See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/336260540

Creep deformation characteristics of microalloyed HP40Nb steel at 950 °C

Article in Journal of Mechanical Science and Technology · October 2019

DOI: 10.1007/s12206-019-0922-9

CITATIONS READS

11 1,682

4 authors, including:

Dao Van Hung Kee Bong Yoon

Korea Research Institute of Standards and Science Chung-Ang University
13 PUBLICATIONS 103 CITATIONS 150 PUBLICATIONS 1,326 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Dao Van Hung on 08 May 2020.

The user has requested enhancement of the downloaded file.

 Journal of Mechanical Science and Technology 33 (10) (2019) 4813~4821
www.springerlink.com/content/1738-494x(Print)/1976-3824(Online)
DOI 10.1007/s12206-019-0922-9

Creep deformation characteristics of microalloyed HP40Nb steel at 950 oC†

Van Hung Dao, Jung Soo Song*, Joo Yong Kim and Kee Bong Yoon
Department of Mechanical Engineering, Chung-Ang University, 84 Heukseok-ro, Dongjak-gu, Seoul, Korea

(Manuscript Received June 23, 2019; Revised July 18, 2019; Accepted July 22, 2019)

----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Abstract

In the present work, several specimens were obtained from microalloyed HP40Nb reformer tubes with the different serviced period.
The creep tests were conducted at a temperature of 950 °C and under stress levels in the range 25-55 MPa. Results indicated that the
degradation of the creep life was due to the increase of the serviced period. The microstructural degradation and creep voids were ob-
served at the grain boundary regions, attributed to the main reason leading to the premature creep failure. The creep exponent started
decreasing after the servicing and approached a constant value after 7.2 years. The creep coefficient was obtained to be a function of time
when the creep exponent was constant. Norton’s creep power law equation was used to derive an equation describing the relationship
between the diametrical expansion and serviced period of the reformer tubes. Larson-Miller curves were obtained from creep test data.
Amount of carbides from the different serviced period was also calculated.

Keywords: HP40Nb alloy; Reformer tube; Carbide; Creep test; Larson-Miller parameter
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

persion of secondary carbides in the matrix of austenite grains.

1. Introduction
They act as a barrier to prevent the propagation of creep dam-
Over the past five decades, reformer tubes of microalloyed age through the microstructure. In addition, researchers have
HP40Nb—an austenitic heat-resistant alloy—have been demonstrated that creep properties of a material can be identi-
manufactured using the centrifugal casting process for high fied from the metallurgical features of the material, type of
strength and high creep resistance [1]. Therefore, the tubes distribution of carbides in the material microstructure, and
could be used at high temperatures and under high stresses, mechanical properties of the material [8, 9].
and they had a long life span under standard operating condi- Reformer tubes operate in extremely harsh environmental
tions. The microstructural features of HP40Nb can be stabi- conditions [10]. The conditions may include high tempera-
lized by adding a microalloying element (Nb and Ti). Microal- tures, high stresses, and corrosive factors. The failure of re-
loying elements have been documented to play an important former tubes has been attributed to creep, which often leads to
role in promoting not only the fragmentation and dispersion of microstructure destabilization. In particular, the coalescence
hard deformation-resistant carbide particles of the as-cast mi- and coarsening of carbides and the formation of creep voids
crostructure but also the partial replacement of chromium (Cr) during service reduce creep deformation [1, 2]. Hence, it
carbide; they form an MC-type carbide (M: Metal with Nb or should recommend investigating creep on the serviced period
Ti), which enhances the creep strength of steel [2-7]. Thus, for microalloyed HP40Nb steel reformer tubes for better and
microalloyed HP40Nb is popular for manufacturing reformer safer design. In the present work, for experiments, the speci-
tubes of heating furnaces. mens were used for comparing creep tests at the temperature
Microalloyed HP40Nb steel has high Cr (25 wt.%) and Ni of 950 °C and for stresses in the range 25-55 MPa. The initial
(35 wt.%) content in its austenitic structure, fulfilling the main results indicated that the variation in creep properties was
requirement for creep-resistant materials. The microstructure related to the serviced period. The levels of microstructural
consists of primary and secondary carbide (M7C3/MC and degradation were studied carefully to explain changes in the
M23C6) precipitates at the grain boundary regions, which con- creep deformation behavior. Furthermore, for a constant creep
tribute to the high creep resistance of the material [7]. In the exponent, the relationship between an increase in the tube
early stages of its use, the small carbide particles are the dis- diameter and the serviced period was determined based on the
*
creep coefficient, which was identified to be a function of time.
Corresponding author. Tel.: +82 2 8211547
E-mail address: songjs@cau.ac.kr The use of the Larson-Miller diagram for assessing predic-
†
Recommended by Editor Chongdu Cho tions of the remaining life is proposed. Furthermore, carbide
© KSME & Springer 2019
4814 V. H. Dao et al. / Journal of Mechanical Science and Technology 33 (10) (2019) 4813~4821

