Materials Today Communications 19 (2019) 114–123

Contents lists available at ScienceDirect

Materials Today Communications

journal homepage: [Link]/locate/mtcomm

Crack propagation behavior of die-cast AlSiMgMn alloys with in-situ SEM T

observation and finite element simulation
⁎
Fei Liua, Haidong Zhaoa, , Runsheng Yanga, Fengzhen Sunb
a
National Engineering Research Center of Near-net Shape Forming for Metallic Materials, South China University of Technology, Guangzhou, 510641, China
b
Department of Mechanical Engineering, Imperial College London, Exhibition Road, London, SW7 2AZ, UK

A R T I C LE I N FO A B S T R A C T

Keywords: Crack propagation of die-cast AlSiMgMn alloys with high strength and ductility was investigated using tensile
AlSiMgMn alloys tests with in-situ scanning electron microscopy (SEM) observation and finite element (FE) simulation. Through
Tensile 3D X-ray micro computed tomography (μ-CT) inspection and reconstruction of pores in the alloys, the pore-scale
Pores simulation of tensile fracture was carried out and the stress distribution around pores during the tensile was
In-situ SEM observation
calculated. The tensile results showed that pores and Fe-rich intermetallics had important effects on the mi-
Finite element simulation
crocracks and the pores played a decisive role in main crack path. It is found that the pore morphologies and
locations had significant influence to stress distribution and microcrack. The pores of lower sphericity were
prone to cause multi microcracks and promote crack deflection. The experiment results demonstrated that the
aggregation of brittle α-Fe intermetallics in the alloys also affected the main crack propagation.

1. Introduction were measured by Archimedes Law and two-dimensional (2D) me-

tallograph analysis [17,18]. Irfan [19] showed that using the porosity
A main way of automobile light weighting is usage of aluminum data measured from 2D surface to judge the mechanical properties can
parts or components. High pressure die casting (HPDC) method is one be inaccurate. With the development of high energy X-ray technology,
of the most popular casting methods for forming the components. the recent inspection of pores with 3D X-ray CT technique shows sig-
AlSiMgMn alloys have been widely used for making high-integrity die nificantly advantages to the conventional methods such as providing
castings with high strength and ductility [1–3]. Especially combined information of pore volume, surface area, sphericity, and curvature etc.
with high vacuum die casting method, the alloys can be solution-treated Felberbaum [20] investigated the 3D characteristics of micropores in
at high temperatures improving mechanical performance further [4]. gravity-cast Al-Cu alloys, in particular, the pore curvatures, with high-
However, due to complexity of various components, sometimes slides resolution X-ray tomography. They showed that the separated pores on
should be used in die casting process. In this case, maintaining high metallographic images were interconnected spatially. Li [21] studied
vacuum in the die cavity is not easily realized, and high injection ve- the stress distribution around pores in cast aluminum alloys by FE si-
locity of molten metal in the process usually causes entrapment of mulation. With the application of Abaqus software, Taxer [22] used
cavity gases resulting in final porosities [5,6]. In addition, solidification random distribution model of pores to analyze the effect of their vo-
shrinkage also contributes to formation of shrinkage pores. Pores in die- lume fraction on elastic-plastic deformation behavior. Dezecot [23]
cast aluminum alloys can be classified as gas, gas-shrinkage, and estimated the effect of pore morphology on local strain/stress levels by
shrinkage pores [7,8]. The pore defects are detrimental to the casting static FE simulations. In our previous work, Wan [24] calculated the
properties and performance [9–14]. Mohd [15] investigated the re- statical stress concentration on pores in ADC12 alloys with considera-
lationship of pore size and fatigue lives, and Wang [16] studied the tion of pore morphology from 3D X-ray CT inspection, and correlated
effect of pore distribution on crack initiation. In our previous work [5], the predicted stresses to experimental fatigue failures.
the relation between tensile properties and pore size and fraction of Tensile properties of AlSiMgMn alloys with conventional die
ADC12 alloys was studied, and the critical pore fraction and maximum casting, and simulation of dynamic fracture with scanned pore mor-
diameter were proposed for the qualified properties. phology have not been studied. In the present work, the pores of die-
Conventionally, the volume fractions of pores in aluminum alloys cast AlSiMgMn alloys were inspected and reconstructed by high

⁎
Corresponding author.
E-mail address: hdzhao@[Link] (H. Zhao).

