DIMENSIONAL CONTROL OF NICKEL-BASED SINGLE CRYSTAL TURBINE BLADE

INVESTMENT CASTING BY PROCESS CONTROL OPTIMIZATION

Fei Qiu, Kun Bu, Jin-Hui Song, Guo-Liang Tian, and Xian-Dong Zhang
School of Mechanical Engineering, Northwestern Polytechnical University, Xi’an, Shaanxi, China

Copyright Ó 2017 American Foundry Society

DOI 10.1007/s40962-017-0180-5

Abstract
The main research is aimed at the deformation control of hierarchy process method. The chill plate temperature,
single crystal turbine blade margin plate. The measure- shell mould preheating temperature, pouring temperature
ment data and numerical results of margin plate are and withdraw rate are selected as the main process
compared. The previous research indicates that the margin parameters to control the margin plate warpage deforma-
plate warpage deformation is related to the directional tion. Using the process optimization, the margin plate
solidification structure and process parameters. However, warpage deformation is reduced from 0.232295 to
the margin plate structure is designed by experience and 0.181698 mm, decreasing 21.8%.
usually the structure can not be changed. Thus, an opti-
mizing method for the process parameters is put forward to Keywords: turbine blade, margin plate warpage
control the deformation of blade margin plate. The char- deformation control, directional solidification, process
acterization model for describing the margin plate war- optimization
page deformation is established by using analytic

Introduction flat plate is used to study the effect of materials and wall
thickness on warpage deformation by Lee et al.9 and the
Nickel based single crystal superalloys are widely used in result shows no matter what kind of material is used, the
aeronautics and astronautics because of their excellent amount of warpage deformation is inversely proportional
mechanical properties at elevated temperatures1–5 and the to the thickness of the flat plate. Xue10 investigated the
turbine blade as the key part for aero-engine is made by it. deformation reasons of thin-walled parts which is formed
Now, with the upgrade of aero engine’s thrust weight ratio by die casting process and pointed out the main reasons for
and the development of cooling air injection, the shape of thin-walled die casting warping deformation are the
the turbine blade becomes more and more complex. The unreasonable design on structure and process. Therefore
method for controlling the quality of turbine blades during the way to decrease the warpage deformation is to modify
investment casting becomes more complex. Currently, the the structure of parts and to optimize the process. How-
dimensional tolerance, out of range, is a significant reason ever, the structure of turbine blades are usually designed
to induce low precision of turbine blades. Therefore, how by experiments and it’s hard to be changed. In order to
to decrease the margin plate warpage deformation becomes satisfy the design needs, the better way to decrease the
a research focus. warpage deformation is to optimize the investment casting
process.
Warpage deformation is one of the common quality defects
in molded products process.6 Numerous factors can affect Lee et al.11,12 proposed a new calculation model of warping
the warpage deformation during the solidification, includ- deformation for thin-walled parts and the results shows the
ing the structure of parts, materials and the process of new model is useful to control the deformation of warpage
casting. According to the research of Baldw7 and Spatola in injection model. However, the model used to describe
et al.8 the structure of parts and materials play the main the warping deformation is not suitable to the other parts.
influence on warpage deformation. In addition, a simple Wu13 proposed an optimization method using a new

International Journal of Metalcasting

optimizing object, the mean square error between the sur-
face temperature and the given temperature, which is
considered as a key component to reflect the degree of the
warpage deformation. The results, optimized using the
boundary element method, show that the thin-walled part’s
warping deformation decreases. Moreover, Zeng14 intro-
duced the neural network method to decrease the warpage
deformation. The samples, built by orthogonal test exper-
iments, was taken into building the neural network model,
and the process parameters were used as input data, then
the warpage deformation was achieved as the output data.

In this paper, the research is aimed at warpage deformation

control of single crystal turbine blade margin plate by
process optimization. A multiscale three-dimensional (3D)
turbine blade with a gating system model was established
for simulating the displacement field during the investment Figure 1. Heating cycle used during investment casting
casting process. The numerical results of the displacement process.
field is thoroughly analyzed comparing with the experi-
mental data. According to the analysis, a margin plate
warpage deformation characteristics model is established shell mould, and withdrawn thereafter at 4.5 mm/min.
and then the optimization of the process parameters is After the mould has passed the baffle, it is cooled to room
conducted. temperature in the furnace, and then taken out.

