Microstructural Restoration by HIP and Heat Treatment Processes in

Document Sample
Microstructural Restoration by HIP and Heat Treatment Processes in Powered By Docstoc
					Chiang Mai J. Sci. 2009; 36(3)                                                                       287


              Chiang Mai J. Sci. 2009; 36(3) : 287-295
              www.science.cmu.ac.th/journal-science/josci.html
              Contributed Paper


Microstructural Restoration by HIP and Heat Treatment
Processes in Cast Nickel Based Superalloy, IN-738
Panyawat Wangyao [a], Gobboon Lothongkum [a], Viyaporn Krongtong [b],
Weerasak Homkrajai [c] and Nutthita Chuankrerkkul* [d]
[a] Department of Metallurgical Engineering, Faculty of Engineering, Chulalongkorn University,
    Bangkok 10330, Thailand.
[b] National Metals and Materials Technology Center (MTEC), Pathum Thani, Thailand.
[c] Electricity Generating Authority of Thailand (EGAT), Nonthaburi, Thailand.
[d] Metallurgy and Materials Science Research Institute (MMRI), Chulalongkorn University,
    Bangkok 10330, Thailand.
*Author for correspondence; e-mail: nutthita.c@chula.ac.th
                                                                                    Received: 9 April 2009
                                                                                    Accepted: 7 May 2009

ABSTRACT
          Inconel 738, as a cast nickel-based superalloy components for land-based gas turbine
blades, utilized at elevated temperatures, expresses microstructural stability with excellent
mechanical properties. In this study, the microstructural refurbishment by a hot isostatic pressing
(HIP) and different heat treatment conditions after long-term service of the alloy was
investigated. Although, the HIP temperature in the HIP process was high enough to act as
solutioning treatment, it could not completely dissolve coarse γ particles into the matrix,
resulting in the residue particles. However, the next step, solutioning treatment, could dissolve
these particles more in some extent depending on each solutioning temperature. The higher
solutioning temperature provided the finer γ particles and new γ precipitation after solutioning
that could take place during primary aging. Secondary aging led to the growth of coarse and/
or very fine γ particles. All heat treatment conditions after HIP process provided much more
uniform γ morphology comparing to as-received microstructure.
Keywords: hot isostatic pressing (HIP), rejuvenation, microstructural repair, and nickel-based
superalloy.

1. INTRODUCTION
     Nickel base superalloys are structural           solid solution strengthening or carbide
materials with chemical composition and               strengthening do. Microstructure, mainly
structure, which have been developed to be            characteristics and morphology of γ particles,
utilized at high temperature applications.            has significant effect on mechanical properties
The alloys, used for high-strength applications       more than on physical properties [1-3]. The
at elevated temperatures such as materials            microstructure and mechanical properties (for
for gas turbine blades, are strengthened              both low and high temperatures) for turbine
significantly by γ precipitated particles in          blades of fixed chemistry are vitally dependent
FCC matrix. The γ particles precipitated in γ         with prior manufacturing processes such as
matrix provided the most strengthening of             casting, mechanical and heat treatments.
superalloys even much higher than the additive        However, the desired mechanical properties
288                                                                Chiang Mai J. Sci. 2009; 36(3)



