Living with Aging Materials – Challenges, and

Opportunities
Willis Whitney Award Lecture, AMPP Corrosion 2025
by

Narasi Sridhar
MC Consult LLC
The Ohio State University
Acknowledgement
Haynes International Aziz Asphahani, Steve Corey, Paul Crook, Lee Flasche, Galen Hodge, Venky Ishwar,
Dwaine Klarstrom, Juri Kolts, Paul Manning, Bill Silence, Jim Wu
Southwest Research Institute Sean Brossia, Gustavo Cragnolino, Jim Dante, Darius Daruwalla, Darrell Dunn, Sheewa
Feng, Vijay Jain, Marta Jakab, Peter Lichtner, Fred Lyle, Julio Maldonado, Oliver
Moghissi, Roberto Pabalan, Yi Ming Pan, Wes Patrick, Herb Pennick, Osvaldo Pensado,
Budhi Sagar, Nabuo Shiratori, Frank Song, Ben Thacker, Garth Tormoen, Tony Torng, Liz
Trillo, Lietai Yang, John Walton

DNV Arun Agarwal, Francois Ayello, John Beavers, Liu Cao, Sandeep Chawla, Hongbo Cong,
Ken Evans, Ayca Ertekin, Shan Guan, Feng Gui, Peter Friis Hansen, Angeire Huggins-
Gonzalez, Mariano Iannuzzi, Swati Jain, Gerry Koch, Xiaoji Li, Guanlan Liu, Stefan
Marion, Brandon Rollins, Andrea Sanchez Sours, Chris Taylor, Ramgo Thodla, Neil
Thompson, Yumei Zhai

Ohio State University Jerry Frankel, Liu Cao, Mariana Georges, Jiheon Jun, Mariano Kappes, Anup Panindre,
Jinwook Seong, Angeire Huggins-Gonzalez, Chenxi Liu, Vijay Srinivasan
OLI Systems Andre Anderko, Deepti Ballal
Sponsors California Energy Commission, DNV, DoD, DOE, EPRI, NASA, NIST, NRC, PHMSA, PRCI,
SwRI, MTI, WRPS (H2C), Batelle Savannah River Alliance, Various Industrial Consortia
2
A special appreciation of my mentor and friend, the late Gustavo Cragnolino

3
Why must we live with aging materials?

• Managing aging materials is critical

to safety

• Longer life of materials Improves

sustainability (Iannuzzi and
Frankel, 2023; Milosev and Scully,
2023)

• Efficient use of materials helps us

meet emissions target (Milford et
al., 2013)
What is NOT aging

• It is not the calendar time

• “Ageing is not about how old your equipment is; it is about its condition, and
how that is changing over time. Ageing is the effect whereby a component
suffers some form of material deterioration and damage (usually, but not
necessarily, associated with time in service).....” - UK HSE,
www.hse.gov.uk/offshore/ageing.htm.

There is no single, quantitative definition of Aging

Aging Management Approaches focus on lagging indicators

A quantitative definition of aging
Aging is defined as increasing hazard rate

𝑓𝑓(𝑡𝑡) Probability of failure at any time

• 𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟, ℎ 𝑡𝑡 =
𝑅𝑅(𝑡𝑡)
𝑡𝑡
Reliability up to that time = 1 − ∫0 𝑓𝑓 𝑡𝑡 𝑑𝑑𝑑𝑑

𝑡𝑡
• Cumulative hazard function, 𝐻𝐻 𝑡𝑡 = ∫0 ℎ 𝑡𝑡 𝑑𝑑𝑑𝑑 = − 𝐿𝐿𝐿𝐿 𝑅𝑅 𝑡𝑡

• More generally: H t = 𝐹𝐹 𝑡𝑡 |𝑅𝑅(𝑡𝑡 − ∆𝑡𝑡) × 𝑅𝑅 𝑡𝑡 − ∆𝑡𝑡

Statistical approaches
Hypothetical Distributions
10

9 Normal
8
Cumulative Hazard

7
Weibull (2.5,45)
6 Varying scale,location
5

3 Weibull(1.5,45)

2 Exponential

1 Weibull (0.5,45)
0
0 10 20 30 40 50 60 70 80 90 100
t, y
SCC of MA 600 steam generators
Staehle and Gorman, Corrosion, 59(11), 931-994

If there is no prior failure what distribution should I choose?