precipitation is determined to explain the tendency of carbide Table 1. Chemical composition of microalloyed HP40Nb according to
grains to coarsen. ASTM A608 (wt.%).

Material C Ni Cr Nb Si Mn Mo Fe
2. Experimental procedures
Microal-
0.38- 24- 0.5- 0.5-
2.1 Reformer tube information and specimen preparation loyed 34-37 ≤1.5 0.50 Bal
0.45 27 1.5 1.5
HP40Nb
Reformer tubes are mainly manufactured through the cen-
trifugal casting process, and their nominal lifetime is ap- Table 2. Detailed information on the reformer tubes.
proximately 100000 h (≈11.4 years). They can operate at tem-
Tube No. #17-1 #19 #17-2 #56
peratures up to 950 °C and are designed for internal pressures
of 10-40 bars [1]. In this study, reformer tubes were obtained Installed 2014 2012.08 2008.01 2008.01
from a furnace containing 72 catalyst-filled vertical tubes ar- Replaced 2014 2013.12 2013.12 2015.03
ranged in two radiant chambers. These vertical tubes with a Serviced
length of 14 m were supported at the bottom and free to ex- period 0 (virgin) 1.75 6.0 7.2
(years)
pand at the top. The outer diameter of each tube was
146.6 mm and the internal diameter was 135.8 mm. The raw
materials were metal gas (CH4) and steam, which entered into
the top of reformer tubes from the top end at a temperature of
around 755 K and with a pressure in the range 15-17 bars.
They flowed downward over the catalyst in the individual
tubes before emerging from the bottom end at a temperature
of around 1173 K.
The material used in this study was microalloyed HP40Nb,
and its chemical composition is presented in Table 1. Virgin (a)
specimens were cut from tubes collected from the bottom part
outside the furnace. Other specimens were obtained from
tubes collected from the heated part of the furnace, and the
serviced periods of the tubes were 1.75, 6, and 7.2 years. De-
tailed information and photographs of the reformer tubes are
presented in Table 2 and Fig. 1. These specimens were taken
out and brought to the laboratory for examination and analysis.

2.2 Creep test

The creep test specimens were fabricated according to the

ASTM E8/E8M standard [11] by wire-cut electrical discharge
machining, and their diameter and gage length were 7 and
60 mm from the longitudinal direction of the tubes (see Fig.
1(b)).
The creep test procedure was in accordance with the ASTM (b)
E139 standard [12], and the tests were conducted creep testing
machines at 950 °C and under stresses in the range 25-55 MPa. Fig. 1. (a) Photographs of the reformer tubes; (b) specimen geometry.
The test conditions are summarized in Table 3; a total of 16
creep tests were conducted. Creep curves showing the creep using an ultrasonic machine for 2 min in ethanol solution.
strain as a function of the elapsed time were directly received Finally, they were etched with a solution of 10 % HNO3, 30 %
and plotted on a graph on the computer screen. The creep HCl, and 20 % glycerol. Metallographic characterization was
strain versus time data was converted to creep strain rate ver- performed using light optical microscopy (LOM) and scan-
sus elapsed time data by using the seven-point incremental ning electron microscopy (SEM), and an energy-dispersive X-
method specified in the ASTM E647 standard. ray spectroscopy (EDS) system.
To identify microstructural changes at the grain boundaries,
rectangular specimens were cut from the tubes so that they 3. Results
could be held in the hand for microstructure analysis (see Fig.
3.1 Creep curves and microstructural degradation
1(b)). The specimens were ground using silicon carbide pa-
pers with a grit size of 320-2400 and water, polished on a Typical creep curves of microalloyed HP40Nb creep tested
cloth with 0.6 and 0.1 µm diamond suspensions, and cleaned at a constant temperature of 950 °C and under stresses in the
 V. H. Dao et al. / Journal of Mechanical Science and Technology 33 (10) (2019) 4813~4821 4815

