- Email: [email protected]
Non-uniform temperature distribution of turbine casing and its effect on turbine casing distortion
Accepted Manuscript Non-uniform Temperature Distribution of Turbine Casing and Its Effect on Turbine Casing Distortion E. Poursaeidi, M. Taheri, A. Farhangi PII:
S1359-4311(14)00567-5
DOI:
10.1016/j.applthermaleng.2014.07.019
Reference:
ATE 5794
To appear in:
Applied Thermal Engineering
Received Date: 25 January 2014 Revised Date:
5 July 2014
Accepted Date: 7 July 2014
Please cite this article as: E. Poursaeidi, M. Taheri, A. Farhangi, Non-uniform Temperature Distribution of Turbine Casing and Its Effect on Turbine Casing Distortion, Applied Thermal Engineering (2014), doi: 10.1016/j.applthermaleng.2014.07.019. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Non-uniform Temperature Distribution of Turbine Casing and Its Effect on Turbine Casing Distortion E. Poursaeidi ,M. Taheri∗ ,A. Farhangi
RI PT
Department of Mechanical Engineering,University of Zanjan , Zanjan, Iran
Abstract
Stress analysis is essential for gaining an understanding of the factors affecting crack on turbine casing arising from temperature gradients. Hence, making determinations of temperature
SC
distribution on gas turbine casing is the first step in stress analysis. The next step is comparison of results with available thermography data related to the casing. In addition, stress and distortion distributions are presented for three test levels of working load on the casing.
M AN U
Comparison of stress concentrations at the eccentric pin hole and observed cracks in these locations validated evaluations for stress distribution.
Keywords: Gas turbine casing, Thermal distribution, Thermal stress, Thermography 1. Introduction
TE D
Achievement of high thermal efficiency in gas turbine systems is strongly related to increased temperature at the turbine inlet, which is accompanied by excess thermal load in the hot components of a gas turbine. Thus, various cooling techniques [1–3] have been used to protect the main hot parts of gas turbines. If an unsuitable cooling method is used, local thermal crack
EP
and structural failure are yielded due to thermal stress and reduced material strength at high temperature. Therefore, failure analyses as well as thermal analyses for temperature, deformation
AC C
and stress are required for effective thermal design and lifetime prediction of hot components. It is noted that failure analysis was investigated in other research only in terms of material, and rarely by thermal analysis [4–8]. Furthermore, temperature gradient in each hot component increases according to an increase of turbine inlet temperature and it generates thermal damage from high thermal stress. It is necessary to estimate temperature distributions in materials of the system in an appropriate thermal environment to predict the life and safety of hot components such as combustors, vanes, blades and casing. In recent years, several investigators [9–12] have
∗
Corresponding author. Tel.: +989109717821 E-mail address: [email protected] 1
ACCEPTED MANUSCRIPT attempted thermal analyses of hot components of gas turbines and made predictions of thermal damage. It has been shown that computed results are useful for inspecting the thermal environment of a gas turbine and to define factors that contribute to operational longevity. Many numerical studies have been done using CFD (computational fluid dynamics) codes [13,14], which are developed by solving Navier–Stokes equations using boundary layer modeling.
RI PT
Among them, TEXTAN [15] has been widely used in industry. With developing computer technology and turbulence models, CFD has become a powerful design tool. Many researchers have performed CFD predictions and compared results with test data obtained in the turbine. Brandts and Wesorick [16] pointed out that most cases of nozzle failure occurred because of
SC
thermal fatigue, whereas there is less impact at absolute high temperature. Moreover they show that conditions of maximum temperature and maximum gradient occurred at two different operating conditions. Analysis of the failure of a high-pressure nozzle of a 70 MW gas turbine
M AN U
reported a similar conclusion Mazur et al [17]. The authors identified the origin of cracks in nozzles by numerical analysis and investigations of alterations to the metal grain. The critical region was identified as the inner part of the coolant holes, corresponding to the highest concentration of thermal stresses. Maintaining reliability is an important issue in thermal power plants as well as considerations of safety under operating conditions that include frequent
TE D
startups and load changes. Unstable states arising during startup, shutdown, and load change give produce unsteady temperature distribution with respect to time in steam turbine components. Thermal stress is caused by a rapid increase in temperature that renders the components susceptible to failure and reduces their operational longevity. The internal stationary and rotating
EP
components of a turbine in a power plant are encased in massive steel-cast casings. These highand intermediate-pressure casings are susceptible to frequent cracking due to thermo-mechanical,
AC C
low-cycle fatigue (LCF) at the nozzle fit corner radius or other stress concentration shapes. The non-uniform temperature field contributes to a high stress gradient appearing in the turbine casing. High temperature reduces strength of the casing material. Failure analysis for turbine engine components has received the attention of several investigations [18–24]. Choi et al [25] defined thermal stress concentration factors for inner surfaces of the casing and valve to account for geometric variations using three-dimensional, finite element analysis. In addition, total strain range was obtained to assess the low-cycle fatigue life according to life assessment procedures conducted in Korea. The model can be used to obtain maximum thermal stress level and strain values related to creep and fatigue damage. Using this model, more accurate data on life consumption can be obtained by using steam turbine inner casings as an input for the Korean simple life assessment procedure without the need for complex time-consuming calculations. 2
ACCEPTED MANUSCRIPT Witek et al [26] described the fracture problem of turbine casing for a helicopter engine. Visual inspection of the defected component was incomplete because the fracture was repaired by welding during a technical inspection of the engine. Authors of this work tried to explain the causes of damage to the turbine casing by application of numerical stress analysis. A geometrically complicated numerical model was created to solve the problem. The finite element
RI PT
method (FEM) was used in computations. Stress and deformation contours were generated from results of nonlinear static analyses performed for both mechanical and thermal loads occurring during operating conditions. High thermal stress gradients were found at the region of casing where cracks were detected in engine operation. Cheong and Karstensen [27] reviews a recent
SC
structural integrity assessment carried out on a high-pressure turbine inner casing that had suffered from temper brittleness. The assessment was made to demonstrate that the casing can be safely returned to service based on revised operating conditions with particular emphasis on
M AN U
temperature control of ramp rates during start-up and shutdown events. A retirement for- cause philosophy was adopted to account for some operational flexibility that is required prior to replacing the casing at the next planned outage. Descriptions of the open cycle gas turbine operation, the operational background and the problem of failure at the Putrajaya Power Station in Malaysia were reported by Rashid et al [28]. According to the report, the main concerns were
TE D
repeated findings of several obvious surface-crack spots on the weld joint zone of a plenum barrier plate of the gas turbine frame.
2. Turbine casing and its performance
EP
The most important role of a gas turbine casing is taking turbine rotor with an axial symmetry and inhibition of elements such as nozzles and Shroud segments in its fixed position with a total high weight. In addition, turbine casing has an important structural role. Besides its own weight,
AC C
it tolerates the moment arising from components such as the exhaust chamber, the combustion chamber and compressor weights. Turbine casing operation is similar to that of pressure vessels that tolerate hot gas pressure caused by products of combustion. Another vital role of gas turbine casing, especially in aircraft engine turbines, is its ability to tolerate failure of its moving parts such as turbine blades, such that if there is a high centrifugal force and failure occurrence, then there should be no possibility that the casing of the turbine is damaged by piercing. Nonetheless, due to the relatively large thickness of different parts of the turbine casing, its vulnerability to high temperature (phenomenon of restructuring and deterioration of the material and casing bending) and high temperature gradient between internal and external surfaces 3
ACCEPTED MANUSCRIPT (creating mechanical and thermal fatigue) is much more than that of any other static or dynamic load. The GE-Frame9 heavy-duty gas turbine was designed on the basis of the GE-Frame7 gas turbine, and has been developed from the B series with 85.2 (MW) outputs to the FA series with output of 251.8 (MW) [16]. GE-Frame9 is an axial type gas turbine engine and includes three
RI PT
stages of turbine section, 17 stages in compressor section and 14 can-type combustion chambers. The gas turbine rotor is single-type and the net output mechanical power is directly transferred to the generator at the speed of 3000 (rpm) via a coupling.