If failure mechanism changes how would the distribution change? 7
Aging of materials is a part of hierarchy of complex systems

Organizations/Nations/Civilizations
Organisms
Complex social
structures of people, Multifunctional Materials
organisms and assemblages of
materials Complex
molecules and assemblages of
materials atoms and
molecules
 Aging of societies, nations, and civilizations

Rome

France
Source: Peter Turchin, Ages of Discord, 2016

China
Certain fundamental mechanisms drive aging of societies that cannot be derived from
historical data alone
Aging in the tree of life

The large diversity in aging rates needs an understanding of

fundamental mechanisms

O. R. Jones, et al., Nature, 2014, 505(7482), 169-173.

Corrosion lifetimes Delhi iron pillar, approx. >1600 y
Josephenite, Ni-25%Fe alloy > 1.6 million y

Hanford AY-102, approx. 31 y

Djoser pyramid, >4700y

Biodegradable
implants. 0.2 y

Alloy C, > 65 y

Aloha airlines, 19 y

10-1 1 10 102 103 104 105 106

Age, Years 11
 Challenges in predicting aging

• The performance of an aging system depends Complex Distributions

on its history (probability curve) – lack of 10 2
Weibull (2.5,5)
knowledge of early history of a structure may 101

lead to failures of currently well-managed 100

Cumulative Hazard
systems. -1
10
Weibull(0.5,45)

-2
10

• Predicting aging requires integration across

-3
10

diverse size and time scales and integration -4

10

across various disciplines. -5

10
-6
10

• Things we don’t know that we don’t know

-7
10
0 10 20 30 40 50 60 70 80 90 100
(unknown unknowns) may affect aging t, y
Two approaches to dealing with time problem

• Model failure rates explicitly (corrosion rate, crack growth rates, etc.)
• Successful in well constrained systems over limited time periods
• Works for metal-environment conditions where 𝑝𝑝𝑝𝑝 ≤ 𝑝𝑝𝑝𝑝(𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑) –
active corrosion
• Does not work for complex systems or when engineered system interacts with
a natural system

• Give the time problem to somebody else

• We establish limit states for corrosion/SCC
• Others determine whether a system exceeds these limit states
• Allows modeling of complex systems and passive materials
Use of limit states in temporal predictions

• In structural reliability, failure is assumed to occur when load exceeds

resistance
• 𝑃𝑃 𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹 = 𝑃𝑃 𝑚𝑚 × 𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 − 𝑛𝑛 × 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 ≥ 0

• In corrosion, there may be other limit states:

• Thermodynamic limit states – Pourbaix diagrams
• pH ≤ Depassivation pH (Galvele, Oldfield-Sutton criteria)
• Corrosion potential ≥ Repassivation potential
• Stress intensity factor ≥ Threshold stress intensity factor
• Temperature ≥ Critical crevice temperature
• Aggressive species ≥ critical concentration
• Inhibitors ≤ critical concentration
 Limit state pH model

pH at tip of crevice/pit 5.0

Ni-20Cr-5Fe alloy
gap 4.5 0.2V SHE
Gaps are stochastic on real mating
2 cm 4.0 surfaces

3.5 4 mm gap

3.0 0.4 mm gap

2.5
-1 0 1 2 3
10 10 10 10 10
Time, hours
pH vs. Current (304L ss)

-5 10
10
2
Current Density, A/cm

8 Bulk
-6
10

pH
6
-7 Mouth
10
4 304L, 600 Grit
Tip
1000 ppm Cl
20 deg. C
-8
10 2
0E+0 2E+5 4E+5 6E+5 8E+5 1E+6 0E+0 2E+5 4E+5 6E+5 8E+5 1E+6
Time, Seconds Time, Seconds
N. Sridhar and D. S. Dunn, Corrosion, 1994, 50(11), 857-872.

• Crevice pH always lags stable current increase

Observation of Corrosion in Crevice

• Moiré fringe measurements

under a crevice with plane glass

• Observations of several micro-

sites of initiation

• An intermediate distance seems

to be favored

• Repassivation potential occurred

at a current density of about
T. Shinohara, N. Mascko, and S. Tsujikawa, 3x10-1 A/m2 regardless of bulk
Corrosion Science, 1993, 35(1-4), 785-789.
chloride
 Importance of potential as limit state
0.4

Gap, mm
0.4 0.3
Gap, mm

0.3 0.2
0.2 0.1
0.1 0
0
0 0.4 0.8 1.2 1.6 2
0 0.4 0.8 1.2 1.6 2 12

6 0.001h Eext = - 0.6V vs. SHE

500h
10h
0.01h 10
5

1h

pH
pH

4
8
0.1h
1h
3 10h 0.01h
500h 6 0.001h
Eext = 0.0V vs. SHE
2

0 0.4 0.8 1.2 1.6 2 0 0.4 0.8 1.2 1.6 2

Distance in Crevice, cm Distance in Crevice, cm
18
N. Sridhar, Corrosion, 2016, 73(1), 18-30.
 The role of salt film on repassivation

Total Dissolved Ni Concentration (molal)

10-1
+1.50 VSCE Pit Base
1.5
+0.50 VSCE
10-2 4

Current Density (A/cm )