Experimental Numerical Simulation

Materials In this paper, the investment casting process is simulated

by finite element method (FEM). The boundary conditions
In this study, the investment casting process is employed to are set referring to Reference 18 and the DD6 alloy thermal
fabricate the turbine blade and single crystal superalloy physical parameters are defined according to References
DD6 is chosen as the base alloy to investigate the influence 19–21, which is shown in Figure. 2. Meanwhile, the
of process parameters on the solidification process simu- mechanical property parameters are shown in Figure 3.
lation of investment turbine blade. The chemical compo- The investment shell unit, the ceramic core and shell mould
sitions of the alloy are shown in Table 1.15 For the shell are assumed as rigid. The thickness of shell mould is about
and core, representative values for alumina and silica-based 7 mm. The thermophysical properties of the shell mold and
shell materials adapted from References 16, 17. ceramic core are given in Table 2. The final model is
illustrated in Figure 4 and the boundary conditions are
Single crystal directional solidification process shown in Table 3.

An industrial-scale investment casting facility at the AECC

Xi’an Aero-Engineer LTD was employed. Moulds were
prepared using wax assembly, ceramic processing and
stem-autoclave dewaxing methods, following the standard
procedures. The single crystal turbine blade was produced
by investment casting and the casting cycle followed is
shown in Figure 1. The molten metal is poured under
vacuum conditions at a temperature of 1500 °C into the

Table 1. DD6 Super Alloy Chemical Composition15

(wt%)

ALloy Cr Co Mo W Ta Re Nb Al Hf Ni

DD6 4.3 9 2 8 7.5 2 0.5 5.6 0.1 Bal

Figure 2. Thermal parameters of DD6.

International Journal of Metalcasting

 Deformation and Numerical Simulation
of Margin Plate

Margin plate deformation analysis

The casting’s solidification sequence makes the margin

plate bear different mechanics during the investment
casting process. Especially for sudden changes in the cross
sectional areas, the warpage deformation of the margin
plate follows the rule, which shows in Figure 5. When the
margin plate area began to solidify, the deformation of
margin plate could be divided into four steps. (a). The
upper and lower part of the margin plate did not solidify
and did not appear deformed as shown in Figure 5a; (b),
The lower part of margin plate started to solidify and the
Figure 3. Mechanical properties of DD6. upper part still stayed in the melting state. At this moment,
due to the shrinkage of the casting, the warpage deforma-
Table 2. Thermal Data for Shell Mold and Ceramic Core tion of the margin plate appear upward as illustrated in
Figure 5b, (c) The lower part of margin plate was already
Material Conductivity Density Specific heat
(W/m k) (kg/m3) (kj/kg/K)
in solidifying state, but the upper part just started to
solidify. The margin plate might stay in a balanced state
Sand silica 0.59 1520 1.20 when the two parts exerted the same contractile force with
Ceramic 1.35 2010 1.103 different direction. Under this condition, the deformation in
margin plate would be eliminated, but it was under a tensile
force, as shown in Figure 5c, (d) The two areas solidified
together, as the upper part of the margin plate has more
volume than the lower part. The upper part showed more
severe shrinkage, then the balance of the whole margin
plate was destroyed, inducing downward warpage, as
shown in Figure 5d.

The margin plate in a single crystal turbine blade is not a

simple shape, but actually is composed of complex curved
cross sectional areas such as a hook and ladder structure.
The assembly structure of these curved panels makes the
margin plate have different thickness, as shown in

Figure 5. The warpage deformation of margin plate

Figure 4. CAE model of numerical simulation. during directional solidification process.