Table 1 Chemical analysis results of casing
W%
Si
Mn
Al
Cu
Ni
V
Cr
M AN U
Element C
SC
Table 1 shows results of chemical analysis for the casing material [29].
Mg
Ti
Co
Fe
3.344 2.7 0.136 0.016 0.351 0.082 0.044 0.12 0.057 0.038 0.033 Balance
These results showed good correlation with ASTM A395 that is a kind of ductile cast iron.
TE D
Table 2 shows some mechanical properties of the ASTM A395 ductile cast iron [30]. Table 2 shows mechanical properties of the ASTM A395 ductile cast iron.
(Brinell)
Tensile Strength,
Modulus of
Poisson's
Ultimate (MPa)
Yield (MPa)
Elasticity(GPa)
Ratio
461
329
165
0.29
AC C
167
Tensile Strength,
EP
Hardness
In this work, evaluations for temperature distribution and corresponding stress and distortion distribution on turbine casing GF-9 (Shahid Rajai power plant) were obtained according to the procedure described as follows: The first step was to determine temperature distribution of the entire casing by CFX software. For this reason temperature was measured by thermocouples placed inside the turbine casing that were used as the thermal boundary conditions. Consideration was also made for the cooling effect of air passing through the cooling channels and the natural heat transfer from casing to the 4
ACCEPTED MANUSCRIPT ambient air. Temperature distribution was compared with results of thermography and results showed acceptable accordance that verifies the numerical results. Then temperature distribution obtained in the CFX software was readouted by Static structural software and stress and distortion distributions were obtained by application of expressed constraints, stress and distortion distributions. This procedure was followed for three different working loads of gas
RI PT
turbine. 3. Thermography results from the outer side of the turbine casing
Thermography results on the upper-half casing were recorded. Two methods; Fluke thermal
SC
imaging and SDS Hotfind cameras did this process. These cameras record infrared radiation from objects and represented them as a colored contour. Thermography operation was conducted from a gas unit of Shahid Rajai power plant, where the unit is located under the load of 82 MW.
M AN U
One of the thermography results is shown in Fig.1. This figure is related to the thermography image of the outer side of the casing at the first row of the turbine. In this figure, white color state of areas with higher temperature. Obviously, temperature corresponds to location of the fixing pins in the first low shroud segment and vents for measurements of blade tip clearance. The physical explanation of this phenomenon is related to the pin roles that act like a heat fin
AC C
EP
TE D
with a high heat flux from the shroud segment to the outer side of the casing.
Fig.1 thermography image of the outer side of the
Fig. 2thermography image of middle part of
casing in the first row of the turbine
bottom half casing near the horizontal turbine casing flange
According to Fig.2 the highest temperature was recorded around the second nozzle’s pin, near the horizontal flange at 314°C . Unlike other parts of the casing, there was no cooling airflow in this area.
5
ACCEPTED MANUSCRIPT
4. Temperature measurement inside the casing using thermocouples At this stage, 10 type K thermocouples were installed, within the range of 0 to 1100°C at 10 different positions in the turbine casing and connected to a paperless recorder, to record
M AN U
SC
RI PT
temperatures inside the casing (Fig 3).
Fig.3 installation of thermocouples inside the turbine casing
TE D
Due to mechanical limitations of the casing, mounting positions of the thermocouple are considered inside the bolt retaining pins of the segment Shroud and pin guide of the segment nozzle.
Fig.4 shows insert location of sensors inside the upper turbine casing. To estimate reasonable
EP
temperature of the inner side of the casing in different rows, two thermocouples are used in each
AC C
row. Thermocouple output data was readout by Eurotherm software.
6
RI PT
ACCEPTED MANUSCRIPT
SC
Fig.4 insert location of sensors inside the upper turbine casing
In the current work, three different working loads of gas turbine were as follows: A, B and C whose values are given in Table 3. These loads were chosen arbitrarily and all temperature and
M AN U
stress analyzes were performed separately for each load. Tables 4-6 show temperature evaluations taken by thermocouples placed inside the turbine casing for each of the three tested loads.
Table 3 three different turbine working load Values (MW)
A
96
B
82
C
87
EP
TE D
Load
AC C
Table 4 temperature recorded by sensors install inside the turbine casing at load of A. Right side of the casing Top of the casing Left side of the casing Track of first row 346.88 ° C nozzles Track of second row 365.33° C 328.54 ° C nozzles Track of third row 345.11° C 286.33° C nozzles Track of first row 363.95° C 347.99° C Shroud Track of second row 365.61° C 354.20° C Shroud Track of third row 287.80° C Shroud
7
ACCEPTED MANUSCRIPT
SC
RI PT
Table 5 temperature recorded by sensors install inside the turbine casing at load of B. Right side of the casing Top of the casing Left side of the casing Track of first row 323.45 ° C nozzles Track of second row 346° C 316.62 ° C nozzles Track of third row 354° C 289.94° C nozzles Track of first row 337° C 331.77° C Shroud Track of second row 347.77° C 338.25° C Shroud Track of third row 290.87° C Shroud
EP
TE D
M AN U
Table 6 temperature recorded by sensors install inside the turbine casing at load of C. Right side of the casing Top of the casing Left side of the casing Track of first row 330.95 ° C nozzles Track of second row 352.15° C 324.76 ° C nozzles Track of third row 344.80° C 296.10° C nozzles Track of first row 346.44° C 340.27° C Shroud Track of second row 352.46° C 344.55° C Shroud Track of third row 291.92° C Shroud
5. Thermal analysis of gas turbine
AC C
In the current study, heat transfer coefficient of the cooling air through the turbine casing passage, changes during channel and cannot be assumed as a fixed number. However, temperature of the cooling air will also change. For this reason CFX was used to simulate airflow through the cooling passage and examine its effect on casing temperature distribution. Figures 5, 6 and 7 show passages, casing and assembled geometry, respectively. 5.1. Thermal boundary conditions inside the turbine casing Due to the low number and positioning of the sensors, temperature inside the casing was not available at all locations. Also, the number of these sensors as boundary conditions was not enough to obtain temperature measurements on the outside of the casing, consequently answers 8
ACCEPTED MANUSCRIPT could not be considered as acceptable. Because of this reason, in addition to the distributed
SC
RI PT
sensor locations, linear temperature distribution between sensors needed consideration.
Fig.6 turbine casing
TE D
M AN U
Fig.5 21 cooling passages
EP
Fig.7 turbine casing assembled by 21 cooling passages
AC C
5.2. Heat losses due to free convection outside the turbine casing The gas turbine is located within the chamber so heat transfer from its outer side will appear in free convection mode. Accordingly heat loss from the turbine casing can easily be calculated by evaluating the heat transfer coefficient and ambient temperature (air temperature of inside chamber and outside casing). Using Churchill and Chu [31] proposed Nusselt number: 1
Nu D = [0.6 +
0.387 Ra 6 9 6
0.559 (1 + ( ) ) Pr
8 27
]2 ,
(1)
where Ra denotes Rayleigh number and is given by: 9
ACCEPTED MANUSCRIPT Ra =
gβ (Ts − T∞ ) D 3
(2)
αυ
Finally heat transfer coefficient is obtained by: kNu D
(3)
RI PT
hD =
All used properties in the above equations are related to air and evaluated in from temperature of the film.
5.3. Force convection inside the cooling passages
SC
Ambient temperature in loads of A, B and C were determined as 50, 44 and 46 ° C, respectively.
M AN U
The SSTk − ω model was used to simulate flow outside the turbine casing, within the cooling passage. In the framework of eddy viscosity models, the hydrodynamic behavior of a turbulent in compressible fluid is governed by RANS equations for velocity u and pressure p
∇u = 0
(4)
TE D
∂u + u∇u = −∇p + ∇((υ + υ T )[∇u + ∇u T ]), ∂t
where υ depends only on physical properties of the fluid, while υT is the turbulent eddy viscosity which is supposed to emulate the effect of unresolved velocity fluctuations u ' .