2
+0.20 VSCE
1.0

Potential (VSCE)
10-3 -0.20 VSCE
10-4 3 -0.30 VSCE
0.5
10-5

10-6 2
0.0
Salt Film No Longer Present
10-7
Repassivation Salt Film
10-8 -0.5 1
0 150000 300000 450000 600000
Time (s)
0
0 2 4 6 8 10
Pit Depth (mm)
• Concentrated metal-chloride solutions affect
repassivation of pits

N. Sridhar and D. S. Dunn, Journal of the Electrochemical Society, 1997, 144(12), 4243-4253.
 Experimental Results - 308SS
1.00 0.5
10-1 Pit Base
+0.90 VSCE

Cr-Cl Complex Concentration (molal)

0.75
0.4 +0.50 VSCE

Current Density (A/cm )

2
10-2 +0.10 VSCE
0.50

Potential (VSCE)
-0.40 VSCE
0.3

10-3 0.25

0.2
0.00
10-4
0.1
-0.25
Repassivation

10-5 -0.50 0.0

0 100000 200000 300000 400000 0 2 4 6 8 10 12 14 16 18
Time (s) Pit Depth (mm)

Brossia et al., ECS, 1998

• Dissolution rate still high even at -0.20 VSCE

• Observed relationship between Cr-Cl concentration and
dissolution rate -- similar to Ni
 Single pit in Al

Al Single pit specimen, Pit depth = 5.25 mm, E= -650 mVSCE, t = 140 hr
307
177

99.998% Al in 0.6 M NaCl, T = 23 oC, N2 deaerated

1640
800
Top of pit
521

694

464
AlCl3 salt

1638
1.50 mm

805
from top

1638
523
Intensity, Arbitrary units

195
335

2.50 mm

Intensity, Arbitrary units

3.0 M AlCl3 from top

1638
5.25 mm
529

from top
2.0 M AlCl3
347

1638
532
346

1.0 M AlCl3

805
1638

1062
Bottom of pit

708

929
DI Water (Angle)

0 400 800 1,200 1,600 2,000 2,400

Raman Shift, cm-1 0 400 800 1,200 1,600 2,000 2,400
Raman Shift, cm-1
21
D. Dunn et al., unpublished
Repassivation Potential Model:
Limiting Behavior as Repassivation is Approached

•The expressions can be solved in the limit E -> Erp

A. Anderko, N. Sridhar and D.S. Dunn, Corrosion Science, 46 (2004) 1583

Modeling long-term performance using limit states

Gui et al. 2016

Sridhar et al., 2017, 13Cr

Dunn et al., Corrosion/2005

Cragnolino et al. 2001, 316L

 Relationship between limit states
Alloy 625

2.0
Alloy 825
Erp
1.8
1.6
Ecorr
1.4
1.2
E (SHE)

Exp. CCT (Hibner,

1.0 1986, MTI method)
0.8 Exp. CCT (Hibner,
0.6 1986, mod. MTI
method)
0.4 Exp. CCT (Garner, Sridhar et al., JES, 2023
0.2 1981)
0.0
-10 0 10 20 30 40 50 60 70 80 90
t (C)

Anderko et al., Corrosion/2005, Paper 05053

Predicting hazard rate from limit states
Current and future work
 Probability of SCC vs. Time using limit states
Initial Crack Length, m
Stress, MPa
8%
Stacking Geometry 6%
Ni 8%
Fault Energy 6% Factor 4%
4% 2%
2%
0.00060.00070.00080.00090.0010.00110.00120.00130.00140.0015

W 100 200 300

Critical K
Initial K, MPa.m^0.5
N 8%
CGR Exponent 8%
6%
6%
4% 4%
2% 2%
Mo
Alloy
0 10 20 30 0 100 200
Equival...

Cr Init Conditions Temporal Plate (10 time steps) Term Conditions

K, MPa.m^0.5
KIC-K(t), MPa.m^0.5

The concept of repassivation

6000
1000
4000 0
governing SCC in chloride
1
Chloride, M 2000 -2000
Repassivation
environments is used in defining Potential 0
0 1 2 3 4 5 6 7 8 9 -4000
0 1 2 3 4 5 6 7 8 9
threshold K
Temperature,C 1
Threshold Elapsed
1 1
CGR,m/s time, s

Corrosion Potential
CGR, m/s
-0.5..-0.2 13.8% Crack Length,m 1
OCP-Erp 0.0002
-0.2..0 6.3%
0.0001 200
0..0.2 19.7% 1
0 100
0.2..0.8 60.1%
-0.0001 0
0 1 2 3 4 5 6 7 8 9
-100
0 1 2 3 4 5 6 7 8 9

1
1
26
O2, ppm
 Growth of fracture limit distribution over time
2000 Time
400
1800
200 Max
1600
0
1400 Mean
-200