Table 3. Boundary Conditions of Numerical Simulation

Interfacial heat transfer coefficient Casting and chill plate hcastingchill ¼ 1500 þ 1241:5t0:52
Casting and shell mould hcastingmold ¼ 200 þ 1145:6t0:6

Radiant heat transfer coefficient Radiation heat transfer coefficient in high temperature 0.9
Radiant heat transfer coefficient at baffle 0.8
Radiation heat transfer coefficient in low temperature 0.5

International Journal of Metalcasting

Figure 6. The complex design makes the heat exchange area A, which is near the body of turbine blade, solidified,
inhomogeneous during the solidification that induces fur- the heat barrier zone starts to disappear along the direction
ther warpage deformation in margin plate. of A ! B and C ! D, respectively. Because the distance
of dA!B is longer than dA!D , the area D is faster to finish
In order to further study the margin plate deformation of a the solidification than B, and the area C is the last one to
single crystal turbine blade, the laser scanner is used to finish solidification. Thus the deformation is accumulated
measure the profile data of turbine blade. The measurement to the area C.
model and the CAD model are registered with 3D software.
To analyze the deformation of margin plate, the margin Numerical Simulation Analysis of Margin Plate
plate is cut off by six cross-section lines (in Figure 7). Then Areas
each cross section lines are discretized into 400 points. The
nearest point iterative function is used to calculate the Temperature Simulation Results
displacement of each two related points, and it is deter-
mined using The shrinkage rate of the margin plate is different when
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi compared to the other areas of the blade, because of the
di ¼ ðxi2  xi1 Þ2 þðyi2  yi1 Þ2 þðzi2  zi1 Þ2 Eqn: 1 differences in cross-sectional areas. And this discontinuous
structure makes the margin plate have a large deformation
where ðxi2 ; yi2 ; zi2 Þ is the coordinates of the measurable during the solidification. In order to calculate the defor-
model discrete points;ðxi1 ; yi1 ; zi1 Þ is the coordinates of the mation of the margin plate, the temperature of the turbine
nearest point on the CAD model. blade should be analysis for the first.
According to the calculated displacements, a 3D dis-
placement is built as shown in Figure 8 and indicates the The temperature field is shown in Figure 10. The solidi-
max deformation occurs at the C area. The design of the fication sequence of the turbine blade is from the bottom
margin plate could not make its solidification occur at the to the top, and it is opposite to the direction of the heat
same time. The casting’s solidification sequence of A, B, flux. All the heat transfer in this system are through the
C, D shown in Figure 9 are determined by the distance to thermal radiation and the heat conduction of the chill
the chill plate where hC \hB \hA ¼ hD , thus the casting’s plate during solidification. The margin plate is far from
solidification sequence is C [ B[ A = D. However, there the chill plate, the heat exchange is dominated by thermal
is a discontinuous section between the margin plate and radiation. The radiant heat exposure of the mold wall near
turbine blade body, and the shell mould at the connection the furnace is smaller than the far mold wall. Thus, the far
region is thicker than the other areas. Therefore, the heat mold wall of the blade will release more heat by thermal
transfer at the connection region is slow and the connection radiation than the wall near the furnace. The isotherm
areas become a heat barrier zone. The solidification of area shows in Figure 10a that left is higher than right. How-
B, C and D are stopped because of the heat barrier zone. As ever, with the move of the casting system, when the
casting moves to low temperature area. The view angle of
the near furnace wall is bigger than the far from the
furnace wall. Thus, the isotherm shows right is higher
than the left in Figure 10b.

Deformation Simulation Results

Figure 11 shows the registration result of CAD model and

simulation model. Most of the areas are green which means
the deformation of these areas are small. However, the
Figure 6. Margin plate diagram. deformation of margin plate shows bigger than the other
areas. Thus the further study of the margin plate defor-
mation is started, the laser scanner is used to measure the
profile data of margin plate. The detail of the deformation
data is shown in Figure 12.

Figure 12 shows the simulation of margin plate deforma-

tion verified by experimental data. The simulation results
and experimental data indicate the same transform trend
and the max deformation occurs at the VI cross section
(corresponding to the C region in the Figure 8).
Figure 7. Extraction of sectional curve.

International Journal of Metalcasting

 Figure 8. Margin plate displacement field distribution.

deformation are chill plate temperature, withdrawing rate,

shell mould preheating temperature and pouring tempera-
ture. Considering the single factor, the effect results of
deformation for four different parameters are shown in
Figure 13, respectively. The range analysis method is used
to calculate the range of deformation for each parameter
and the results are shown in Table 4. The results indicate
the effect of four investment casting process for margin
plate is not a simple linear model but a complex coupling
model. Thus, the four different process parameters are
selected to control the margin plate deformation.
Figure 9. Heat barrier zone and solidification sequence.