[Link]
Received 4 November 2018; Received in revised form 11 January 2019; Accepted 11 January 2019
Available online 14 January 2019
2352-4928/ © 2019 Elsevier Ltd. All rights reserved.
F. Liu et al. Materials Today Communications 19 (2019) 114–123

Table 1 Table 2
Chemical compositions of the AlSiMgMn alloy (wt, %). Governing equations of the ductile damage model in simulation.
Si Mg Mn Fe Cu Zn Ti Sr Al Function Equation

10.14 0.18 0.59 0.075 0.0009 0.027 0.076 0.0095 Bal Stress tensor σ=(1-D) σ
The stress triaxiality −p
η=
q
State variable ¯p
resolution 3D X-ray μ-CT technique. Tensile tests were conducted with ∫ ¯p
dε
¯p
=1
in-situ SEM observation. With consideration of the internal pore char- εD (η, ε )
Effective plastic displacement u¯˙ p
acteristics, 3D FE simulation of crack propagation during tensile was Ḋ = p
u¯ f
carried out. The computational and experimental results were com-
pared, where the microstructures affecting cracks were analyzed and
the discrepancy was discussed. The effects of porosity on stress dis-
microscopy (SEM-Servopluse) test machine attached under ambient
tribution and crack growth were emphasized.
temperature. The tensile speed is at 1 μm/s to observe the fracture
processing. During the tensile test, the images of sampling surfaces
2. Experimental procedure were recorded at the propagation of 0.3 mm.

2.1. Material and specimen

2.4. FE simulation

The AlSiMgMn alloy chemical compositions measured by ARL4460

To simulate the tensile fracture, a FE model integrated with the
are presented in Table 1. A die-cast automotive board made of Al-
internal micro-pores was developed. Ductile damage model was chosen
SiMgMn alloy was fabricated by 2000 tons cold chamber die casting
for the simulation due to excellent ductility of AlSiMgMn alloys. The
machine. In the die casting process, the pouring and die temperatures
model equations [25] are summarized in Table 2. For the damage in
are 650 ℃ and 200 ℃, respectively. Tensile samples as shown in Fig. 1
ductile metals, the degradation of the stiffness of the material was
were directly machined from the board. In the middle of the samples, a
modeled using a scalar damage variable, D. σ is the effective (or un-
v-notch with angle of 60° and depth of 0.3 mm was machined on one
damaged) stress tensor computed in the current increment. The stress
side. The surfaces of tensile samples were polished by abrasive papers
triaxiality is defined as η ; p is compression stress; q is Mises equivalent
up to 2000# and washed by distilled water and alcohol. Finally, the ¯ pl
samples were etched by 0.5% hydrofluoric acid for better in-situ SEM stress. Equivalent plastic strain at the onset of damage, εD , is a function
¯ pl
observation during tensile experiment. of stress triaxiality (η) and strain rate (ε ) . ūfp is the effective plastic
displacement at failure. When u¯ p = u¯ fp , the material stiffness will be
2.2. Pore inspection with 3D μ-CT fully degraded (D = 1). Abaqus software was used as the platform to
simulate the ductile-type fracture.
The 3D X-ray μ-CT (nanoVoxel-2000 type of Tianjin Sanying Material property was set as isotropic hardening, which is based on
Precision Instrument Co., Ltd) with the highest resolution of 0.5 μm was Von Mises type elastoplastic constitutive law. Extracted from the results
used to detect pores in the samples. The inspection domain was of of the tensile experiment, the fracture strain was set to 0.07 (please see
4 × 2.5 × 1.3mm3 as indicated by the dashed rectangle in Fig. 1, which the tensile curve in Section 3.2). The stress-strain curve of the corre-
is larger than that in the literature [23]. During the CT scanning, the sponding sample tensile was taken as input for the simulation. Crack tip
current and voltage were 80 μA and 120 kV in X-ray tube, respectively. opening displacement (CTOD) is an important criterion in fracture
The samples were rotated with the palstance of 0.5°/min for a series of analysis. Through calculating the value of CTOD with Eq. (1), dis-
X-ray tomography raw images at 1024 × 1024 pixels. To obtain the placement at failure is determined:
pore characteristics in the whole domain, the inspection resolution was
Kc2
at 3.88 μm. Pores were reconstructed and their 3D morphologies and CTOD =
σy Em (1)
spatial distributions were determined via Volume Graphics (VG) Studio
Max 2.2 software. where, Kc is stress intensity factor, σy is yield strength, and Em is young’s
modulus. For AlSiMgMn alloys, Kc is 5.5-9 MPa· m1/2, σy is 162 MPa
2.3. In-situ observation of crack propagation according to tensile curve in Section 3.2 and Em is 76 GPa. Thus, the
CTOD value with Eq. (1) is 2.46–6.57 × 10−3 mm. In the simulation,
The tensile tests were conducted with a Shimadzu scanning electron the value was set to 0.004 according to the literature [26]. Namely, if
the displacement of an element/node exceeds the critical value, 0.004,
the element including the node will be automatically deleted. The alloy
density was set to 2.64 kg/m3, and young’s modulus and poisson ratio
were 76 GPa and 0.33, respectively.
The morphology and location of pores inside the samples were ob-
tained through the 3D X-Ray μ-CT inspection and reconstruction with
VG software. The data of serial sections orthogonal to the Z direction
were imported into Mimics software. Large internal pores and pores
adjacent to surface are mostly detrimental [27–29]. Therefore, the
pores that were bigger than 1 × 10−4 mm3 and adjacent to surface
were retained as well as for the sake of computational efficiency. For
the mesh quality, the ratio of height to base (H/B) in the triangle ele-
ments was controlled no more than 0.4. The sizes of the triangle ele-
ment were adapted to the curvature of geometry and they were be-
tween 3 and 100 μm. This is to improve accuracy and efficiency of the
FE simulation [23]. Fig. 2 shows the obtained volumetric tetrahedral
Fig. 1. Shape and dimensions of the in-situ tensile sample (in mm). meshes.

