Corrosion Science 233 (2024) 112063

Contents lists available at ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

Experimentally verified thermodynamic framework for corrosion

K.P. Lijesh , Ali Mahmoudi , M.M. Khonsari *
Department of Mechanical and Industrial Engineering, Louisiana State University, Baton Rouge, LA, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Utilizing the foundational principles of thermodynamics, this study characterizes material degradation resulting
Corrosion from corrosion, substantiating the outcomes through experimental verification. The proposed methodology es­
Degradation entropy generation tablishes a unified corrosion model, effectively correlating entropy generation with measurable parameters,
Degradation coefficient
including potential difference, corrosion current, and temperature. The Degradation Entropy Generation (DEG)
theorem is applied to establish a linear relationship between entropy generation and the degradation rate
through a constant material property known as the degradation coefficient (B). Material degradation, quantified
by assessing the material loss, is systematically examined with a focus on low-carbon steel (LCS 1018) and
Al6061-T6. Corrosion experiments were conducted using a three-electrode electrochemical corrosion cell
immersed in a 3.5 wt% NaCl aqueous solution. The degradation parameter, B, for both materials was assessed
through multiple accelerated corrosion tests, where various magnitudes of potentials were applied to specimens
with differing exposed areas. The presented results unveil a consistent degradation coefficient of (30±2)×10− 9
m3K/J for LCS 1018 and (25±2)×10− 9 m3K/J for Al6061-T6. To ascertain the generalizability of these findings
under diverse environmental conditions, additional experiments are conducted on LCS 1018, exposing it to
varying environmental parameters such as different percentages of NaCl solution, acidic conditions, and elevated
solution temperatures. Remarkably, the degradation coefficient remains constant across all tested corrosive
environments, attesting to its independence. This study offers a novel approach to understanding material
degradation through a thermodynamic lens and demonstrates the applicability of the proposed corrosion model
in diverse and challenging operational conditions.

1. Introduction approaches to explore the damage caused by corrosion [12–14].

Corrosion is an electrochemical degradation mechanism that involves
Material degradation, specifically corrosion, is rooted in complex multiple irreversible oxidation and reduction reactions. It is an irre­
chemical and mechanical interactions between materials and their versible, dissipative process and causes permanent degradation. Ac­
environment [1–3]. Conventional quantification of material degrada­ cording to the second law of thermodynamics, the processes occurring
tion over time (t) due to corrosion (C) typically employs a power func­ during material degradation, encompassing damage or microstructural
tion C = At n [4–6], where parameters A and n are determined by fitting changes, are inherently irreversible. These processes entail the dissipa­
the data measured at different environmental conditions. However, this tion of energy, a phenomenon quantifiable through the measurement of
equation explicitly applies to specific environmental conditions [7]. entropy generation [15]. In the current study, we leverage the principles
Established statistical methods such as conventional regression, analysis of thermodynamics to develop a physics-based model that quantifies
of variance (ANOVA), and the hierarchical linear model (HLM) are often material degradation due to corrosion across different environmental
applied in an attempt to characterize material loss attributed to corro­ conditions.
sion across diverse environmental settings [7,8]. Among these statistical A general description of the thermodynamics of degradation intro­
models, the HLM model provides a more reliable estimation of corrosion duced by Bryant, Khonsari, and Ling [16] has revealed that for a single
loss [9–11]; however, it necessitates the determination of multiple co­ dissipative mechanism, the relationship between entropy generation
efficients to model corrosion under varying environmental conditions. (S’) and the magnitude of degradation (w) is linear and remains inde­
Recently, there has been an interest in utilizing thermodynamic pendent of the operating conditions. Coined as the Degradation Entropy

* Corresponding author.
E-mail address: [email protected] (M.M. Khonsari).

https://doi.org/10.1016/j.corsci.2024.112063
Received 19 February 2024; Received in revised form 31 March 2024; Accepted 16 April 2024
Available online 17 April 2024
0010-938X/© 2024 Elsevier Ltd. All rights reserved.
K.P. Lijesh et al. Corrosion Science 233 (2024) 112063