range 25-55 MPa are shown in Figs. 2(a)-(d). A normal creep

curve shows three regions (primary, secondary, and tertiary
creep), and the secondary and tertiary creep regions are more
dominant than the primary creep region. The steady-state
stage is characterized by the secondary creep region. In addi-
tion, the creep life often depends strongly on the applied stress
and temperature. The tertiary stage marks the start of micro-
structural degradation. Figs. 3(a)-(d) show plots of creep strain
rate versus elapsed time for different stress levels (25-
55 MPa); roughly three creep regimes are visible. At the first
stage, the creep strain rate initially decreases with an increase
in the creep strain. The secondary stage, creep strain rate tends (a) As-cast (virgin)
to be constant which is dominant throughout the creep life,
and in the final stage, the creep strain rate increased with a
further increase in the creep strain. The minimum creep strain
rate values were obtained from the secondary creep rate by
determining the average value of the strain rate with data in
the range of 30 %-60 % of creep rupture time for each speci-
men, shown in Table 3. These values are also directly propor-
tional to the applied stress, and this relationship is plotted (log-
log) in Fig. 4. A, n are creep coefficient and creep exponent
values were determined using Norton’s power law creep equa-
tion, as follows:

e& = As n . (1)
(b) Specimen with serviced period of 1.75 years

For the serviced periods 0 (virgin), 1.75, 6.0, and 7.2 years,
the creep exponent was 11.19, 9.14, 6.40, and 6.28, respec-
tively, confirming that power law hardening decreased with an
increase in the serviced period. It should be noted that the
creep exponent value of the virgin specimen was considerably
higher than the values for the other specimens at 950 °C. The
creep exponent values also indicates that the secondary-
regime creep deformation in microalloyed HP40Nb is con-
trolled by the dislocation creep mechanism. According to past
studies [13-15], while dislocation creep is dominant for creep
exponent values in the range 3-12 because of the movement of
dislocations, diffusion creep and grain-boundary sliding are
dominant for the creep exponent values of 1 and 2, respec- (c) Specimen with serviced period of 6 years
tively. Fig. 5 depicts the detailed microstructure at two magni-
fications, ×100 and ×1000, for the different serviced periods
for the specimens from the tubes. The virgin microstructure of
microalloyed HP40Nb in Fig. 5(a) shows a continuous net-
work of primary carbides forming a dendritic pattern during
solidification. However, in the microstructure of the specimen
with a serviced period of 1.75 years in Fig. 5(b), a large quan-
tity of small precipitates are observed, they are more apparent
at the grain and neighborhood boundaries. The interdendritic
carbide content decreased for the serviced periods of 6 and 7.2
years (see Figs. 5(c) and (d)), and the blocky-shaped interden-
dritic carbides dispersed and formed large granular carbides.
EDS analysis of the virgin specimen, shown in Fig. 6(a), re-
(d) Specimen with serviced period of 7.2 years
vealed that the white precipitates correspond to Nb-rich car-
bides that combined with carbon atoms to form fine precipi- Fig. 2. Plot of time versus strain obtained in the creep test at 950 °C for
tates of new MC-type carbides (M: Metal with Nb or Ti) that microalloyed HP40Nb under different stresses.
4816 V. H. Dao et al. / Journal of Mechanical Science and Technology 33 (10) (2019) 4813~4821

Table 3. Parameters obtained at different stress levels for each of the

specimens in creep test for microalloyed HP40Nb at 950 oC.

Serviced Applied Time to Minimum

Temperature
period stress rupture creep strain
(°C)
(years) (MPa) (h) rate (%/h)
40 225.03 2.82E-02
45 94.48 1.08E-01
Virgin
50 40.97 3.45E-01
55 20.98 1.00E+00
35 205.41 6.36E-02
40 90.17 2.11E-01
1.75
(a) As-cast (virgin) 45 31.37 6.78E-01
50 14.98 1.60E+00
950
30 252.43 1.06E-01
35 103.97 3.08E-01
6
40 47.37 6.19E-01
45 22.63 1.51E+00
25 554.30 5.24E-02
30 148.60 1.43E-01
7.2
35 71.92 4.09E-01
40 32.31 9.83E-01

(b) Specimen with serviced period of 1.75 years

(a)

