Upload
others
View
4
Download
0
Embed Size (px)
Citation preview
Hydrogen Effects in Materials
Brian SomerdaySandia National Laboratories
Livermore, CA, USA
3rd European Summer School on Hydrogen SafetyUniversity of Ulster, Belfast, UK
July 28, 2008
Outline
•Introduction–H2 interactions with metal components– Safety concern: hydrogen embrittlement (HE)
•Methods to assess material performance
•Variables affecting HE– Material– Environmental– Mechanical
Outline
•Considerations for materials selection
•Component design–Design approach–Material property measurement– Example design analysis
•Recommendations
H2 containment components
On-board fuel tanks Manifold components
H2 containment components
PipelinesTransport and stationary tanks
H2 interactions with metal components
•H2 gas
1) H2 gas2) H2 adsorption3) H2 dissociation to H
4) H absorption5) H diffusion
Two principal safety concerns
• H2 permeation through wall of structure– Effectively results in H2 leak
• Hydrogen embrittlement (HE) of structural metal– Crack propagation through wall of structure
– HE mechanisms involve H in solution• Temperatures < 200 oC• Other H-related degradation mechanisms
operate >200 oC
H2H2
H2
H2
H2
H2 H2
stress
stress
defect
component surface
H2H2
H2H2
H2
HH
H HH2 H2 H2H2
H absorbsinto surface
H2
H2
H2H
H
H HH2 H2 H2H2
H concentratesat defect
H
HHHH
H2H2
H2H2 H2H2 H2 H2H2 HH H
H H HH
H
H causes embrittlementand crack propagation
HE enables crack propagation
Hydrogen embrittlement mechanisms
Hydrides
Hydrogen-EnhancedLocalized Plasticity(H.E.L.P.)
Hydrogen-EnhancedDecohesion
H
H
HH
HH
σ
σ σ
σ
H
σ
σ
τ1 τ2
HHHHH
HHH
HH
HH
H
H
H
τ1 > τ2
σ
σ
HHH
H
HH
HH
HH
H HH
σ
σ
void
particle
hydrides
C. Beachem, H. Birnbaum, I. Robertson, P. Sofronis
A. Troiano, R. Oriani
D. Westlake, H. Birnbaum
HE due to hydrogen-enhanced decohesion
“Intergranular” HE can be devastating
Barthélémy, 1st ESSHS, 2006
Material performance: test methods
strength of materials:σu, σy, εf, RAS-N
fracture mechanics:KIH, KTHda/dN vs ΔK
dl/dt > 0
H2H2
H2
H2H2
H2
H
HHH
HH
HH
dl/dt > 0
l
HHH
H H
H H
H
HH
dδ/dt ≥ 0
H
H
H
HH H
H
H
H
H
HH H
H
δ HHHH2 H2H2H2H2
H2
H2H2
dδ/dt ≥ 0
Material performance: test methods
HE manifested by reduced fracture toughness
H2 gas pressure (kpsi)0 5 10 15 20 25
Stre
ss in
tens
ity fa
ctor
, K (
ksi-i
n1/2)
0
50
100
150
200
250
300
H2 gas pressure (MPa)0 25 50 75 100 125 150
Stre
ss in
tens
ity fa
ctor
, K (
MP
a-m
1/2 )
0
50
100
150
200
250
300KJIcKTH
X100 ferritic steelWOL specimensC-L orientation25 oC
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
Material performance: test methods
HE manifested by increased fatigue crack growth rate
Stress intensity factor range, ΔK (ksi⋅in1/2)1 10 100
Fatig
ue c
rack
gro
wth
rate
, da/
dN (
in/c
ycle
)
10-8
10-7
10-6
10-5
10-4
10-3
10-2X42 ferritic steelfrequency = 1 Hz
1000 psi H2, R=0.11000 psi N2, R=0.1
HHH
dδ/dt > 0
H2H2H2H2
H2
H2
H2H2
HHH
dδ/dt > 0
H2H2H2H2
H2
H2
H2H2
Cialone and Holbrook, Met Trans A, 1985
•Fuel tanks– Aluminum or polymer lined composite
– H2 pressure < 70 MPa
•Manifold components– Austenitic stainless steel
– H2 pressure < 138 MPa– Low material strength– No welds
Material performance: service experience
Material performance: service experience
•Storage tanks– Low-alloy ferriticsteel
– H2 pressure < 42 MPa– Limited material strength
– No welds
•Pipelines– C-Mn ferritic steel– H2 pressure < 14 MPa– Low material strength– No pressure cycling
H2 compatibility guidelines for metals
•Materials favored for H2 service– austenitic stainless steels, aluminum alloys, low-alloy ferritic steels, C-Mn ferritic steels, copper alloys