115
F. Liu et al. Materials Today Communications 19 (2019) 114–123

Fig. 4. Reconstructed pores inside the sample.

is discussed in details.

3. Result

3.1. Microstructure

The microstructures of AlSiMgMn alloys are composed of α-Al, Al-Si

eutecticand intermetallic compounds with pores as shown in Fig. 3a
Fig. 2. 3D FE enmeshment of pores and boundary condition set-up. and b. The Si particles are characterized as short and small rod or fi-
brous in eutectic zones, and the distribution of α-Al grains and Si par-
Calculating the fracture under monotonic loading, the displacement ticles is relatively homogeneous. The EDS analysis was carried out to
of top surface was set to zero and the bottom one was with uniform obtain the compositions of Fe-rich intermetallics. The result is given in
motion at the velocity of 1 μm/s. The simulation aims to calculate the Fig. 3c, confirming the appearance of α-Fe (Al15(Fe,Mn)3Si). The above
effect of pores size and morphology on stresses and crack growth during phases agree with the results of AlSiMgMn alloys by Niklasa and Do-
the tensile. The sample is composed of a homogeneous AlSiMgMn alloy minik [30,31].
matrix (alpha cells, eutectics, and intermetallics) with pores, where The reconstructed pores are shown in Fig. 4. The pores with larger
individual phases and their properties in the matrix were not con- volumes (> 10−6 mm3) are in red. It can be seen that the pores have
sidered in the simulation. One representative of all investigated samples various shapes and volumes. This is contributed to the different for-
mation mechanism of pores including gas entrapment during mold

Fig. 3. Microstructure of the AlSiMgMn alloy: (a) ×100, (b) ×1000, (c) EDS result of α-Fe phase.

116
F. Liu et al. Materials Today Communications 19 (2019) 114–123

Fig. 5. (a) SEM images of fracture of tensile sample, (b) strain and stress curve of in-situ tensile test.

Fig. 6. Crack propagation during the in-situ tensile process.

filling and shrinkage within solidification. Agreed with our previous the notch region, however, the primary crack dominates the damaging
work [8], here the gas-shrinkage pores are larger and are with complex behavior. With the propagation of the primary crack, it was deflected to
shapes. the upper direction (to be discussed below). The slip bands formed at
45 °s from the direction of the stress axis. Microcracks can be observed
which were initiated at the α-Fe intermetallics as indicated in Fig. 6a.
3.2. In-situ observation of crack propagation behavior As the crack further propagates, a few small pores are observed near the
crack tip as shown in Fig. 6b. The details about effect of these micro-
Fig. 5 shows the fracture path on sample surface and obtained structures on crack initiation and propagation will be discussed in
strain-stress curve from the tensile test. The curve can be divided into Section 4.1.
four stages: elastic stage; slip band stage; crack propagation stage and
final fracture stage, as shown in Fig. 5b. During the slip band stage, the
plastic deformation occurred and crack wasn’t observed on the sample 3.3. Simulation results
surfaces. When approaching to the ultimate tensile strength, the crack
was initiated and started to propagate. The sudden changes in the Fig. 7 shows the projected locations of the pores in the FE model
strain-stress curve were caused by the suspending of load during the in- from the top (tensile axis direction) and front views. Based on their
situ tensile to record the sample damage. When the crack began to volumes, the pores are numbered from P1 to P24. P1 to P11 pores are
propagation, nominal stress was decreased dramatically during the with bigger volume. Note that P1, P3, P4, and P6 pores are near the
stage III and IV as can be seen Fig. 5b. sample back surface and P19, P21, and P22 pores are close to the
Fig. 6 displays the crack at the initiation and propagation stages, sample front surface. P1, P2, P4, P6, P10, P11, P12, P14 and P18 pores
respectively. As can be seen in Fig. 6a, secondary cracks are observed at are far away from the minimum section, and P7, P23 and P24 pores are

