1、ORIGINAL ARTICLEMoldstructure designandcasting simulationofthe high-pressuredie casting for aluminum automotive clutchhousing manufacturingSeong Il Jeong1& Chul Kyu Jin1& Hyung Yoon Seo2&Jong Deok Kim2& Chung Gil Kang3Received: 23 February 2015 /Accepted: 5 July 2015#Springer-Verlag London 2015Abstr
2、act This study aims ata molddesignbased onthe cast-ingsimulationandmoldstructuralsimulationinordertoman-ufacture automotive clutch housing aluminum parts in a high-pressure die casting process. For melt to flow into the moldandfilltotheproductsectionevenly,agatingsystemwithfivegates was designed. MA
3、GMAsoft, a casting simulation pro-gram, was used to predict or prevent the possibility of castingdefects that may occur in the filling process and solidificationprocess. The way to select proper casting devices for theclutch housing product is also presented. A structural simula-tion of the mold bas
4、e was conducted by means of ANSYSWorkbench to predict possible damage to the vulnerable partof the mold right from the step of mold design. In the actualshot test, five clutch housing products in total weremanufactured. All of them were fully filled and involved nodefectonthesurface.Thespotofshrinka
5、geporositypredictedin the casting simulation and the spot of actual occurrencewerenot exactly the same but quite close.The degreeofhard-nessoftheseproductswasabout84HVregardlessoflocation.Keywords High-pressurediecasting.Moldstructure.Castingsimulation.Castingdefects.Aluminumalloy1 IntroductionDie c
6、asting began when Doehle manufactured die castingproducts by using aluminum alloys in 1915, and since then,the demands have gradually increased as the automotive in-dustry developed 1. Among casting methods of aluminumautomotiveparts,themostcommonlyusedonesaredividedtolow-pressure die casting, gravi
7、ty die casting, and high-pressure die casting. The low-pressure casting and gravitycasting are applicable to automotive parts difficult to makesuch as cylinder head because the internal structure is formedin one mold. Recently, many of automotive parts that havebeen produced usually in such ways as
8、low-pressure castingand gravity casting started to be light-weighted in applicationof high-pressure die casting for mass production through onesingle mold 27. High-pressure die casting (HPDC) is ad-vantageous in that it can mass-produce complicated and pre-cise shapes within a short time from tens o
9、f seconds or100180 s.In the past, die casting mold design was challenging as thetrial and error method was applied by mold designers and on-sitetechnicians.Themolddesignprocess ofhigh-pressure diecasting, in contrast, adopts computer-aided engineering(CAE), includes the filling and solidification pr
10、ocesses in theearly development stage to predict and evaluate the quality,andthusestablishestheoptimalwayofmolddesign5,810.Today, a number of designers adopt mold design methods thatcombine CAE and their practical experiences, which reducesfraction defective, saves costs, and shortens the developmen
11、tperiod.Inaddition,sinceitispossibletopredictmoldstickinganddeformation,thismethodisofgreathelpwhenitcomestostructural design of molds.This study includes a casting simulation for clutch housingproducts that control or deliver the dynamic force of an auto-motive engine using MAGMA. Predicted or prev
12、ented are* Chung Gil Kangcgkangpusan.ac.kr1Graduate School of Mechanical and Precision Engineering, PusanNational University, San 30 Chang Jun-dong, Geum Jung-Gu,Busan 609-735, South Korea2Department of Computer Science and Engineering, Pusan NationalUniversity, San 30 Chang Jun-dong, Geum Jung-Gu,B
13、usan 609-735, South Korea3School of Mechanical Engineering, Pusan National University, San30 Chang Jun-dong, Geum Jung-Gu, Busan 609-735, South KoreaInt J Adv Manuf TechnolDOI 10.1007/s00170-015-7566-4castingdefectspossibleinthefillingandsolidificationprocessof a casting simulation. Efforts were als
14、o put forth into mini-mizing casting expenses. The gating system design and prod-uctshapewereappliedtomolddesignwiththeaimtoproduceoptimized quality products with minimal casting defects. Forcasting equipment selection and structural simulation of themold base, ANSYS Workbench was utilized to predic
15、t molddamages to the vulnerable section of a mold right from thestep of mold design, save mold production costs through theoptimal mold design, and to reduce other mold modificationsand expenses that might additionally occur by minimizingtimes of experiments.2 Conditions of simulation and experiment
16、2.1 Mold modelingSolidmodelingfor clutchhousingproductswas conducted bymeans of Pro/ENGINEER, a commercial 3D modelingprogram, and then other sections such as biscuit, runner, gate,and overflow were also modeled in the order. Figure 1 showsan example of mold modeling to manufacturer clutch housingpa
17、rts in application of the high-pressure die casting method.Five gates were used, and the thickness was 3 mm.Figure 2 shows the 3D image of a mold in actual sizemodeled by using Pro/ENGINEER. The total weight of themold weighed in 3D modeling was about 12.52 ton, the fixedmold about 3.83 ton, and the