mainly derived from the end of the processing       which is the most similar to initial one,
sequences, which provided the final micro-          and should maximize the phase stability of
structure. One of these processes is heat           precipitated strengthening gamma prime
treatment. Solution treatment in most cases is      particles after long-term exposure. The aim
followed by a single or a double aging sequence     of this research work is to determine the
to precipitate homogeneous distributions of         most suitable and practicable repair-condition,
either cuboidal or spherical gamma prime            which can give the proper microstructural
within the grains interior as well as discrete      characteristics by rejuvenation method of hot
grain boundary carbides [4-6]. Full or partial      isostatic pressing (HIP), followed by various
solution treatment temperatures, including          heat treatments for long term exposed gas
aging treatments, have been developed and           turbine blades, casting nickel-based superalloy
modified to optimize the completed precipita-       grade IN-738 after 70,000-hour service
tion of gamma prime phase in matrix.                operated by Electricity Generating of Thailand
      Practically, the alloy’s standard reheat      (EGAT).
treatment for cast superalloys does not always
work perfectly when applied to the long-term        2. MATERIALS AND METHODS
serviced microstructure to restore mechanical            The cast nickel-based superalloy in this
properties as well as to the welded superalloy      study was IN-738, with the composition
components or to HIPed superalloy parts to          (mass%) of 15.84%Cr, 8.5%Co, 3.47%Ti,
rejuvenate and establish optimal microstruc-        3.46%Al, 2.48%W, 1.88%Mo, 1.69%Ta,
ture [1, 5-12]. For example, practical standard     0.92%Nb, 0.11%C, 0.07%Fe, 0.12%B,
heat treatment cycles used for the most             0.04%Zr and balance nickel. Rectangular
common industrial turbine rotating blade            plates, having a dimension of 1 1 cm2, were
material, a cast polycrystalline IN-738 is in       cut from the most severe degradation zone
the following step: 1) Solution treatment at        of turbine blades. For the HIP condition,
1120 oC for 2-4 hr followed by a rapid gas          specimens were HIPed at pressure of 100
quench (25 oC to 55 oC/minute) to below             MPa for 5 hr at 1,200 oC. Next, the HIPed
650 oC and 2) Precipitation aging at 845 oC         specimens were heat treated according to
for 20 hr followed by a rapid gas quench to         heat treatment conditions, including solution
room temperature, which is similar to the one       treatment and primary and secondary
of another cast superalloy, GTD-111. The            precipitate aging treatments in vacuum furnace.
reason why the standard heat treatment does         Detail of the experimental heat-treatment is
not often work well is that the γ solution          shown in Table 1. Heat treated plates were
temperature for the alloy range from 1175 oC        cross sectioned in order to observe micro-
to 1,190 oC [1]. Furthermore, the heat treatment    structure comparing to those of parallel
condition resulted only in a partial homogeni-      grinded and polished surface of turbine
zation by exclusive dissolution of dendritic        blades. All sectioned samples were polished
core g’ particles [1-8]. Therefore, the alloy was   using standard metallographic techniques
recommended to be solutioning treated at            and were subsequently etched in marble
higher temperatures, such as at 1,200 oC, to        etchant, i.e. 10 g CuSO4, 50 ml HCl, and
fully restore the microstructure after long-term    50 ml H 2 O. The microstructures were
service.                                            observed by scanning electron microscope
      The proper re-heat treatment condition        with secondary electron mode and image
should provide the optimum microstructure,          analyzer.
Chiang Mai J. Sci. 2009; 36(3)                                                                289



Table 1. Heat treatment conditions applied to long-term exposed IN-738.
                                           Primary precipitate         Secondary precipitate
   No.         Solution annealing
                                                 aging                        aging
     1          --------------------         --------------------       845 oC / 20 hr (AC)
     2          --------------------        925 oC / 1 hr (AC)          845 oC / 20 hr (AC)
     3          --------------------        1055 oC / 1 hr (AC)         845 oC / 20 hr (AC)
     4*        1125 oC / 2 hr (AC)           --------------------       845 oC / 20 hr (AC)
     5         1125 oC / 2 hr (AC)          925 oC / 1 hr (AC)          845 oC / 20 hr (AC)
     6         1125 oC / 2 hr (AC)          1055 oC / 1 hr (AC)         845 oC / 20 hr (AC)
     7         1175 oC / 2 hr (AC)           --------------------       845 oC / 20 hr (AC)
     8         1175 oC / 2 hr (AC)          925 oC / 1 hr (AC)          845 oC / 20 hr (AC)
     9         1175 oC / 2 hr (AC)          1055 oC / 1 hr (AC)         845 oC / 20 hr (AC)
    10         1205 oC / 2 hr (AC)           --------------------       845 oC / 20 hr (AC)
    11         1205 oC / 2 hr (AC)          925 oC / 1 hr (AC)          845 oC / 20 hr (AC)
    12         1205 oC / 2 hr (AC)          1055 oC / 1 hr (AC)         845 oC / 20 hr (AC)
* Standard heat treatment condition


3. RESULTS AND DISCUSSION                          have spheroidized and secondary gamma
3.1 The microstructure of as-received              prime coarsened in the airfoil samples.
alloy
     SEM micrographs of the transverse             3.2 HIPed microstructure
sections at about mid blade height of the                SEM investigation in specimens found
airfoil is shown in Figure 1. The microstructure   that the microvoids were very rarely to be
of as-cast alloy generally consists of non-        detected. This fact is very important to
uniform precipitation of ordered L1 2 γ            consider about the advantage of HIP process
intermetallic phase within dendrite core and       for refurbishment of superalloy components.
in the interdendritic region. Coalescence of       HIP process resulted in a reduction of internal
the primary and secondary gamma prime              voids and/or microcracks due to the alloy
particles, as a result of long-term service,       yielding in compression under the action of
seems to occur causing larger and rounded          the applied hot pressure [13-15]. Figure 2
particles. The degree of degradation, as           illustrates the etched microstructure of HIPed
measured by the gamma prime particle size,         specimen with the uniform dispersion of
increases with exposed time and service            some smaller coarse gamma prime particles,
temperature. In this study, however, the           which previously were partially dissolved into
average coarse gamma prime particle size was       the matrix during HIP at high temperature.
approximately 1 mm. The airfoil microstructure     It should be noted that the previous coarse
shows significant degradation in service           gamma prime particles could not be
comparing to the microstructure of the root        completely solutioned at 1,200 oC for 5 hr.
section. The primary gamma prime particles
290                                                           Chiang Mai J. Sci. 2009; 36(3)