117
F. Liu et al. Materials Today Communications 19 (2019) 114–123

Fig. 7. Distribution of pores in the FE model: (a) view from Z axis direction, (b) view from Y axis direction.

Table 3 pores, are found on the fracture surface. The simulated fracture surface
Characteristics of pores in the FE model. is shown in Fig. 8b, P1, P4, P5, P7, P8, P9, P13, P17, P19 and P21 pores
Pore Volume (10−6 x sphericity Pore Volume (10−6 x sphericity can be observed. Comparing the two results, most of pores except P13
mm3) mm3) and P19 pores in the simulated fracture were experimentally observed.
Fig. 9 shows that the simulated result of fracture path. It shows that
P1 1118 0.47 P13 61 0.45
the crack first grew towards left, down and then deflected upwards.
P2 987 0.46 P14 61 0.54
P3 179 0.32 P15 57 0.47
During the late period of the tensile, it moved downwards again.
P4 170 0.48 P16 56 0.46 Therefore, the simulation result agrees with the experimental ob-
P5 144 0.36 P17 53 0.65 servation (Fig. 5a) on both the cracking path and pores on the fracture
P6 129 0.64 P18 50 0.42 surface. This qualitatively validated the numerical model and also in-
P7 126 0.46 P19 41 0.6
dicates that the simulation mimicked the expected behavior of cracking
P8 119 0.6 P20 33 0.42
P9 114 0.38 P21 32 0.58 propagation. By comparing the fracture morphology in Fig. 9b and,
P10 112 0.51 P22 31 0.5
P11 104 0.58 P23 30 0.49
P12 93 0.62 P24 18 0.48 4. Discussion

4.1. Phases affecting crack initiation and propagation

on the minimum section. Table 3 lists the volumes and sphericities of 24
pores in the FE model.
Fig. 10 shows the crack propagation of AlSiMgMn alloys observed
Fig. 8a shows the fracture morphology of the sample from SEM
from the in-situ SEM. With applied load (655 N), the microcracks were
observation. Eight pores, namely P1, P4, P5, P7, P8, P9, P17 and P21
initiated at subsurface pores and rough surface in the notch due to

Fig. 8. (a) Pores found on the sample fracture, (b) pores on the simulated fracture surface.

118
F. Liu et al. Materials Today Communications 19 (2019) 114–123

Fig. 9. Simulated fracture of the tensile: (a) fracture path in 3D, (b) fracture on the surface.

Fig. 10. SEM images of in-situ tensile samples during crack propagation: (a) stress/ load: 241 MPa/ 655 N; (b) stress/ load: 225 MPa/ 637 N; (c) stress/ load:
198 MPa/ 527 N; (d) stress/ load: 181 MPa/ 400 N.

machining. Slip bands were formed on angle of approximately 45 ° with deflection is caused by the location and morphology of internal pores
the loading direction, as can be seen in Fig. 10a. (Note that the darker (to be discussed below).
region in the figure was caused by over-etched.) As shown in Fig. 10b Fig. 6 and 10 from the SEM observations only show the fracture
with 637 N load, since the Si particles in eutectics are characterized as information on the sample surface. In Al alloy castings, several defects
small fibrous shape due to Si-modification (see Table 1) and high so- such as pores, oxide film, and Fe intermetallics, have important effect
lidification rate of die casting, microcracks are hardly observed on on tensile properties. The oxide films are formed on free surfaces of
them. However, the microcracks are found on the α-Fe intermetallics. liquid Al during mold filling and broken in the collision of free surfaces.
Due to the differences of plasticity of α-Al grains and eutectics, small Prior study has shown that the higher collision velocities, the smaller
voids are formed at their interfaces as illustrated in Fig. [Link] coa- film size [32]. Therefore, for Al alloy gravity and low pressure die
lescence of these small voids resulted in the growth of the main crack. castings with low mold filling velocity (usually no more than 0.5 m/s
Fig. 10c-d show the fracture morphology at the loads of 527 N and [33]), the sizes of final oxide films were in the order of one hundred
400 N, respectively. In the Fig. 10c, the α-Al grain marked by yellow microns [34–36]. Hence, their influence on the properties has to be
dotted line is located between the main crack and a microcrack between investigated. On the contrary, in high pressure die casting process, Al
α-Fe intermetallic and matrix. Since the two cracks are almost ortho- liquid fills the die cavity with velocity in level of tens meter per second.
gonal to the loading direction, the section in the grain between the two As a result, oxide films in the free surfaces were broken into relative
cracks can cause significant stress concentration. As a result, the main smaller size. Meanwhile, the turbulent filling usually causes gas en-
crack grew through the grain forming the transgranular fracture as trapment and forms pores with size of several tens to several hundreds
shown in Fig. 10d. On the contrary, the α-Al grain marked by blue of microns. Therefore, pores have been chosen as the primary focus
dotted line is located before the main crack, the different deformation [37–44]. The internal pores also played an important role in the crack
rate of the grain and neighbored eutectics caused microcracks and the propagation, as confirmed by Fig. 8. Therefore, the FE simulation of
intergranular fracture occurred there. After the intergranular fracture, stress distribution and crack growth was used to investigate effects of
the crack stopped and then deflected upwards as shown in Fig. 10d. The the internal pores. Fig. 11 shows the stress concentration around the

