaw 1mm deep x fully circumferential (360 ) internal surface flaw K solution (varies by assessment case) Reference stress solution (varies by assessment case) Section thickness, B = 6mm (i.e. thickness of the clad layer only) Fracture toughness, J 0.2mm = 45N/mm Inferred from tests on similar materials in sour environment Fatigue crack growth law for steels in air from BS 7910 Tensile properties for 625 at 4 C Yield strength = 479MPa Tensile strength = 965MPa Young s Modulus for alloy 625 = 207500MPa Poisson s ratio = 0.3 Primary membrane stress (at 1mm depth) Based on uncracked FE model Axial load applied after hydrotest + pressure only (worst case from FE) P m = 360.7MPa Assume pure membrane since taking stress at flaw depth Cyclic stresses (linearized over clad section thickness) Based on uncracked FE model Pressure only (worst case) P m = 240.0MPa P b = 199.0MPa Assume zero residual stresses Perform sensitivity analysis 38
Failure Criterion Failure defined as breach of the clad layer For internal flaw (Location 1), reference stress solution based on clad layer only 39
Assessment Cases 1. Fracture check Calculate crack driving force (and critical flaw height) a) Standard K solution - BS 7910 (Crackwise software) b) Standard K solution - API/ASME 579 (Signal FFS software) c) K from FEA 2. Fracture check Calculate collapse load a) Standard reference stress solution - BS 7910 (Crackwise software) b) Standard reference stress solution - API/ASME 579 (Signal FFS software) c) Reference stress from FEA 3. Fatigue and fracture assessment Cyclic operating stresses + fracture check above Calculate fatigue life a) BS 7910 (Crackwise software) standard K and reference stress solutions b) API/ASME 579 (Signal FFS software) standard K and reference stress solutions Compare with FEA nb cyclic stresses linearized from initial crack depth in this case 40
Results Assessment Case 1 1. Fracture check a) Standard K solution - BS 7910 (Crackwise software) K = 786.9Nmm -3/2 (1mm flaw) Critical flaw height = 2.1mm b) Standard K solution - API/ASME 579 (Signal FFS software) K = 780.6Nmm -3/2 (1mm flaw) Critical flaw height = 2.1mm c) K from FEA K = 856.4Nmm -3/2 (1mm flaw) 41
Results Assessment Case 2 1. Fracture check a) Standard reference stress solution - BS 7910 (Crackwise software) L r = 0.9036, σ ref = 432.8MPa (1mm flaw) b) Standard reference stress solution - API/ASME 579 (Signal FFS software) L r = 0.895, σ ref = 428.9MPa (1mm flaw) c) Reference stress from FEA L r = 0.5618 (1mm flaw) 42
BS 7910-1mm flaw 43
API/ASME 579 1mm flaw 44
Results Assessment Case 3 3. Fracture and fatigue a) BS 7910 (Crackwise software) standard K and reference stress solutions Fatigue life = ~1,549 cycles (1mm initial flaw height) Add safety factor b) API/ASME 579 (Signal FFS software) standard K and reference stress solution Fatigue life = ~1,432 cycles (1mm initial flaw height) Add safety factor 45
BS 7910-1mm flaw 46
API/ASME 579 1mm flaw 47
Comparison: BS 7910 and API/ASME 579 vs. FEA 3. Fracture and fatigue Used reduced cyclic stresses to more closely match FEA fatigue calculation a) BS 7910 (Crackwise software) standard K and reference stress solutions Fatigue life = 15,520 cycles b) API/ASME 579 (Signal FFS software) standard K and reference stress solution Fatigue life = 15,567 cycles c) FEA Fatigue = 16,340 cycles (at 2.1mm flaw depth) Note this does not represent failure in FE-based assessment due to lower L r 48
Key Findings from Assessment of Internal Surface Flaw in Fully Clad Component BS 7910 and API/ASME 579 similar for fully circumferential internal surface flaw K at initial flaw depth (1mm) lower than FEA (i.e. FEA is more conservative for initial flaw depth) L r much lower in FEA (i.e. FEA is less conservative than Standard solutions) Linearizing cyclic stresses over section thickness is more conservative than FEA FEA would give longer fatigue life for internal flaw located in clad layer, e.g. due to lower L r 49