(c) Specimen with serviced period of 6 years

(b)
(d) Specimen with serviced period of 7.2 years
Fig. 4. Relationship of the minimum strain rate with stress and serviced
Fig. 3. Plot of time versus creep strain rate obtained in the creep test at period after the creep testing of microalloyed HP40Nb specimens: (a)
950 °C for microalloyed HP40Nb. Before; (b) after retaining n value as a constant.
 V. H. Dao et al. / Journal of Mechanical Science and Technology 33 (10) (2019) 4813~4821 4817

(a) As-cast (virgin)

(a) Virgin specimen

(b) Specimen with serviced period of 1.75 years

(c) Specimen with serviced period of 6.0 years (b) Serviced specimen (7.2 years)

Fig. 6. EDS analysis results for the dark and bright carbide precipitate
region.

(indicated by point 5), as reported in a previous study [4].

The microstructural changes highlighted the progressive re-
duction in the material’s stress resistance capability; this re-
duction is termed creep damage. The results of the creep tests
(d) Specimen with serviced period of 7.2 years. performed for a constant temperature of 950 °C and a stress
Fig. 5. LOM and SEM images of the surface of microalloyed HP40Nb level of 40 MPa for all specimens showed that damage growth
for the specimens with different serviced periods (magnifications of manifests as an increase in the minimum creep rate during
×100 and ×1000). secondary creep and decreases the rupture time (see Table 3).
Results showed the creep deformation decreasing with an
are stable at high temperatures (indicated by point 1). Thus, increase in the serviced period; this finding also confirmed
the formation of the fine precipitates enhanced the creep that the creep exponent values could show little change and
strength. By contrast, the precipitates in black contrast could remain constant after 7.2 years, as shown in Fig. 4(a). This
be identified as Cr-rich carbides (M7C3 and/ or M23C6; M: observation is important for deriving the relation between the
Metal with high Cr content) (indicated by point 2) that have diameter extension and the serviced period for predicting as
adverse effects on the mechanical properties of the material [1, well as evaluating the residual life of reformer tubes. Fig. 4(b)
2, 8]. For long serviced periods, the transformation of primary shows plots of the minimum creep rate versus stress level for a
carbides into secondary carbides is detected in EDS analysis, creep exponent of 8.25, which is the average of the creep ex-
as shown in Fig. 6(b). The transformation from the initial ponent values for all the specimens, and a high R2 value of
M7C3 carbides into M23C6 at the grain boundaries can be ob- 0.98. The creep coefficient A was 2.78×10-15, 1.36×10-14,
served mainly in the case of Cr-rich carbides with dark grey 4.70×10-14, and 8.85×10-14 for the serviced periods 0 (virgin),
precipitates (indicated by point 3), while the observation of 1.75, 6.0, and 7.2 years. The relation between the creep coeffi-
Nb-rich carbides with some amount of nickel and silicon sug- cient and the serviced period obtained in this study was a sec-
gests that the primary carbide MC (NbC) tends to transform ond order polynomial equation, and the accuracy expressed in
into G-phase, which consists of Ni16Nb6Si7 (indicated by point terms of the coefficient of determination (R2) was 0.9743; the
4). In particular, (NbTi)C carbides are more stable than NbC relation is nonlinear and the relation obtained is as follows:
4818 V. H. Dao et al. / Journal of Mechanical Science and Technology 33 (10) (2019) 4813~4821

Log10 A = -0.0223t S2 + 0.3509 t S - 14.5060 (2)

where tS is the serviced period of the tube (in years). This

equation was derived using data for the serviced period range
0-7.2 years, and it can be used to predict the creep coefficient.
In practice, one of the main reasons for changing the tube
diameter is the hydrocarbon cracking reaction that occurs
under high-temperature conditions and the stress concentra-
tion results from internal pressure. In general, in most cases,
tube destruction occurs as a result of strain propagation in the
longitudinal direction. The tangential and radial stress are
illustrated by Lame’s equation, can be expressed as follows
for exclusively considering the effect of the internal pressure
Fig. 7. Larson-Miller diagram for microalloyed HP40Nb with C =
on tube failure: 22.9.