119
F. Liu et al. Materials Today Communications 19 (2019) 114–123

Fig. 11. Effect of pores on the main crack propagation.

Fig. 12. (a) 3D morphology of P8 pore, (b–d) stress changes around P8 pore during the tensile.

pores during the tensile process. In Fig. 11a–b, the clustered P8, P13, those pores, Fig. 11c shows the stress concentration and microcracks on
and P17 pores caused local stress concentration and microcracks, and P5 and P9 pores where the main crack grew and connected the mi-
hence, led the main crack growing towards them. As a result, the main crocracks on the pores (Fig. 11d). This is also confirmed by the two
crack propagated towards down and left from the notch at the begin- pores on the fracture as shown in Fig. 8a. Since the two pores are lo-
ning of the tensile damage (Fig. 6a) and the three pores were found on cated in the upper and left to P8, P13, and P17 pores as shown in
the fracture (Fig. 8a). as the crack progressed, the clustered pores above Fig. 7(b), this is the understood reason for the crack deflection upwards
the section with minimum area as shown in Fig. 7b determined the in Fig. 10d (without the reference of pores, it is difficult to judge the
direction of main crack propagation. As an example of the effect of accurate location of Fig. 10d in simulation). As can be seen in Fig. 5a,

120
F. Liu et al. Materials Today Communications 19 (2019) 114–123

Fig. 13. (a) 3D morphology of P5 pore, (b–f) stress changes around P5 pore during the tensile.

the main crack grew downwards to the notch section at the tensile final With the proceeding of tensile process, the microcracks formed at the
stage, which is induced by P23 and P24 pores located at the notch lower regions and progressed upwards, seen in Fig. 13c-d. In the later
section in Fig. 7b. Therefore, this further confirms that the distribution stage, the microcrack at the lower part was connected with the main
of pores played decisive role in the crack main deflection. crack (Fig. 13e). The main crack exited the pore from the middle por-
tion (Fig. 13f) due to the stress concentration and microcracks, making
the slight deflection. Therefore, irregular pores can cause stress con-
4.2. Stress concentration around different pores
centration on multi locations and deflect the growth of main crack.
In die-cast aluminum alloys, pores, which are caused by gas en-
trapment during mold filling and solidification shrinkage, are called gas 4.3. Fracture from aggregated Fe-rich intermetallics
and gas-shrinkage pores. The former pores are characterized as regular
shape and high sphericity while the later ones have irregular shape and As shown in Fig. 8, the simulation shows the main crack passing
low sphericity [24]. According to Table 3, P8 and P5 pores, with through P19 and P13 pores. However, they are not experimentally
sphericity value of 0.6 and 0.36, are typical gas and gas-shrinkage observed on the fracture surface. In Fig. 7b, P13 pore is next to P8 and
pores, respectively. Fig. 12 shows the 3D morphology and the contour P17 pores. From the view of the loading direction, P13 pore is in the
maps of Mises stress of P8 pore. The maximum stress is located in the region 1 in Fig. 14a. The morphology of the region at high magnifica-
maximum cross section, namely the minimum bearing area, which is tion, Fig. 14b, clearly shows the agglomerated α-Fe intermetallics on
perpendicular to the loading direction. With increasing tension, the the fracture surface. The results by Bösch [45] and the discussion in
microcracks initiated in that section (Fig. 12c). Fig. 12(d) illustrates Section 4.1 indicate that the intermetallics have very poor plasticity and
that the main crack moved into the pore by connecting the microcrack can initiate microcracks. The similar results are found for P19 pore as
in the right-hand side of the pore part, and exited the pore from the shown in Fig. 14c-d. Therefore, the main crack propagation in these two
microcrack in the opposite direction. regions were determined by the aggregated α-Fe intermetallics. Due to
Fig. 13 shows the 3D morphology of gas-shrinkage P5 pore and the the current limitation of the simulation, the alloy matrix including α-Al
Mises stress distribution during the tensile test. Three stress con- grains, eutectics, and the intermetallics were treated as homogeneous,
centration regions were found in the upper, middle, and lower parts. which can cause this disagreement between the simulation and

