Computer Simulation and Experimental Validation of ...

7 downloads 0 Views 808KB Size Report
effect of gating system geometries, iron chills, and the melt temperature on the shrinkage and porosity distribution of LPSCed Mg engine blocks have been ...
Computer Simulation and Experimental Validation of Low Pressure Sand Casting Process of Magnesium Alloy V6 engine block Yingxin WANG2, Liming PENG1,2, Penghuai FU2, Wenjiang DING1,2 1 Key State Laboratory of Metal Matrix Composite, Shanghai Jiao Tong University, Shanghai 200240 2 National Engineering Research Center of Light Alloy Net Forming, Shanghai Jiao Tong University, Shanghai 200240 Abstract:

292Mpa and 10% respectively.

1. Introduction

Aluminum alloys have commonly used for engine blocks in the past ten years, and continues to rise in automotive industry. In order

With

the

increasing

demands

to

control

to decrease the dioxide carbon emission and fuel

inter-national fuel consumption and reduce

consumption of cars in future, using magnesium

harmful

alloys instead of aluminum alloys for fabricating

automobile manufacturers are being pressured

engine blocks are very attractive for automotive

into developing more fuel efficient vehicles.

industry. This paper reports the development of

Reducing the overall weight of the vehicles is

Mg alloy engine blocks produced by low

the key to achieving this goal. A major

pressure sand casting (LPSC) process with a new

contributor to the weight of vehicle is the engine

heat-resistant JDM1 alloy. For the purpose of

itself, and the most significant component of an

predicating and eliminating the casting defects in

engine is the block, which makes up 20–25% of

Mg engine blocks, the LPSC process is modeled

the total engine weight. Aluminum alloys have

TM

and simulated by Anycasting

emission

into

the

atmosphere,

software. The

commonly used for engine blocks in the past ten

effect of gating system geometries, iron chills,

years, and continues to rise in automotive

and the melt temperature on the shrinkage and

industry for about 50% significant weight

porosity distribution of LPSCed Mg engine

savings made by introducing an aluminum block

blocks have been investigated, and then the

to replace the traditional grey iron block, which

optimum gating system design and LPSC

improves the fuel consumption efficiency and

parameters are obtained. Several engine block

power performance obviously. Further about

castings are practically casted with different

40%reduction

LPSC parameters, and the result of dissecting

magnesium alloy, which could withstand the

the block casting validates the simulation

temperature and stresses generated during the

could

be

achieved

if

a

accuracy very well. Finally, the Mg V6 engine

engine operation, were used for engine blocks[1].

block castings without defects go through GM

This will further improve the fuel consumption

leakage

efficiency and power performance and is very

test

standard

and

shows

good

mechanical properties on bulkhead, with yield

attractive for automotive industry.

tensile

tensile

To devlop a new magnesium alloy engine block,

strength(UTS) and elongation of 151Mpa,

three main problems are needed to be solved: i) a

strength(YTS),

ultimate

suitable material for magnesium engine block; ii)

predicate and eliminate the casting defects in

the structural design for magnesium engine

magnesium engine blocks. Effects of gating

block and iii) the fabrication process of

system geometries, iron chills, and the melt

magnesium engine block. For the first problem,

temperature on the shrinkage and porosity

JDM1 alloy (Mg-3wt%Nd-0.2wt%Zn-0.4wt%Zr)

distribution of LPSCed magnesium engine

[2],

Engineering

blocks were investigated. Finally, using the

Research Center for Light Alloy Net Forming,

LPSC process parameters optimized by the

Shanghai Jiao Tong University, has a better

above

comprehensive mechanical properties used for

magnesium

the sand casting in the fully heat-treated T6

fabricated and their leakage tests and mechanical

condition,

critical

properties on bulkheads were also investigated.

performance of the engine block, as shown in

The casting experimental result validated the

Table 1. For the second problem, through

simulation accuracy very well.

developed

by

which

National

can

meet

the

repeative

simulation

engine

block

work,

sound

castings

were

cooperation with GM Powertrain Lab, a new magnesium

engine

block

structure

was

2. Experimental process

re-designed. And for the third problem, the fabrication process of magnesium engine block

2.1 Simulation process

will be investigated in this paper.