ri 2 Pi - ro2 Po ( Pi - Po ) ri ro
2 2
From Eq. (7), the relationship between diameter extension
sq = + (3)
ro2 - ri 2 ( ro2 - ri 2 ) r 2 and serviced period (year) can be studied by considering the
equivalent stress distribution and the creep coefficient as func-
ri 2 Pi - ro2 Po ( Pi - Po ) ri ro
2 2

sr = - (4) tions of time. Recently, many researchers have conducted

2
ro - ri 2
( ro2 - ri 2 ) r 2 failure analysis of serviced exposed primary reformer tubes
made of microalloyed HP40Nb steel. They measured and
where σθ and σr are the tangential and radial stress at a point calculated tube diameters extremely precisely to obtain the
on the tube wall; these stresses are directed circumferentially maximum limits preceding tube failure [19]. In a couple of
and radially, respectively [16]. The maximum and minimum past studies [2, 20], the authors found that all failed tubes
stress values are identified at r = ri and r = ro, respectively, showed a large longitudinal crack that opened up over a length
where ri and ro are the internal and external radii of the tube (in of 300 to 880 mm. The diameter extension is in the range 3 %-
millimeters). Pi and Po denote the internal and external pres- 4.5 % which can be considered as the standard for replacing
sure (in megapascals), it is considered that Po = 0. Eqs. (3) and tube components for avoiding the risk of damage to the entire
(4) indicate that the strong influence of the tube diameter on system and the time to rupture can be obtained from Eq. (7).
the equivalent stress should be determined based on Tresca
theory [17]; this influence is described by Eq. (5) below.
3.2 Creep life measurement by using Larson-Miller curve
2 2
2 Pr r
i i o The Larson-Miller curve is extensively used in industries to
s = sq - s r = . (5)
(ro2 - r )r
i
2 2
predict the creep life of materials since remaining-life assess-
ment is important for choosing the right material for certain
From a microstructural viewpoint, the authors of a past engineering applications. In the most extreme instances, it
study [18] observed that the stress concentration usually helps prevent component failure, which could otherwise lead
causes the condensation of atomic vacancies by breaking to catastrophic effects. In most engineering applications, it is
down the sustainable network structure, and the condensation necessary to choose materials and design components by con-
of vacancies significantly affects creep voids formation. Thus, sidering creep damage. The prediction of the creep life of
void formation is strongly dependent on stress. components subjected to high temperatures is necessary to
The integration of Eq. (1) was performed to describe the re- avoid catastrophic effects.
lationship between diameter extension and serviced period. The Larson-Miller curve shown in Fig. 7 was obtained from
an analysis of the results of creep tests in Table 3. The remain-
es ts
n
ing creep life was calculated using the Larson-Miller parame-
ò de = ò A(t ).s dt .
eo to
(6)
ter (LMP) curve and substituting the stress rupture time ob-
tained in each test in the LMP equation, which is given by:
here, ε is the diameter extension (in percentage), and σ is the
equivalent stress (in megapascals). Substituting Eq. (5) into Eq. P = (T + 273) x (22.9 + Log t r ) (8)
(6) gives the following equation:

es ts 2 2
n where P is the LMP parameter, tr is the stress rupture time, T
é 2 Pr i i ro
ù
òe d e = òt A(t ). ê 2 2 2 ú
dt . (7) is the temperature (in kelvin), and C = 22.9 is the Larson-
o o ë (ro - ri )r û Miller constant. Comparison of the curves (shown in Fig. 7)
 V. H. Dao et al. / Journal of Mechanical Science and Technology 33 (10) (2019) 4813~4821 4819

(a) Specimen with serviced period of 1.75 years

Fig. 9. Influence of serviced period on amount of carbide of microal-

loyed HP40Nb for the specimens with different serviced periods.

3.3 Amount of carbides measurement

The microstructure of the specimens was observed at rela-

tively low magnification (×50). The morphology of the car-
bides at the grain boundaries could be seen and their charac-
teristics could be accurately analyzed by the image analysis
method, as shown in Figs. 8(a)-(c). For the serviced periods
(b) Specimen with serviced period of 6 years 1.75, 6.0, and 7.2 years, amount of carbides was 15.21 %,
16.83 %, and 17.78 %, respectively. The result indicated that
the carbides increased with the serviced period. This explains
why the coarsening of carbides became clear after a long ser-
viced period. This result is in agreement with the result of a
past study [23].