EP
the SSTk − ω model is employed for evaluations of turbulent kinetic energy and dissipation rate.
AC C
5.4. Mesh generation
Tetrahedrons meshes in three sizes of 0.02, 0.018 and 0.016 were used to analyze temperature of the casing. Mesh generation size of 0.018 is shown in Fig. 8. In Figures 9 and 10, mesh generation around the inlet and outlet of the cooling passage is shown. Also, as it can be seen in these figures, clustering near the passage wall was used to consider the big gradients near the wall. Note that in all the following evaluations, mesh independency has been considered and mesh size of 0.018 was applied.
10
RI PT
ACCEPTED MANUSCRIPT
TE D
M AN U
SC
Fig.8 mesh size of 0.018 (top view of casing)
Fig.9 mesh generation of the cooling channel (inlet)
Fig.10 mesh generation of the cooling channel
AC C
EP
(outlet)
5.5. Thermal analysis results Thermography was on the outer surface of the casing, at 17 points (Fig. 11). These results were obtained for working load B. Fig. 12 shows results for temperature distribution obtained for three tested working loads. As can be seen, in parts of the casing with cooling channels, the effect of cooling effect corresponds to a low temperature and there are higher temperatures in parts that where less thick with and without a cooling effect.
11
RI PT
ACCEPTED MANUSCRIPT
a
AC C
EP
TE D
M AN U
SC
Fig.11 17 selected points on the turbine casing
b
c
Fig. 12 the temperature distribution in the turbine casing at the three tested turbine working loads of a) A, b) B and c) C 12
ACCEPTED MANUSCRIPT Results for temperatures measured by thermocouples for working loads A and C were not much different from those for load B, thermography result for load B was used to verify temperature distribution for loads A and C. Comparison between thermography and CFX results are reported in Fig.13. Comparison between thermography and CFX results confirms results of the thermal
M AN U
SC
RI PT
analysis and guaranteed accuracy of temperature distribution.
AC C
EP
TE D
a
b
c
Fig.13 comparison between thermography and CFX results at three working loads of a) A, b) B and c) C
6. Thermal stresses and distortions 6.1. Constraints It is noteworthy that the obtained temperature distribution of the casing was related to the steady state condition of the turbine. These evaluations for temperature distribution for the three different working times correspond to the three different turbine loads. In this section, thermal 13
ACCEPTED MANUSCRIPT stress distribution of the casing was obtained for the three working loads. Asymmetric distribution of temperature causes thermal stresses. The turbine casing is constrained by other components so this stress will have considerable effect. Hence it is important to impose constrains in a real way. Now imposing constrains by Static Structural software is going to be expressed to obtain thermal stresses. As it can be seen in Fig.14, the turbine casing had four sides
RI PT
and there were holes on each side to constrain the casing. Results shown in red in Fig.14, demonstrate that each side did not move in a direction perpendicular to itself. However, when the thermal load was imposed, those plates in green in Fig. 15 and Fig. 16 did not move relative to each other. So some constraints were imposed to avoid any rotation of these plates around the
SC
non-perpendicular direction of their own. For example, if the orthogonal direction in the green plate is displayed by w, this plate should not rotate around other two directions. The same rule
EP
TE D
M AN U
would be applied in green plate in Fig.16.
AC C
Fig.14 Zero displacement boundary condition normal to red plates
Fig.15 No rotation boundary condition on green
Fig.16 No rotation boundary condition on green
plate in x and v directions
plate in x and v directions
14
ACCEPTED MANUSCRIPT 6.2. Results and discussion Several arbitrary points on the casing were chosen in order to check results of mesh independency. These points are shown in Fig.17. This figure is relates to a mesh size of 0.0155m. In this figure shows selected points in one of the holes located at the second row slot of the shroud. The same points were chosen in the other mesh sizes to investigate mesh
RI PT
independency and stress results are reported in Table 7. Results of stresses for selected points clearly show stress points close together and only a small difference between them. It is noteworthy that the value of the stress is magnified in fine mesh and there is less difference
TE D
M AN U
SC
between them. Finally mesh size of 0.014 (m) was applied to all simulations.
AC C
EP
Fig.17 stress in selected points of one of the second row shroud groove’s holes.
Table 7 stress values for selected points for different mesh sizes (MPa) Stress Mesh size Point 1 Point 2 Point 3 Point 4
0.02 (m)
0.018 (m)
0.0165 (m)
0.0155 (m)
0.014 (m)
0.0136 (m)
0.0132 (m)
127.48 141.1 175.01 117.32
131.04 90.132 177.4 79.467
125.06 107.73 200.46 108.25
140.12 106.84 208.91 97.111
144.19 110.52 202.61 103.65
144.2 111.1 203.21 101.14
144.18 110.94 203.04 102.92
15
ACCEPTED MANUSCRIPT The next step was to apply the mentioned boundary conditions to the model of turbine casting for analysis by Static Structural software. This analyze was applied in three different working loads and the corresponding results of stress distribution and distortion are presented in the following figures. The General Electric Company claims that it has a general policy in gas turbine manufacturing
RI PT
that aims to facilitate the work by simplifying it. For example, common gray cast iron is used for casing and low alloy steel is used for compressor and turbine discs in order to use casting and forging for the manufacturing process [16]. However, statistics have shown that there are still numerous weaknesses in the gas units manufactured by this company. For example, a relatively
SC
common technical problem in gas turbine casing of the series GE-Frame 9 is that deep cracks appear from inside the casing.
M AN U
Initial inspection of turbine casing of gas units GE-F9 showed that in all cases at least one crack among those in pinholes was off-center. In addition, other cracks have been observed in some holes of mounting tracks of first and second row segments of the shroud. The existences of such cracks determined by thermography images of the outer surface of the casing (Figs. 1 and 2) are predictable. There are white points in the thermography images that represent points with a
TE D
higher temperature. Obviously, temperature of these points corresponds to the location of fixing pins in the first low shroud segment and vents for measurements of blade tip clearance. A physical explanation of this phenomenon is the effective role of pins which act as a thermal
EP
fin with a large heat flux from the segments Shroud that are directly exposed in the flame to the outer surface of the casing. This can lead to a concentration of heat in the pin hole that increases probability of thermal fatigue and germination of crack around the holes. Normally, what causes
AC C
stress concentration and crack initiation is high temperature gradient around a point or very small region, that edge of the installation pin hole satisfies these conditions. While, relatively large areas with high temperatures due to a lack of local temperature gradients will not encounter problems generated from stress concentration and cracks. This means that for the initiation of crack in the casing, wide areas with a maximum temperature are not necessarily subjected to damage, but in areas with high temperature gradient and even the lower temperature range will be more prone to initiation crack. As previously mentioned, most of the cracks has been observed in the eccentric pin hole and thus to confirm the validity of the obtained stress distribution, these holes have been selected as validity criterion. Fig. 18, shows location of observed crack in the eccentric mounting hole. Figs. 19, 20 and 21, show obtained stress 16
ACCEPTED MANUSCRIPT distribution in the eccentric pin hole for the three working loads A, B and C, respectively. As can be seen, the obtained stress concentration in the figures of 19, 20 and 21, are in good accordance with observed cracks (Fig. 18) and this confirms accuracy of the obtained stress distribution in
M AN U
SC
RI PT
turbine casing.