Sensitivity Analysis Residual Stresses BS 7910 a) Zero residual stresses K = 786.9Nmm -3/2 (1mm flaw) Critical flaw height = 2.1mm Fatigue life = 1,549 cycles b) PWHT (20% of yield strength) K = 995.9Nmm -3/2 (1mm flaw) Critical flaw height = 1.8mm Fatigue life = 1,260 cycles 50
Material Issues 51
Material Issues Identify key materials for the industry 718? F22? Clad F22? Identify primary failure modes Fracture Fatigue Identify key parameters that control performance Quantify the performance of the key materials against these failure modes Potentially develop database of material properties in environments Material Environment F22 718 SW + CP (low T) Sour (HPHT) Failure Modes Fatigue (FCGR, S-N) Fracture (FT) 52
Materials Challenges for Subsea HPHT Applications HPHT Sour Environments Low ph/high H 2 S/High Cl - High T SCC and Corrosion Fatigue Fabrication Challenges Welds/Clad layers (625) Alloy Selection (718/945/625+) Cu Plating issues leading to low T H embrittlement Nickel Based Alloys/CRA s Low T CP Issues Lower T (~40F) Cathodic Protection Fatigue & Fracture Issues Low T Shut in Conditions High H 2 S/Low ph Lower T (~40F) Fracture Issues 53
Current Acceptance Limits in ISO 15156-3 SCC behavior of precipitation hardened alloys has been evaluated in various environments using C-ring tests 54
Recent Work on PH Nickel Alloys to Develop a Robust Test Method Corrosion2015-5497 55
Typical J-R curve 400 300 400 F K-rate Air : 16Nmm -3/2 /s K-rate Env :0.016Nmm -3/2 /s 100psia H 2 S 500psia CO 2 J (N/mm) 200 100 J-R CMOD - Air J-R CMOD - Air J-R 25wt% NaCl J-R 2.5wt% NaCl J-R 0.25wt% NaCl J-R 25wt% NaCl w/0.5wt% Acetic Acid J-R 2.5wt% NaCl w/0.5wt% Acetic Acid There is substantial decrease in the fracture toughness of IN718 at high chloride concentration at 400 F 0 0 1 2 3 4 a (mm) 56
Fracture toughness of IN718 Effect of Chloride and Temperature 120 100 K Jint (MPa m) 80 60 40 300F 350F 400F 400F w/0.5wt% Acetic Acid IN718 100psis H 2 S 200/500psia CO 2 K-rate: 0.016Nmm -3/2 /s 0.1 1 10 Nacl (wt%) K Jint is determined from J 0.2mm, a sharp drop in K Jint occurs in the presence of high chlorides in 718 at 400 F The effect of temperature at low chloride concentration is not significant 57
Connecting Micro Process to Macroscopic Measurements Micro process that lead to stabilization of pits or cracks appear to be similar MCl x FPZ H 1 2 3 4 Macroscopic Measurements (E rp ) Repassivation potential which is associated with the driving for stabilization/repassivation of localized process Ni/Fe - Matrix Micro Processes During Cracking 1. Establishment of crack tip chemistry 2. Local acidification 3. Development of crack tip strain rate. 4. Crack Propagation Macroscopic Measurements for SCC K int and CGR (da/dt) Can the macroscopic measurements be related? 58
Relation between Localized Corrosion & SCC 718 K int (MPa m) 120 100 80 60 40 Susceptible to localized corrosion at OCP IN718 100psis N 2 500psia CO 2 400 F Not susceptible to localized corrosion at OCP 1E-4 1E-5-100 0 100 200 300 1E-6 da/dt (mm/s) at K J = 133MPa m There is a strong corelation between K Jint and da/dt with the difference of E rp and OCP When the E rp is lower than the OCP, K Jint is substantially lower than when E rp >OCP Susceptibility to cracking appears to be co-related with susceptibility to localized corrosion (E rp - OCP) w/o H2S (mv vs RE at Temperature) 59
Relationship between Localized Corrosion & SCC 625+ K J (MPa m) 180 160 140 120 Susceptible to localized corrosion at OCP IN625+ 100psis N 2 500psia CO 2 400 F Not susceptible to localized corrosion at OCP 1x10-5 8x10-6 6x10-6 4x10-6 2x10-6 da/dt (mm/s) at K J = 133MPa m There is a strong corelation between K Jint and da/dt with the difference of E rp and OCP. As E rp approaches OCP, K Jint is begins to drop. Susceptibility to cracking appears to be co-related with susceptibility to localized corrosion. 100-100 0 100 200 300 E rp -OCP (mv vs RE) 60
