Aging investigation of metals of the pipes in Lithuanian Power Station/ Lietuvos elektrines vamzdynij metalij senejimo prognozavimas.
Daunys, M. ; Dundulis, R. ; Karpavicius, R. 等
1. Introduction
In order to supply hot steam in the thermal power stations from
heat-exchanger to turbine, long pipes are used. Therefore, the
efficiency of such pipeline depends on the load, temperature and
aggressive influence of hydrogen from the supplied hot steam. The
technological parameters of the supplied hot steam are very high:
temperature reaches up to 570[degrees]C, the working pressure of [empty
set] 219 pipe when thickness of the wall is 28.5 mm-13.2 MPa and the
working pressure of [empty set] 245 pipe when thickness of the wall is
45 mm-25.4 MPa. These pipelines can be damaged by temperature, tension
strengths, own weight (including isolation) and vibrations caused by the
changes in vapour pressure and dynamic loads from unbalanced rotors of
pumps [1-6]. Walls of the pipes are mechanically treated during
manufacture and the pressure results in thickness decreasing of the
wall. Thus the state of residual tension strengths initiates the cracks.
This is the reason, why special attention is paid to the residual
tension strengths, which appear during manufacture of thick-walled
pipes.
The attention in this work was concentrated on the investigation of
mechanical characteristics of the material of straight part of the pipe
depending on exploitation life and operating temperature. Mechanical
properties of the material in pipeline depends on temperature, because
when the temperature is increasing, strength of the steel is decreasing
and plasticity increases, while the increasing exploitation life causes
degradation of metal's structure and increased saturation of the
metal with hydrogen, which increases metal fragility and reduces its
plasticity [7-9]. The hydrogen's diffusion into metal is happening
within limits of its structure's grains. Such saturation of the
metal with hydrogen weakens the interaction forces between metal's
grains and stimulates disintegration of the boundaries of mosaic blocks.
Even small changes of tension strengths in such local areas induce
cracks.
[FIGURE 1 OMITTED]
2. Methodology of investigation of mechanical characteristics
While performing this work, mechanical characteristics of the
material, steel 12Ch1MF, of hot steam supply pipeline for AB
"Lietuvos elektrine" (Lithuanian Power Station Ltd.) (outer
diameter 219 mm, wall's thickness 28.5 mm) were analyzed: the limit
of proportionality - [[sigma].sub.pl], yield strength - [R.sub.p0,2],
tension strength - [R.sub.m], fracture stress - [[sigma].sub.f],
reduction of cross-section area - Z .
The listed mechanical characteristics were determined in the
straight part of the pipeline in a new pipe and the pipes after the
exploitation of 45000 and 16000 h at 20[degrees]C temperature and at
operating temperature of 550[degrees]C. The specimens were cut in the
longitudinal direction, as it is shown in Fig. 1. The tension tests were
performed using the 25 kN testing stand [10].
The force was measured with strain gauge attached to circular
cross-section dynamometer fixed in the top catch of the testing machine,
and the displacement of specimens was measured using the transverse
deformometer [11, 12]. The testing stand was calibrated in the State
Metrological Centers in Vilnius and Kaunas, while the tests were
performed using the Ignalina NPP certificate. Mechanical characteristics
of the specimens were determined using usual standard methodology.
The tension tests were performed at the velocity of machine's
catch of 0.8 mm/min. The allowable velocity of growth of tension
strength during the test, according to the standard EN10002-1 [11-14] is
2-20 MPa/s. The velocity of the testing machine's catch that we
were used 0.8 mm/min-corresponds to the required velocity of tension
stresses: [[sigma].sub.1] [approximately equal to] 20 MPa/s.
When the tests were performed at the elevated temperature,
inductive heating of the specimen was used [10]. The scheme of inductive
heating is shown in Fig. 2.
The specimen 1 heats up from the inductor 2, through which electric
current of high frequency is passed from generator VCH4-10 6. The
thermocouples 3 are welded to the specimen, which signals get into the
potentiometer KSP-9 that registers temperature 5. The difference in
temperature in length of the testing part of the specimen at optimal
form of the inductor does not exceed 2%, while there is no
temperature's gradient in the thickness of specimen. The chromel
and copel wires of 0.2 mm thickness is used to measure temperature. The
thermocouples are attached to the specimen using the impulse electric
welding. The accuracy of temperature's measurement is [+ or -]
0.5%, while the regulation accuracy is [+ or -] 1.5%.