Generation (DEG) theorem, this theorem introduces a proportionality processes are stress, temperature gradient, electric potential gradient,
constant known as the degradation coefficient (B). The degradation etc. Examples of thermodynamic flow are plastic strain, heat flow,
coefficient, B, has been successful in quantifying material degradation electric current, etc. [28]. From Eqs. (1) and (2),
across diverse scenarios, ranging from complex tribo-pair systems
dζ
operating in both dry [17,18] and lubricated conditions [19,20] to J= (3)
dt
materials subjected to cyclic loads [21] and, more recently, in the realm
of Li-ion battery systems [22,23]. dS′ dpj
Examining material degradation resulting from the irreversible and X= (4)
dpj dζ
dissipative electrochemical reactions inherent in the corrosion process,
we employ the degradation coefficient (B) to quantify the extent of dw dpj
material degradation under various environmental conditions. We posit Y= (5)
dpj dζ
that B remains constant across diverse environmental conditions and
helps to generalize the results. Determination of B requires quantifying Note that for any dissipated process, the definition of entropy sug­
the entropy generation during the electrochemical reaction by gests dS′/dp = 1/T [16], where T is the temperature in Kelvin and dp/dζ
measuring pertinent parameters such as voltage, corrosion current, and represents the dominant driving force during the dissipative process
temperature. [16]. According to DEG theorem, the degradation coefficient, B, is
In evaluating material fracture induced by cyclic loading [24,25], defined as the ratio of Y and X.
coating failure [17], and transitions in wear regimes [18], an impactful ⃒ / ⃒
application of thermodynamics lies in the postulation that the accu­ Y⃒ dw dS′ dpj dw⃒
B = ⃒⃒ = ′ |P = ′⃒⃒ (6)
mulation of entropy generation up to the point of failure remains con­ X p dS dpj dζ dS p
stant. This constant accumulation of entropy serves as a valuable metric
Parameter B evaluates how entropy generation and degradation
in understanding the degradation process and for determining the
interact and should be a non-negative coefficient. Now, from Eq. (6), the
remaining useful life of the material. To extend this concept to electro­ ( ∫t )
chemical corrosion processes and establish a comparable failure cumulative degradation w = 0 ẇdt = w and entropy generation S′ =
∫t ′
threshold, it becomes imperative to quantify B and verify its constancy. Ṡ dt until time t, is as depicted in Eq. (7).
0
The verification is carried out for different materials under different
w⃒⃒
environmental conditions. B = ′⃒ (7)
S p
This paper embarks on a thermodynamic exploration into material
degradation induced by corrosion, focusing on applying the DEG theo­
2.2. B for corrosion
rem to find the degradation coefficient B as a fundamental parameter.
Through rigorous experimentation on low-carbon steel (LCS 1018) and
In this study, we apply the DEG theorem to establish the degradation
aluminum alloy (Al6061-T6), a direct correlation between material
coefficient B, which relates the rate at which corrosion degradation oc­
degradation and entropy generation is established, unveiling consistent
curs to entropy generation.
B values for each material under varying environmental conditions. This
Accelerated corrosion experiments are conducted using potentio­
research delves into the theoretical underpinnings of degradation pro­
static polarization, wherein a constant potential (E) is applied to the
cesses, presents novel experimental methodologies, and offers insights
working electrode with respect to a reference electrode and the resulting
into the applicability of thermodynamic principles in corrosion science.
reaction current (i) is monitored. The entropy generation attributed to
By elucidating the intricate relationship between entropy generation
material degradation caused by corrosion is assessed using Eq. (2) and
and material deterioration, this study not only contributes to the
adhering to the methodology outlined by Bryant et al. [16]. During the
fundamental understanding of corrosion phenomena but also holds
corrosion process, the charge flow (q) between the electrode is consid­
practical implications for corrosion damage prediction.
ered the phenomenological variable and the overpotential (η) of elec­
trochemical reaction is considered the driving force and the dominant
2. Theoretical background
dissipative process [29,30]. The overpotential is the difference between
the open circuit potential (OCP) EOCP or equilibrium potential and the
2.1. DEG theorem [16]
applied potential E [31], i.e., η = (E − EOCP ). During an electrochemical
reaction, η determines the kinetics of the reaction. If E > EOCP (η is
A material or system undergoing degradation shall comprise of
positive), the direction of the reaction will be oxidation, e.g.,
dissipative degradation process, p, which describes a dissipative mech­
Fe→Fe2+ +2e2− and if E < EOCP (η is negative), the direction of the re­
anism such as work, heat, energy, etc, and depends on time-dependent
action will be a reduction, e.g., Fe→Fe2+ +2e2− [31].
phenomenological variables ζ(t). The degradation process p yields a
′ From the above discussion the rate of entropy generation yields:
non-negative rate of entropy generation Ṡ ≥ 0, where the superscript ˙
represents the time rate, and degradation w is non-negative. Following ′
Ṡ =
dS′ dp dζ dS′ dp dq 1
= = (E − EOCP )
dq
(8)
the notation of Bryant et al. [16], the rate of degradation (ẇ) and entropy dp dζ dt dp dq dt T dt
( ′) ( dq)
generation Ṡ , can be determined using the chain rule: Note that in Eq. (8), the rate of flow of charge represents the
dt
dq
{ } ( ) reaction current (i) [32]. Substituting dt
= i in Eq. (8) will provide the
dw dpj dζ
ẇ pj [ζ(t)] = = YJ (1) ′
dpj dζ dt final equation for determining Ṡ .
( ) ′ 1
′{ } dS′ dpj dζ Ṡ = (E − EOCP )i (9)
Ṡ pj [ζ(t)] = = XJ (2) T
dpj dζ dt
′
Cumulative entropy generation is determined by summating the Ṡ
Bejan [26] and Kondepudi & Prigogine [27] expressed Ṡ as the
′
until time t (see Eq. (9)).
product of thermodynamic force (X) and conjugate thermodynamic flow
(J) and ẇ as the product of degradation force (Y) and J. Examples of
thermodynamic forces encompass factors leading to irreversible