CP Effects Effect of applied potential and HT on the fracture toughness of UNS N07718 has been explored using slow rising displacement tests K th has been performed at low ph as opposed to higher ph s associated with seawater The role of various parameters in seawater + CP needs to be explored to develop guidelines for Ni-Based alloys 61
Fracture Toughness - Shut In Conditions 250 200 IN718 3.5wt% NaCl 40 F (4.4 C) -1050mV SCE Numbers indicate charging time J (N/mm) 150 100 718 Fracture toughness decreases with K-rate and IGC increases No data on fatigue behavior in shut in conditions F22 Limited / No data on FT and fatigue behavior 7 5.5 J 0.2mm - Air 50 J 5 0.2mm - Environment 3 J maxload - Air J maxload - Environment Increasing K-rate 0 0.01 0.1 1 10 100 K-rate (Nmm -3/2 /s) 62
Fracture Toughness Shut In Conditions - Rising Displacement vs. Step-Load J (N/mm) 65 60 55 50 45 40 IN718 3.5wt% NaCl 40 F (4.4 C) -1050mV SCE 2 day soak 2h hold 4h hold Step load: ASTM F1624 Slow rising displacement: ASTM E1820 (modified) Nominally same loading rate in all tests ~10 times slower than recommended in API 17TR8 35 30 25 K-rate 0.005Nmm -3/2 /s 0 10 20 30 40 Step size (lb) Which toughness value to use? What is representative? What is conservative Rising Displacement J int < Small Step Load J int < Large Step Load J int 63
Summary of Key Issues for CRA s PH Nickel Based CRAs used in HPHT Applications are susceptible to Environmentally Assisted Cracking. Fracture toughness appears to drop in environment Mechanism appears to be one of SCC and electrochemically driven Intergranular cracking is observed High strength CRAs also appear to be susceptible to H embrittlement at low temperature Sharp decrease in toughness with applied CP Mechanism is likely associated with the crack tip hydrogen generation and diffusion A number of issues to be addressed in this area HPHT SCC & fracture toughness behavior Effect of environments on wrought and clad 625? Behavior of other PH Nickel Based CRAs 625 + /945? Low T behavior of PH Nickel Based CRAs What controls the H embrittlement impurity content of Alloys? What is the impact of low T sour environments/coupled to steel? Others? 64
Rapid Material Characterization for HPHT 65
Typical Operating Conditions Pressure 15ksi/20ksi 350F/400F Temperature 20 50 days Primary loading scenarios Fatigue loading from pressure transients which are relatively quick (~hours) Fatigue loading from thermal transients which are relatively long (~1-2days) Static loading associated with long steady operations (~20-30days) ~a few days ~10 s min to hours Ambient Time 66
Loading Scenarios Load Low cycle fatigue and static crack growth behavior may not be discrete phenomenon but part of a continuum of the same phenomenon Development of a single specimen methodology to capture all of the critical design parameters Environmentally Assisted Crack Propagation Fatigue Controlled ~a few days ~10 s min to hours Static Crack Growth Controlled 20 50 days Ambient Time Loading scenario s involve low cycle fatigue which in the presence of environments can lead to environmentally assisted fatigue crack growth Constant load in environment can lead to static crack growth Primary environments of interest Production environment Seawater + CP environment Characterize the FCGR and static crack growth behavior in environment is essential for design 67
FCGR Screening Methodology K Decreasing Frequency Decreasing Frequency K min, R = 0.13 K max K min, R = 0.4 Time Decreasing Frequency K min, R = 0.6 Perform tests at constant K under decreasing frequency Decrease K by increasing R- ratio but keeping K max constant This allows for keeping the size of the static fracture process zone fixed Frequency changes can be made in a sequence so as not to disturb the crack chemistry 68