121
F. Liu et al. Materials Today Communications 19 (2019) 114–123

Fig. 14. Fracture caused by aggregated Fe-rich intermetallics: (a) region projected from P13 pore high; (b) magnification of the region 1, (c) region projected from
P19 pore high, (d) magnification of the region 2.

experiment. This leads to further simulations to take into account of Asso. Prof. L. Yuan with University of South Carolina.
heterogeneity of alloy matrix.
References
5. Conclusion
[1] S. Otarawanna, C.M. Gourlay, H.I. Laukli, et al., Microstructure formation in
AlSi4MgMn and AlMg5Si2Mn high-pressure die castings, Metall. Mater. Trans. A 40
The crack propagation behavior of die-cast AlSiMgMn alloys was (2009) 1645–1659, [Link]
studied by both the in-situ SEM and FE simulation. The modeling of [2] J. Medved, S. Kores, M. Vončina, Development of Innovative Al–Si–Mn–Mg Alloys
With High Mechanical Properties// TMS Meeting & Exhibition, Springer, Cham,
tensile fracture integrated the realistic pore characteristics from the 3D 2018, pp. 373–380, [Link]
X-ray μ-CT with ductile damage model. The two results were compared [3] G. Timelli, O. Lohne, L. Arnberg, et al., Effect of solution heat treatments on the
and discussed. It shows that microstructure and mechanical properties of a die-cast AlSi7MgMn alloy, Metall.
Mater. Trans. A 39 (2008) 1747–1758, [Link]
(1) The simulation results show good agreement with the experi-
9527-0.
mental observations on crack path. Most of pores on the simulated [4] A. Niklas, S. Orden, A. Bakedano, et al., Effect of solution heat treatment on gas
fracture surface were experimentally observed. This confirms the va- porosity and mechanical properties in a die cast step test part manufactured with a
lidity of current simulation. new AlSi10MnMg(Fe) secondary alloy, Mater. Sci. Eng. A 667 (2016) 376–382,
[Link]
(2) The pores caused serious stress concentration and local micro- [5] H.D. Zhao, F. Wang, Y.Y. Li, et al., Experimental and numerical analysis of gas
cracks, hence, played the decisive effect on the crack propagation and entrapment defects in plate ADC12 die castings, J. Mater. Process Tech. 209 (2009)
deflection. Both the transgranular and intergranular fracture were 4537–4542, [Link]
[6] M. Okayasu, S. Takeuchi, K. Aizawa, In situ observation of material failure in cast
found on α-Al grains. The microcracks were found in the α-Fe inter- aluminum alloy during monotonic loading observed by a high-speed camera, Exp.
metallics and the interface between them and matrix. Microcracks were Mech. 54 (2014) 1479–1489, [Link]
hardly observed on the small fibrous eutectic Si particles. [7] S.G. Lee, A.M. Gokhale, Formation of gas induced shrinkage porosity in Mg-alloy
high-pressure die-castings, Scripta. Mater. 55 (2006) 387–390, [Link]
(3) The simulation results show that the pore shapes have an im- 1016/[Link].2006.04.040.
portant effect on the stress distribution and formation of microcracks. [8] Q. Wan, H. Zhao, C. Zou, Three-dimensional characterization and distribution of
Stress concentrations were found at multiple locations in irregular gas- micropores aluminum alloy high pressure die casting, Acta Metall. Sin. 49 (2013)
284–290, [Link]
shrinkage pores, which can lead the deflection of the crack. [9] Y. Iwata, S. Dong, Y. Sugiyama, et al., Compression behavior of entrapped gas in
(4) The discrepancy between the simulation and experiment was high pressure diecasting, Mater. Trans. 53 (2012) 483–488, [Link]
possibly caused by the aggregated α-Fe intermetallics, which are prone 1016/[Link].2008.10.028.
[10] Aziz, A.K.M. Ahamed, Influence of casting defects on tensile properties of ADC12
to form the microcracks. This leads to further simulations to take into
aluminum alloy die-castings, Mater. Trans 49 (2008) 1621–1628, [Link]
account of heterogeneity of alloy matrix. 10.2320/matertrans.F-MRA2008814.
[11] R.A. Hardin, C. Beckermann, Integrated design of castings: effect of porosity on
mechanical performance, Materials Science and Engineering Conference Series vol.
Acknowledgements 8, (2012) 12069–12076 [Link]
[12] M. Tiryakioğlu Jr, J.T.S.J. Campbell, The effect of structural integrity on the tensile
deformation characteristics of A206-T71 alloy castings, Mater. Sci. Eng. A 487
The authors would like to thank the support of National Industry (2008) 383–387, [Link]
Base Enhanced Project (TC160A310-10-01), Guangdong Science and [13] H.R. Ammar, A.M. Samuel, F.H. Samuel, Porosity and the fatigue behavior of hy-
Technology Project (2015B090926002, 2013B090600118), and poeutectic and hypereutectic aluminum–silicon casting alloys, Int. J. Fatigue 30
(2008) 1024–1035, [Link]
Guangzhou Science and Technology Project (2014Y2-00214). Also, the [14] M. Leitner, C. Garb, H. Remes, et al., Microporosity and statistical size effect on the
authors are thankful to the help of manuscript English improvement by