Fig.18 PT test from crack corner at
in the eccentric pin hole for load A
AC C
EP
TE D
the eccentric pin hole of half-casing
Fig.19 the obtained stress distribution
Fig.20 the obtained stress distribution in the
Fig.21 the obtained stress distribution in the
eccentric pin hole for load B
eccentric pin hole for load C
Figures 22a, 22b and 22c, show the obtained stress distribution in part of the mounting track of second row segment Shroud holes for three working loads of the turbine. As can be seen, there is stress concentration in most of the holes that predicts the observed cracks in this track. Figures 23a, 23b and 23c, show stress distribution in one hole of the mounting track of the second row segment Shroud for three working loads of the turbine. There are similar stress concentrations in some holes of the mounting track of the first row segment Shroud. 17
M AN U
SC
a
RI PT
ACCEPTED MANUSCRIPT
b
c
AC C
EP
TE D
Fig.22 the obtained stress distribution in the part of the mounting track of second row segment Shroud at three working loads of a) A, b) B and c) C
18
RI PT
ACCEPTED MANUSCRIPT
M AN U
SC
a
b
c
TE D
Fig.23 the obtained stress distribution in the one holes of the mounting track of second row segment Shroud at three working loads of a) A, b) B and c) C
EP
It should be noted that the main cause of extension of cracks in the turbine casing is by operating the turbine in its transient mode (turn on, turn off and trip mode). Nonetheless, stress
AC C
distribution at steady-state operation identifies the locations where cracks may appear. Another consideration is that creep cracks are created in the steady-state operation of the turbine and these may supplement fatigue cracks. Figures 24a, 24b and 24c, show stress distribution in the turbine casing for the three working loads tested for the turbine.
19
M AN U
SC
a
RI PT
ACCEPTED MANUSCRIPT
c
TE D
b
Fig.24 the obtained stress distribution in turbine casing at three working loads of a) A, b) B and c) C
EP
Figures 25a, 25b and 25c depicted the distortion distribution for the three turbine working loads. As it is evident that the maximum distortion is located at the turbine inlet and it reduces
AC C
toward the turbine outlet.
20
b
TE D
M AN U
SC
a
RI PT
ACCEPTED MANUSCRIPT
c
Fig.25 the distortion distribution in turbine casing at three turbine working loads of a) A, b) B and
7. Conclusion
EP
c) C
AC C
This paper has presented an analysis of thermal stress that occurred at steady operation of the turbine. Initially, with the available temperature measurements recorded by thermocouples inside the turbine casing and applying the boundary conditions, temperature distribution of the turbine casing was obtained and its accuracy has been confirmed by thermography. These results demonstrate that the turbine casing temperature was high at the turbine inlet that was connected to the combustion chamber and whatever was going to the turbine outlet had a decreased temperature. This temperature reduction is caused by cooling channels that conduct the air passing from the inlet to the outlet of the turbine. Next, using the obtained evaluations for temperature distribution, corresponding stress distribution was obtained. Comparison of the stress concentration at the eccentric pin hole and 21
ACCEPTED MANUSCRIPT observed crack in this place validates evaluations for stress distribution. More cracks have been observed in the eccentric pinhole that in all examples of casing there at least one crack has been identified in this area. In addition, there are other stresses concentrations in casing that require investigation due to being located in a sensitive area. Most of these stresses are in the hole of mounting tracks of first and second row segments Shroud that in operation also, cracks have
RI PT
been observed in this part of the turbine casing. Also, according to results of stress distribution, another place that will experience high stress is the end of the left hook.
Finally, distortion distribution of the stress distribution has been obtained that its value
M AN U
SC
decreases from the turbine inlet to its outlet.
References
AC C
EP
TE D
[1] Goldstein RJ. Film cooling. Adv Heat Trans 1971;7:321–79. [2] Han JC, Dutta S, Ekkad SV. Gas turbine heat transfer and cooling technology. New York: Taylor and Francis; 2000. [3] Kim KM, Kim YY, Lee DH, Rhee DH, Cho HH. Influence of duct aspect ratio on heat/mass transfer in coolant passages with rotation. Int J Heat Fluid Flow 2007;28(3):357–73. [4] Tryon RG, Cruse TA, Mahadevan S. Development of a reliability-based fatigue life model for gas turbine engine structures.EngFractMech 1996;53(5):807–28. [5] Gallardo JM, Rodriguez JA, Herrera EJ. Failure of gas turbine blades. Wear 2002;252:264–8. [6] Chang JC, Yun YH, Choi C, Kim JC. Failure analysis of gas turbine buckets. Eng Fail Anal 2003;10:559–67. [7] Carter TJ. Common failures in gas turbine blades. Eng Fail Anal 2005;12:237–47. [8] Kawasaki A, Watanabe R. Thermal fracture behavior of metal/ceramic functionally graded materials. EngFractMech 2002;69:1713–28. [9] Cunha FJ, Dahmer MT, Chyu MK. Thermal–mechanical life prediction system for anisotropic turbine components. ASME J Turbomach 2006;128:240–50. [10] Tinga T, Kampen JF, Jager B, Kok JBW. Gas turbine combustor liner life assessment using a combined fluid/structural approach. ASME J Eng Gas Turb Power 2007;129(1):69–79. [11] Hou NX, Wen ZX, Du ZX, Yue ZF. Crystallographic failure analysis of film near cooling hole under temperature gradient of nickel-based single crystal superalloys.TheorApplFractMech 2007;47:164–70. [12] Kim KM, Lee DH, Cho HH. Thermal analysis of a film cooling system with normal injection holes using experimental data. Int J Fluid Mach Syst 2009;2(1):55–60. [13] Giel PW, Van Fossen GJ, Boyle RJ, Thurman DR, Civinskas KC. Blade heat transfer measurements and predictions in a transonic turbine cascade. NASA technical note, No. TM1999-209296; 1999. [14] Kim KM, Yun N, Jeon YH, Lee DH, Cho HH, Kang S-H. Conjugated heat transfer and temperature distributions in a gas turbine combustion liner under base-load operation. J MechSciTechnol 2010;24(9):1939–46. 22
ACCEPTED MANUSCRIPT [15] Dunn MG. Convective heat transfer and aerodynamics in axial flow turbine. ASME J Turbomach 2001;123(4):637–86.
AC C
EP
TE D
M AN U
SC
RI PT
[16]Brandts, D.E., Wesorick, R.R., 1994. GE Gas turbine design philosophy, GE Industrial & Power Systems. GER-3434D. [17]Mazur, Z., Hernandez-Rossette, A., Garcia-Illescas, R., Luna-Ramirez, A., 2008.Failure analysis of a gas turbine nozzle. Journal of Engineering Failure Analysis 15, 913–921. [18] Daniel N. Musili, Stress and strain analysis of the PZL-10W turbine casing, Rzesz َ◌w University of Technology, Faculty of Mechanical Engineering and Aeronautics, Rzesz َ◌w; 2007. [19] Hakl J et al. Residual life assessment of steam turbine casing containing crack defect.Int J PresVes Pip 2001;78:977–84. [20] Price John WH. The failure of the Dartmouth turbine casing.Int J PresVes Pip 1998;75:559– 66. [21] Witek L. Failure analysis of turbine disc of an aero engine. Eng Fail Anal 2006;13(1). [22] Witek L, Kowalski T, Mamrowicz J. Numerical stress and fatigue analysis of the first stage of turbine for helicopter engine. In: Proceedings of international conference on aeronautical fatigue, Napoli; 2007. [23] Witek L, Poznan´ ska A, Wierzbin´ ska M. Fracture analysis of a compressor blade of a helicopter engine. Eng Fail Anal 2009;16:1616–22. [24] Witek L. Experimental crack propagation and failure analysis of the first stage compressor blade subjected to vibration. Eng Fail Anal 2009;16(7):2163-70. [25] Choi w, Fujiyama K, Kim B, Song G. Development of thermal stress concentration factors for life assessment of turbinecasings. International Journal of Pressure Vessels and Piping 98 (2012) 1-7. [26]Witek L, Orkisz M, Wygonik P, Musili DN, Kowalski T. Fracture analysis of a turbine casing. Engineering Failure Analysis 18 (2011) 914–923. [27] Cheong KS, Karstensen AD. Integrity assessment of an embrittled steam turbine casing. International Journal of Pressure Vessels and Piping 86 (2009) 265–272. [28] Rashid AZ, Purbolaksono J, Ahmad A, Ahmad SA. Thermal fatigue analysis on cracked plenum barrier plate of open-cycle gas turbine frame.EngFailAnal 2010;17:579–86. [29] Technical reports of "Montazer Qa`em" power plant, IRAN, 2005. [30] www.matweb.com [31] F. P. Incropera, D. P. DeWitt,T. L. Bergman, A. S. Lavine, Fundamentals of heat and mass transfer, sixth edition
23
S1359-4311(14)00567-5
DOI:
10.1016/j.applthermaleng.2014.07.019
Reference:
ATE 5794
To appear in:
Applied Thermal Engineering
Received Date: 25 January 2014 Revised Date:
5 July 2014
Accepted Date: 7 July 2014
Please cite this article as: E. Poursaeidi, M. Taheri, A. Farhangi, Non-uniform Temperature Distribution of Turbine Casing and Its Effect on Turbine Casing Distortion, Applied Thermal Engineering (2014), doi: 10.1016/j.applthermaleng.2014.07.019. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Non-uniform Temperature Distribution of Turbine Casing and Its Effect on Turbine Casing Distortion E. Poursaeidi ,M. Taheri∗ ,A. Farhangi
RI PT
Department of Mechanical Engineering,University of Zanjan , Zanjan, Iran
Abstract
Stress analysis is essential for gaining an understanding of the factors affecting crack on turbine casing arising from temperature gradients. Hence, making determinations of temperature
SC
distribution on gas turbine casing is the first step in stress analysis. The next step is comparison of results with available thermography data related to the casing. In addition, stress and distortion distributions are presented for three test levels of working load on the casing.