Fig.1 shows the 3D model of Mg engine block cast with LPSC pouring system, as the primary

Table 1: Critical performance of the magnesium

casting case for simulation, named Case 1,

engine block [1]

where 26 inner gates are designed to feed the Mg Room temperature

engine block cast. There is just engine block cast 150℃

177℃

investigate the full defects of the engine block

Yield strength

and gating system but no chills, which used to

120

110

cast. The thermo-physical properties of JDM1

(MPa)

alloy are shown in Fig.2. The initial conditions

Creep

of Case 1 are shown in Table 2 and the LPSC

strength

110

90

Based on the simulation result of Case 1, the

(MPa) Fatigue limit (MPa) Thermal conductivity

processing parameters are shown in Table 3. other several casting cases (i.e. different chills

50

and gating systems) are designed to modify the LPSC parameters.

115W/Km

Creep strength: stress to give 0.1% creep strain after 100 hours. This target was set to be equivalent to the performance of A319.

Cast

Fatigue limit: R= -1,n=5×107 cycles In this paper, magnesium engine blocks were

Gate

fabricated by low pressure sand casting (LPSC) process with a new heat-resistant JDM1 alloy. The process was modelled and simulated by use

Figure 1: 3D model of Mg engine block cast

of AnycastingTM software and self-developed

with pouring system (Simulation case 1)

thermo-phyical property databse, in order to

Table 2: The initial conditions during simulation

Some JDM1 alloy engine blocks were casted

process

according to the LPSC processing parameters Materials

Casting

Initial conditions

JDM1

740℃

suggested by the simulations and dissected for defects validation. Leakage test of the sound JDM1 alloy engine block was done according to

temperature Mold

Silica sand

air

Heat transfer

80℃

the GM HFV6_SOR_Leaktest specification. The

25℃

bulkheads (as shown in Fig.3) are the key

air-mold

5

positions of the engine block which suffers a

air-cast

11

huge load as the engine works and their mechanical properties were tested. After the

cast-mold

coefficient

400

solution treatment (540℃,10h) and the aging

100

treatment (200℃,time varied from 0.5h to 24h),

(W/m2K) fractio 640℃

the tensile test samples were cut into rectangular tensile specimens with an electric-sparking

Table 3: LPSC processing parameter for Mg

wire-cutting machine and details were shown in Fig.4. Tensile testing was carried out on a

engine block using JDM1 alloy during

Zwick/Roell

simulation process LPSC stage

Pressure/MPa

Time/s

Rising stage

0.015

20

Filling stage

0.025

17

Increasing stage

0.065

15

Dwelling stage

0.065

300

Total time

material

test

machine

at

cross-head speed of 0.2 mm/min at room temperature.

352

Figure 3: The location of the mechanical properties test on the bulkhead

Figure 4: The shape and size of the tensile Figure 2: Thermophysical properties of JDM1 alloy 2.2

specimen

3. Results and discussion Defect

property test

validation

and

mechanical

a

3.1 simulation and defects validation results

Fig.5 shows the filling sequence of JDM1 alloy

time of engine block cast is about 365s, which is

engine block for Simulation Case 1. The full

a little larger than that shown in Table 3.

basin runner, almost all of 26 ingates and a small part of engine block cast are filled with a filling fraction of 45% at 7.8s, as seen in Fig.5a. The melt surface has a little fluctuation for large difference of wall thickness of engine block cast as 55% to 75% filling fraction, as shown in Fig.5b and c. The engine block cast is filled soundly at 17.3s (as shown in Fig.5d). As a whole, the JDM1 alloy melt enters into the 26

(a) solidification time=61s and a solidification fraction of 5%

(b) solidification time=108s and a solidification fraction of 10%

ingates through the basin runner and then fill the whole engine block cast smoothly, which presents the characteristics of LPSC process. The full filling time shows a good agreement with the set filling stage in Table 2. The solidification sequence of JDM1 alloy engine block

cast

in

Fig.6

shows

a

sequential

solidification basically. The bottom part (i.e. the

(c) solidification time=156s and (d) solidification time=203s and a solidification fraction of 15% a solidification fraction of 20%

top side of figure) of engine block cast solidifies firstly (Fig.6a and b), and then the bulkhead (Fig.6c and d), finally, the whole engine block cast except for some isolated melt regions (Fig.6e and f), which leads to shrinkage defects of

engine

block

cast.

Fig.7

shows

the

probabilistic shrinkage defects estimated by Retained Melt Modulus. The total solidification

(e) solidification time=251s and (f) solidification time=298s and a solidification fraction of 25% a solidification fraction of 30%

Figure 6: The solidification sequence of JDM1 alloy engine block for Simulation Case 1

It is easy to see that, the shrinkage defect-1 of Fig.7a

connects

to

one

of

the

ingates,

defects-2,3,4,5 and 10 occur on the positions of thick-thin wall junctions, defects-6,7,8 and 9, on (a) filling time=7.8s and a fill

(b) filling time=10.4s and a fill fraction of 60%

the thick positions of engine block cast. All those defects come from the lack of feeding of melt during the solidification of the casting.

