SAND2014-16479PE
Comparison
of
Stainless
Steels
for
High-‐Pressure
Hydrogen
Service
Chris San Marchi and Brian Somerday
Sandia National Laboratories
ASME Pressure Vessels and Piping Conference
Anaheim CA
July 24, 2014
Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin
Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000.
Mo;va;on
• Annealed type 316 & 316L alloys remain the
primary “material of choice” for tubing, fittings
and valves in hydrogen fuel applications
– Low strength and high cost
– Are there opportunities to lower cost and
maintain reliability?
• There exists an extensive database of properties for
austenitic stainless steels in hydrogen environments
– What does this data tell us?
– Do other materials meet the performance needs of high-
pressure hydrogen applications?
2
Materials:
austeni;c
stainless
steels
alloy Cr Ni Mn Mo C N
304L 18.3 8.7 1.4 0.34 0.016 0.08
316/316L 16.8 11.2 1.6 2.0 0.02 0.02
316 17.8 12.1 1.2 2.1 0.046 0.02
201LN 16.2 4.1 6.6 0.34 0.024 0.14
XM-11 20.4 6.2 9.5 NR 0.033 0.26
Nitronic 60 16.5 8.0 7.4 NR 0.071 0.14
XM-19 21.0 13.5 6.0 2.1 0.01 0.33
A-286 13.9 24.3 0.11 1.2 0.04 NR
3
Why
are
materials
such
as
304L
and
XM-‐11
not
considered
for
hydrogen
service?
100 Tensile ductility for a variety of austenitic stainless steels
non-charged
H-precharged
80
Reduction of Area (%)
60
Aluminum
40
Thermal
20 precharging:
300˚C
140 MPa H2
0
316
Nitronic 60
7475
316/316L
304L
201LN
A-286
XM-11
XM-19
4
Why
is
A-‐286
considered
appropriate
for
hydrogen
service?
100 Tensile ductility for a variety of austenitic stainless steels
293K
80
Reduction of Area (%)
60
40
XM-11
20 A286
201LN Nitronic 60 304 & 316 XM-19
0
4 6 8 10 12 14
Ni content (wt %)
5
Tensile
duc;lity
of
austeni;c
stainless
steels
do
not
scale
with
strength
100
293K
Open = air
80 Closed = H-precharged
Reduction of Area (%)
60 316/316L
Nitronic 60
XM-11 XM-19
40
Thermal
precharging:
20 304L 300˚C
A-286
140 MPa H2
201LN
0
300 400 500 600 700 800 900 1000
Yield Strength (MPa)
• Tensile ductility is not used directly in design
• If there is no design criteria associated with tensile ductility, what
tensile ductility is necessary for pressure applications?
6
Fracture
data
suggests
other
stainless
alloys
perform
similar
to
316
alloys
Fracture resistance measured in hydrogen environments
200
XM-19
Stress Intensity Factor, K H (MPa m 1/2)
316 • Fracture resistance in
XM-11 hydrogen environments
150
XM-19 XM-11 depends on strength
JBK-75
and microstructure
316 - not necessarily
100 IN903
XM-19 A-286 JBK-75 composition
2507
XM-19
50 JBK-75
IN903
0
500 600 700 800 900 1000 1100 1200
Yield Strength (MPa)
Fracture mechanics (and fracture properties) can be
used directly in the design of pressure components 7
Fatigue analysis is necessary to evaluate the
performance of materials in cyclic environments
S-N curves of H-precharged materials
Rotating
beam fatigue
R = -1
f = 50 Hz
Thermal
precharging:
300˚C
140 MPa H2
Fatigue performance of austenitic stainless steels does
not appear to be affected by H-precharging
Performance-‐based
assessment
suggests
that
life
is
not
limited
by
fa;gue
• For moderate design life, the limiting fatigue stress is greater than the
yield strength
• Design stresses are typically < yield strength
• Result: very conservative designs
HH
HH
HH
300
HH
HH
HH
10 MPa He
HH
HH
10MPa H2
250 KT~3
HH
HH
HH
Stress amplitude (MPa)
T = 223K
HH
R = 0.1
HH
HH
200 f = 1 Hz HH
HH
HH
HH
150
peak stress ~ Sy Tension-tension fatigue
of standard notched
100 tensile specimen
(after ASTM G142)
50 3
10 104 105 106
Cycles to Failure
from: Michler et al. IJ Fatigue 51 (2013) 1-7. 9
How
do
we
take
advantage
of
fa;gue
performance?
• By increasing the strength, higher fatigue stresses can be
accommodated in design
– Higher stress = less material
– Less material = lower cost
Stress
Stress
solid = air solid = air
dashed = hydrogen dashed = hydrogen
Annealed Strain-hardened
yield (high-strength)
limiting fatigue stress limiting fatigue stress
yield (annealed)
design life Cycles to Failure design life Cycles to Failure
of 30,000 of 30,000
10
Preliminary
fa;gue
results
700
f = 1 Hz
600 R = 0.1 XM-11 austenitic
Maximum cyclic stress (MPa)
stainless steel
500
HH
HH
400
HH
HH
300
HH
HH
HH
HH
200
HH
HH
HH
air
HH
HH
100 103 MPa H 2 HH
HH
HH
0 HH
4 5 6
1000 10 10 10
HH
Cycles to Failure
• High fatigue stress can be achieved with cycles to failure
greater than 10,000 cycles
• Broader evaluation of methodology requires testing under
combination of low temperature and high pressure
11
Preliminary
results:
internal
versus
external
H
700
f = 1 Hz XM-11 austenitic
600 R = 0.1 stainless steel
Maximum cyclic stress (MPa)
500
400
HH
HH
HH
300
HH
HH
HH
air, KT ~ 3
HH
103 MPa H 2, KT ~ 3
HH
200
HH
air, KT ~ 6
HH
HH
100 H-precharged, KT ~ 6
HH
HH
HH
HH
HH
0
1000 104 105 106 HH
HH
Cycles to Failure
• Available data is incomplete (inconsistency of notch acuity and
environments)
• Initial results suggest some correlation between internal and external H
• Data at low temperature is needed
12
Conclusions
• Tensile properties have limited utility for materials selection
for hydrogen service
– Results can be misleading
• Fracture properties suggest 316 alloys perform similar to
other austenitic stainless steels
– Wider range of alloys and strength conditions should be considered
• Fatigue performance in hydrogen environments suggests
that some hydrogen fueling applications may not be fatigue
limited
– Higher-strength alloys/conditions may enable more efficient designs
13