4. Discussion
The creep exponent tends to be influenced by microstruc-
tural degradation, which results from the coalescence and
coarsening of primary and secondary carbides, as shown in
(c) Specimen with serviced period of 7.2 years Fig. 5. The precipitate of small carbide particles at the inter-
dendritic interface could be explained by the transformation of
Fig. 8. Images analysis on the surface of microalloyed HP40Nb for the
primary carbides into intergranular- or intragranular-type sec-
specimens with different serviced periods (magnifications of ×50).
ondary carbides. The small secondary carbide particles formed
acted as a barrier, preventing creep deformation, and the creep
for manufacturer data [2], previously reported creep life val- strength of microalloyed HP40Nb increased. Similar results
ues [8, 19], and the virgin and serviced period specimens lead have been reported previously [9, 24]. The microstructural
to the following inferences: changes appeared to reach saturation because of the coales-
Creep test results for the virgin specimen were within the cence and coarsening of carbides being dominant for these
limits of the manufacturer data. The curve for the virgin serviced periods. In addition, creep voids occurred, which can
specimen is located above that for previously reported creep be verified from material degradation resulting from high
test values for a HP40Nb modified material The superior operating temperatures [20, 25, 26]. This occurrence is a pre-
creep properties of microalloyed HP40Nb was obtained which requisite for microcracking and the propagation of macro-
allows to manufacture tubes with thinner wall or, larger di- cracks. Furthermore, high-temperature operation (above
ameter. The creep properties can be explained by the presence 950 °C) for prolonged durations accompanied by internal
of the microelement Ti in microalloyed HP40Nb; the micro- pressure concentration can lead to microstructural changes and
element enhances the creep property [1, 2, 7, 21, 22]. severe damage to structural tube components as a result of
The curves for the specimens with serviced periods of 1.75, temperature-sensitive plastic deformation and surface degra-
6, and 7.2 years are below the virgin and manufacturer creep dation processes. Consequently, the creep resistance is re-
curves. This is because of creep damage accumulated in the duced. Similar results were obtained in a previous report [2].
specimens, and it will be discussed in a later section. Also, the degree of coarsening of grain boundaries should be
4820 V. H. Dao et al. / Journal of Mechanical Science and Technology 33 (10) (2019) 4813~4821