•Materials commonly avoided in H2 service–high-alloy ferritic steels, nickel alloys, titanium alloys
HE susceptibility can be sensitive to many variables
Variables affecting HE
•All metals can be susceptible to HE depending on–Material variables– Environmental variables–Mechanical variables
•Trends demonstrated from laboratory test methods–Emphasize fracture mechanics data– Emphasize ferritic steels, austenitic stainless steels, aluminum
Material variable: strength (hardness)
Yield strength (MPa)600 700 800 900 1000 1100
KTH
(M
Pa-
m1/
2 )
0
20
40
60
80
100
120
Yield strength (ksi)90 105 120 135 150
K TH (
ksi-i
n1/2)
0
20
40
60
80
100Pressure Vessel SteelsH2 gas pressure = 41 MPa4130 steel4145 steel4147 steel
Loginow and Phelps, Corrosion, 1975
•Strength promotes HE in all materials•Technologically important trend
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
Material variable: alloy composition
Alloy composition affects intergranularHE in ferritic steels
[Mn + 0.5Si + S + P] (wt%)0.0 0.2 0.4 0.6 0.8 1.0
KTH
(M
Pa-
m1/
2 )
0
20
40
60
80
100Ni-Cr-Mo ferritic steelσYS = 1450 MPa110 kPa H2 gas296 K
B7Mn=0.007Si=0.002P=0.003S=0.003Mn=0.02
Si=0.01P=0.014S=0.003
Mn=0.09Si=0.01P=0.012S=0.005
Mn=0.02Si=0.27P=0.0036S=0.005
Mn=0.23Si=0.01P=0.009S=0.005
B2Mn=0.68Si=0.08P=0.009S=0.016
Mn=0.72Si=0.01P=0.008S=0.005
B6Mn=0.72Si=0.32P=0.003S=0.005
Mn=0.75Si=0.20P=0.006S=0.004
Bandyopadhyay et al., Met Trans A, 1983
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
Material variable: alloy composition
Nickel affects deformation and/or martensite formation in stainless steels
San Marchi et al., IJHE, 2008
H-precharged
non-charged 316 stainless steels
HH
HH
HH
dl/dt > 0
Material variable: welding
•Weld microstructure can promote HE•Hardness, residual stress also affect HE
δγ
Stainless steel weld Weld tested after H2 exposure
δγ
base metal weld
Environmental variable: H2 pressure
HE more severe at higher H2 pressure for all materials
H2 gas pressure (MPa)0 30 60 90 120 150
K TH (
MP
a-m
1/2 )
0
30
60
90
120
150
180
H2 gas pressure (ksi)0 5 10 15 20
KTH
(ks
i-in1
/2)
0
20
40
60
80
100
120
140
160Pressure Vessel Steels
4130 steel (σYS=634 MPa)4145 steel (σYS=669 MPa)4147 steel (σYS=724 MPa)
Loginow and Phelps, Corrosion, 1975
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
HE can be maximum at ambient temperature
HH
HH
HH
dl/dt > 0
Caskey, Hydrogen Compatibility Handbook for Stainless Steels, 1983
Environmental variable: temperature
Gas impurities can inhibit or intensify HE in ferritic steels
Environmental variable: H2 purity(d
a/dN
) H2+
addi
tive /
(da/
dN) H
2
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
O20.10%
2.25 Cr-1 Mo ferritic steelσYS = 430 MPa1.1 MPa H2 gasΔK = 24 MPa√mfrequency = 5 HzR = 0.1293 K
CO0.99%
SO21.10%
H2O0.03%
CH40.98%
CO21.01%
CH3SH1.04%
H2S0.10%
Fukuyama et al., Pressure Vessel Technology, 1992
HHH
dδ/dt > 0
H2H2H2H2
H2
H2
H2H2
HHH
dδ/dt > 0
H2H2H2H2
H2
H2
H2H2
Water vapor promotes HE in aluminum alloys
Environmental variable: H2 purity
Speidel, Hydrogen Embrittlementand Stress Corrosion Cracking, 1984
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
Mechanical variable: loading rate
HE more severe at lower loading rates in all materials
Loading rate, dK/dt (MPa-m1/2/min)0.1 1 10 100
KIH
(M
Pa-
m1/
2 )
0
20
40
60
80
1004340 ferritic steelσYS = 1235 MPa550 kPa H2 gas297 K
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
HHH
dδ/dt ≥ 0
H2H2H2H2
H2
H2
H2H2
Clark and Landes, ASTM STP 610, 1976
HE more severe at lower load cycle frequency in all materials