       Figure 1. As-received microstructure after long-term service showing the
       coalescence of γ’ particles.




      Figure 2. HIPed microstructure showing partially dissolved gamma prime particles.
Chiang Mai J. Sci. 2009; 36(3)                                                                 291



3.3 Heat treated microstructure after HIP               Figure 3b shows the HIPed microstructure
process                                           followed with both primary aging at 925 oC
      According to the previous work [1],         / 1 hr and secondary aging at 845 oC for
repeating the standard heat treatment sequence    20 hr. It consists of homogeneous distribution
does not always work well. The structure and      of precipitated gamma prime particles with
properties were not fully recovered by this       slightly coarser size in the matrix comparing
refurbishment treatment applied to these          than those of heat treated microstructure of
IN-738 blades. It is reported that the            program No.1. The average diameter size
microstructure was only partially recovered       of these coarse precipitated particles is
by such simple re-heat treatment. However,        approximately 0.18 μm. However, it should
most reheat treatments after HIP process in       be noted that the volume fraction of
long-term serviced turbine blade IN-738,          precipitated phases is also slightly higher
all provide the various microstructural           than that of microstructure in program No.1.
restoration characteristics with more             It can be noted that the primary aging had an
homogeneous structure comparing to the            effect on the obtained microstructure
initial one, as shown in Figure 3. This was due   characteristics. During primary aging at
to the fact that the HIP temperature was high     925 oC , the gamma prime phase could early
enough to dissolve previous coarse gamma          reprecipitate throughout the matrix and
prime particles acting as partial solutioning     continue to grow under secondary aging
process. It was found that the heat treated       period, resulted in slightly coarser size and
microstructure according to program No. 1,        higher volume fraction of γ particles, see
which followed by only secondary aging,           Figures 4 and 5.
consists of the uniform distribution of gamma           Figure 3c shows the HIPed microstructure
prime particles precipitating in the matrix as    followed with primary aging at 1,055 οC / 1
mixing between cubic and round shape,             hr and then secondary aging at 845 οC for 20
see Figure 3a. It should be noted that the        hr. The microstructure consists of gamma
microstructure contains only single size of       prime particles in mixing round-cubic shape
precipitated gamma prime particles having the     precipitating uniformly throughout the matrix.
average diameter size about 0.15 μm. In this      The average size of diameter of precipitated
structure, it could be counted that the heat      gamma prime particles is about 0.17 μm with
during HIP process worked as a partial            slightly less volume fraction than those of
solution treatment at 1,200 oC for 5 hr and       conditions 1 and 2. The primary aging at
then followed by aging at 845 oC for 20 hr.       1,055 οC/1 hr could have an influence on the
The microstructure after HIP process              morphology of γ particles, which is different
was partially prepared for the next step,         from the lower temperature aging (at 925 οC ).
precipitation aging, which previous coarse        During this primary aging with too high
gamma prime particles were partially              temperature, gamma prime phase could
dissolved into the matrix in some degree          precipitate into the matrix but in lower
resulting in smaller size of gamma prime          rate and/or amount comparing to the micro-
particles, see Figure 2. When precipitation       structure after aging at lower temperature
(secondary) aging was applied, uniform            (925 οC) in the alloy. Therefore, the obtained
gamma prime particles precipitated again          volume fraction of gamma prime phase is
from existed particles with higher volume         slightly less than the microstructure of condition
fraction and diameter size.                       No.2. After secondary aging at 845 οC for
292                                                           Chiang Mai J. Sci. 2009; 36(3)




Figure 3. Microstructures of specimens after HIP and heat treatment: a) Condition No. 1, b)
Condition No. 2, c) Condition No. 3, d) Condition No. 4, e) Condition No. 5, f) Condition
No. 6, g) Condition No. 7, h) Condition No. 8, i) Condition No. 9, j) Condition No. 10, k)
Condition No. 11 and l) Condition No. 12.

