Vol.20 No.2Joumal of Wuhan University of Technology - Mater. Sei. Ed.Jun. 2005Process Optimization for AZ91 Mg- alloy Low-pressure EPC ProcessWU Hebao FAN Zitian Huang Naiyu DONG Xuanpu TIAN Xuefeng( State Key Laboratory of Plastic Forming Simulation and Die and Mold Technology , Schoo of Materials Science and Engineering ,Huazhong University of Science and Technology , Wuhan 430074 , China )Abstract : The influence of a key process tauriable on the mold. filing characeristics of AZ91 Mg alloy uus stud-ied in the low pesure EPC procss. The applied flonv quanity of insert gas from 1 to 5 m'/h associated rvith the pres-surizing rate in the low pressure EPC casting process 1uus considered for rectangle and L- shaqpe plate casting. The experi-mental resuls show that there is an optimal flon quantity of insert gas for good mold flling characeristics in AZ91 Mg-alloy low pessure EPC process . The optimal flow quantity of insert gas for the specimensis 3 to4 m'/h. Either lessor higher than the optimal flov quantity of insert gas would lead to misrun defects or folds , blisters and porosity de-fects . The practice of hub casting confirmned that the lou-pressure EPC process vuith an optimal proessing uariable erem-plfied as 4 m'/h gas flow quantity uas capable of producing complicated magnesim castings rithout misn defects .Key words: magnesium alloy ; lou-pressure casting ; expandable pattern casting ; process optimization1 IntroductionThe molten magnesium alloy flls the mold cavity at a lowpressure in counter-gravity manner ，forming a laminarflow profile. This will reduce the expose time in air forA dramatic increase in the production and utilizationmagnesium melt and overcome the fill problems for mag-of magnesium castings over the world in last ten yeas isnesium melt in conventional EPC process due to its excel-primarily driven by the demand in automotive industryent thermal conductivity and less latent heat. As a newlywilling for increasing the vehicle fuel efficiency by usingdeveloped casting process , there are lttle open documentslighter materials 11. Because of the lower density , highto describe its characteristics and lttle information tostrength-to-weight ratio ，high modulus , superior dampingguide foundrymen in practice. Meanwhile , the most im-characteristic , good machinability and availability , mag-portant initiatives for the development of such a process inesium alloys have been considered as a desirable alterna-ive to aluminum and steel for the production ofthe improvement of mold fling during casting，fromwhich it is expected to significantly reduce the casting de-lightweight components 21. Magnesium alloys are mainlyfects . Producing magnesium alloy castings with low- pres-produced via sand casting ，permanent mold casting andsure EPC process also has the advantages of easy au-high- pressure die-casting process 33 . The magnesiumtomatigation and saving labors as well as better castingproducts from sand casting primarily serve the military andquality and higher yield. Therefore , this paper aimed ataerospace industry. The high-pressure die-casting processinvestigating the optimal processing parameter for low-is generally used to produce the magnesium componentspressure EPC process to produce magnesium productswith relatively simple geometry and lower mechanicalwithout any defects .properties. As the magnesium alloy market expanded sig-nificantly in past few years , it is essential to develop al-2 Experimentalterative casting methods , which could be employed toproduce the magnesium castings with complicated geome-2.1 Experimental equipmenttry and higher mechanical properties that is impossible forThe low-pressure EPC process consists of two sys-either. high-pressure die-casting or sand casting pro-tems , low-pressure casting and vacuum-sealed EPC cast-cess 61ing process. A schematic diagram of a typical low-pres-The expandable pattern casting( EPC ) process hassure EPC casting is shown in Fig. 1. The low-pressurenow been developed into a well-defined manufacturingsystem includes a pressurized crucible and a feeding tubetechnology and currently recognized as a cost-effective vi-that can guide melt flow from the crucible to the bottom oftal option to the conventional casting process for producingthe flask. When a proper pressure is applied on the sur-of near-net shape casting with high quality and high in-face of the molten metal in the crucible , the molten metaltegrityf 7-10]. The low-pressure expandable pattern castingwould be, forced to rise alone the feeding tube and thenprocess is a new precision method to make castings withflow中國(guó)煤化工n via a gating system ,the advantages of EPC process and low-pressure process .whiccNMHGithdrysandandvacumsysten. v uo1 u CApaiuauic pattern is fully replaced ，( Received :Sep. 12 2004 Accepted Feb.20 2005 )the pressure on melt is further increased to reduce feedingWU Hebao(吳和保): Senior Engineer E-mail ivu_ hust @ 163. comshrinkage during solidification. The pressure is released* Funded by the National Natural Science Foundation of Chinaafter the casting is completely solidified. At this time , the( No.50275058 )molten metal in the riser of casting and in the transferringtube is still in liquid state and therefore it can flow backVol.20 No.2WU Hebao et al Process Optimization for AZ91 Mg- aloy Low- pressure EPC Pocss....43to crucible by the action of gravity .5 mmX 40 mm. The two patterms were connected directly2.2 Expendable patternto the horizontal runner with cross-section of 20 mm X 60The experimental plate casting pattern and its gating mm and length of 180 mm. The size of sprue was 46 mmxsystem were prepared using the most widely used expand- 80 mm. The cluster was rigged using commercial glue , asable polystyrene( EPS ) board by a hot wire cutter , which shown in Fig. 2. On one side of each pattern , 18 electricalhad a bulk density of 0.016 g/cm’. In these tests , L- probes were arranged to monitor the flling time of meltshape and rectangle shape of plate patterm with 5 mm in front. Each probe gave an electrical signal once meltthickness were prepared. The ingate was directly cut to- touched the tip of the probe.gether with plate pattern , so the cross-section waso160Probes~EPS pattemnDry rand mold--VacuunPattern-Insert gas邑|Riser tube.GateMolten metal. Crueibl、Runnerig.1 Schematic diagram of experimental equipmentFig.2 Schematic diagram of expendable patterm used2.3 Coating3.1 Mold flling profileThe glued clusters were coated with specially con-Using the times at which the molten metal front con-fected refractory slurries with Baume 45. The coated clus-tacted each probe , isochronal lines were used to constructters were then put into an oven to dry for 4 hours at 40-60metal flling profiles for the experimental castings. Fig.3C. The clusters were coated and dried twice or evenshows the shapes of the metal front at each 0.5 ,0.3 andmore times until the coating thickness reached 0.5 mm.0.2 seconds interval after metal passed through the in-2.4 Molding , melting and pouringgate of experimental castings.After the probes were positioned on one side of theFig.3( a) shows the flling patterm of Mgalloy in thepatterm and connected to the data acquisition system , thelow- pressure EPC process with 1 m'/h flow quantity of ircoated patterns were placed into an top-open( 400 mm xsert gas. It is indicated that metal-foam interface exhibit-400 mm ) steel flask with a plenum chamber on its base ,ed a smooth convex shape when the flow quantity of insertand subsequently flling dry silica sand ( 40/70 size ) togas was lower , but the molten Mg-alloy flled mold at aenclose the coated pattern and then compacting the dryslow filling rate , leading to form misrun defects , or in-sand using a three-dimension compaction table. After thecomplete fill.coated patter was properly embedded and packed withAs the flow quantity of insert gas increased , thedry sand ,a plastic film of 0.1 mm in thickness was useddriving force for mold filing increased. When the flowto cover the open-top of the flask .quantity of insert gas rised 3 m'/h , the liquid metalPre-made AZ91 magnesium alloy ingots were meltedflowed in form of convex or arrow shape , as shown irand superheated in the resistance furmace. The melt tem-Fig.3 b), the flling time of liquid metal reduced greatlyperature was measured with a chormel-alumel thermocou-and no misrun or other pyrolysis defects occurred in theple. When the alloy began to melt , a mixture gas ofinternal and the surface of plate castings .0.2% -0.5% sulphur hexafluoride( SF。) in nitrogen wasIf the flow quantity of insert gas was too high( eg ,flowed to the melt surface to protect magnesium alloy from5 m?/h) ,however , the metal front did not move smooth-oxidizing and burning. Once the melt was heated to 720ly ,as shown in Fig.3( c). A large number of entrappedC ,0. 