2
K.P. Lijesh et al. Corrosion Science 233 (2024) 112063

Table 1
The chemical composition of LCS 1018 and Al 6061-T6.
Material C Mn P S Si Cr Cu Mg Zinc Fe Al

LCS 1018 0.13–0.20 0.3–0.9 0.04 Max 0.5 Max 0.15–0.30 - - - - Remainder -
Al 6061-T6 - 0–0.15 - - 0.4–0.8 0.4–0.8 0.05–0.4 0.8–1.2 0–0.25 0–0.7 Remainder

Fig. 1. Procedure for achieving circular area for corrosion. (a) 75 × 25 mm specimen, (b) specimen with selected circular shape tape for corrosion site (c) a forming
polymer was applied over the tape, (d) Ceramic coating paint was applied over the polymer, (e) tape was removed to achieve the desired area for perform­
ing corrosion.

∫t 3D profilometer, the roughness of the specimens was measured. The

1
S′ = (E − EOCP )idt (10) final surface roughness of the specimens was measured to be ~ 0.2 µm.
T
0 The electrochemical corrosion experiments were performed on
specimens with dimensions 75 mm × 25 mm × 3 mm (Fig. 1a) by
The sign of i follows the same sign of η. i is positive for oxidation
exposing different circular areas of substrate material (A1 = 0.5 × 102
reaction (η > 0), negative for reduction reaction (η < 0), and zero in the
mm2, A2 = 1.27 × 102 mm2, and A3 = 3.14 × 102 mm2) to the elec­
equilibrium condition (η = 0) [31]. The current i measured during the
trolyte. The desired circular areas on the test coupons were selected by
corrosion experiment by the potentiostat is the sum of anodic (ia) and
affixing a masking tape of corresponding dimensions onto the substrate
cathodic (ic) currents; nevertheless, at a higher magnitude of η, one of
material (Fig. 1b) and a layer of corrosion resistance polymer was
the currents dominates. For example, Tait [33] reported 50 times larger
applied (Fig. 1c), followed by a ceramic coating paint (Fig. 1d), masking
anodic current magnitude than the corresponding cathodic current
the taped area. The coating was applied to ensure that only the desig­
magnitude for a η = + 50mV. The equation presented in Eq. (10) shares
nated circular regions on the test coupons are susceptible to corrosion.
the same form used to characterize the entropy generation of Li-ion
Subsequently, the specimen was allowed to dry under atmospheric
batteries during their discharge [22,32].
conditions. Upon drying, the tape was removed, leaving behind the
Considering the material loss (Wvol ) during the corrosion experiment
circular substrate area for performing the corrosion experiment
until time t is considered as the measure of degradation (w), the equation
(Fig. 1e).
of B is determined from Eq. (7) as