Optimization Model Building

Margin Plate Deformation Control
The deformation of margin plate is concentrated on cor-
Margin Plate Deformation Sensitive Factors
ners. In order to describe the whole deformation of margin
of Extraction
plate, the optimization model is built by analytic hierarchy
process (AHP).10 The margin plate deformation equation is
Generally the directional solidification process needs to
control shell mould preheating temperature, pouring tem- F ¼ ajXj þ bjYj þ cjZj Eqn: 2
perature, chill plate temperature, withdrawing rate, thick-
ness of shell mould, and the heat preservation temperature, where, X is the deformation of the max point; Y is the
etc. However, the main parameters to adjust turbine blade deformation of the first 10% points average; Z is the

International Journal of Metalcasting

 Figure 10. Isotherm of turbine blade at different time during solidification. (a) T =
2550s isotherm. (b) T = 4150s isotherm.

deformation of margin plate, the weight coefficient for X, Y

and Z is a  b  c.

To determine the weight coefficient by AHP has four steps.

a. Establish structure model. Compose their deci-

sion problem into a hierarchy of more easily
comprehended sub-problems, each of which can
be analyzed independently. The elements of the
hierarchy can relate to any aspect of the decision
problem, such as measured or roughly estimated
data. For the warpage deformation model, the
model only have targeted layer and objected
layer. So the warpage deformation is set as the
target layer. X, Y, Z are set as the object layer.
b. Establish comparison matrix. Systematically
evaluate the variation by comparing them to
one another two at a time, with respect to their
impact on an element above them in the hierar-
chy. The decision makers can use concrete data
about the elements to evaluate the degree of
importance for each element. For the warpage
model, the importance degree is divided 1–9 and
Figure 11. The registration of CAD model and simulation the higher the number, the higher the impor-
model. tance.22 There are three factors in the object
layer. With the help of engineering experience,
deformation of the whole points; F is the warpage the factor evaluation index is shown in Table 5.
deformation of margin plate. a, b, c are the weight 0 1
1 5 9
coefficient and a ? b ? c = 1 and a [ 0, b [ 0, c [ 0.
The comparison matrix A ¼ @ 1=5 1 3 A
The weight coefficient is determined by the AHP method. 1= 1= 1
9 3
Statistical analysis of the deformation of all the points
obtained in the simulation experiment is shown in Fig- c. Calculate the weight of a column. When these
ure 14. According to the statistical results, with almost evaluations are converted to numerical values, a
50% deformation concentrate on 0–0.05 mm, because X numerical weight is derived for each element of
accounts for most of the deformation for margin plate. In the hierarchy. Usually the geometric average
order to make the results more precise, the weight coeffi- method and specifications listed average method
cient of X should be the maximum value during the three- are used to calculate the weight of judgment
weight coefficient. And with difference contribution to the matrix. The results with two methods are

International Journal of Metalcasting

 Figure 12. Simulation and experimental comparison of margin plate deformation.

Figure 13. The variation of deformation at different parameters (a) chill plant temperature and
surface deformation; (b) shell mould preheating temperature and surface deformation; (c) pouring
temperature and surface deformation; (d) withdraw rate and surface deformation.

approximate, so any weight calculation method method to deal with the comparison metric in this
can be used to calculate the weight. Method was paper. And, the weight of a column C is
used to deal with Specifications listed average calculated by the Eqns. 3 and 4.