[FIGURE 2 OMITTED]
3. Investigation of mechanical characteristics
While performing the tests, 13 specimens were tested at
20[degrees]C temperature, where 4 were from new, unused pipes, 4 - from
the pipe after 45000 h of exploitation, and 5-from the pipe after 160000
h of exploitation. 9 specimens were tested at the operating temperature
of power station (550[degrees]C), where 3 specimens were from a new
pipe, 3-from the pipe after 45000 h of exploitation and 3 specimens from
the pipe after 160000 h of exploitation.
The results from tension tests are shown in Tables 1, 2; the
results of tension tests are shown graphically in Fig. 3, using the
coordinates "tension strength-strain".
The average tension curves in the Fig. 3 were compared depending on
the exploitation life and testing temperature. The tension strength
curves are expressed taking into account real tension stresses, when the
force is divided from the momentary cross-section area of the specimen
(dotted lines), and taking into account so called engineering tension
stresses, when the force is divided from in the initial cross-section
area of the specimen (continuous lines). Besides, the tension curves up
to strength's limit [R.sub.m] is shown.
According to the Tables 1, 2 and Fig. 3 the main mechanical
characteristics, such as yield limit [R.sub.p0,2] and strength limit
[R.sub.m] strongly depend on the exploitation life and testing
temperature, because when the exploitation life and testing temperature
increase from 20 to 550[degrees]C, these characteristics become
significantly smaller.
For example, at 20[degrees]C temperature [R.sub.p0,2] = 425 MPa in
case of a new pipe, while in case of the pipe after 160000 h of
exploitation [R.sub.p0,2] = 241 MPa, respectively [R.sub.m] = 601 and
453 MPa. At 550[degrees]C testing temperature, [R.sub.p0,2] = 217 MPa in
case of a new pipe, while in case of the pipe after 160000 h of
exploitation [R.sub.p0,2] = 159 MPa, respectively [R.sub.m] = 276 and
230 MPa.
[FIGURE 3 OMITTED]
The plasticity or reduction of cross-section area Z practically do
not depend on the exploitation life, however if the testing temperature
increases from 20 to 550[degrees]C, it also increases a little, about
8%.
These changes of [R.sub.p0,2], [R.sub.m] and Z depending on
temperature are characteristic to all grades of steel, because when the
temperature is increasing, the steel becomes weaker and more plastic.
But in our case growth of plasticity is stopped by structural changes in
the pipe's material, because in case of the new pipe when the
temperature increases from 20 to 550[degrees]C, Z increases from 70.63
to 84.90%, and in case of the pipe after 160000 h of exploitation Z
practically is not changing, because at 20[degrees]C Z = = 76.20%, and
at 550[degrees]C Z = 75.11%.
As it has been already mentioned, the strength characteristics
[R.sub.p0,2] and [R.sub.m] are decreasing with regard to the
exploitation life and temperature. Such reduction depending on testing
temperature can be partly explained by the influence of temperature,
because in case of the new pipe, at 20[degrees]C temperature
[R.sub.p0,2] = 425, and [R.sub.m] = 601 MPa, while at 550[degrees]C
[R.sub.p0,2] = 217 and [R.sub.m] = 276 MPa, thus the changes of these
characteristics with regard to exploitation life at 20 and 550[degrees]C
depend only on structural changes of the metal during exploitation, and
these changes are considerable. At 20[degrees]C temperature, yield limit
decreases from 425 down to 241 MPa during 160000 h of exploitation,
while strength limit decreases from 601 to 453 MPa. At 550[degrees]C
temperature, yield limit decreases from 217 to 159 MPa, while strength
limit decreases from 276 to 230 MPa. This shows that during
exploitation, significant structural changes in the metal take place,
and they change significantly strength characteristics, and a
little--metal's plasticity.
The attention should also be paid to the change of fracture
stress-[[sigma].sub.f] if a new pipe is used. This stress essentially
does not depend on the exploitation life at 20[degrees]C temperature
(Tables 1, 2); while at 550[degrees]C it decreases from 1057 to 541 MPa.