w
∫t
T 3.2. Test procedure
B= = Wvol (11)
S′ i(EOCP − E)
0 Initial corrosion experiments were performed at room temperature
(25 ◦ C) and by developing a three-electrode electrochemical corrosion
3. Experiments and methods cell with a 3.5 wt% NaCl aqueous solution having an approximate pH
value of 6.0 as the electrolyte solution, a specimen with an exposed
3.1. Material and test equipment circular area as the working electrode (WE), graphite with 6.25 mm
diameter and 100 mm length as the counter electrode (CE), and Ag(s)/
Specimens made of low-carbon steel (LCS 1018) and aluminum alloy AgCl(s)/Cl as the reference electrode (RE) (see Fig. 2).
(Al 6061-T6) were prepared for corrosion testing. The density of LCS The degradation parameter, B, for both materials was determined by
1018 and Al 6061-T6 materials is 7850 kg/m3 and 2700 kg/m3, performing nine accelerated corrosion tests involving the application of
respectively, and their chemical compositions are provided in Table 1. three distinct potentiostat voltages of − 0.1, − 0.2, and − 0.3 V to speci­
The specimens were ground using 400, 800, 1500, and 2000 grit silicon mens with three different exposed areas (0.5, 1.27, and 3.14 × 102
carbide abrasive sheets, and then the specimens were ultrasonically mm2). The corrosion tests were electrochemically monitored using a
cleaned in de-ionized water, followed by rinsing with acetone. Using a potentiostat/galvanostat/ZRA device, and during each test, data were

Fig. 2. Three-electrode system for electrochemical corrosion test.

3
K.P. Lijesh et al. Corrosion Science 233 (2024) 112063

Fig. 3. Calculation of degradation coefficient for LCS 1018. (a) OCP followed by the applied voltage of − 0.2 V for an hour for an exposure area of A3, (b) Corrosion
current for different applied voltage for an exposure area of A3 (c) Volumetric weight loss for different applied voltage and area of exposure, (d) Plot between
volumetric weight loss and entropy generation, (e) Degradation coefficient B calculated for different applied voltage and area of exposure.

acquired at the rate of 1 data per second. The corrosion experiments 4. Results and discussion
were performed for one hour after immersing the specimen in the so­
lution and achieving a stable OCP of 0.01 V/s. To stamp the corrosion 4.1. LCS 1018 as a working electrode
behavior in the elevated temperature, experiments were performed at
35 ◦ C and 50 ◦ C. A heater controlled by a thermostat was used to achieve The evolution OCP values recorded for the LCS 1018 specimen
higher temperatures. During the potentiostat experiments, the corrosion having an area of exposure of A3 area is provided in Fig. 3a. For all nine
current was recorded, and the degradation due to corrosion was quan­ experimental conditions, the value of OCP ranged between - 0.60 to
tified by measuring the weight loss of the specimens using a digital scale − 0.64 V. The corrosion currents obtained for - 0.1, - 0.2, and - 0.3 V
with an accuracy of 0.1 mg. applied voltage to the A3 area of exposure are shown in Fig. 3b. Fig. 3b
Next, the extended experiments were conducted in five different shows that the magnitude of the corrosion current increases with the
environmental conditions: (i) Case 1–1.75 % NaCl solution at 25◦ C, (ii) increase in the applied potential. The measured weight loss for all nine
Case 2–7.0 % NaCl solution at 25◦ C, (iii) Case 3–3.5 % NaCl and 3.5 HCl experiments is shown in Fig. 3c, which shows that the weight loss in­
solution at 25◦ C, (iv) Case 4–3.5 % NaCl at 37.5◦ C and (v) Case 5–3.5 % creases with the applied voltage and the exposure area. Fig. 3d shows
NaCl at 50◦ C. the calculated entropy generation values plotted as a function of the