20 hr, the size and amount of gamma prime      characteristics. The solution treatment at the
particles precipitated increasingly and        temperature provided morphology with
uniformly.                                     bimodal γ phase precipitation, as shown in
    The solution treatment at 1,125 οC/2 hr    Figures 3d-3f of conditions 4-6. The average
has a great effect on microstructure           diameter size of γ particles were about 0.37
Chiang Mai J. Sci. 2009; 36(3)                                                               293




  prime particle diameter (Micron)
      Average size of gamma




                                         Heat treatment condition

Figure 4. The relationship between heat treatment condition and average size of γ particle
diameter..
  Volume fraction of gamma prime
           particles (%)




                                         Heat treatment condition

Figure 5. The relationship between heat treatment condition and volume fraction of γ particles.

and 0.07 μm for coarse and very fine particles,   to the coarser size and more cubic shape.
respectively. This solution treatment also        Hence, after final aging at 845 oC / 20 hr with
resulted in more cubic shape of coarse γ          or without primary aging, these γ particles
particles. The solutioning at 1,125 oC / 2 hr     could continuously grow up and/or repreci-
led to more dissolution into matrix of            pitate into the matrix with coarser size and
previous or residual γ particles after HIP        slightly higher volume fraction of total γ
process. However, this solutioning temperature    precipitated phase comparing to those final
with short heating time also resulted the rapid   microstructures according to conditions No.
growth of some previous γ particles leading       1-3. The addition of primary aging (at both
294                                                                Chiang Mai J. Sci. 2009; 36(3)



925 oC and 1,055 oC for 1 hr) had a slight         845 oC / 20 hr, the γ particles could uniformly
effect on precipitation of very fine γ particles   reprecipitate in single size and round shape
but also provided great effect on volume           throughout the matrix. It should be pointed
fraction of coarse γ particles. These coarse       out that no γ particle in big size was observed,
cubic γ precipitated particles with primary        Figure 3j. However, when inserting primary
aging could lead to the less volume fraction       aging process between solutioning and final
of very fine γ particles during final aging.       aging (or secondary aging), this could lead to
     Figures 3g-3i of conditions 7-9 show the      more precipitation of γ particles in bigger
effect of higher solutioning temperature of        size during primary and secondary agings, as
1,175 oC for 2 hr on final microstructures.        shown in Figures 3k and 3l. The higher
Comparing to microstructures with the lower        temperature (1,055 oC ) of primary aging in
solutioning temperature (at 1,125 oC) of           condition No. 12 resulted in more coarsening
conditions No. 4-6, it was found that this         of γ particles than that of the lower primary
higher temperature of solution treatment           aging temperature of 925 oC in condition
provided the smaller average size of diameter      No. 11.
of precipitated γ particles, see Figure 4. This
higher solutioning temperature could better        4.1 CONCLUSIONS
dissolve the previous γ particles into the              4.1 The most proper solutioning tempera-
matrix comparing to that of lower one. This        ture in this experimental program that should
could lead to the more homogeneous                 be used after HIP process is 1,125 oC, which
reprecipitation during primary and secondary       could provide the final microstructures with
aging comparing to those with lower solutioning    highest volume fractions as well as proper
temperature or without solutioning. The            shape and size of coarse γ particles.
primary aging at 925 oC / 1 hr did not provide          4.2 In most cases, the addition of primary
significant effect on microstructure but only      aging could aid more uniform distribution
resulted in the precipitation of very fine γ       of both coarse and very fine γ particles as
particles in higher volume fraction as well as     well as the increase of volume fractions, in
the less volume fraction of coarse γ particles     comparison with those without primary aging.
comparing to microstructures without primary       However, when applying the primary aging
aging. However, when primary aging at higher       after too high solutioning temperatures
temperature of 1,055 oC for 1 hr was applied,      (of 1,125 oC, 1,175 oC and 1,205 oC), the com-
the γ precipitation could take place rapidly,      bination with the aging with highest
resulting in higher volume fraction and            temperature would lead to fast and uniform
coarser size of coarse γ particles after final     precipitation of coarse γ particles, conditions
aging comparing to those of conditions             No. 6, 9 and 12.
No. 7 and 8.                                            4.3 The most proper heat treatment
     The solutioning at the highest tested         conditions after HIP process for the alloy
temperature of 1,205 oC / 2 hr of conditions       should be condition No. 6 due to its highest
10-12 could cause very high dissolution of         volume fractions of γ particles. The total
previous or residual coarse γ particles after      volume fraction of γ’ particles was nearly 60%
HIP process into the matrix. The obtained          and coarse γ particles precipitated uniformly
microstructures after this solutioning were        and densely. However, mechanical testing,
expected to be very highly homogeneous.            especially at elevated temperatures, such as
Therefore, during next step long term aging at     creep, fatigue and thermal fatigue should
Chiang Mai J. Sci. 2009; 36(3)                                                                     295