KIC - K(t), MPa.m0.5

1200 -400
Count

1000 -600
-800
800
-1000
600
-1200
400 Min.
-1400
200 -1600

0 -1800
0 1 2 3 4 5 6 7
-200 -120 -40 40 120 200
Time,d
KIC-K(t), MPa.m0.5 27
Hallmarks of Aging
Leading indicators

28
Hallmarks of aging in biological systems

C. Lopez-Otin et al., Cell, 2023, 186(2), 243-278

 Can we identify hallmarks of aging for materials?
Domains Factors and Events Hallmarks of Aging Failure Modes
Temperature Changes in corrosion (open-circuit)
Pressure potential
Gas-liquid reactions
External Deliquescence
Increase in corrosion products Large area nonuniform corrosion
Environment Speciation, ion exchange, micellization
Radiolysis
Microbiological colonies Rapid changes in electrochemical
Pitting
Flow/erosion impedance
Stress/Strain rates

Alkalinization at cathodes Increase in current variations between Crevice corrosion

Acidification from hydrolysis nominally identical electrodes
Local Biofilm environment
Environment Inclusion reactions Stress corrosion cracking
Increase in hydrogen uptake
Chemistry change - % saturation
Scaling
Evaporative concentration Crack colonies increase and link up Hydrogen stress cracking

Adsorption/desorption Increasing segregation of alloying elements

Passivity/depassivation Hydrogen Induced Cracking
Slip steps/fresh metal
Surface Point defect generation and transport Changes in surface composition and
Segregation element distribution Hydrogen reaction cracking
Dealloying
Increase in macro or microhardness
Alloying element depletion Corrosion fatigue
and decrease in toughness
Bulk Anti-phase domains/coherent phase – planar slip
Microstructure Deformation mode/ twinning
Dislocation pileup at boundaries Increase in precipitates or second phase Embrittlement
Trapping formation

Point defect generation and transport

Acceleration of increase in strain rate or
Nanostructure decrease in load
Nanovoid formation

Grain boundary void formation/depletion

 Corrosion potential vs. time (Carbon steel – nitrate-nitrite
solutions)

K. Evans, N. Sridhar, B. Rollins, S. Chawla, J. Beavers, and J. Page, Corrosion, 2019, 75(1), 106-119.
 Corrosion potential vs. time

Shorter time period Long time period

t3 t2
t1 t5
t4
Potential

OCP

Potential
OCP
OCP

t4
Time
t5

Log (current) Log (current)

32
 Interpretation of corrosion potential must be done carefully
Reference electrode drift
Crevice corrosion

D. S. Dunn, G. A. Cragnolino, and N. Sridhar, Corrosion, 2000, 56(1), 90-104.

S. Chawla, K. Evans, S. Feng, and N. Sridhar, Corrosion, 2024, 80(5), 472-488.

Unknown unknowns
“Thus, the category of scenarios not yet thought of may be handled by the same
process as any other scenario category: The relevant evidence is assembled and
quantitatively assessed using Bayes’ theorem.” – Kaplan and Garrick
Unknown unknowns

External triggers out of

our control
(treated stochastically)

Known corrosion
Unknown
and materials
sequences of
processes
events
(Treated with
(Use Bayesian
uncertainty
networks)
propagation)
Event sequences
Factors and Events Failure Modes Hallmarks of
External environment
Bulk Nanostructure
Aging
Local environment Surface
Microstructure
Aggressive OCP change
Inhibitive ions
ions
Temperature
Local pH Adsorption corrosion
Bulk Alloy Nonuniform
Composition products
Bulk pH corrosion
Pressure

Biofilm Impedance
Passivation change
Speciation
Gases Initiial OCP Localized
Corrosion
Inclusion Current
Reactions depassivation Variations
Deliquescence
Ionic Concentration

Alloy Depletion Void formation Hydrogen

% saturation SCC Uptake
Solvent

Scaling Crack colonies

Vacancies/ HSC
Radiation Slip steps/fresh APB/GP Zones
metal Interstitials

HIC GB Segregation
Microbial
Evaporation
Colonies
Point defect
transport deformation mode
HRC Surface
Composition
Flow

Dealloying CF
Increase in
Loading hardness

Crack tip strain rate

Embrittlement
Precipitation

strain rate/ load

change

GB Voids
Summary

• Aging involves increasing hazard rate – but hazard rates cannot be predicted using only data-
based models

• Our knowledge is the key to our ability to predict future performance of aging systems

• Predicting the behavior of aging systems should involve an understanding of fundamental

mechanisms – aggregate data should be used to validate predictions, but not as inputs

• Limit state models provide us flexibility in considering a variety of materials and systems

• A complete identification of hallmarks of materials aging will be beneficial in monitoring aging

systems proactively

• Unknown unknowns can be represented as unknown sequences of events governed by known

laws of corrosion
37