Fig. 4. Calculation of degradation coefficient for Al 6061-T6. (a) Plot between volumetric weight loss and entropy generation, (b) Degradation coefficient (B)
calculated for different applied voltage and area of exposure.

4
K.P. Lijesh et al. Corrosion Science 233 (2024) 112063

Fig. 5. Volumetric wear versus entropy generation for different environmental conditions (a) Case 1, (b) Case 2, (c) Case 3, (d) Case 4, and (e) Case 5. (f) B values
determined for Cases 1–5.

measured volumetric weight loss. This figure reveals the existence of a polished to 2000 grit, slight variations in surface roughness may still
linear correlation between weight loss, i.e., the measure of degradation exist, which could serve as initiation sites for corrosion pits, contributing
′
(ẇ) and entropy generation (Ṡ ). This observation indicates the propor­ to the observed differences.
tionality degradation coefficient, B, is constant in all experimental
conditions tested. The calculated B value for all nine experimental
4.2. Al 6061-T6 as a working electrode
conditions is shown in Fig. 3e. This figure shows that the B value varies
within a narrow band of 28.4–32 × 10− 9 m3K/J. The average value of B
Considering the same area of exposure and magnitude of potentiostat
from the nine experiments for the LCS 1018 with the chemical compo­
voltages, nine corrosion experiments were repeated on the Al 6061-T6
sition provided in Table 1 is BLCS1018 ~ 30 × 10− 9 m3K/J.
specimen. The magnitude of OCP from all the experiments was
The minor fluctuations shown in Fig. 3e could potentially stem from
observed to be ranging from - 0.70 to - 0.74 V. From the measured
surface imperfections, both visible and subsurface, which may slightly
corrosion current and using OCP and applied voltages, the entropy
influence the results. Moreover, despite the specimens being hand-
generation is calculated for an hour of the experiment. Fig. 4a presents

Fig. 6. Corrosion current with time for different environmental conditions for exposure area of A1 (a) different NaCl percentage in the electrolyte solution, (b) for
different electrolyte temperatures.

5
K.P. Lijesh et al. Corrosion Science 233 (2024) 112063

Fig. 7. Application B in prediction of degradation of material (a) three, six and nine slots of exposure area of A1 , (b) Corrosion current for all five cases, (c)
Comparison of measure and predicted weight loss.

the results plotted as a function of the measured volumetric weight loss graphite as the CE, and Ag(s)/AgCl(s)/Cl as RE. The experiments are
in Fig. 4a. This figure also shows a linear correlation between the repeated at 3.5 % NaCl electrolyte solution, supplying a potential of
degradation of the specimen and the entropy generation during the − 0.2 V.
corrosion. The proportionality coefficient B values determined for all the Fig. 7b shows the measured current for all five cases is provided.
experimental conditions are provided in Fig. 4b. This figure shows that Comparing Cases 1–3, it is observed that the corrosion current increases
the B value for Al 6061-T6, with the chemical composition provided in with an increase in the number of areas of exposure. Comparing Cases 1,
Table 1, varies between ~23 to ~27 × 10− 9 m3K/J. The average value of 4, and 5, it is again observed that with the rise in temperature, corrosion
nine experiments is BAl6061 ~25.5 × 10− 9 m3K/J. current increases. Now, considering the average value of B for LCS 1018
Comparing the average degradation coefficients for LCS 1018 and Al material as BLCS1018 ~ 30 × 10− 9 m3K/J, the weight loss is predicted
6061 material, BLCS1018 > BAl6061 indicates that the corrosion rate of Al from the measured entropy generation values using Eq 11. The com­
6061 is lower than that of low-carbon steels, as reported elsewhere [34]. parison of the predicted and measured weight loss for different cases is
In general, a material with a higher magnitude of B value degrades faster provided in Fig. 7c. These results reveal using B and entropy generation,
for the same entropy generation. Similar findings have been reported for the material degradation can be predicted with > 8 % accuracy.
tribo-pairs [18,35] and Li-ion batteries [22], where a softer material and
battery with lower capacity had higher B values. 5. Conclusions