M AN U
Comparison of stress concentrations at the eccentric pin hole and observed cracks in these locations validated evaluations for stress distribution.
Keywords: Gas turbine casing, Thermal distribution, Thermal stress, Thermography 1. Introduction
TE D
Achievement of high thermal efficiency in gas turbine systems is strongly related to increased temperature at the turbine inlet, which is accompanied by excess thermal load in the hot components of a gas turbine. Thus, various cooling techniques [1–3] have been used to protect the main hot parts of gas turbines. If an unsuitable cooling method is used, local thermal crack
EP
and structural failure are yielded due to thermal stress and reduced material strength at high temperature. Therefore, failure analyses as well as thermal analyses for temperature, deformation
AC C
and stress are required for effective thermal design and lifetime prediction of hot components. It is noted that failure analysis was investigated in other research only in terms of material, and rarely by thermal analysis [4–8]. Furthermore, temperature gradient in each hot component increases according to an increase of turbine inlet temperature and it generates thermal damage from high thermal stress. It is necessary to estimate temperature distributions in materials of the system in an appropriate thermal environment to predict the life and safety of hot components such as combustors, vanes, blades and casing. In recent years, several investigators [9–12] have
∗
Corresponding author. Tel.: +989109717821 E-mail address: [email protected] 1
ACCEPTED MANUSCRIPT attempted thermal analyses of hot components of gas turbines and made predictions of thermal damage. It has been shown that computed results are useful for inspecting the thermal environment of a gas turbine and to define factors that contribute to operational longevity. Many numerical studies have been done using CFD (computational fluid dynamics) codes [13,14], which are developed by solving Navier–Stokes equations using boundary layer modeling.
RI PT
Among them, TEXTAN [15] has been widely used in industry. With developing computer technology and turbulence models, CFD has become a powerful design tool. Many researchers have performed CFD predictions and compared results with test data obtained in the turbine. Brandts and Wesorick [16] pointed out that most cases of nozzle failure occurred because of
SC
thermal fatigue, whereas there is less impact at absolute high temperature. Moreover they show that conditions of maximum temperature and maximum gradient occurred at two different operating conditions. Analysis of the failure of a high-pressure nozzle of a 70 MW gas turbine
M AN U
reported a similar conclusion Mazur et al [17]. The authors identified the origin of cracks in nozzles by numerical analysis and investigations of alterations to the metal grain. The critical region was identified as the inner part of the coolant holes, corresponding to the highest concentration of thermal stresses. Maintaining reliability is an important issue in thermal power plants as well as considerations of safety under operating conditions that include frequent
TE D
startups and load changes. Unstable states arising during startup, shutdown, and load change give produce unsteady temperature distribution with respect to time in steam turbine components. Thermal stress is caused by a rapid increase in temperature that renders the components susceptible to failure and reduces their operational longevity. The internal stationary and rotating
EP
components of a turbine in a power plant are encased in massive steel-cast casings. These highand intermediate-pressure casings are susceptible to frequent cracking due to thermo-mechanical,
AC C
low-cycle fatigue (LCF) at the nozzle fit corner radius or other stress concentration shapes. The non-uniform temperature field contributes to a high stress gradient appearing in the turbine casing. High temperature reduces strength of the casing material. Failure analysis for turbine engine components has received the attention of several investigations [18–24]. Choi et al [25] defined thermal stress concentration factors for inner surfaces of the casing and valve to account for geometric variations using three-dimensional, finite element analysis. In addition, total strain range was obtained to assess the low-cycle fatigue life according to life assessment procedures conducted in Korea. The model can be used to obtain maximum thermal stress level and strain values related to creep and fatigue damage. Using this model, more accurate data on life consumption can be obtained by using steam turbine inner casings as an input for the Korean simple life assessment procedure without the need for complex time-consuming calculations. 2
ACCEPTED MANUSCRIPT Witek et al [26] described the fracture problem of turbine casing for a helicopter engine. Visual inspection of the defected component was incomplete because the fracture was repaired by welding during a technical inspection of the engine. Authors of this work tried to explain the causes of damage to the turbine casing by application of numerical stress analysis. A geometrically complicated numerical model was created to solve the problem. The finite element
RI PT
method (FEM) was used in computations. Stress and deformation contours were generated from results of nonlinear static analyses performed for both mechanical and thermal loads occurring during operating conditions. High thermal stress gradients were found at the region of casing where cracks were detected in engine operation. Cheong and Karstensen [27] reviews a recent
SC
structural integrity assessment carried out on a high-pressure turbine inner casing that had suffered from temper brittleness. The assessment was made to demonstrate that the casing can be safely returned to service based on revised operating conditions with particular emphasis on
M AN U
temperature control of ramp rates during start-up and shutdown events. A retirement for- cause philosophy was adopted to account for some operational flexibility that is required prior to replacing the casing at the next planned outage. Descriptions of the open cycle gas turbine operation, the operational background and the problem of failure at the Putrajaya Power Station in Malaysia were reported by Rashid et al [28]. According to the report, the main concerns were
TE D
repeated findings of several obvious surface-crack spots on the weld joint zone of a plenum barrier plate of the gas turbine frame.
2. Turbine casing and its performance
EP
The most important role of a gas turbine casing is taking turbine rotor with an axial symmetry and inhibition of elements such as nozzles and Shroud segments in its fixed position with a total high weight. In addition, turbine casing has an important structural role. Besides its own weight,
AC C
it tolerates the moment arising from components such as the exhaust chamber, the combustion chamber and compressor weights. Turbine casing operation is similar to that of pressure vessels that tolerate hot gas pressure caused by products of combustion. Another vital role of gas turbine casing, especially in aircraft engine turbines, is its ability to tolerate failure of its moving parts such as turbine blades, such that if there is a high centrifugal force and failure occurrence, then there should be no possibility that the casing of the turbine is damaged by piercing. Nonetheless, due to the relatively large thickness of different parts of the turbine casing, its vulnerability to high temperature (phenomenon of restructuring and deterioration of the material and casing bending) and high temperature gradient between internal and external surfaces 3
ACCEPTED MANUSCRIPT (creating mechanical and thermal fatigue) is much more than that of any other static or dynamic load. The GE-Frame9 heavy-duty gas turbine was designed on the basis of the GE-Frame7 gas turbine, and has been developed from the B series with 85.2 (MW) outputs to the FA series with output of 251.8 (MW) [16]. GE-Frame9 is an axial type gas turbine engine and includes three
RI PT
stages of turbine section, 17 stages in compressor section and 14 can-type combustion chambers. The gas turbine rotor is single-type and the net output mechanical power is directly transferred to the generator at the speed of 3000 (rpm) via a coupling.