proportional to the serviced period or temperature. It is also 500060), granted financial resource from the Ministry of
consistent with previous results [23, 27, 28]. Trade, Industry & Energy (MOTIE), Republic of Korea. This
Fig. 9 also shows a comparison of the amount of carbides work was also supported by the KETEP granted from MOTIE
between the specimens of this study and a HP40Nb modified (No. 20141010101850). Authors deeply appreciate the com-
material of a previous study at the temperature of 900 °C [23]. pany providing the serviced reformer tubes and relevant in-
Apparently, amount of carbides in HP40Nb steel is greater formation.
than that in microalloyed HP40Nb containing titanium. This
can be explained a follows. The addition of titanium possibly Nomenclature------------------------------------------------------------------------
controls and restrains the coarsening. Moreover, titanium at-
oms remain in their positions without changing the crystalline A : Creep coefficient
volume during the transformation of carbides under high tem- C : Larson-Miller constant
peratures in the microalloyed material. In the EDS analysis, as n : Creep exponent
shown in Fig. 6, a small amount of Ti (0.52; 0.22 wt.%) was σ : Applied stress
detected on the white precipitates, which included fine pre- e& : Minimum creep strain rate
cipitation of the new MC-type carbides ((NbTi)C); the pres- P : Pressure
ence of Ti appears to have enhanced the creep properties. Po : External pressure
Similar arguments have been presented in previous studies [1, Pi : Internal pressure
2, 7, 21, 22]. In other words, the carbide content is depended r : Radii of tube
on the serviced period, which ranges from 1.75-7.2 years. The ro : External radii of tube
remaining life of tubes after the serviced period should proba- ri : Internal radii of tube
bly be evaluated after the amount of carbides of the tube is R2 : Coefficient of determination
determined. T : Temperature
tS : Serviced period
tr : Rupture time
5. Conclusions
Variations in the creep behavior at the temperature of
References
950 °C, remaining life assessment, and the carbide precipitate
of microalloyed HP40Nb steel for different service periods [1] L. Bonaccorsi, E. Guglielmino, R. Pino, C. Servetto and A. Sili,
were investigated. The following conclusions were obtained: Damage analysis in Fe-Cr-Ni centrifugally cast alloy tubes for
(1) On the basis of microstructural degradation with the ser- reforming furnaces, Engineering Failure Analysis, 36 (2014)
viced period, creep life and creep deformation were reduced. 65-74.
The creep exponent was found in the range of 6.28-11.19, [2] J. Swaminathan, K. Guguloth, M. Gunjan, P. Roy and R.
indicated the existence of the dislocation creep mechanism in Ghosh, Failure analysis and remaining life assessment of ser-
microalloyed HP40Nb steel. vice exposed primary reformer heater tubes, Engineering Fail-
(2) The creep exponent decreased non-linearly and it was ure Analysis, 15 (2008) 311-331.
constant for serviced periods of 7.2 years under different [3] A. R. Andrade, C. Bolfarini, L. A. M. Ferreira, A. A. A. Vilar,
stresses at 950 °C. For a constant creep exponent value of 8.25, C. D. Souza Filho and L. H. C. Bonazzi, Influence of niobium
the creep coefficient was a function of time. The equation addition on the high temperature mechanical properties of a
proposed for predicting the diameter extension (%) and ser- centrifugally cast HP alloy, Materials Science and Engineering
viced period (year) was used to assess the residual life of tube A, 628 (2015) 176-180.
components. [4] L. H. de Almeida, A. F. Ribeiro and I. Le May, Microstructural
(3) Larson-Miller curves were plotted by using most of the characterization of modified 25Cr-35Ni centrifugally cast steel
test data obtained in this study. The results indicated the supe- furnace tubes, Materials Characterization, 49 (2002) 219-229.
rior creep properties of microalloyed HP40Nb compared to [5] Y. Zhang, Y. Sun, S. Guan, X. Deng and X. Yan, Effect of
that of HP40Nb modified which allows to manufacture tubes titanium and tungsten on the structure and properties of heat-
with lower thin wall and/or larger diameter. Also, the curves abrasion resistant steel, Materials Science and Engineering A,
are in good agreement with the general tendency of standard 478 (1-2) (2008) 214-220.
curves documented in the literature. Amount of carbides de- [6] G. Pilloni, E. Quadrini and S. Spigarelli, Interpretation of the
termined to be 15.21 %, 16.83 %, 17.78 % with serviced peri- role of forest dislocations and precipitates in high-temperature
ods of 1.75, 6, and 7.2 years, respectively. It explained the creep in a Nb-stabilised austenitic stainless steel, Materials Sci-
tendency of carbide grains to coarsen. ence and Engineering A, 279 (1) (2000) 52-60.
[7] A. Alvino, D. Ramires, A. Tonti and D. Lega, Influence of
chemical composition on microstructure and phase evolution of
Acknowledgments two HP heat resistant stainless steels after long term plant-service
This work was supported by the KETEP (No. 2013 2010 aging, Material at High Temperature, 31 (1) (2014) 2-11.
 V. H. Dao et al. / Journal of Mechanical Science and Technology 33 (10) (2019) 4813~4821 4821