4.3. B for LCS 1018 under different environmental conditions In this paper, a thermodynamic approach is applied to investigate the
material degradation due to corrosion using the degradation coefficient
Extended experiments were performed to understand the depen­ B developed in accordance with the laws of thermodynamics. The lab­
dence of B under different corrosive environments. For this purpose, the oratory experiments were conducted on LCS 1018 and Al6061-T6 by
graphite electrode is considered CE, the LCS 1018 specimen is a WE, and providing three different potentiostat voltages and for three other areas
Ag(s)/AgCl(s)/Cl is the RE. For each condition, three corrosion experi­ of exposure in a 3.5 % NaCl solution at 25 ◦ C. For all nine experiments,
ments were performed for 1 h by applying a voltage of − 0.2 V to three the degradation of the material was observed to be proportional to the
different exposure areas (A1 , A2 , A3 ). The values of volumetric wear entropy generation. The B values for LCS 1018 and Al6061-T6 were
obtained for different entropy generations for Cases 1–5 are plotted in recorded to be 30 ± 2 × 10− 9 m3K/J and 25 ± 2 × 10− 9 m3K/J,
Fig. 5a-e, respectively. This figure shows that the entropy generation is respectively. The comparison indicates that LCS 1018, with a higher
proportional to the volumetric wear rate. The calculated B values for all magnitude of the degradation coefficient, degrades faster than Al6061-
environmental conditions for LCS 1018 are shown in Fig. 5f and the T6 for the same entropy generation.
calculated B value ranged between 29.8 and 31.6 m3K/J, which is in the Five extended experiments were performed on LCS 1018 material
range predicted for 3.5 % NaCl solution at 25 ◦ C. These results indicate with three different exposure areas to study the influence of environ­
that magnitude B is independent of the environmental conditions. The mental conditions on B. from the experimental results. For all the
reason can be attributed to the counteracting impact between the environmental conditions, the material degradation was observed to be
change entropy generation and volumetric wear with environmental proportional to the entropy generation, and the B values were recorded
conditions. To illustrate, the change in the corrosion current with to be inside the range of ~ 30 ± 2 × 10− 9 m3K/J. Subsequently,
different NaCl percentages and temperature for exposure is A1 is shown leveraging the established average value of B = 30 × 10− 9 m3K/J, the
in Fig. 6a and b, respectively. These results suggest that the corrosion practical application of this degradation coefficient was demonstrated in
current increases as the NaCl percentage or the solution temperature predicting material weight loss based on entropy generation across
increases, resulting in more entropy generation and material varying numbers of exposure areas and elevated temperatures. It was
degradation. observed that across five different scenarios, the predicted weight loss
closely aligns with experimental findings, exhibiting a margin of error of
4.4. Applications less than 8 %. This demonstrates the reliability and efficacy of utilizing
the degradation coefficient B for accurate predictions of material
This section illustrates how one can use the results presented for the degradation under diverse environmental conditions.
degradation coefficient (B) to determine the corrosion rate. For this
purpose, we perform five different experiments on specimens with three Author statement
different numbers of areas of exposure of A1 (see Fig. 7a) and at three
different temperatures (25, 37.5, 50 ◦ C). The five cases are: Case 1: three The authors confirm that this is an original research manuscript and
slots of the exposed area at 25 ◦ C, Case 2: six slots of the exposed area at has not been submitted elsewhere. All authors have contributed to this
25 ◦ C, Case 3: nine slots of the exposed area at 25 ◦ C, Case 4: three slots manuscript.
of the exposed area at 37.5 ◦ C, and Case 5: three slots of exposure area at
50 ◦ C. These experiments are performed on LCS 1018 as the WE,