International Journal of Metalcasting

 Table 4. Effect of Investment Casting Process Parameters for Margin Plate by Range Analysis

Process Pouring temperature Shell mould preheating Chill plate temperature Withdrawing rate V
parameters T1 (°C) temperature T2 (°C) T3 (°C) (mm/min)

Range value 0.033124 mm 0.018017 mm 0.022093 mm 0.028281 mm

Table 6. The Value Range of Four Process Parameters

Effect factor T1 (°C) T2 (°C) V(mm/ T3

min) (°C)

Range of 1450–1600 1450–1500 2.0–3.0 10–30

values

The Eqn. 6 is used to compute consistency index (CI), and

CI ¼ 0:01465.
ðkmax  mÞ
CI ¼ Eqn: 6
m1
The CI is used to test consistency ratio (CR). If the
Figure 14. Statistical data distortion of margin plate CR B 0.1, the weight of a column C is fine. The CR is
nodes. calculated by Eqn. 7. Where the RI (Average random
consistency index) is get by RI table and n = 3, RI = 0.52.
Table 5. Factor Evaluation Index
CI
CR ¼ Eqn: 7
Factors X Y Z RI
X 1 5 9 The CR = 0.02817 B 0.1, So the warpage deformation is
Y 1/5 1 3 determined by Eqn. 8.
Z 1/9 1/3 1 F ¼ 0:7482X þ 0:1804Y þ 0:0714Z Eqn: 8
With the Eqn. 8 the warpage deformation of experimental
aij turbine blade’s margin platform is 0.232295 mm (pouring
bij ¼ P
m ; and i; j ¼ 1; 2; 3; m ¼ 3 Eqn: 3 temperature T1 ¼ 1550  C; shell mould preheating
aij temperature T2 ¼ 1450  C; withdraw rate V ¼ 3 mm/min;
i¼1
Pn chill plate temperature T3 ¼ 20  C).
bij
j¼1
Wi ¼ Pm ; and i ¼ 1; 2; 3; m ¼ 3 Eqn: 4 According to the engineering experience, the value range
0
Wi of four process parameters is determined in Table 6.
i¼1 0
1
0:7482
And the C ¼ W ¼ @ 0:1804 A; And the optimization model is Eqn. 9.
0:0714 8
>
> Objective function : min F ¼ CT D
a ¼ 0:7482; b ¼ 0:1804; c ¼ 0:0714 >
>
>
> Design variables : DðX; Y; ZÞ ¼ f ðT1 ; T2 ; T3 ; VÞ
>
>
d. Consistency check. In general, the judgment < s:t:
matrix is inconsistent, but the inconsistency 1450  C  T1  1600  C
>
>
should be within the scope of the permit. So the >
> 1450  C  T2  1500  C
>
>
maximum eigenvectors of corresponding >
> 10  C  T3  30  C
:
characteristic root could be used to compute 2 mm/min  V  3 mm/min
consistency index. The Eqn. 5 is used to compute Eqn: 9
the maximum eigenvalue, and kmax ¼ 3:0293.
Pn P
m where DðX; Y; ZÞ—Deformation 2 matrix
3 under different
aij Wj ki X
j¼1
ki ¼ ; kmax ¼ i¼1 Eqn: 5 influence factors, namely D ¼ 4 Y 5 Cða; b; cÞ—weight of
Wi m Z
0 1
a
a column, namely C ¼ @ b AT1 —pouring temperature, T2
c

International Journal of Metalcasting

 Table 7. Orthogonal Simulation Experiment algorithm input data shows in Table 8. And the optimized
results shows in Figure 15. The warpage deformation of
No. T1 (°C) T2 (°C) T3 (°C) V (mm/min) F/mm
the margin plate keeps stable after iteration number ups to
1 1520 1450 10 2 0.209124 120. Thus the minimum warpage deformation of margin
plate is 0.174024 mm. The optimum process parameters
2 1520 1470 10 2.5 0.217031
is T1 ¼ 1534:642254  C, T2 ¼ 1489:718323  C, T3 ¼
3 1520 1500 10 3 0.219212
24:327215  C, V ¼ 2:046137 mm/min.
4 1550 1450 10 2.5 0.213534 ,
Pi ¼ i P fi ; fi ¼ F1i ; Fi
f N
5 1550 1470 10 3 0.223243 And is the warpage
6 1550 1500 10 2 0.203932 deformation. j¼1