FCGR Response da/dn (mm/cycle) Operating Frequency In-Air Type 1 f (Hz) Type 1 No environmental effects i.e. no frequency dependence Perform tests at any convenient frequency Type 2 Cycle dependent environmentally assisted crack growth rate, strong dependence on frequency followed by a plateau FCGR Perform tests at plateau frequency, if very low could perform at higher frequency and knock down the measured curve by an appropriate amount Knock up the measured curve can be estimated by ratio of FCGR plateau /FCGR in-ar Type 3 Crack growth rate dominated by static crack growth rate with no evidence of plateau upto to low frequencies (1mhz/0.3mHz). Can also be identified by a FCGR 1/f 69
Transition from Fatigue to Static Crack Growth K K max 1h 3h 24h Increasing hold time Increasing hold time Constant at K max K min, R = 0.6 Time FCGR is typically a function of the cyclic loading however static CGR is a function of the stress intensity (or perhaps more accurately crack tip strain rate) Transition from fatigue to static crack growth rate by introducing periods of static holds 70
Crack Growth Rate Measurements Under K-control a (mm) da/dt (mm/s) Note Y-axis is CGR Detection Limit Type 1 Time (s) Smooth transition from long hold periods to constant K behavior and a linear a vs Time response i.e. a steady CGR is maintained Type 2 Change in crack length varies with time under constant K and a non-linear a vs Time response is obtained a decrease in CGR with time Type 3 No detectable crack growth is observed Time (s) 71
72da/dt (mm/s) Potential Options Under K-control Constant K at K 1 Note Y-axis is CGR K 3 >K 2 >K 1 Rising K under dk/da control Low CGR Constant K at K 2 Rising K under dk/da control Intermediate CGR Constant K at K 3 High CGR K th will be identified based on a threshold value of CGR. K The CGR at constant K will be used to determine the strategy for subsequent testing. The factors that will influence subsequent testing will be based on the measured CGR and the K-level If the CGR is low K will be increased under dk/da control to a higher value of K where the CGR will be measured at a constant K. This process will be repeated until the CGR reaches a high value. If the CGR when transitioning to constant K is high K will be decreased under dk/da control to lower value of K and the CGR measured at constant K. The process will be repeated until the CGR reaches a low value.
Static & Ripple Crack Growth Rate Measurements da/dt (mm/s) High CGR Low CGR CGR K n K th da/dt (mm/s) K max K K K K th will be determined based on an increase above low values of CGR with increasing K Static CGR measurements at different K-values to determine the K vs CGR relationship Time Determine effect of small ripples to establish CGR at low K Re-establish constant K behavior before transitioning to next K level 73
Constant K tests - No K-rate effect Development of test methodology to transition from a fatigue pre-crack to a statically loaded crack Decreasing K Decreasing frequency Increasing hold time Measure CGR 46.5 50 R = 0.3/0.33Hz R = 0.5/0.05Hz R = 0.5/0.01Hz R = 0.7/0.001Hz R = 0.7/0.001Hz 3600s R = 0.6/0.001Hz R = 0.6/0.001Hz 3600s R = 0.6/0.001Hz 9000s R = 0.6/0.001Hz 864000s Constant K 718 40F -1050mV SCE 3.5wt% NaCl, ph = 8.2 t hold
718 Overall Record W/Reference a (mm) 19 18.8 18.6 18.4 18.2 18 5.7564e-009 mm/s1.8572e-007 mm/s 9.9861e-007 mm/s 4.0929e-006 mm/s 4.0419e-005 mm/s 6/15/2016 @ 0 hours 17.8 0 200 400 600 800 1000 1200 1400 1600 1800 Overall record of crack length as a function of time 2.6156e-008 mm/s 5.8514e-008 mm/s 5.83e-008 mm/s 8.9433e-008 mm/s Time (h) 718 Kmax = 50ksiin 3.5wt% NaCl -1050mV SCE 40F 8/22/2016 @ 0 hours Transition to static crack growth by cycling at high K and high frequency initially followed by decreasing K and frequency Low frequency low K followed by introduction of hold times at K max