Mechanical variable: load cycle frequency
ΔK (MPa-m1/2)0 15 30 45 60
da/d
N (μ m
/cyc
le)
0.1
1
10
100SA 105 ferritic steel103 MPa H2 gasload ratio = 0.1
1.0 Hz35 MPa He gas
1.0 Hz
0.1 Hz
0.01 Hz
0.001 Hz
HHH
dδ/dt > 0
H2H2H2H2
H2
H2
H2H2
HHH
dδ/dt > 0
H2H2H2H2
H2
H2
H2H2
Walter and Chandler, Effect of Hydrogen on Behavior of Materials, 1976
Considerations for materials selection
1) HE data or service experience should not be extrapolated
H2 gas pressure (MPa)0 30 60 90 120 150
KTH
(M
Pa-
m1/
2 )
0
30
60
90
120
150
180
H2 gas pressure (ksi)0 5 10 15 20
KTH
(ks
i-in1
/2)
0
20
40
60
80
100
120
140
160Pressure Vessel Steels
4130 steel (σYS=634 MPa)4145 steel (σYS=669 MPa)4147 steel (σYS=724 MPa)
Yield strength (MPa)600 700 800 900 1000 1100
KTH
(M
Pa-
m1/
2 )
0
20
40
60
80
100
120
Yield strength (ksi)90 105 120 135 150
KTH
(ks
i-in1
/2)
0
20
40
60
80
100Pressure Vessel SteelsH2 gas pressure = 41 MPa4130 steel4145 steel4147 steel
•H2 compatibility must be established for specific conditions
Considerations for materials selection
2) HE can be severe in alloys considered compatible with H2
HH
HH
HH
dl/dt > 0
316 stainless steels
•Intersections of variables important
Considerations for materials selection
3) Materials selection may balance H2compatibility with other constraints
highcryogenic to lowintermediate
to highintermediatealuminum
low to intermediate
ambient to highlow to highlow
C-Mn/low-alloy steels
highcryogenic to highlow to highhighaustenitic SS
HE resistanceTemperature rangeSpecific strengthCostMaterial
Considerations for materials selection
4) Materials must be qualified for H2 service based on– Material property measurements under
relevant conditions• Material variables• Environmental variables• Mechanical variables
– Design approach for component
Design based on fracture mechanics
Local stress drives crack extension
σlocalH2H2
H2
H2
H2
H2 H2
stress (σ)
defect
component surface
H2H2
H2H2
H2H2 H2H2 H2 H2H2
σlocal = σc
stress (σ)
)(2
θπ
σ ijij frK
=
•Local stress described by stress-intensity factor, K
•Crack extension governed by critical K values
Design based on fracture mechanics
Component design: sustained-load cracking
H2 induces time-dependent crack propagation under static loading
pa
pa
H2H2H2
N2N2
N2
time
a
p
crack in H2
crack in N2
Sustained-load cracking proceeds when K > KTH
paH2H2
K = p * f(a/t,Ri,Ro)
RiRo
t
time
a
K
KTH
measured in laboratory
Component design: sustained-load cracking
Crack growth under cyclic loading accelerated by H2
Δpa
Δpa
H2H2H2
N2N2
N2
a
Number of pressure cycles, N (time = N/f)
p
crack in H2
crack in N2
Δp
Component design: fatigue crack growth
Fatigue crack growth rates (da/dN) are a function of ΔK
ΔpaH2H2
ΔK = Δp * f(a/t,Ri,Ro)
RiRo
t
da/dN
ΔK
measured in laboratory H2 gas
measured in laboratory N2 gas
da/dN = C[ΔK]m
Component design: fatigue crack growth
Component design analysis
•Objective: calculate number of pressure cycles, Nc, to grow crack to critical length, ac
Calculation requires material property measurements: KTH and da/dN vs ΔK
a t
ac
Number of pressure cycles, N Nc
critical crack depth forsustained-load cracking cycles tocritical crack depth
K = p * f(a/t,Ri,Ro)
ΔpaH2H2
RiRo
t
a
Measurement: fatigue crack growth
HHH
dδ/dt > 0
H2H2H2H2
H2
H2
H2H2
HHH
dδ/dt > 0
H2H2H2H2
H2
H2
H2H2
Measurement gives crack length vsnumber of cycles in design analysis
Δa = C[Δp * f(a/t,Ro,Ri)]m
Number of pressure cycles, N
a
ΔK (MPa-m1/2)0 15 30 45 60
da/d
N (μ m
/cyc
le)
0.1
1
10
100SA 105 ferritic steel103 MPa H2 gasload ratio = 0.1
1.0 Hz35 MPa He gas
1.0 Hz
0.1 Hz
0.01 Hz
0.001 Hz