When the exploitation life increases up to 160000 h, the fracture
stress-[[sigma].sub.f] decreases from 541 to 279 MPa.
Dispersion of mechanical characteristics depend on the type of
characteristics. The most precise measurement is the reduction of
cross-section's area Z, which coefficient of variation was changing
from 0.01 to 0.04. Besides, quite steady are the strength limit
[R.sub.m] and yield limit [R.sub.p0,2], because at 20[degrees]C, the
[R.sub.m] variation coefficient is up to 0.05, and that of [R.sub.p0,2]
- up to 0.09. The dispersion at 550[degrees]C temperature is bigger,
because instabilities of the testing temperature and its gradients in
the testing part of the specimen create additional errors, and in this
case, the [R.sub.m] variation coefficients are up to 0.15, and
[R.sub.p0,2] up to 0.20.
Bigger variation coefficients are of the limit of proportionality
[[sigma].sub.pl] and of the fracture stress-[[sigma].sub.f]. Higher
accuracy is needed to accomplish the experiment to determine the limit
of proportionality [[sigma].sub.pl], while the variation in dispersion
for fracture stress-[[sigma].sub.f] is big because of instability of
disintegration process.
4. Calculation stresses for thick-walled pipes
In order to check strength of the straight part of the pipe, the
analytical calculations and calculations based on the method of finite
elements are performed. The analytical calculations were based on Lame
theory, whereas the model of finite elements was formed from
multilayered elements so that the distribution of tension stresses in
the inner layer wall of the pipe could be determined more precisely.
The LS-DYNA preprocessor was used to form the model of finite
elements. It is meant for calculations of nonlinear dynamics. In our
case the static load of pressure was imitated with the occurring dynamic
fluctuation of several tenths of percent using 164 Solid element as
"fully integrated S/R solid[#2]" material model. The
"piecewise linear plasticity" is indicated and both ends of
the pipe are fixed tight. It should be noted that the calculations were
performed for two types-[empty set] 219 mm pipe with the wall of 28.5 mm
thickness and working pressure was 13.2 MPa, and the [empty set] 245 mm
pipe with wall of 45 mm thickness and working pressure was 25.4 MPa.
According to Lame theory, the following stresses appear in the wall
of thick-walled pipe: [[sigma].sub.R]-radial stress,
[[sigma].sub.H]-circumferential stress, [[sigma].sub.L]-longitudinal
stress. When the pressure acts only on the inner surface of the pipe,
the tension stresses in the wall of thick-walled pipe are calculated in
the following way
[[sigma].sub.R] = p [R.sup.2.sub.2]/[R.sup.2.sub.1] -
[R.sup.2.sub.2] (1 - [R.sup.2.sub.1]/[R.sup.2.sub.x]) (1)
[[sigma].sub.H] = p [R.sup.2.sub.2]/[R.sup.2.sub.1]-
[R.sup.2.sub.2] (1 + [R.sup.2.sub.1]/[R.sup.2.sub.x]) (2)
where [R.sub.1], [R.sub.2], Rx are external, central and internal
radiuses of the pipe, p is working pressure of the pipe (Fig. 4).
Results of the calculation are shown in Tables 3 and 4.
[FIGURE 4 OMITTED]
The equation of circumferential and longitudinal stress in
thin-walled pipes is as follows
[[sigma].sup.Thin..sub.H] = p [R.sub.2]/h (4)
[[sigma].sup.Thin..sub.L] = p [R.sub.2]/2h (5)
where h is thickness of the wall, p is pressure.
While working with LS-DYNA, the distribution field of stresses is
received. In order to determine precise stresses at the outside, in the
center and inside, we indicate three type of elements, which numbers are
accordingly H35621 at interior, H35624 at center and H35625 at
exterior--they are shown in Fig. 5.
The calculation result as presented in Figs. 6-11 were obtained
after processing the calculation data by Excel program (curves in the
pictures), and the obtained characteristics are presented in Tables 5
and 6.
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
The survey of received values of stresses using the method of
finite elements resulted in high conformity between [[sigma].sub.R],
[[sigma].sub.H] values and lower conformity [[sigma].sub.L] values,
calculated on the basis of Lame theory method. Thus it is possible to
state that the performed calculations are correct and can be used to
determine durability of the pipes.