7 1580 1450 10 3 0.235306

8 1580 1470 10 2 0.205912 Optimization Model Validation by Simulation
9 1580 1500 10 2.5 0.220421 Experiment
10 1520 1450 20 2.5 0.216484
The input parameters are set as an integer because of the
11 1520 1470 20 3 0.227756
software parameter setting precision limit. With
12 1520 1500 20 2 0.204688 T1 ¼ 1535  C, T2 ¼ 1490  C, T3 ¼ 24  C, V¼
13 1550 1450 20 3 0.220214 2:05 mm/min the warpage deformation is
14 1550 1470 20 2 0.200396 F = 0.181698 mm. The results of the simulation experiment
15 1550 1500 20 2.5 0.205251 indicate that the warpage deformation optimum model
16 1580 1450 20 2 0.204344 F = 0.174024 mm is similar with the simulation experi-
ment. Therefore, the optimum model could reflect the effect
17 1580 1470 20 2.5 0.214697
of the four process parameters for margin plate warpage
18 1580 1500 20 3 0.226148
deformation. The margin plate warpage deformation com-
19 1520 1450 30 3 0.223104 parison of computation results with and without optimization
20 1520 1470 30 2 0.190684 is shown in Figure 16. Figure 16 indicates that the warpage
21 1520 1500 30 2.5 0.210336 deformation is under control and the maximum deformation
22 1550 1450 30 2 0.189113 is reduced. The F is changed from 0.232295 to 0.181698 mm
23 1550 1470 30 2.5 0.202796 and the warpage deformation decrease 21.8%.
24 1550 1500 30 3 0.213116
25 1580 1450 30 2.5 0.215032 Conclusion
26 1580 1470 30 3 0.226013
27 1580 1500 30 2 0.203642 1. The reasons for the margin plate deformation are
the studying between the measurement data and
CAD model. The measurement data and numer-
shell mould preheating temperature, T3 chill plate ical results of margin plate are compared. The
temperature, V withdraw rate. previous research indicates that the margin plate
The BP (back propagation network) mapping model warpage deformation is related to the directional
DðX; Y; ZÞ ¼ f ðT1 ; T2 ; T3 ; VÞ is established by orthogonal solidification structure and process parameters.
simulation experiment and genetic algorithm is used to get 2. A characterization model of warpage deformation
the optimum solution. The details of orthogonal simulation is built to describe the margin plate deformation.
experiments are shown in Table 7. The AHP is used to determine the weight
coefficient. Moreover, the achieved characteriza-
The BP mapping include four input data and its pouring tion model is shown in Eqn. 8.
temperature, shell mould preheating temperature, chill 3. The range analysis method is used to calculate
plate temperature and withdraw rate. The hidden layer the range of deformation for each parameter and
include nine nodes. The output is warpage deformation. the chill plate temperature, shell mould preheat-
And the predict data compared with the measurement data, ing temperature, pouring temperature, withdraw
the error is 2.9%. What’s more, the parameter of genetic rate are selected as the control process

Table 8. Genetic Algorithm Parameter Setting Table

Algorithm parameters Selection probability Cross probability Mutation probability Population size Genetic algebra

Value Pi 0.1 0.1 40 150

International Journal of Metalcasting

 Figure 15. Process parameters optimization by genetic neural network.

Figure 16. With and without optimization of margin plate of each section warping deformation distribution.

parameters of margin plate warpage deforma- Acknowledgements

tion. Considering the four control parameters,
the optimize model is established. In addition, This study was co-supported by the National Natural
the margin plate warping deformation is Science Foundation of China (No. 51371152).
F ¼ 0:174024 mm.
4. The results of the simulation verification indicate
that the warpage deformation optimum model REFERENCES
F = 0.174024 mm is similar with the simulation
experiment F = 0.181698 mm. The warpage 1. Z. Kan, S. Fida, F. Nisar et al., Investigation of
deformation is under control and the maximum intergranular corrosion in 2nd stage gas turbine blades
deformation is reduced. Moreover, the F is of an aircraft engine. Eng. Fail. Anal. 68, 197 (2016)
changed from 0.232295 to 0.181698 mm, the 2. R. Wang, K. Jiang, F. Jing et al., Thermomechanical
warpage deformation decrease 21.8%. fatigue failure investigation on a single crystal nickel