[8] R. Voicu, J. Lacaze, E. Andrieu, D. Poquillon and J. Furtado, tubes, Engineering Failure Analysis, 6 (2) (1999) 101-112.
Creep and tensile behaviour of austenitic Fe-Cr-Ni stainless [26] W. Z. Wang, F. Z. Xuan, Z. D. Wang, B. Wang and C. J. Liu,
steels, Materials Science and Engineering A, 510 (2009) 185- Effect of overheating temperature on the microstructure and
189. creep behavior of HP40Nb alloy, Material and Design, 32 (7)
[9] D. Alessio, G. Gonzalez, V. F. Pirrone, L. Iurman and L. Moro, (2011) 4010-4016.
Variation of creep properties in HP steel by influence of tem- [27] V. H. Dao, K. B. Yoon, G. Yang and J. S. Oh, Determination
perature, Procedia Materials Science, 1 (2012) 104-109. of creep constitutive model for 28-48WCo alloy based on ex-
[10] J. Lee, J. H. Han, J. H. Moon, C. H. Jeong, M. Kim, J. Y. Kim perimental creep tests at 817-982 °C, Journal of Mechanical
and S. H. Lee, Characteristics of heat transfer and chemical re- Science and Technology, 32 (2018) 4201-4208.
action of methane-steam reforming in a porous catalytic me- [28] J. M. Gong, S. T. Tu and K. B. Yoon, Damage assessment
dium, Journal of Mechanical Science and Technology, 30 (1) and maintenance strategy of hydrogen reformer furnace tubes,
(2016) 473-481. Engineering Failure Analysis, 6 (1999) 143-153.
[11] ASTM Standard E8.8M, Standard test methods for tension
testing of metallic materials.
[12] ASTM3-01-E139, Standard test method for conducting creep, Van Hung Dao received his M.S degree
creep-rupture, and stress-rupture test of metallic materials. in Mechanical Engineering from Chung-
[13] S. Latha, M. D. Mathew, P. Parameswaran, K. B. S. Rao and Ang University, Korea in 2016. He is
S. L. Mannan, Thermal creep properties of alloy D9 stainless currently a Ph.D. candidate at Chung-
steel and 316 stainless steel fuel clad tubes, Int. of Pressure Ang University. His research interests
Vessels and Piping, 85 (2008) 866-870. are microstructural analysis and applica-
[14] A. Ghatak and P. S. Robi, A comparative study of constitutive tion of high temperature fracture me-
equations for the creep deformation of HP40Nb micro-alloyed chanics to life assessment of structural material. He is extend-
steel, Materials Science and Engineering A, 648 (2015) 418-427. ing research to behavior of additive manufactured materials.
[15] A. Ghatak and P. S. Robi, High-temperature deformation
behavior of HP40Nb micro-alloyed reformer steel, Metallogra- Jung Soo Song received his Master
phy, Microstructure and Analysis, 4 (6) (2015) 508. degree in Mechanical Engineering from
[16] B. N. Phalguna, Stress and failure analysis of thick walled Chung-Ang University, Korea in 2002.
cylinder with oblique hole, Int. Journal of Engineering Re- He is currently a Ph.D. candidate at
search and Technology, 6 (2017) 2278-0181. Chung-Ang University. And he is also a
[17] https://en.wikipedia.org/wiki/Henri_Tresca. Senior Researcher in Energy Safety
[18] W. D. Nix, Introduction to the viewpoint set on creep cavita- Research Institute at Chung-Ang Uni-
tion, Scripta Metallurgica, 17 (1983) 1-4. versity. His research interests are high
[19] E. Guglielmino, R. Pino, C. Servetto and A. Sili, Handbook of temperature fracture mechanics, remaining life assessment
Materials Failure Analysis with Case Studies from the Chemi- and risk-based inspection for industrial facilities.
cals, Concrete and Power Industries, Elsevier Ltd. (2016) 69-
91. Joo Yong Kim received his M.S. degree
[20] A. K. Ray, S. K. Sinha, Y. N. Tiwari, J. Swaminathan, G. Das, in Mechanical Engineering from Chung-
S. Chaudhuri and R. Singh, Analysis of failed reformer tubes, Ang University, Korea in 2016. He is
Engineering Failure Analysis, 10 (3) (2003) 351-362. now working on Korea Company. His
[21] A. R. Andrade, C. Bolfarini, L. A. M. Ferreira, C. D. Souza research interests are microstructural
Filho and L. H. C. Bonazzi, Titanium micro addition in a cen- analysis, reliability and life assessment
trifugally cast HPNb alloy: High temperature mechanical prop- of structural material such as reformer
erties, Materials Science and Engineering A, 636 (2015) 48-52. tube.
[22] B. Piekarski, Effect of Nb and Ti additions on microstructure,
and identification of precipitates in stabilized Ni-Cr cast austen- Kee Bong Yoon received his M.S. in
itic steels, Materials Characterization, 47 (2001) 181-186. Mechanical Engineering from KAIST
[23] C. J. Liu and Y. Chen, Variations of the microstructure and and Ph.D. from Georgia Institute of
mechanical properties of HP40Nb hydrogen reformer tube with Technology. He is currently a Professor
time at elevated temperature, Materials and Design, 32 (4) at Chung-Ang University. His research
(2011) 2507-2512. interests are high temperature fracture
[24] M. H. Shariat, A. H. Faraji, A. Ashraf Riahy and M. M. Ali- mechanics, failure analysis and risk
pour, In advances creep failure of HP modified reformer tubes based management of facilities in semi-
in an ammonia plant, Journal Corrosion Science and Engineer- conductor industry as well as the conventional power and
ing (2003). process plants. He is extending research to fracture of additive
[25] K. B. Yoon and D. G. Jeong, Oxidation failure of radiant heater manufactured materials.

View publication stats