6
K.P. Lijesh et al. Corrosion Science 233 (2024) 112063

CRediT authorship contribution statement [11] J.W. Osborne, The Advantages of Hierarchical Linear Modeling. ERIC/AE Digest,
2000.
[12] H.H. Uhlig and R.W. Revie, Corrosion and corrosion control, 1985.
Ali Mahmoudi: Writing – original draft, Methodology, Investiga­ [13] A. Imanian, M. Modarres, A thermodynamic entropy approach to reliability
tion. K. P. Lijesh: Writing – original draft, Validation, Methodology, assessment with applications to corrosion fatigue, Entropy vol. 17 (10) (2015)
Investigation, Formal analysis, Conceptualization. Michael Khonsari: 6995–7020.
[14] H.W. Lee, H. Fakhri, R. Ranade, C. Basaran, H. Egner, A. Lipski, M. Piotrowski,
Writing – review & editing, Supervision, Project administration, S. Mroziński, Modeling fatigue of pre-corroded body-centered cubic metals with
Investigation. unified mechanics theory, Mater. Des. 224 (2022) 111383.
[15] C. Basaran, C.Y. Yan, A thermodynamic framework for damage mechanics of solder
joints, J. Electron. Packag. vol. 120 (4) (1998) 379–384, https://doi.org/10.1115/
Declaration of Competing Interest 1.2792650.
[16] M. Bryant, M. Khonsari, F. Ling, On the thermodynamics of degradation, Proc. R.
The authors have no conflict of interest to declare. Soc. A: Math., Phys. Eng. Sci. vol. 464 (2096) (2008) 2001–2014.
[17] K.P. Lijesh, M.M. Khonsari, On the useful life of tribo-pairs experiencing variable
loading and sliding speed, Wear 416 (2018) 103–114.
Data Availability [18] K.P. Lijesh, M. Khonsari, Characterization of multiple wear mechanisms through
entropy, Tribol. Int. vol. 152 (2020) 106548.
[19] J. Lin, X. Fan, P. Wang, Y. Li, Z. Shi, U. Olofsson, Gear wear prediction based on the
Data will be made available on request. theorem of degradation entropy generation, Tribol. Int. (2023) 109175.
[20] K. Doelling, F. Ling, M. Bryant, B. Heilman, An experimental study of the
Acknowledgements correlation between wear and entropy flow in machinery components, J. Appl.
Phys. vol. 88 (5) (2000) 2999–3003.
[21] B. Hajshirmohammadi, M. Khonsari, On the entropy of fatigue crack propagation,
K. P. Lijesh thankfully acknowledges a seed grant support through Int. J. Fatigue vol. 133 (2020) 105413.
the US National Science Foundation under grant number OIA-1946231 [22] K.P. Lijesh, M.M. Khonsari, A thermodynamic approach for characterizing the
and the Louisiana Board of Regents for the Louisiana Materials Design degradation of Li-ion batteries, J. Energy Storage 82 (2024) 110565.
[23] J.A. Osara, M.D. Bryant, A thermodynamic model for lithium-ion battery
Alliance (LAMDA). degradation: application of the degradation-entropy generation theorem,
Inventions vol. 4 (2) (2019) 23.
References [24] M. Naderi, M. Amiri, M. Khonsari, On the thermodynamic entropy of fatigue
fracture, Proc. R. Soc. A: Math. Phys. Eng. Sci. vol. 466 (2114) (2010) 423–438.
[25] T. Temfack, C. Basaran, Experimental verification of thermodynamic fatigue life
[1] J.E. Silva, Y. Garbatov, C. Guedes Soares, Ultimate strength assessment of prediction model using entropy as damage metric, Mater. Sci. Technol. vol. 31 (13)