122
F. Liu et al. Materials Today Communications 19 (2019) 114–123

fatigue strength of cast aluminum alloys EN AC-45500 and 46200, Mater. Sci. Eng. defects on the mechanical properties of secondary AlSi10MnMg (Fe) test parts
A (2017) 707, [Link] manufactured by vacuum assisted high pressure die casting technology, Mater.
[15] S. Mohd, Y. Mutoh, Y. Otsuka, et al., Scatter analysis of fatigue life and pore size Today Proceed 2 (2015) 4931–4938, [Link]
data of die-cast AM60B magnesium alloy, Eng. Fail. Anal. 22 (2012) 64–72, https:// 059.
[Link]/10.1016/[Link].2012.01.005. [31] B. Dominik, P. Stefan, H. Marc, et al., Secondary Al-Si-Mg high-pressure die casting
[16] Q.G. Wang, D. Apelian, D.A. Lados, Fatigue behavior of A356/357 aluminum cast alloys with enhanced ductility, Metall. Mater. Trans. A 46 (2015) 1035–1045,
alloys. Part II–Effect of microstructural constituents, J. Light Metals 1 (2001) 85–97, [Link]
[Link] [32] Haidong Zhao, Itsuo Ohnaka, Yukiko Sako, et al., Gyeongju, KoreaProceedings of
[17] Z. Živcová-Vlčková, J. Locs, M. Keuper, et al., Microstructural comparison of porous the 65th World Foundry Congress2002, Proceedings of the 65th World Foundry
oxide ceramics from the system Al2O3-ZrO2, prepared with starch as a pore-forming Congress (2002).
agent, J. Eur. Ceram. Soc. 32 (2012) 2163–2172, [Link] [33] J. Mi, R.A. Harding, J. Campbell, The tilt casting process, Nt. J. Cast. Metal. Res. 14
jeurceramsoc.2012.02.005. (2002) 325–334, [Link]
[18] J. Fan, D.L. Mcdowell, M.F. Horstemeyer, et al., Cyclic plasticity at pores and in- [34] M. Divandari, J. Campbell, Oxide film characteristics of Al–7Si–Mg alloy in dy-
clusions in cast Al–Si alloys, Eng. Fract. Mech. 70 (2003) 1281–1302, [Link] namic conditions in casting, Nt. J. Cast. Metal. Res 17 (2004) 182–187, [Link]
org/10.1016/S0013-7944(02)00097-8. org/10.1179/136404604225017546.
[19] M.A. Irfan, D. Schwam, A. Karve, et al., Porosity reduction and mechanical prop- [35] Q.G. Wang, D. Apelian, D.A. Lados, Fatigue behavior of A356-T6 aluminum cast
erties improvement in die cast engine blocks, Mater. Sci. Eng. A 535 (2012) alloys. Part I. Effect of casting defects, J. Light. Met. 1 (2001) 0–84, [Link]
108–114, [Link] 10.1016/S1471-5317(00)00008-0.
[20] M. Felberbaum, M. Rappaz, Curvature of micropores in Al–Cu alloys: an X-ray to- [36] M. Avalle, G. Belingardi, M.P. Cavatorta, R. Doglione, Casting defects and fatigue
mography study, Acta Mater. 59 (2011) 6849–6860, [Link] strength of a die cast aluminium alloy: a comparison between standard specimens
actamat.2011.07.005. and production components, Int. J. Fatigue 24 (2002) 1–9, [Link]
[21] P. Li, P.D. Lee, D.M. Maijer, et al., Quantification of the interaction within defect 1016/S0142-1123(01)00112-8.
populations on fatigue behavior in an aluminum alloy, Acta Mater. 57 (2009) [37] J. Linder, A. Arvidsson, J. Kron, The influence of porosity on the fatigue strength of
3539–3548, [Link] high-pressure die cast aluminium, Fatigue Fract. Eng. Mater. Struct. 29 (2006)