18、 moving mold about 8.69 ton,respectively. Two slide cores were operated with five tunnelpins. Specially applied was a chill block with no vacuumdevice. To prevent dispersion toward the chill block uponcasting, a vent insert was added between the moving coreandchillblockasashock-observingdevicetodece
19、leratemelt-ing. A slipper was added around the four guide pins to handleproblems that might occur to the mold. To prevent the slidecore from being pushed toward the hydraulic cylinder uponclosingthemold,aretrograding-preventiveplatewasinstalledonthefixedmoldbase.Adistributorincontactwiththesleevewas
20、designedinaplatetype.ThehydrauliccylinderappliedtoFig. 1 Gating system design of clutch housingFig. 2 3D mold modeling: afixed mold and b movable moldFig. 3 Cooling line of moldInt J Adv Manuf Technolthe slide was of a diode FA type. A rectangular thin hydrauliccylinderwasappliedtotunnelpins.Thelimi
21、tswitcheswereoflever and push types, rather than those for proximity sensors,for the operators safety. As for the ejector of die casting, theejector bar is connected directly to the plate instead of using aC-plate clamp. Figure 3 shows the cooling channel installedfor the mold. Spot cooling and line
22、 cooling methods weremixed around the mold.2.2 Selection of casting devicesForcastingdeviceselection,the working force was calculatedbased on Eq. (1) below, which indicates the relation betweenthe projected area of casting and casting pressure applied tothe mold. In fact, however, the mold clamping
23、force actuallynecessary could be larger than the force applied to the mold,and thus the mold clamping force was calculated by usingEq. (2)11Fs PmC2Ap1where Fsindicates the working force applied to the mold, Pmis the casting pressure, and Apis the general area of projectionT PmC2 ApC2 CC0C1=1000 2whe
24、re T indicates the actual mold clamping force and Cis the safety coefficient.For Al alloy die casting, the mechanical properties wouldimprovedrasticallyupto49MPaofcastingpressure,butover78.45 MPa, the porosity removing effect is insignificant.Thus, the casting pressure is set to 4978 MPa in general1
25、2. In consideration of the shocking force of a plunger, it isappropriate to select a die casting device with mold clampingforce 2025 % greater than the working force (Fs). When thesafety rate to the mold clamping force of a 1600-ton castingdevice is 20 %, the general area of projection of the clutch
26、housing is 1706.4 cm2. When the casting pressure is78.4 MPa, the mold clamping force is 1638.1 ton, which ex-ceeds 1600 ton of the applied equipment. Thus, the castingpressure was decreased down to 73.5 MPa, and the moldclamping force was 1535.4 ton (safety rate 20 %) for the diecasting process.The
27、force ineachquadrantapplied toeachtiebarofthecastingequipmentwascalculatedontheassumptionthatthesafetyrateofthemoldclampingforcewouldbe20%.Figure 4 shows the moldclampingforce applied toa quadrantofthe mold. The force ineachquadrantapplied toeachtie barwasevenashighas392372kgf.2.3 Structural simulat
28、ion of a mold baseTo confirm the safety of the mold structure, a structural sim-ulation was conducted by means of ANSYS Workbench. Theanalysis of mold base deformation at the moving section wasfollowed by the assessment of structural safety.The chemical composition and mechanical properties ofductil
29、e iron (GCD500) for the mold base are presented inTable 1 and Table 2,respectively.Table3 and Fig. 5 showmold structure information and Youngs modulus andPoissons ratio of ductile iron in reflection of the mold designfor structural simulation.To calculate the secure thickness of the mold base,the ba
30、sic thickness was determined in application of thesimplified calculation expression. After the safety coef-ficient 1.5 was applied to the result of the calculation,Fig. 4 Locking force in each quadrant of movable moldTable 1 Chemical composition of GCD500 (wt%)CSiMnPSMgFe2.5 2.7 0.4 0.08 0.02 0.09 b
31、alTable 2 Mechanical properties of GCD500Tensile strength Yield strength Elongation Hardness500 MPa 320 MPa 7 % 150230 HBTable 3 Conditions for structure analysis of movable mold baseClassification Conditions Classification ConditionsCast area (A) 1197.56 cm2Mold width (b) 113 cmDie base legsdistanc
32、e (l)57 cm Cast pressure (P) 73.54 MPaMold basethickness (h)1st 18 cm Youngs modulus 150109N/m22nd 20 cm3rd 22 cm Poissons ratio 0.284th 24 cmInt J Adv Manuf Technoldeformationasmuchas0.20.25mmwasappliedtothe design. This data was reanalyzed to the structuralsimulation in the most realistic conditio
33、ns by usingANSYS Workbench. When the max. deformation valuewas 0.23 mm or higher in the structural simulation, itwas judged as “fail,” and then a reanalysis was conduct-ed by changing the value from the first to the second.When the max. deformation was lower than 0.23 mm,the thickness was reset to o