rectangular steel plates subjected to a random localised corrosion degradation, (2015) 1627–1632.
Eng. Struct. 52 (2013) 295–305. [26] A. Bejan, Advanced engineering thermodynamics, John Wiley & Sons, 2016.
[2] El Ibrahimi, B., J.V. Nardeli, and L. Guo, Sustainable Corrosion Inhibitors I: [27] D. Kondepudi, I. Prigogine, Modern thermodynamics: from heat engines to
Fundamentals, Methodologies, and Industrial Applications, by CM Hussain and C. dissipative structures, John Wiley & Sons, 2014.
Verma, ASC Symp. Ser, 2021. [28] B.T. Lu, J.L. Luo, Synergism of electrochemical and mechanical factors in erosion−
[3] V.P. Singh, D. Kumar, R.P. Mahto, B. Kuriachen, Microstructural and mechanical corrosion, J. Phys. Chem. B 110 (9) (2006) 4217–4231.
behavior of friction-stir-welded AA6061-T6 and AZ31 alloys with improved [29] Robert G. Kelly, John R. Scully, David Shoesmith, Rudolph G. Buchheit,
electrochemical corrosion, J. Mater. Eng. Perform. 32 (9) (2023) 4185–4204. Electrochemical techniques in corrosion science and engineering, CRC Press, 2002.
[4] S. Sun, Q. Zheng, D. Li, J. Wen, Long-term atmospheric corrosion behaviour of [30] Kruger, Jerome, Electrochemistry of corrosion, Electrochemistry Encyclopedia
aluminium alloys 2024 and 7075 in urban, coastal and industrial environments, (2001).
Corros. Sci. 51 (4) (2009) 719–727. [31] Marcel Pourbaix, Thermodynamics and corrosion, Corros. Sci. 30 (10) (1990)
[5] Y. Cai, Y. Zhao, X. Ma, K. Zhou, Yuan Chen, Influence of environmental factors on 963–988.
atmospheric corrosion in dynamic environment, Corros. Sci. 137 (2018) 163–175. [32] Jude A. Osara, Michael D. Bryant, Thermodynamics of lead-acid battery
[6] Yu.M. Panchenko, A.I. Marshakov, T.N. Igonin, V.V. Kovtanyuk, L.A. Nikolaeva, degradation: application of the degradation-entropy generation methodology,
Long-term forecast of corrosion mass losses of technically important metals in J. Electrochem. Soc. 166 (16) (2019) A4188.
various world regions using a power function, Corros. Sci. 88 (2014) 306–316. [33] William S. Tait, Electrochemical corrosion basics. Handbook of environmental
[7] P.R. Roberge, R.D. Klassen, P.W. Haberecht, Atmospheric corrosivity modeling—a degradation of materials, William Andrew Publishing, 2018, pp. 97–115.
review, Mater. Des. 23 (3) (2002) 321–330. [34] L. Calabrese, V. Brancato, V. Palomba, E. Proverbio, An experimental study on the
[8] M. Morcillo, B. Chico, I. Díaz, H. Cano, D. De la Fuente, Atmospheric corrosion data corrosion sensitivity of metal alloys for usage in PCM thermal energy storages,
of weathering steels. A review, Corros. Sci. vol. 77 (2013) 6–24. Renew. Energy vol. 138 (2019) 1018–1027.
[9] Y. Cai, Y. Zhao, X. Ma, K. Zhou, H. Wang, Application of hierarchical linear [35] A. Aghdam, M. Khonsari, Application of a thermodynamically based wear
modelling to corrosion prediction in different atmospheric environments, Corros. estimation methodology, J. Tribol. vol. 138 (4) (2016) 041601.
Eng. Sci. Technol. vol. 54 (3) (2019) 266–275.
[10] M.V. Hernández-Lloreda, F. Colmenares, R.M. Arias, Application of hierarchical
linear modelling to the study of trajectories of behavioural development, Anim.
Behav. vol. 66 (3) (2003) 607–613.