Table 1 Chemical analysis results of casing
W%
Si
Mn
Al
Cu
Ni
V
Cr
M AN U
Element C
SC
Table 1 shows results of chemical analysis for the casing material [29].
Mg
Ti
Co
Fe
3.344 2.7 0.136 0.016 0.351 0.082 0.044 0.12 0.057 0.038 0.033 Balance
These results showed good correlation with ASTM A395 that is a kind of ductile cast iron.
TE D
Table 2 shows some mechanical properties of the ASTM A395 ductile cast iron [30]. Table 2 shows mechanical properties of the ASTM A395 ductile cast iron.
(Brinell)
Tensile Strength,
Modulus of
Poisson's
Ultimate (MPa)
Yield (MPa)
Elasticity(GPa)
Ratio
461
329
165
0.29
AC C
167
Tensile Strength,
EP
Hardness
In this work, evaluations for temperature distribution and corresponding stress and distortion distribution on turbine casing GF-9 (Shahid Rajai power plant) were obtained according to the procedure described as follows: The first step was to determine temperature distribution of the entire casing by CFX software. For this reason temperature was measured by thermocouples placed inside the turbine casing that were used as the thermal boundary conditions. Consideration was also made for the cooling effect of air passing through the cooling channels and the natural heat transfer from casing to the 4
ACCEPTED MANUSCRIPT ambient air. Temperature distribution was compared with results of thermography and results showed acceptable accordance that verifies the numerical results. Then temperature distribution obtained in the CFX software was readouted by Static structural software and stress and distortion distributions were obtained by application of expressed constraints, stress and distortion distributions. This procedure was followed for three different working loads of gas
RI PT
turbine. 3. Thermography results from the outer side of the turbine casing
Thermography results on the upper-half casing were recorded. Two methods; Fluke thermal
SC
imaging and SDS Hotfind cameras did this process. These cameras record infrared radiation from objects and represented them as a colored contour. Thermography operation was conducted from a gas unit of Shahid Rajai power plant, where the unit is located under the load of 82 MW.
M AN U
One of the thermography results is shown in Fig.1. This figure is related to the thermography image of the outer side of the casing at the first row of the turbine. In this figure, white color state of areas with higher temperature. Obviously, temperature corresponds to location of the fixing pins in the first low shroud segment and vents for measurements of blade tip clearance. The physical explanation of this phenomenon is related to the pin roles that act like a heat fin
AC C
EP
TE D
with a high heat flux from the shroud segment to the outer side of the casing.
Fig.1 thermography image of the outer side of the
Fig. 2thermography image of middle part of
casing in the first row of the turbine
bottom half casing near the horizontal turbine casing flange
According to Fig.2 the highest temperature was recorded around the second nozzle’s pin, near the horizontal flange at 314°C . Unlike other parts of the casing, there was no cooling airflow in this area.
5
ACCEPTED MANUSCRIPT
4. Temperature measurement inside the casing using thermocouples At this stage, 10 type K thermocouples were installed, within the range of 0 to 1100°C at 10 different positions in the turbine casing and connected to a paperless recorder, to record
M AN U
SC
RI PT
temperatures inside the casing (Fig 3).
Fig.3 installation of thermocouples inside the turbine casing
TE D
Due to mechanical limitations of the casing, mounting positions of the thermocouple are considered inside the bolt retaining pins of the segment Shroud and pin guide of the segment nozzle.
Fig.4 shows insert location of sensors inside the upper turbine casing. To estimate reasonable
EP
temperature of the inner side of the casing in different rows, two thermocouples are used in each
AC C
row. Thermocouple output data was readout by Eurotherm software.
6
RI PT
ACCEPTED MANUSCRIPT
SC
Fig.4 insert location of sensors inside the upper turbine casing
In the current work, three different working loads of gas turbine were as follows: A, B and C whose values are given in Table 3. These loads were chosen arbitrarily and all temperature and
M AN U
stress analyzes were performed separately for each load. Tables 4-6 show temperature evaluations taken by thermocouples placed inside the turbine casing for each of the three tested loads.
Table 3 three different turbine working load Values (MW)
A
96
B
82
C
87
EP
TE D
Load
AC C
Table 4 temperature recorded by sensors install inside the turbine casing at load of A. Right side of the casing Top of the casing Left side of the casing Track of first row 346.88 ° C nozzles Track of second row 365.33° C 328.54 ° C nozzles Track of third row 345.11° C 286.33° C nozzles Track of first row 363.95° C 347.99° C Shroud Track of second row 365.61° C 354.20° C Shroud Track of third row 287.80° C Shroud
7
ACCEPTED MANUSCRIPT
SC
RI PT
Table 5 temperature recorded by sensors install inside the turbine casing at load of B. Right side of the casing Top of the casing Left side of the casing Track of first row 323.45 ° C nozzles Track of second row 346° C 316.62 ° C nozzles Track of third row 354° C 289.94° C nozzles Track of first row 337° C 331.77° C Shroud Track of second row 347.77° C 338.25° C Shroud Track of third row 290.87° C Shroud
EP
TE D
M AN U
Table 6 temperature recorded by sensors install inside the turbine casing at load of C. Right side of the casing Top of the casing Left side of the casing Track of first row 330.95 ° C nozzles Track of second row 352.15° C 324.76 ° C nozzles Track of third row 344.80° C 296.10° C nozzles Track of first row 346.44° C 340.27° C Shroud Track of second row 352.46° C 344.55° C Shroud Track of third row 291.92° C Shroud
5. Thermal analysis of gas turbine
AC C
In the current study, heat transfer coefficient of the cooling air through the turbine casing passage, changes during channel and cannot be assumed as a fixed number. However, temperature of the cooling air will also change. For this reason CFX was used to simulate airflow through the cooling passage and examine its effect on casing temperature distribution. Figures 5, 6 and 7 show passages, casing and assembled geometry, respectively. 5.1. Thermal boundary conditions inside the turbine casing Due to the low number and positioning of the sensors, temperature inside the casing was not available at all locations. Also, the number of these sensors as boundary conditions was not enough to obtain temperature measurements on the outside of the casing, consequently answers 8
ACCEPTED MANUSCRIPT could not be considered as acceptable. Because of this reason, in addition to the distributed
SC
RI PT
sensor locations, linear temperature distribution between sensors needed consideration.
Fig.6 turbine casing
TE D
M AN U
Fig.5 21 cooling passages
EP
Fig.7 turbine casing assembled by 21 cooling passages
AC C
5.2. Heat losses due to free convection outside the turbine casing The gas turbine is located within the chamber so heat transfer from its outer side will appear in free convection mode. Accordingly heat loss from the turbine casing can easily be calculated by evaluating the heat transfer coefficient and ambient temperature (air temperature of inside chamber and outside casing). Using Churchill and Chu [31] proposed Nusselt number: 1
Nu D = [0.6 +
0.387 Ra 6 9 6
0.559 (1 + ( ) ) Pr
8 27
]2 ,
(1)
where Ra denotes Rayleigh number and is given by: 9
ACCEPTED MANUSCRIPT Ra =
gβ (Ts − T∞ ) D 3
(2)
αυ
Finally heat transfer coefficient is obtained by: kNu D
(3)
RI PT
hD =
All used properties in the above equations are related to air and evaluated in from temperature of the film.
5.3. Force convection inside the cooling passages
SC
Ambient temperature in loads of A, B and C were determined as 50, 44 and 46 ° C, respectively.
M AN U
The SSTk − ω model was used to simulate flow outside the turbine casing, within the cooling passage. In the framework of eddy viscosity models, the hydrodynamic behavior of a turbulent in compressible fluid is governed by RANS equations for velocity u and pressure p
∇u = 0
(4)
TE D
∂u + u∇u = −∇p + ∇((υ + υ T )[∇u + ∇u T ]), ∂t
where υ depends only on physical properties of the fluid, while υT is the turbulent eddy viscosity which is supposed to emulate the effect of unresolved velocity fluctuations u ' .
EP
the SSTk − ω model is employed for evaluations of turbulent kinetic energy and dissipation rate.