5. Analysis of mechanical characteristics
The mechanical characteristics from static tension test
accomplished with the specimens of different work resources are shown in
Figs. 12 and 13. They are described by equations
[[sigma].sub.pl] =-14.055 ln(t) + 398 (6)
[R.sub.p0,2] =-15.758 ln(t) + 425 (7)
[R.sub.m] =-13.161 ln(t) + 601 (8)
[[sigma].sub.f] =-42.2 ln(t) +1235 (9)
where 398 is the limit of proportionality of not used steel, 425 is
the yield strength of not used steel, 601 is the tension strength of not
used steel and 1235 is the fracture stress of not used steel.
The determination coefficient of the equations (8) and (9)
[R.sup.2] is 0.97 and 0.99 respectively.
[FIGURE 12 OMITTED]
The equation of linear dependence of reduction of cross-section
area on work time (determination coefficient [R.sup.2] = 0.99)
Z = 4.[10.sup.-51] + 70.63 (10)
where 70.63 is reduction of cross-section area of not used steel.
[FIGURE 13 OMITTED]
6. Investigation of microstructures
The specimens were made from the steel 12Ch1MF pipes, that had been
working different time. The microstructure seen in the photos are,
ferrite, perlite. During the carbon diffusion in perlite the carbides
are formed, perlite grains disappear, the tension strength decreases and
the plasticity increases. In order to see the changes of microstructure
better, we present the samples corroded by HN[O.sub.3] 10% alcoholic
solution.
In Fig. 14, a-c microstructures of not used, steel 12Ch1MF are
shown. Perlite (black area), ferrite (light area) and nonmetal inserts
are seen. The microstructures correspond to grade 3, according to the
standard TS 14-3-560 scale, while the ferrite grains correspond to grade
6, the grains of perlite phase correspond to grade 3 according to the
standard GOST 5639-82 scale.
In Fig. 15, a-c microstructures of steel 12Ch1MF used for 45000 h
are shown. The presented microstructures correspond grade 6, according
to standard scale TS 14-2-460. The first changes are visible in the
photos. The carbide particles start to separate within the limits of
ferrite grain. The ferrite grains correspond to grade 8, the grains of
perlite phase correspond to grade 2, according to the standard GOST
6539-82 scale.
In Fig. 16, a-c microstructures of steel, 12Ch1MF, used for 160000
h are shown. The nonmetal inserts of ~20 [micro]m are seen, as well as
small carbide particles within the limits ferrite grains. The changes of
microstructure affect mechanical characteristics, the strength
characteristics are decreasing and plasticity is increasing.
After the investigation of specimens the conclusion was done that
difference between of microstructures in the specimens are seen. We see
that when steel 12Ch1MF is used at 550[degrees]C temperature long time,
intensive carbon diffusion takes place. Thus carbon contained in perlite
diffuses and forms carbides, which results in worsening of steel's
properties, increase of plasticity and decrease of strength
characteristics.
[FIGURE 14 OMITTED]
[FIGURE 15 OMITTED]
[FIGURE 16 OMITTED]
7. Conclusions
The investigation of mechanical characteristics of the material of
hot steam supply pipeline for AB "Lietuvos elektrine"
(Lithuanian Power Station Ltd.) with regard to the straight part of the
new pipe, the pipe after 45000 and after 160000 h of exploitation at 20
and 550[degrees]C temperature allows making the following conclusions:
1. The mechanical characteristics of the pipe material:
[[sigma].sub.pl], [R.sub.p0,2], [R.sub.m], [[sigma].sub.f], Z depend on
the exploitation life and testing temperature, because when the
exploitation life grows from 0 to 160000 h and testing temperature
increases from 20 to 550[degrees]C, the strength mechanical
characteristics [[sigma].sub.pl], [R.sub.p0,2], [R.sub.m] and Of
decrease. For example, in case of the new pipe at 20[degrees]C
temperature [R.sub.p0,2] = 425 MPa, while after the 160000 h
exploitation Rp02 = 241 MPa, respectively [R.sub.m] = 601 and 453 MPa.