[22] T.K. Taxer, C. Schwarz, W. Smarsly, et al., A finite element approach to study the 357–363, [Link]
influence of cast pores on the mechanical properties of the Ni-base alloy MAR- [38] H. Mayer, M. Papakyriacou, B. Zettl, et al., Influence of porosity on the fatigue limit
M247, Mater. Sci. Eng. A 575 (2013) 144–151, [Link] of die cast magnesium and aluminum alloys, Int. J. Fatigue 25 (2003) 245–256,
2013.02.067. [Link]
[23] S. Dezecot, V. Maurel, J.Y. Buffiere, et al., 3D characterization and modeling of low [39] X. Sun, K.S. Choi, D.S. Li, Predicting the influence of pore characteristics on duc-
cycle fatigue damage mechanisms at high temperature in a cast aluminum alloy, tility of thin-walled high pressure die casting magnesium, Mater. Sci. Eng. A 572
Acta Mater. 123 (2017) 24–34, [Link] (2013) 45–55, [Link]
[24] Q. Wan, H. Zhao, C. Zou, Effect of micro-porosities on fatigue behavior in aluminum [40] O. Kuwazuru, Y. Murata, Y. Hangai, et al., X-ray CT inspection for porosities and its
die castings by 3D x-ray tomography inspection, ISIJ Int. 54 (2014) 511–515, effect on fatigue of die cast aluminium alloy, J. Metc. Phys. Soilds 2 (2008)
[Link] 1220–1231, [Link]
[25] Abaqus/CAE User’s Manual, (2017). [41] X. Li, S.M. Xiong, Z. Guo, Failure behavior of high pressure die casting AZ91D
[26] M. Vaz, D.R.J. Owen, V. Kalhori, et al., Modelling and simulation of machining magnesium alloy, Mater. Sci. Eng. A 672 (2016) 216–225, [Link]
processes, Arch. Comput. Method E. 14 (2007) 173–204, [Link] 1016/[Link].2005.10.013.
s11831-007-9005-7. [42] S.G. Lee, A.M. Gokhale, G.R. Patel, et al., Effect of process parameters on porosity
[27] H. Uchikawa, K. Tsukiyama, Studies on the influence of the shape of the pore and distributions in high-pressure die-cast AM50 Mg-alloy, Mater. Sci. Eng. A 427
aggregate on the stress-strain distribution hardened cement paste by finite element (2006) 99–111, [Link]
method, J. Ceram. Soc. Jpn. 83 (1975) 117–121, [Link] [43] M.A. Irfan, D. Schwam, A. Karve, et al., Porosity reduction and mechanical prop-
001384. erties improvement in die cast engine blocks, Mater. Sci. Eng. A 535 (2012)
[28] M. Kobayashi, Y. Dorce, H. Toda, et al., Effect of local volume fraction of micro- 108–114, [Link]
porosity on tensile properties in Al-Si-Mg cast alloy, Met. Sci. 26 (2014) 962–967, [44] J.P. Weiler, J.T. Wood, R.J. Klassen, et al., Relationship between internal porosity
[Link] and fracture strength of die-cast magnesium AM60B alloy, Mater. Sci. Eng. A 395
[29] J.P. Weiler, J.T. Wood, R.J. Klassen, et al., Relationship between internal porosity (2005) 315–322, [Link]
and fracture strength of die-cast magnesium AM60B alloy, Mater. Sci. Eng. A 395 [45] D. Bösch, S. Pogatscher, M. Hummel, et al., Secondary Al-Si-Mg high-pressure die
(2005) 315–322, [Link] casting alloys with enhanced ductility, Metall. Mater. Trans. A 46 (2015)
[30] A. Niklasa, A. Bakedanoa, S. Ordenb, et al., Effect of microstructure and casting 1035–1045, [Link]

123