Effect of Hold Time on CGR 18.9 a (mm) 18.85 18.8 18.75 18.7 18.65 8.9433e-008 mm/s 5.83e-008 mm/s 5.8514e-008 mm/s 2.5599e-008 mm/s 718 Kmax = 55MPam0.5 R = 0.6 3.5wt% NaCl -1050mV SCE 40F 1mHz + 86400s 1mHz + 9000s 1mHz + 3600s 1mHz 800 900 1000 1100 1200 1300 1400 1500 1600 1700 Time (h) Addition of hold time at 1mHz leads to a slight decrease in the CGR, increasing hold times doesn t to decrease the CGR Suggests that the CGR at the long hold times is associated with static crack growth
Effect of K max on CGR 22 21.5 2.6156e-007 mm/s 1.91e-006 mm/s 1.9613e-007 mm/s 1.17e-007 mm/s 21 a (mm) 20.5 20 19.5 19 18.5 18 17.5 2.4469e-005 mm/s 1.602e-006 mm/s 718 3.5wt% NaCl ph = 8.2-1050mV SCE 40F Kmax = 60ksiin 1900 2000 2100 2200 2300 2400 Time (h) Raw 500/500 + 86400s 500/500 + 900s 500/500 50/50 1.5/1.5 50/50 At higher K max, the CGR at the low frequencies is significantly higher 77
Effect of Hold Time on CGR at K max of 72ksi in 28 10/12/2016 @ 1 h a (mm) 27 26 25 24 5.6312e-007 mm/s 2.6474e-006 mm/s 9.048e-006 mm/s 5.1905e-005 mm/s 2.2439e-007 mm/s 718 3.5wt% NaCl ph = 8.2-1050mV SCE 40F Kmax = 72ksiin Raw 500/500 + 9000s 500/500 50/50 10/10 1.5/1.5 5/5 23 1.5329e-005 mm/s 22 2500 2550 2600 2650 2700 2750 2800 2850 Time (h) No transition to static CGR was possible at these levels of K max. Further tests are needed to establish a K value at which stable static crack growth can be established.
SEM of Fracture Surfaces Fracture surface exhibits evidence of intergranular cracking. 79
Summary of the effect of Frequency da/dn (mm/cycle) 0.01 1E-3 1E-4 K max = 66MPa m K max = 55MPa m K max = 79.2MPa m 1E-5 1E-4 1E-3 0.01 0.1 1 f (Hz) K = 34.6MPa m, R = 0.2 K = 25.3MPa m, R = 0.5 K = 22.2MPa m, R = 0.6 K = 26.4MPa m, R = 0.6 K = 31.4MPa m, R = 0.6 718 3.5wt% NaCl -1050mV SCE 40 F Significant effect of hold time on the CGR The effect of frequency is more apparent at the lower frequencies. More so when hold times were introduced Static CGR at constant K was difficult to establish. Low CGR was observed under 1day unload cycles.
FCGR and Static CGR K max Effects on 625+ 26.8 26.75 2.6512e-008 mm/s a (mm) 26.7 26.65 26.6 26.55 8.5496e-008 mm/s 4.3772e-008 mm/s 625+ 3.5wt% NaCl ph = 8.2-1050mV SCE Kmax = 50ksiin0.5 R = 0.6 Raw 500/500 + 86400s 500/500 + 9000s 500/500 + 3600s 500/500 Crack growth rate as a function of frequency and hold time. Increasing rise time leads to a decreasing crack growth rate. No Steady State CGR under constant K conditions can be established at 50ksi in. 2.1124e-007 mm/s 200 300 400 500 600 700 800 900 1000 Time (h)
FCGR and Static CGR K max Effects on 625+ a (mm) 27.1 27 26.9 26.8 26.7 2.0084e-007 mm/s 1.6524e-006 mm/s 1.0481e-007 mm/s 3.9197e-008 mm/s 625+ 3.5wt% NaCl ph = 8.2-1050mV SCE Kmax = 60ksiin0.5 R = 0.6-1.7456e-008 mm/s Raw Constant K 500/500 + 86400s 500/500 + 9000s 500/500 50/50 26.6 Crack growth rate as a function of frequency and hold time. Increasing rise time leads to a decreasing crack growth rate. No Steady State CGR under constant K conditions can be established at 60ksi in. 1200 1300 1400 1500 1600 1700 1800 1900 2000 Time (h)
Effect of Frequency da/dn (mm/cycle) 1E-3 500/500 + 86400s Hold Times No Hold Times 500/500 + 9000s 500/500 K max = 60ksi in R = 0.6 3.5wt% NaCl ph = 8.2-1050mV SCE 40 F 50/50 1E-4 1E-5 1E-4 1E-3 0.01 Frequency (Hz) da/dn (mm/cycle) 1E-3 500/500 + 86400s Hold Times No Hold Times 500/500 + 9000s 500/500 + 3600s K max = 50ksi in R = 0.6 3.5wt% NaCl ph = 8.2-1050mV SCE 40 F 500/500 1E-4 1E-5 1E-4 1E-3 Frequency (Hz) FCGR increases sharply with decreasing frequency, with FCGR increasing by about 10x as the frequency decreases from 1mHz to 0.01mHz. The effect is apparent at two different K max values.