In case of the new pipe at 550[degrees]C temperature, [R.sub.p0,2] = 217
MPa, while after the 160000 h of exploitation- [R.sub.p0,2] = 159 MPa,
respectively [R.sub.m] = 276 and 230 MPa. The plasticity and reduction
of cross-section area Z practically do not depend on the exploitation
life, however if the testing temperature increases from 20 to
550[degrees]C, it also increases a little, about 8%.
2. The dependency of mechanical characteristics on structural
changes of the metal during exploitation is high, because the yield
strength after 160000 h of exploitation decreases from 425 MPa (for new
pipe) to 241 MPa, while the tension strength decreases from 601 to 453
MPa at 20[degrees]C testing temperature, while at 550[degrees]C
temperature the decrease is accordingly from 217 to 159 MPa and from 276
to 230 MPa. There have not been noticed any significant changes in the
material's plasticity with regard to exploitation life.
3. Fracture stress [[sigma].sub.f] after 160000 h of exploitation
decreases from 541 MPa (for new pipe) to 279 MPa at 20[degrees]C testing
temperature, while at 550[degrees]C [[sigma].sub.f] of the new pipe
decreases from 1235 to 728 MPa.
4. It was determined that the reliability of mechanical
characteristics mostly depends on the type of characteristics, testing
temperature and number of specimens. In our tests the smallest variation
coefficient at 20 and 550[degrees]C temperature was in the reduction of
cross-section area Z (changes from 0.01 to 0.10). Besides, the tension
strength and yield strength were also sufficiently stable, because at
20[degrees]C, their variation coefficients were changing up to 0.05, and
up to 0.09 respectively. The dispersion of results at 550[degrees]C
temperature is bigger, because instabilities of testing temperature and
its gradients in the testing part of the specimen create additional
errors, and in this case, the [R.sub.m] variation coefficient was
changing up to 0.015, and [R.sub.p0,2] up to 0.20.
5. When steel 12Ch1MF is used at 550[degrees]C temperature and long
working time, intensive carbon diffusion takes place. Thus carbon
contained in perlite diffuses and forms carbides, which results in
worsening of steel's properties, increase of plasticity and
decrease of strength characteristic.
6. Calculations results made according to thick walled pipes theory
(Lame) and thin walled pipes theory were compared with the results from
finite element method and can be drawn a conclusion, that Lame
theoretical equation describes stresses in the pipes more precisely,
than equations of thin walled pipe.
Acknowledgement
This work was supported by Lithuanian State Scientific and Study
fund, project T-92/09.
Received June 03, 2010
Accepted February 07, 2011
References
[1.] Antikain, P.A. 1990. Metals and Strength Calculation of
Boilers and Pipelines. Moscow: Energoatomizdat. 368p. (in Russian).
[2.] Nahalov, V.A. 1983. Reliability of Pipe Bends of Power
Generating Equipment. Moscow: Energoatomizdat. 216p. (in Russian).
[3.] Melehov, R.K.; Pokhmurskij, V.I. 2003. Structural Materials of
Power Generating Equipment. Kiev: Naukova Dumka. 382p. (in Russian).
[4.] Nelson, H.G.; Williams, D.P.; Tetelman, A.S. 1971.
Embrittlement of ferrous alloys in a partially dissociated hydrogen
environment, Met. Trans 2, No.4: 953-959.
[5.] Jankovski V.; Atkociunas J. 2010. Saosys toolbox as matlab
implementation in the elastic-plastic analysis and optimal design of
steel frame structures, Journal of Civil Engineering and Management
16(1): 103-121.
[6.] Balevicius R.; Dulinskas E. 2010. On the prediction of
non-linear creep strains, Journal of Civil Engineering and Management
16(3): 382-386.
[7.] Student, O.; Tkachuk, Yu.; Sydor, P. 2009. Technical expertise
of damaged structural elements of plant steam turbine. Proc. of the 14th
international conference "Mechanika--2009": 396-400.
[8.] Svirska, L.; Markov, A. 2009. Mechanical properties of
exploited 12Kh1MF steel from different zones of the elbow bend on steam
main power plant. -Proc. of the 14th international conference
"Mechanika--2009": 401-406.
[9.] Shariati, M.; Sedighi, M.; Saemi, J.; Eipakchi, H.R.;
Allahbakhsh, H.R. 2010. Numerical and experimental investigation on
ultimate strength of cracked cylindrical shells subjected to combined
loading, Mechanika 4(84): 12-19.