AC C
5.4. Mesh generation
Tetrahedrons meshes in three sizes of 0.02, 0.018 and 0.016 were used to analyze temperature of the casing. Mesh generation size of 0.018 is shown in Fig. 8. In Figures 9 and 10, mesh generation around the inlet and outlet of the cooling passage is shown. Also, as it can be seen in these figures, clustering near the passage wall was used to consider the big gradients near the wall. Note that in all the following evaluations, mesh independency has been considered and mesh size of 0.018 was applied.
10
RI PT
ACCEPTED MANUSCRIPT
TE D
M AN U
SC
Fig.8 mesh size of 0.018 (top view of casing)
Fig.9 mesh generation of the cooling channel (inlet)
Fig.10 mesh generation of the cooling channel
AC C
EP
(outlet)
5.5. Thermal analysis results Thermography was on the outer surface of the casing, at 17 points (Fig. 11). These results were obtained for working load B. Fig. 12 shows results for temperature distribution obtained for three tested working loads. As can be seen, in parts of the casing with cooling channels, the effect of cooling effect corresponds to a low temperature and there are higher temperatures in parts that where less thick with and without a cooling effect.
11
RI PT
ACCEPTED MANUSCRIPT
a
AC C
EP
TE D
M AN U
SC
Fig.11 17 selected points on the turbine casing
b
c
Fig. 12 the temperature distribution in the turbine casing at the three tested turbine working loads of a) A, b) B and c) C 12
ACCEPTED MANUSCRIPT Results for temperatures measured by thermocouples for working loads A and C were not much different from those for load B, thermography result for load B was used to verify temperature distribution for loads A and C. Comparison between thermography and CFX results are reported in Fig.13. Comparison between thermography and CFX results confirms results of the thermal
M AN U
SC
RI PT
analysis and guaranteed accuracy of temperature distribution.
AC C
EP
TE D
a
b
c
Fig.13 comparison between thermography and CFX results at three working loads of a) A, b) B and c) C
6. Thermal stresses and distortions 6.1. Constraints It is noteworthy that the obtained temperature distribution of the casing was related to the steady state condition of the turbine. These evaluations for temperature distribution for the three different working times correspond to the three different turbine loads. In this section, thermal 13
ACCEPTED MANUSCRIPT stress distribution of the casing was obtained for the three working loads. Asymmetric distribution of temperature causes thermal stresses. The turbine casing is constrained by other components so this stress will have considerable effect. Hence it is important to impose constrains in a real way. Now imposing constrains by Static Structural software is going to be expressed to obtain thermal stresses. As it can be seen in Fig.14, the turbine casing had four sides
RI PT
and there were holes on each side to constrain the casing. Results shown in red in Fig.14, demonstrate that each side did not move in a direction perpendicular to itself. However, when the thermal load was imposed, those plates in green in Fig. 15 and Fig. 16 did not move relative to each other. So some constraints were imposed to avoid any rotation of these plates around the
SC
non-perpendicular direction of their own. For example, if the orthogonal direction in the green plate is displayed by w, this plate should not rotate around other two directions. The same rule
EP
TE D
M AN U
would be applied in green plate in Fig.16.
AC C
Fig.14 Zero displacement boundary condition normal to red plates
Fig.15 No rotation boundary condition on green
Fig.16 No rotation boundary condition on green
plate in x and v directions
plate in x and v directions
14
ACCEPTED MANUSCRIPT 6.2. Results and discussion Several arbitrary points on the casing were chosen in order to check results of mesh independency. These points are shown in Fig.17. This figure is relates to a mesh size of 0.0155m. In this figure shows selected points in one of the holes located at the second row slot of the shroud. The same points were chosen in the other mesh sizes to investigate mesh
RI PT
independency and stress results are reported in Table 7. Results of stresses for selected points clearly show stress points close together and only a small difference between them. It is noteworthy that the value of the stress is magnified in fine mesh and there is less difference
TE D
M AN U
SC
between them. Finally mesh size of 0.014 (m) was applied to all simulations.
AC C
EP
Fig.17 stress in selected points of one of the second row shroud groove’s holes.
Table 7 stress values for selected points for different mesh sizes (MPa) Stress Mesh size Point 1 Point 2 Point 3 Point 4
0.02 (m)
0.018 (m)
0.0165 (m)
0.0155 (m)
0.014 (m)
0.0136 (m)
0.0132 (m)
127.48 141.1 175.01 117.32
131.04 90.132 177.4 79.467
125.06 107.73 200.46 108.25
140.12 106.84 208.91 97.111
144.19 110.52 202.61 103.65
144.2 111.1 203.21 101.14
144.18 110.94 203.04 102.92
15
ACCEPTED MANUSCRIPT The next step was to apply the mentioned boundary conditions to the model of turbine casting for analysis by Static Structural software. This analyze was applied in three different working loads and the corresponding results of stress distribution and distortion are presented in the following figures. The General Electric Company claims that it has a general policy in gas turbine manufacturing
RI PT
that aims to facilitate the work by simplifying it. For example, common gray cast iron is used for casing and low alloy steel is used for compressor and turbine discs in order to use casting and forging for the manufacturing process [16]. However, statistics have shown that there are still numerous weaknesses in the gas units manufactured by this company. For example, a relatively
SC
common technical problem in gas turbine casing of the series GE-Frame 9 is that deep cracks appear from inside the casing.
M AN U
Initial inspection of turbine casing of gas units GE-F9 showed that in all cases at least one crack among those in pinholes was off-center. In addition, other cracks have been observed in some holes of mounting tracks of first and second row segments of the shroud. The existences of such cracks determined by thermography images of the outer surface of the casing (Figs. 1 and 2) are predictable. There are white points in the thermography images that represent points with a
TE D
higher temperature. Obviously, temperature of these points corresponds to the location of fixing pins in the first low shroud segment and vents for measurements of blade tip clearance. A physical explanation of this phenomenon is the effective role of pins which act as a thermal
EP
fin with a large heat flux from the segments Shroud that are directly exposed in the flame to the outer surface of the casing. This can lead to a concentration of heat in the pin hole that increases probability of thermal fatigue and germination of crack around the holes. Normally, what causes
AC C
stress concentration and crack initiation is high temperature gradient around a point or very small region, that edge of the installation pin hole satisfies these conditions. While, relatively large areas with high temperatures due to a lack of local temperature gradients will not encounter problems generated from stress concentration and cracks. This means that for the initiation of crack in the casing, wide areas with a maximum temperature are not necessarily subjected to damage, but in areas with high temperature gradient and even the lower temperature range will be more prone to initiation crack. As previously mentioned, most of the cracks has been observed in the eccentric pin hole and thus to confirm the validity of the obtained stress distribution, these holes have been selected as validity criterion. Fig. 18, shows location of observed crack in the eccentric mounting hole. Figs. 19, 20 and 21, show obtained stress 16
ACCEPTED MANUSCRIPT distribution in the eccentric pin hole for the three working loads A, B and C, respectively. As can be seen, the obtained stress concentration in the figures of 19, 20 and 21, are in good accordance with observed cracks (Fig. 18) and this confirms accuracy of the obtained stress distribution in
M AN U
SC
RI PT
turbine casing.
Fig.18 PT test from crack corner at
in the eccentric pin hole for load A
AC C
EP
TE D
the eccentric pin hole of half-casing
Fig.19 the obtained stress distribution
Fig.20 the obtained stress distribution in the
Fig.21 the obtained stress distribution in the
eccentric pin hole for load B
eccentric pin hole for load C
Figures 22a, 22b and 22c, show the obtained stress distribution in part of the mounting track of second row segment Shroud holes for three working loads of the turbine. As can be seen, there is stress concentration in most of the holes that predicts the observed cracks in this track. Figures 23a, 23b and 23c, show stress distribution in one hole of the mounting track of the second row segment Shroud for three working loads of the turbine. There are similar stress concentrations in some holes of the mounting track of the first row segment Shroud. 17
M AN U
SC
a
RI PT
ACCEPTED MANUSCRIPT
b
c
AC C
EP
TE D
Fig.22 the obtained stress distribution in the part of the mounting track of second row segment Shroud at three working loads of a) A, b) B and c) C
18
RI PT
ACCEPTED MANUSCRIPT
M AN U
SC
a
b
c
TE D
Fig.23 the obtained stress distribution in the one holes of the mounting track of second row segment Shroud at three working loads of a) A, b) B and c) C
EP
It should be noted that the main cause of extension of cracks in the turbine casing is by operating the turbine in its transient mode (turn on, turn off and trip mode). Nonetheless, stress
AC C
distribution at steady-state operation identifies the locations where cracks may appear. Another consideration is that creep cracks are created in the steady-state operation of the turbine and these may supplement fatigue cracks. Figures 24a, 24b and 24c, show stress distribution in the turbine casing for the three working loads tested for the turbine.