34、ptimize the design and thena structural simulation was conducted again.To simplify the hardness calculation process, appliedwas Eq. (3) below on the max. deflection (max)ofthesimple beams distributed load:max 5WL4=384EI 3where Windicates the evenly distributed load, L is the simplebeam distance, E i
35、s Youngsmodulus,andI is the sectionssecond moment.2.4 Casting simulation conditionsThe governing equation used for the filling and solidifyinganalysisofMAGMAsoftisonthe assumptionthatfluid flowsand heat transfer are according to mass conservation law, mo-mentum conservation law, and energy conservat
36、ion law.The MAGMA soft package used in this study has the fol-lowing characteristics 5:(1) Ease of physical interpretation of various steps ofalgorithms(2) Conservation of physical properties(3) Better convergence than pure finite element or finite dif-ference methods (FEM or FDM)(4) Reduction of so
37、lution timeGoverning equations used for fluid and solidification anal-yses of 3D incompressible flows include continuity equation,Navier-Stokesequation,energyequation,andvolumeoffluid.Filling analysis involves finite differential method (FDM) 5.For the accuracy and casting analysis time minimization
38、 in acasting simulation, a manual element dividing method wasadopted for Mesh Generation with 37,000,000 volumes asthe goal. Mesh for Solver5, applicableto thin-walled projects,was used. The number of divided metal cells (except forFig. 5 Shape condition ofmovable mold base for structuralanalysisTab
39、le 4 Conditions of high-pressure die castingClassification Conditions InitialtemperatureMaterial Cast ALDC12 650 CFixed die STD 61 250 CMoving die STD 61Cooling material Water 30 CMachine Machine type 1600 ton Tip diameter 120 mmActive length ofshot sleeve860 mmFilled rate of sleeve 40.47 %Cast cond
40、itions Cast pressure 73.5 MPa Slow shot velocity 0.3 m/sFast shot velocity 3.0 m/sLength of slow-speed region625 mmLength of fastspeed region235 mmTable 5 Chemical composition of ALDC12 alloy (wt%)Si Cu Mg Mn Zn Ni Sn Fe Al9.611.5 23 0.1 0.5 3.0 0.3 0.2 1.3 balInt J Adv Manuf Technolmolds)was988,817
41、,and thenumberofelementswas 37,160,832 volumes total in the casting simulation.The casting simulation conditions are presented in Table 4.The pouring temperature is 650 C, and all mold materialswere STD 61. As for the heat transfer definitions betweenmaterial groups of MAGMAsoft, temperature-depende
42、ntHTC value ofMAGMAsoftwas applied tothe castand mold.Thevaluebetweenmoldandmoldwas3500Wm2Kandthatof between mold and cooling channel was 7000 Wm2K.In this casting simulation, even mold erosion and die stick-ing that might occur in an actual shot test were predictedthrough the analysis. To obtain ca
43、sting simulation results sim-ilartothoseinactualon-siteconditions,theinitialtemperatureof mold parts was set to 25 C, and the number of cycles inwhich mold temperature started to be stabilized with no pre-heating applied was checked. As a result, it turned out thatmold temperature was stabilized in
44、the seventh cycle, whichwas applied to the actual shot test.De casting device applied to the casting simulation is the1600-ton cold chamber die casting machine. The plunger tip120 mm and the shot sleeves length 860 mm. The workingpressure was set to 73.5 MPa to prevent it from exceeding thedevice ca
45、pacity. As for the plunger tip, the speed remained at0.3m/sinthelow-speedareaupto625mm,andfromthenon,the filling was conducted at the speed of 3.0 m/s. To enhancethe quality, the chill vent was applied instead of the gas vent.ALDC 12 die casting aluminum alloy was used in the simu-lation and experim
46、ent process. Table 5 shows the chemicalcomposition of ALDC 12.2.5 Shot test conditionsFigure 6 shows the fixed and moving sections of the moldinstalled at the 1600-ton die casting. The injection conditionsare the same asinthe castingsimulationofMAGMApresent-ed in Table 4. For 3 to 5 min before the i
47、njection, the moldsurfacewaspre-heatedbymeansofatorch.Thiswasrepeatedfivetosixtimesatlowspeedof0.3m/sinthefirstpre-heatingworkandthenthreemoretimesathighspeedof3.0m/sinthesecond pre-heating work. After five injections in total, fiveclutch housing parts were produced. After casting, the prod-ucts wer
48、e pulled out of the mold and went through thequenching process in water.The microstructure and Vickers hardness of the producedclutch housing parts were measured. The microstructure wasmeasured at the thickest part and average-thick part. Vickershardness was measured five times at each location with
49、 thefitting load set to 200 g.3 Results and discussion3.1 Structure simulation of mold baseTable 6 presents the structural simulation results and calcula-tion of simplified Eq. (3). To calculate max. deflection (max)with18cmasthestandard,whichisthefirstconditionofmoldbase thickness, 0.258 mm was multiplied with the safety co-efficient 1.5. The result was about 0.387 mm. Since the firstcondition 18 cm exceeded the standard range of deformation0.200.25 mm, the first condition was judged as “failure”according to the expre