Transition from FCGR to Static CGR a (mm) 30.5 2.573e-007 mm/s 3.2198e-007 mm/s 8.2561e-007 mm/s 1.303e-006 mm/s 30 29.5 29 28.5 28 27.5 5.316e-006 mm/s 6.0753e-005 mm/s 2.1376e-005 mm/s 625+ 3.5wt% NaCl ph = 8.2-1050mV SC Kmax = 90ksiin0.5 R = 0.6 Raw Constant K 900/100 + 86400s 900/100 + 9000s 900/100 90/10 3/1 9/1 90/10 27 6.7077e-006 mm/s Crack growth rate as a function of frequency and hold time. Increasing rise time leads to a decreasing crack growth rate. Steady State CGR under constant K conditions appears to be established at 90ksi in. Steady State CGR at constant K is about 2.57 10-7 mm/s. 2100 2150 2200 2250 2300 2350 Time (h)
Summary of Frequency Effect da/dn (mm/cycle) 0.01 900/100 + 86400s 1E-3 900/100 + 9000s 900/100 90/10 1E-4 1E-5 1E-4 1E-3 0.01 0.1 Frequency (Hz) K max = 90ksi in R = 0.6 3.5wt% NaCl ph = 8.2-1050mV SCE 40 F 9/1 3/1 Decreasing frequency leads to an increasing CGR per cycle. At a high K max the increase in FCGR is about 100x higher than the values at low frequency. This appears to be a result of the combination of K max and K effect. More work is needed to establish the K vs CGR behavior.
Open Questions Environmental Variables What is the effect of applied potential? What is the effect of temperature? Effect of loading variables What is the K vs CGR behavior in cases where steady state CGR can be established? What is the effect of dk/da? Is there is an effect of sample size on the K vs CGR behavior i.e. is plasticity a critical factor in these results? Can lower CGR on the order of about 1 10-8 mm/s be established. Metallurgical Variables Is there an effect of grain boundary precipitation like sigma phase on the susceptibility of high strength nickel based alloys. What is the deformation mode in associated with crack propagation in these systems Does it change based on the nature of the precipitation 718 vs 625+? Can the overall behavior be modelled based on a crack tip strain rate formulation?
Summary Hydrogen embrittlement of 718 in seawater + CP conditions appears to be sensitive to test methodology Constant displacement tests do not exhibit any evidence of cracking in long term exposure tests in seawater + CP conditions. Rising displacement tests on pre-cracked specimens appear to show susceptibility to environmentally assisted cracking. Decreasing K-rate lead to a decreasing K th value at a CGR is sensitive to K-rate in the rising displacement tests Static CGR tests were performed by transitioning from fatigue pre-cracking in environment to static crack growth with introduction of hold times CGR is significantly lower than those observed in the rising displacement tests at the same K-values.
Static Crack Growth Analysis Fracture mechanics used to model static crack growth through life Static crack growth rate law (C1, n(scc)) 1 88
Open Forum Questions? 89
Thank you Dr. Colum Holtam colum.holtam@dnvgl.com +1 281-396-1000 Dr. Rajil Saraswat rajil.saraswat@dnvgl.com +1 281-396-1000 Dr. Ramgo Thodla ramgopal.thodla@dnvgl.com +1 614-761-1214 www.dnvgl.com SAFER, SMARTER, GREENER 90