19
M AN U
SC
a
RI PT
ACCEPTED MANUSCRIPT
c
TE D
b
Fig.24 the obtained stress distribution in turbine casing at three working loads of a) A, b) B and c) C
EP
Figures 25a, 25b and 25c depicted the distortion distribution for the three turbine working loads. As it is evident that the maximum distortion is located at the turbine inlet and it reduces
AC C
toward the turbine outlet.
20
b
TE D
M AN U
SC
a
RI PT
ACCEPTED MANUSCRIPT
c
Fig.25 the distortion distribution in turbine casing at three turbine working loads of a) A, b) B and
7. Conclusion
EP
c) C
AC C
This paper has presented an analysis of thermal stress that occurred at steady operation of the turbine. Initially, with the available temperature measurements recorded by thermocouples inside the turbine casing and applying the boundary conditions, temperature distribution of the turbine casing was obtained and its accuracy has been confirmed by thermography. These results demonstrate that the turbine casing temperature was high at the turbine inlet that was connected to the combustion chamber and whatever was going to the turbine outlet had a decreased temperature. This temperature reduction is caused by cooling channels that conduct the air passing from the inlet to the outlet of the turbine. Next, using the obtained evaluations for temperature distribution, corresponding stress distribution was obtained. Comparison of the stress concentration at the eccentric pin hole and 21
ACCEPTED MANUSCRIPT observed crack in this place validates evaluations for stress distribution. More cracks have been observed in the eccentric pinhole that in all examples of casing there at least one crack has been identified in this area. In addition, there are other stresses concentrations in casing that require investigation due to being located in a sensitive area. Most of these stresses are in the hole of mounting tracks of first and second row segments Shroud that in operation also, cracks have
RI PT
been observed in this part of the turbine casing. Also, according to results of stress distribution, another place that will experience high stress is the end of the left hook.
Finally, distortion distribution of the stress distribution has been obtained that its value
M AN U
SC
decreases from the turbine inlet to its outlet.
References
AC C
EP
TE D
[1] Goldstein RJ. Film cooling. Adv Heat Trans 1971;7:321–79. [2] Han JC, Dutta S, Ekkad SV. Gas turbine heat transfer and cooling technology. New York: Taylor and Francis; 2000. [3] Kim KM, Kim YY, Lee DH, Rhee DH, Cho HH. Influence of duct aspect ratio on heat/mass transfer in coolant passages with rotation. Int J Heat Fluid Flow 2007;28(3):357–73. [4] Tryon RG, Cruse TA, Mahadevan S. Development of a reliability-based fatigue life model for gas turbine engine structures.EngFractMech 1996;53(5):807–28. [5] Gallardo JM, Rodriguez JA, Herrera EJ. Failure of gas turbine blades. Wear 2002;252:264–8. [6] Chang JC, Yun YH, Choi C, Kim JC. Failure analysis of gas turbine buckets. Eng Fail Anal 2003;10:559–67. [7] Carter TJ. Common failures in gas turbine blades. Eng Fail Anal 2005;12:237–47. [8] Kawasaki A, Watanabe R. Thermal fracture behavior of metal/ceramic functionally graded materials. EngFractMech 2002;69:1713–28. [9] Cunha FJ, Dahmer MT, Chyu MK. Thermal–mechanical life prediction system for anisotropic turbine components. ASME J Turbomach 2006;128:240–50. [10] Tinga T, Kampen JF, Jager B, Kok JBW. Gas turbine combustor liner life assessment using a combined fluid/structural approach. ASME J Eng Gas Turb Power 2007;129(1):69–79. [11] Hou NX, Wen ZX, Du ZX, Yue ZF. Crystallographic failure analysis of film near cooling hole under temperature gradient of nickel-based single crystal superalloys.TheorApplFractMech 2007;47:164–70. [12] Kim KM, Lee DH, Cho HH. Thermal analysis of a film cooling system with normal injection holes using experimental data. Int J Fluid Mach Syst 2009;2(1):55–60. [13] Giel PW, Van Fossen GJ, Boyle RJ, Thurman DR, Civinskas KC. Blade heat transfer measurements and predictions in a transonic turbine cascade. NASA technical note, No. TM1999-209296; 1999. [14] Kim KM, Yun N, Jeon YH, Lee DH, Cho HH, Kang S-H. Conjugated heat transfer and temperature distributions in a gas turbine combustion liner under base-load operation. J MechSciTechnol 2010;24(9):1939–46. 22
ACCEPTED MANUSCRIPT [15] Dunn MG. Convective heat transfer and aerodynamics in axial flow turbine. ASME J Turbomach 2001;123(4):637–86.
AC C
EP
TE D
M AN U
SC
RI PT
[16]Brandts, D.E., Wesorick, R.R., 1994. GE Gas turbine design philosophy, GE Industrial & Power Systems. GER-3434D. [17]Mazur, Z., Hernandez-Rossette, A., Garcia-Illescas, R., Luna-Ramirez, A., 2008.Failure analysis of a gas turbine nozzle. Journal of Engineering Failure Analysis 15, 913–921. [18] Daniel N. Musili, Stress and strain analysis of the PZL-10W turbine casing, Rzesz َ◌w University of Technology, Faculty of Mechanical Engineering and Aeronautics, Rzesz َ◌w; 2007. [19] Hakl J et al. Residual life assessment of steam turbine casing containing crack defect.Int J PresVes Pip 2001;78:977–84. [20] Price John WH. The failure of the Dartmouth turbine casing.Int J PresVes Pip 1998;75:559– 66. [21] Witek L. Failure analysis of turbine disc of an aero engine. Eng Fail Anal 2006;13(1). [22] Witek L, Kowalski T, Mamrowicz J. Numerical stress and fatigue analysis of the first stage of turbine for helicopter engine. In: Proceedings of international conference on aeronautical fatigue, Napoli; 2007. [23] Witek L, Poznan´ ska A, Wierzbin´ ska M. Fracture analysis of a compressor blade of a helicopter engine. Eng Fail Anal 2009;16:1616–22. [24] Witek L. Experimental crack propagation and failure analysis of the first stage compressor blade subjected to vibration. Eng Fail Anal 2009;16(7):2163-70. [25] Choi w, Fujiyama K, Kim B, Song G. Development of thermal stress concentration factors for life assessment of turbinecasings. International Journal of Pressure Vessels and Piping 98 (2012) 1-7. [26]Witek L, Orkisz M, Wygonik P, Musili DN, Kowalski T. Fracture analysis of a turbine casing. Engineering Failure Analysis 18 (2011) 914–923. [27] Cheong KS, Karstensen AD. Integrity assessment of an embrittled steam turbine casing. International Journal of Pressure Vessels and Piping 86 (2009) 265–272. [28] Rashid AZ, Purbolaksono J, Ahmad A, Ahmad SA. Thermal fatigue analysis on cracked plenum barrier plate of open-cycle gas turbine frame.EngFailAnal 2010;17:579–86. [29] Technical reports of "Montazer Qa`em" power plant, IRAN, 2005. [30] www.matweb.com [31] F. P. Incropera, D. P. DeWitt,T. L. Bergman, A. S. Lavine, Fundamentals of heat and mass transfer, sixth edition
23