5wt% grain refiner was added to the melt whilstpores were found at the inner and the top end of castings .continuously stirring the melt for 5-8 mins for an uniform3.2 Mold illing velocitydistribution of the grain refiner. After settling down andThe flling velocity of liquid Mgalloy was measuredslag-cleaning , the flask with coated EPS patterm was posi-by using several sets of probes that were pre-positioned a-tioned on the top of furmace and vacuum was applied tolong the centerline of the plate casting. The results indi-the fask for fixing the dry sand. Then the flling proce-cate that the metal velocity was greatly influenced by thedure began to operate through a mold-filling control sys-flow quantity of insert gas ,as shown in Fig.4. The aver-age velocitv was slower for low flow quantity of insert gasless中國(guó)煤化工misrun defects. As the3 Results and DiscussionflowC N M H Gsed , the average velocityincreascu arld 10 uciccs 11 uie Internal and on the surfaceThe flling time at different positions on EPS pattermof the rectangle and L-shape plate castings were discov-was variably measured in experiment . Based on the experi-ered for the flow quantity of insert gas at 3 m'/h. Howev-mental results , we can define the optimal process parame-ter of low-pressure EPC process and pour the Mg- alloyer , when the flow quantity of insert gas was over 3 m*/h，the increase of the metal velocity was not notable .hub produ2方數(shù)據(jù)44Jourmal of Wuhan University of Technology - Mater. Sci. Ed.Jun.20053.251.080.9人324272.2t .01.4.301.95.00.8 )1.2 )1.41.68 .0.71.7.70.60.51.20.4| 0.4.3, 0.70.1.100.2氣廠b)(C)( a )Magallay flled the mold with a smooth convex shape ,but leading to misrun defects( b )Mg- allay flled the mold at appropriate rate to obtain a solid casting a solid casting( c )Mg- aloay flledl the mold at higher fling rate which would result in pore defects in the cating partsFig.3 Schematic diagram of the isochronal lines of molten Mg alloy front200 r100 |50-+L-shape plate casing- D-Rectangle plate ceastingFlux of inserting gas /(m/h)Fig4 The influence of flow quantity of insert gas on the metal velocity Fig.5 The expandable polysyrene pttern (a) and Mgalloy casting (b) of hubThe flling profile and velocity of liquid Mg-alloy inunstable metal front , which resulted in casting defects in-the low-pressure EPC process suggest that neither low flowcluding folds ，blisters and internal porosity. The optquantity nor high flow quantity of insert gas was in favor ofmized flow quantity of insert gas for experimental platethe formation of magnesium products. The optimal flowcasting was 3 to 4 m'/h in Mg-alloy low-pressure EPCquantity of insert gas was 3-4 m'/h for the production ofprocess .plate magnesium casting. When the flow quantity is lowerIn addition , the casting trial of hub suggests that thethan 3 m'/h or higher than 4 m'/h , the magnesium cast-low- pressure EPC process with optimal processing parame-ing has misrun or pores defects.ters was capable of making complicated magnesium casting3.3 Casting trialwithout misrun defects.In order to evaluate the reliability of the optimal pro-cessing parameter in Mg-alloy low-pressure EPC process ,Referencesa test hub pattern was selected for the casting trial. Con-[1] R B Brown. International Magnesium Association 55th Annualsidering the density of hub pattern was higher than the ex-World Conference. Light Metal Age ,1988 8 ) 86-93perimental plate casting , the AZ91 ingots were melted and[2] B L Mordike T Ebert. Magnesium Properties ,Application andheated to 760 C which was higher than the pouring tem-Potential . Materials Science and EngineringA 001 302 37-45perature of plate casting. After drossing , degassing , and[3 ] H Fridrich and S Schumann. Research for a New Age of Mag-grain-refining，the hub casting was poured in counternesium in the Automotive Industry. J. Meter. Proc. Technol .gravity manner at a low pressure with a flow quantity of 42001 ,17 276-281m'/h. The cleaning hub casting had good surface perfor-[4] Anon. Metals Handbook ,10h ed. ASM Intenatonal Handbookonmitte ASM Intemational Materials Park 0H 1990 2 :496mances ,as shown in Fig.5. The results suggest that the[5] J Thomson ,F A Fasoyinu. Casting Characteristics of Perma-nent Mold Cast Mg-alloy AZ91E. AFS Transaction ,2002 ,122 :1425- 1438without misruns or any safety issues .[6] AFS Magnesium Lost Foam Casting omite. Genesis of aNew Process :Magnesium Lost Foam Casting. Modern Casting ,4 Conclusions2003 93(4) 26-28[7] M A Tschopp. Fluidity of Aluminum A356 in the Lost FoamThis work confirms that there is an optimal process-Casting Process . AFS Transaction 2002 ,122 :1211-1222ing parameter in order to obtain an appropriate flling ve-中國(guó)煤化工Shape for Casting . Materialslocity and smoth convex pattem of the metal front in Mg--80alloy low-pressure EPC process . The results indicate that[9]YHC N M H Ging of Aluminum Aly Com-ponents. JOM 42 38-44a low flow quantity of insert gas leds to a slow fllig ve-[ 10] EN Pan K Y Liao. Study on the Flling Behavior of the E-locity and smooth convex pattern of metal front , but mis-vaporative Pattem Casting A356 Alloy. AFS Transaction ,un defects occurred. On the other hand ,a high flow1998 ,108 751-760quantity of insert gas leds to a higher metal velocity and