(c) filling time=13.0s and a fill fraction of 75%

(d) filling time=17.3s and a fill fraction of 100%

Figure 5: The filling sequence of JDM1 alloy engine block for Simulation Case 1

Chill-2 Defect-1

Chill-1

Defect-3 Defect-2

Figure 8: Simulation case 2 of JDM1 alloy engine block cast Therefore, based on the simulation results of Case 1, Simulation Case 2 is designed and shown in Fig.8: the bulkhead chill (named

Defect-4

chill-1 in Fig.8) and bottom chill (named chill-2 in Fig.8) are added on the thick parts of the bottom of engine block cast to see if the shrinkage defects can be removed. The initial Defect-5

conditions and the LPSC processing parameters for JDM1 alloy engine block during this simulation process is same to that in Simulation

Defect-6

Defect-7

Case 1 and the initial conditions of chills are listed in Table 4. Meanwhile, the JDM1 alloy engine block cast was casted according to the LPSC processing parameters of Simulation Case

Defect-8

2 to make the defect validations. The filling sequence of Simulation Case 2 is similar to that of Simulation Case 1 except for the final fill, as shown in Fig.9d. The engine block cast has not been filled soundly, that is, formation of cold shut on the bottom (bulkhead) of engine block, which is brought by the addition of chills. The

Defect-9

total solidification time of engine block cast is about 319s, which is smaller than that shown in Table 3 due to the addition of chills. Fig.10 Defect-2

shows the comparison between the probabilistic shrinkage defects predicted by Retained Melt Modulus method and the real casting defects in

Defect-10 Defect-7

the block which is fabricated according to the parameters in Simulation Case 2 . It can be concluded that, the shrinkage defects 1,2,3,4,5

Figure 7: Probabilistic shrinkage defects of

and 10 in Fig.7 have no change, the shrinkage

engine block cast as a simulation case 1

defects 6,8 and 9 disappear, the shrinkage defect

7 changes to porosity and a new porosity defect Defect-1

11 occurs. From dessecting block cast, it can be can that Defects 1,2,3,4,5 and 10 occur in real engine block cast, and cold shuts are also found on the bottom of engine block cast, which validates the simulation accuracy. Table 4: The initial conditions of chills Materials chill Heat

Initial conditions

iron

80℃ Defect-3

chill-cast 6000

Defect-2

transfer coefficient (W/m2K)

1000 580℃

chill-mold

640℃

500

Defect-10

Defect-5 Defect-4

(a) filling time=7.7s and a fill fraction of 45%

(b) filling time=10.3s and a fill fraction of 60%

Defect-7 Defect-11

(c) filling time=13.1s and a fill fraction of 75%

(d) filling time=17.2s and a fill fraction of 100%

Figure 9: The filling sequence of JDM1 alloy engine block as a simulation case 2

Figure10: Probabilistic and real shrinkage defects of JDM1 alloy block cast in Case 2

Due to the formation of cold shut and still existing shrinkage defects in Simulation Case 2, another 12 ingates are added and ingate connecting to shrinkage defect-1 is enlarged to enhance the melt feeding capacity during the solidification, named Simulation Case 3 as shown in Fig.11. The initial conditions and the LPSC processing parameters for JDM1 alloy engine block during simulation process is same

Figure 11: Simulation case 3 of JDM1 alloy

to that in Simulation Case 2.

engine block cast

The engine block cast can be filled soundly and

Passages

the total solidification time is about 400s, larger

Oil drains

15cc/min@25kPa

than that shown in Table 3. The probabilistic shrinkage defects estimated by Retained Melt Modulus method show that shrinkage defects-1

Defect-2

Defect-5

Defect-2

and 3 disappear and other defects in Simulation Case 2 still exist in Case 3. Hence, simulation

Defect-4

case 4 is designed base Case 3 as following: the cast pouring temperature is changed to 760℃,

Defect-2

the dwelling stage is 400s for a total time of 452s, other LPSC parameters are same to Defect-10

simulation case 3. The engine block cast can be filled

soundly

and

the

total

calculated Figure 12: Probabilistic

solidification time is about 456s, similar to the

shrinkage

designed. The probabilistic shrinkage defects

defects

estimated by Retained Melt Modulus method are

Case 4

shown in Fig.12. Shrinkage defect-2 has no change, but shrinkage defects 4,5 and 10 decrease and defects 7 and 11 disappear.