be performed in further works in order                  Gamma Prime Particle Coarsening
to investigate the relationship between                 Behavior at Elevated Temperatures in
the rejuvenated microstructures and the                 Cast Nickel Base Superalloy, GTD-111
mechanical properties.                                  EA, High Temp. Mater. Processes, 2008; 27:
                                                        41-50.
5.ACKNOWLEDGEMENTS                                 [7] Wangyao P., Chuankrerkkul N., Polsilapa
      This research work was financially                S., Sopon P. and Pornprasertsuk R.,
supported by the postdoctoral research grant            Gamma Prime Phase Stability after
                                                        Long-Term Thermal Exposure in Cast
(No. MGR 4880088) from The Thailand
                                                        Nickel Based Superalloy, IN-738, Chiang
Research Fund (TRF) and The Commission
                                                        Mai J. Sci., Accepted 2009; 36(3): 287-295.
of Higher Education of Thailand. Special
                                                   [8] Kim M.T., Chang S.Y. and Won J.B.,
thank is also extended to Electricity Generating
                                                        Effect of HIP process on the micro-
Authority of Thailand (EGAT), Nonthaburi,
                                                        structural evolution of a nickel-based
Thailand for material support and technical             superalloy, Mat. Sci. Eng., A, 2006; 441:
help.                                                   126-134.
                                                   [9] Kim M.T., Kim D.S. and Oh O.Y., Effect
REFERENCES                                              of γ’ precipitation during hot isostatic
[1] Daleo J.A., Ellison H.A. and Boone D.H.,            pressing on the mechanical property of
    Metallurgical considerations for life               a nickel-based superalloy, Mat. Sci. Eng.,
    assessment and the safe refurbishment               A, 2008; 480: 218-225.
    and requalification of gas turbine blades,
                                                   [10] Lvova E. and Norsworthy D., Influence
    J. Eng. Gas Turbines Power, 2002; 124: 571-
                                                        of service-induced microstructural
    579.
                                                        changes on the aging kinetics of
[2] Kvackaj T., Zrnik J., Vrchovinsky V. and
                                                        rejuvenated Ni-based superalloy gas
    Wang yao P., Controlled rolling of
                                                        turbine blades, J. Mater. Eng. Perform.,
    Hastelloy-N, High Temp. Mater. Processes,
                                                        2001; 10: 299-312.
    2002; 21: 351-359.
[3] Kvackaj T., Zrnik J., Vrchovinsky V. and       [11] Kim M.T. and Oh O.Y., Effect of high
    Wang yao P., Influence of plastic                   pressure on the solid-liquid phase change
    deformation on creep behaviour of                   of a nickel base superalloy during hot
    NiMoCr alloy, High Temp. Mater. Processes,          isostatic pressing, J. Alloy. Compd., In Press
    2003, 22: 57-62.                               [12] Tawancy H.M. and Al-Hdhrami L.,
[4] Lothongkum G., Khuanleang V.,                       Failure of refurbished turbine blades
    Hromkrajai W. and Wangyao P., Effect                in a power station by improper heat
    of aging on stress relaxation of Inconel            treatment, Eng. Fail. Anal., 2009; 16:
    X-750 bolt at 923 and 1033 K, High Temp.            810-815.
    Mater. Processes, 2006; 25: 175-185.           [13] Stevens R.A. and Flewitt P.E.J., Hot
[5] Wangyao P., Krongtong V., Panich N.,                isostatic pressing to remove porosity
    Chuankrerkkul N. and Lothongkum G.,                 & creep damage, Mater. Design, 1982; 3:
    Effect of 12 heat treatment conditions              461-469.
    after HIP process on microstructural           [14] Widmer R., Current status of HIP
    refurbishment in cast nickel-based                  technology and its prospects for the 1990’
    superalloy, GTD-111, High Temp. Mater.              s, Mater. Design, 1991; 12: 243-250.
    Processes, 2007; 26: 151-160.                  [15] Loh N.L. and Sia K.Y., An overview of
[6] Wangyao P., Polsilapa S., Chaishom P.,              hot isostatic pressing, J. Mater. Process Tech.,
    Zrnik J., Homkrajai W. and Panich N.,               1992; 30: 45-65.