[10.] Daunys, M. 2005. Cyclic Strength and Durability of
Constructions. Kaunas: Technologija. 286p. (in Lithuanian).
[11.] Lithuanian Standard LST EN 10002-1. Metals. Tension testing.
Part 1. Testing method in the environmental temperature, 1998. (in
Lithuanian).
[12.] Lithuanian Standard LST EN 10002-5. Metals. Tension testing.
Part 5. High-temperature testing method, 2000. (in Lithuanian).
[13.] Rudzinskas, V.; Janavicius, A. 1999. Investigation of used
energy block's pipeline. Proc. of the second republican conference
of young scientists "Lithuania without science--Lithuania without
future" held in Vilnius on October 12-15, 1999. Mechanics and
material engineering. Vilnius: Technika: 75-80 (in Lithuanian).
[14.] Daunys, M.; Dundulis, R. 1999. Investigation of armature of
ferroconcrete constructions of accident localization system in the 1st
block of Ignalina NPP. Proc. of the international conference
"Mechanika-99". Kaunas: Technologija: 34-38 (in Lithuanian).
M. Daunys *, R. Dundulis **, R. Karpavicius ***, R. Bortkevicius
****
* Kaunas University of Technology, Kqstucio str. 27, 44312 Kaunas,
Lithuania, E-mail:
[email protected].
** Kaunas University of Technology, Kqstucio str. 27, 44312 Kaunas,
Lithuania, E-mail:
[email protected]
*** Kaunas University of Technology, Kqstucio str. 27, 44312
Kaunas, Lithuania, E-mail:
[email protected]
**** Kaunas University of Technology, Kqstucio str. 27, 44312
Kaunas, Lithuania, E-mail:
[email protected]
Table 1
Mechanical characteristics of not used pipes, pipes after
45000 and 160000 h of exploitation, T = 20[degrees]C
Hours Mechanical characteristics, MPa, %
[[sigma]. [R.sub. [[sigma].
sub.p1] p0.2] [R.sub.m] sub.f] Z
0 398 425 601 1235 70.63
45000 284 292 494 786 72.91
160000 232 241 453 728 76.20
Table 2
Mechanical characteristics of not used pipes, pipes after
45000 and 160000 h of exploitation, T = 550[degrees]C
Hours Mechanical characteristics, MPa, %
[[sigma]. [R.sub. [[sigma].
sub.p1] p0.2] [R.sub.m] sub.f] Z
0 183 217 276 541 84.90
45000 128 175 237 292 86.44
160000 116 159 230 279 75.11
Table 3
Results of Lame and thin-walled pipes method for the pipe
[empty set] 0219 x 28.5 when p = 13.2 MPa
Stresses in the layers, MPa
Outer Middle Inner
[[sigma].sub.R] 0.00 -5.21 -13.20
[[sigma].sub.H] 31.90 37.59 45.10
[[sigma].sub.L] 15.95 15.95 15.95
[[sigma].sup.Thin..sub.H] 37.51
[[sigma].sup.Thin..sub.L] 18.76
Table 4
Results of Lame and thin-walled pipes method for the pipe
[empty set] 245 x 45 when p = 25.4 MPa
Stresses in the layers, MPa
Outer Middle Inner
[[sigma].sub.R] 0.00 -8.19 -25.40
[[sigma].sub.H] 33.90 40.89 59.30
[[sigma].sub.L] 16.95 16.95 16.95
[[sigma].sup.Thin..sub.H] 42.19
[[sigma].sup.Thin..sub.L] 21.10
Table 5
Results of finite elements method for the pipe [empty set] 219 x
28.5 when p = 13.2 MPa
Stresses in the layers, MPa
Outer Middle Inner
[[sigma].sub.R] -0.82 -5.07 -11.15
[[sigma].sub.H] 32.80 37.08 43.23
[[sigma].sub.L] 9.01 9.01 9.01
Table 6
Results of finite elements method for the pipe [empty set] 245 x 45
when P = 25.4 MPa
Stresses in the layers, MPa
Outer Middle Inner
[[sigma].sub.R] -1.29 -8.53 -21.08
[[sigma].sub.H] 36.52 44.06 57.09
[[sigma].sub.L] 10.77 10.77 10.77