International Journal of Metalcasting

 superalloy turbine blade. Eng. Fail. Anal. 66, 284–295 14. J.L. Zeng, R.J. Wang, X.X. Feng, Effects of process
(2016) parameters on warpage and shrinkage based on neural
3. C. Panwisawas, J.C. Gebelin, R.C. Reed, Analysis of network. China Plast. Ind. 07, 51–55 ? 72 (2013)
the mechanical deformation arising from investment 15. P. Rong, N. Wang, L. Wang et al., The influence of
casting of directionally solidified nickel-based super- grain boundary angle on the hot cracking of single
alloys. Mater. Sci. Technol. 29(7), 843–853 (2013) crystal superalloy DD6. J. Alloy. Compd. 676,
4. D. Szeliga, K. Kubiak, M. Motyka et al., Directional 181–186 (2016)
solidification of Ni-based superalloy castings: thermal 16. J.R. Li, J.Q. Zhao, S.Z. Liu, M. Han, Effects of low
analysis. Vacuum 131, 327–342 (2016) angle boundaries on the mechanical properties of sin-
5. Z. Li, Q. Xu, B. Liu, Experimental investigation on gle crystal superalloy DD6, in Superalloys 2008, ed.
recrystallization mechanism of a Ni-base single crystal by R.C. Reed, K.A. Green, P. Caron, T.P. Gabb, M.G.
superalloy. J. Alloy. Compd. 672, 457–469 (2016) Fahrmann, E.S. Huron (TMS, Champion, 2008),
6. C. Liu, C. Feng, S. Shi, Effect of injection molding pp. 443–451
parameters on warpage deformation of automobile 17. Y. Dong, K. Bu, Y. Dou et al., Determination of
light lens made of PC. China Plast. 8, 037 (2014) interfacial heat-transfer coefficient during investment-
7. N.C. Baldwin, CAD/CAM/CAE in mold making. casting pro cess of single-crystal blades[J]. J. Mater.
Plast. Design. Process. 25, 12–23 (1974) Process. Technol. 211(12), 2123–2131 (2011)
8. K. Spatola, B. Mindelb, A new generation of warp-free 18. Chinese Society for Metals (CSM), China Aeronauti-
thermoplastic alloys electrical/electronic applications. cal Materials Handbook, vol. 2 (Standards Press of
Eng. Plast. 13(1), 143–148 (1985) China, Beijing, 2011)
9. C.S. Lee, A. Dubin, K. Sarkar. Shrinkage and warpage 19. J. Yu, J.R. Li, J.Q. Zhao et al., Orientation dependence
behavior of injection molded nylon 6 and PET bars of creep properties and deformation mechanism in
and plastics, in ANTEC (1990) DD6 single crystal superalloy at 760 °C and 785 MPa.
10. Y.J. Xue, Measure to prevent deformation of sheet Mater. Sci. Eng. A 560, 47–53 (2013)
metal castings. Foundry 02, 30–32 (1997). (in 20. Q.H. Wu, J. Zhang, Y.S. Luo, Composition and
Chinese) Mechanical Property of DD6 Superalloy Revert/Ma-
11. B.H. Lee, B.H. Kim, Optimization of part wall thick- terials Science Forum, vol. 788 (Trans Tech Publica-
ness to reduce warpage of injection moled part based tions, Zürich, 2014), pp. 488–492
on the modified complex method. Polym. Plast. 21. Y. Fang, Y. Li, W. He et al., Numerical simulation of
Technol. Eng. 34(5), 793–811 (1995) residual stresses fields of DD6 blade during laser shock
12. B.H. Lee, Variation of part wall thickness to reduce processing. Mater. Des. 43, 170–176 (2013)
warpage of injection-molded part: robust design 22. T.L. Saaty, How to Make and Justify a Decision: The
against process variability. Polym. Plast. Technol. Analytic Hierarchy Process (AHP) (RWS Publication,
Eng. 36(5), 701–807 (1997) Pittsburgh, 2001)
13. J.J. Wu, Z.E. Ma, On cooling lines of injection mold
cavity approximatable with 2d model. J. Northwest.
Polytech. Univ. 04, 554–559 (1998)

International Journal of Metalcasting