Measurement: fatigue crack growth
da/dN = C[ΔK]m
Measurement: cracking threshold
Load
(P)
Time in H2
Po ∝ Ko
Loading bolt
Load cell
13 mm
PTH ∝ KTH
Wedge Opening Load (WOL) specimen
Measurement gives critical crack depth (ac) in design analysis
ac/t = g(KTH,p,Ro,Ri)ac
Number of pressure cycles, N Nc
a
H2 gas pressure (MPa)0 30 60 90 120 150
KTH
(M
Pa-
m1/
2 )
0
30
60
90
120
150
180
H2 gas pressure (ksi)0 5 10 15 20
K TH (
ksi-i
n1/2)
0
20
40
60
80
100
120
140
160Pressure Vessel Steels
4130 steel (σYS=634 MPa)4145 steel (σYS=669 MPa)4147 steel (σYS=724 MPa)
Measurement: cracking threshold
Example design analysis for H2 pipeline
•Parameters for H2 pipeline(ASME, Hydrogen Standardization Interim Report, 2005)
– X42 ferritic steel– Inner radius, Ri = 15 cm– Wall thickness, t, from 0.5 to 1.3 cm
•Maximum pressure, p = 10 MPa
•Existing defect with depth aoand length parallel to pipe axis
p
RiRo
aot
• Case 1: t=0.8 cm (σh=65% SMYS) and ao/t=0.10
• Case 2: t=1.3 cm (σh=43% SMYS) and ao/t=0.10
• Case 3: t=1.3 cm (σh=43% SMYS) and ao/t=0.05
p
10 MPa
N
1 MPa
cycle 1ΔK1=f(ao,Δp,Ro,Ri,t)Δa=(2.51x10-12)ΔK16.56a1=ao+Δa
cycle 2ΔK2=f(a1,Δp,Ro,Ri,t)Δa=(2.51x10-12)ΔK26.56a2=a1+Δa
ao a1 a2
Example design analysis for H2 pipeline
p
RiRo
aot
σh
Material/component performance can be quantified with design analysis
number of pressure cycles0 50000 100000 150000 200000 250000
crac
k de
pth
/ wal
l thi
ckne
ss, a
/t
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
hydrogent=1.3 cmσh=43% SMYS
nitrogent=1.3 cmσh=43% SMYS
X42 Pipelinepressure cycle = 1 - 10 MPainner diameter = 30 cm
X
X
critical crack depthcalculated from KIc
critical crack depthcalculated from KTH
Stress intensity factor range, ΔK (ksi⋅in1/2)1 10 100
Fatig
ue c
rack
gro
wth
rate
, da/
dN (
in/c
ycle
)
10-8
10-7
10-6
10-5
10-4
10-3
10-2X42 ferritic steelfrequency = 1 Hz
da/dN=(3.55x10-11)ΔK3.83
da/dN=(2.51x10-12)ΔK6.56
1000 psi H2, R=0.11000 psi N2, R=0.1
Example design analysis for H2 pipeline
Performance depends on both material properties and component design
number of pressure cycles0 5000 10000 15000 20000 25000 30000 35000 40000 45000
crac
k de
pth
/ wal
l thi
ckne
ss, a
/t
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
case 1t=0.8 cmσh=65% SMYS
case 2t=1.3 cmσh=43% SMYS
case 3t=1.3 cmσh=43% SMYS
X42 PipelineH2 gaspressure cycle = 1 - 10 MPainner diameter = 30 cm
X
X X
critical crack depthscalculated from KTH
Stress intensity factor range, ΔK (ksi⋅in1/2)1 10 100
Fatig
ue c
rack
gro
wth
rate
, da/
dN (
in/c
ycle
)
10-8
10-7
10-6
10-5
10-4
10-3
10-2X42 ferritic steelfrequency = 1 Hz
da/dN=(2.51x10-12)ΔK6.56
1000 psi H2, R=0.1
case 1 initial ΔK
case 2 initial ΔK
case 3 initial ΔK
Example design analysis for H2 pipeline
1) Austenitic stainless steels, aluminum alloys best candidates for H2 service– Based on service experience and testing– Do not extrapolate performance or test data
• Define temperature, alloy composition for austenitic stainless steels
• Define water vapor content of gas for aluminum alloys
– Design analysis can quantify safety margins
Recommendations
2) Ferritic steels are susceptible to HE under wide range of conditions– Service experience may be adequate but
design analysis required in many cases– Measure properties under specific conditions
• Material: strength, alloy composition, fabrication (e.g., welding)
• Environment: H2 pressure, H2 composition, temperature
Recommendations
3) Nickel alloys, titanium alloys are generally not recommended for H2 service– Design analysis required– May require other measures such as H2
permeation barrier
Recommendations