Chill-4

To eliminate the shrinkage defects 2,4,5 and 10, Simulation Case 5 is designed and shown in

Chill-3

Fig.13: chill-3 for shrinkage defect-2, bore chill-4 for defects 4,5 and 10 are added, other LPSC parameters are same to simulation case 4.

Figure 13: Simulation Case 5 of JDM1 alloy

Fig.14 shows the probabilistic shrinkage defects

engine block cast

estimated by Retained Melt Modulus. It can be seen that all of shrinkage defects are eliminated. Sound JDM1 alloy engine block is casted successfully according to the LPSC parameters in simulation case 5, as shown in Fig.15. Figure 14: Probabilistic shrinkage defects of 3.2 Leakage test of JDM1 alloy engine block

JDM1 alloy engine block cast in Simulation

Leakage test of the sound JDM1 alloy engine

Case 5

block

was

made

according

to

the

GM

HFV6_SOR_ Leaktest specification, as shown in Table 5. The sound JDM1 alloy engine block can meet

the

requirement

of

the

HFV6_SOR_Leaktest specification. Table 5: Leaktest specification of engine block Feature

Allowable Leak Rate

Water Jacket/Coolant

10cc/min@140kPa

High

Pressure

Figure 15: The sound JDM1 alloy engine block 3.3 Mechanical properties of bulkhead The mechanical properties of JDM1 alloy engine

Passages (4) Oil

12cc/min@140kPa

of

block cast in Simulation

block bulkhead under different conditions at

room temperature is shown in Fig.16, it

1. The sound JDM1 alloy engine blocks were

demonstrated that the alloy has a lower

casted successfully according to the optimum

mechanical strength as cast and T4 condition.

gating system,

The as-cast YTS (Yield Tensile Strength), UTS

processing parameters obtained through the

(Ultimate Tensile Strength) and elongation is

casting simulation process using AnycastingTM

77Mpa, 156MPa and 8.4%. After T4 heat

software. And the probabilistic shrinkage defects’

treatment, those increase to 87Mpa, 206MPa and

validation were confirmed by the real cast

19.4%. The UTS and YTS increase gradually

dissection.

with increasing aging time form 0.5h to 12h, and

2. The sound JDM1 alloy engine block can meet

reach the peak value of 292MPa and 151Mpa

the requirement of the HFV6_SOR_Leaktest

respectively. Both of UTS and YTS have a little

specification. The yield strength of bulkhead can

decrease for a long aging time of 24h. The

meet the critical yield strength of the engine

elongation keeps 10% nearly in the whole aging

block and the YTS, UTS and elongation are

process. It can be seen that the yield strength can

151MPa, 292MPa and 10% respectively.

chills position and

LPSC

meet the critical yield strength of the engine

Acknowledgments

block as shown in Table 1. The fatigue property test of the bulkhead area were also done according to ASTM E466-07 specification, and

This work was conducted under an alliance

shows

fatigue

between National Engineering Research Center

strength of the block castings at RT and 150°C:

good

tension-compression

of Light Alloy Net Forming, Shanghai Jiao Tong

90MPa and 70MPa respectively[3].

University and Powertrain Lab, General Motor, US. The authors appreciate the provision of the 3D structure model of V6 engine block from GM.

References 1.

C.J.Bettles,

C.T.Forwood,

D.S.Jones,

J.R.Griffiths, M.T.Frost, D.H.St.John, Ma Qian, G-L.Song,

J.F.Nie,

AMC-SC1:

A

New

Figure 16: Tensile properties of JDM1 alloy bulkhead

Magnesium Alloy Suitable for Powertrain

at room temperature in different conditions: F(as cast),

Applications, SAE paper No. 2003-01-1365.

T4(solution treated, 540℃×10h), 0.5h (solution and

2. Fu Penghuai, Peng Liming, Jiang Haiyan,

aging treated, 540 ℃ ×10h+200 ℃ ×0.5h) to 24h

Chang Jianwei, Zhai Chunquan, Effects of heat

means T6 treatment.

treatments

on

the

microstructures

and

mechanical properties of Mg–3Nd–0.2Zn–0.4Zr

Conclusions

(wt.%) alloy.Mater Sci Eng A, 486(2008), pp183-192.

The effect of gating system geometries, iron

3. Z.M. Li, Q.G. Wang, A.A. Luo, L.M. Peng*,

chills, and the melt temperature on the shrinkage

P.H. Fu,Y.X. Wang Improved

and porosity distribution of LPSCed JDM1 alloy

fatigue properties of a new magnesium alloy.

engine blocks have been investigated, it can be

Mater. Sci.Eng. A,

concluded that:

high

cycle

A 582 (2013) 170–177.