《外文翻译--对通过塑料注射成型零件的选择性激光融化生产得到的功能梯度材料插入棒的评估 英文版》由会员分享，可在线阅读，更多相关《外文翻译--对通过塑料注射成型零件的选择性激光融化生产得到的功能梯度材料插入棒的评估 英文版（11页珍藏版）》请在金锄头文库上搜索。
1、Evaluating the use of functionally graded materialsinserts produced by selective laser melting on theinjection moulding of plastics partsV E Beal1*, P Erasenthiran2, C H Ahrens1, and P Dickens21Universidade Federal de Santa Caterina, Floriano polis, Brazil2Wolfson School of Mechanical Engineering, L
2、oughborough University, Loughborough, UKThe manuscript was received on 19 October 2006 and was accepted after revision for publication on 9 March 2007.DOI: 10.1243/09544054JEM764Abstract:The demand for productivity and shape complexity on the injection mouldingindustry necessitates new research to i
3、mprove tool design, material, and manufacturing.A research field is the development of functionally graded materials (FGMs) to build injectionmoulds. For example, moulds built with the FGMs technique can have distinctive regionswith higher heat conduction. Higher rates of heat transfers from thicker
4、 regions of theinjected part can be useful to produce better and cheaper injection moulded polymer parts.It is possible to obtain moulds with differential conductivity by adding locally, during thefabrication of the mould, copper to the mould base material such as tool steel. In this work,an investi
5、gation into the effect of FGM copper (Cu)-tool steel mould insert over polymerinjected parts is presented. The work is divided in two parts: a numerical thermal analysiscomparison between Cu-tool steel graded and tool steel inserts and an injection mouldingexperiment with comparisons between mould s
6、urface temperature and degree of crystallinityof polypropylene parts. The numerical model was used to compare different behaviour of themould heat transfer according to the mould insert material. Thereafter, a bolster was built tohold FGMs and tool steel inserts obtained by a selective laser fusion
7、process. Polypropylenewas injected over the inserts to compare with the numeric results. To observe the effect ofthe cooling rate in the polypropylene parts using the graded inserts, the degree of crystallinityof the parts was measured by differential scanning calorimetry (DSC) test. The temperature
8、of the mould was also evaluated during the injection cycles. The results showed that thegraded Cu-tool steel inserts tested had lower capacity to store heat energy. As Cu was addedto the tool steel, the mixture proved to transfer heat more efficiently but it had less capacityto absorb heat.Keywords:
9、functionallygradedmaterials,injectionmoulding,rapidmanufacturing,polypropylene crystallinity1INTRODUCTIONThe benefit of injection moulded parts depends onthree general aspects: tool cost, injection mould-ing raw material, and productivity of the tool. Thistrio makes it difficult to change part/mould
10、 designwithout affecting productivity and material. Hence,durability of the tool, productivity, and costs must beachieved by the optimal material raw selection andpart and mould designs 1. Unfortunately, there arerestraints that make it difficult to find the best com-promising solution. As the compl
11、exity of a moderninjection mould is high, the mould designers concernis how to solidify the part without causing distortionsand keeping the mould with high rates of parts pro-duced per hour. A complex channels network isdesigned to enable cooling liquid to extract the heatfrom the mould. The design
12、of the channels is difficultas it is necessary to keep the ejection system in place.*Corresponding author: Engenharia Meca nica CIMJECT,Universidade Federal de Santa Catarina, Caixa Postal 476,Campus Universita rio Trindade, Floriano polis, SantaCatarina 88040-900, Brazil. email: JEM764? IMechE 2007
13、Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture945Ejector pins, slides, and air stream gates are used toeject the part from the mould cavity avoiding marksin the aesthetic side of the part. Depending on thecomplexity and shape of the part, the space left bythe cooling system is small and it
14、 is not feasible tomanufacture without leaving marks in the mould imp-ression. In many cases, when the heat extraction equi-librium for homogeneously extracting the heat fromcavity and productivity are not achieved, it might benecessary to redesign the part geometry to fit mouldlimitations. An alter
15、native to solve complex thermalissues is the use of copperberyllium (CuBe) inserts1. As CuBe inserts have higher thermal conductivitythan the usual steel alloys, they are used to extractheat from regions where the cooling channels do nothave an effect during the injection moulding cycles.Nevertheles
16、s, they are not environmentally friendly,as beryllium is cited as a highly carcinogenic element2. Another limitation is that inserts in the mouldimpression might leave marks in the part as the mouldsurface has a visible interface between the basematerial and the insert. In addition, the insert featu
17、resthat are needed to attach it to the base contribute toreducing the space left for the cooling channels.In the mid 1980s, new manufacturing technologiesknown as SFF (solid free-form fabrication) emerged3. The main difference of these technologies, as oppo-sed to the traditional ones, was that they
18、 were basedon the layer additive principle. Also known as rapidprototyping (RP), these technologies can produceparts in low-volume production in virtually anyform or material. The variety of available materialsis limited; however, RP processes can build parts inmetals, ceramics, and polymers 3. RP t
19、echnologiesare highly automated and they are also called three-dimensional (3D) printers as the machines almostprint solid parts from data generated from compu-ter software. Designers and engineers can build andverify designed parts without misunderstandings,inaccuracies, and delays. The basic princ
20、iple of RPtechnologies is to build, layer-by-layer, material cor-responding to the data of the designed part. Rawmaterials can be liquid resins, wires, pastes, powders,and sheets. The way to shape these materials andbond layers can be diverse including ultravioletlasers, lamps, power lasers, spray o
21、f glue, depositionof fused material, and others. These additive-layeredmanufacturing technologies (LMTs) have also beenused to produce tools for injection moulding. Depend-ing on the technology and material used in cons-truction, the complexity of the mould impression(injected part), and the injecte
22、d material, thesemoulds can be competitive to traditional cast/milledmoulds. It is possible to build moulds from 12 to10000 parts according to the technology, material,and application 4, 5. One interesting techniqueused with SFF to build injection moulds is conformalcooling channels. The channels ar
23、e designed in themould impression without the concerns of the lim-itations from the traditional manufacturing method.The conformal cooling channels might follow themould impression surface, passing by the ejector sys-tem with fewer limitations than the usual moulds.Unfortunately, they are still limi
24、ted by the ejectionsystem and some part features such as deep grovesmight not be affected by the cooling capabilities ofthe channels. To overcome some of these constraints,it is possible to use functionally graded materials(FGMs) to build injection moulds by SFF technologies.FGMs have been the subje
25、ct of research for the last25 years 6. Most of the natural materials such asminerals and tissues have a gradual change fromone functional region to another. This example ofnature inspires integrated form and function designall in the same component/unit. FGM is not comple-tely new to the manufacturi
26、ng processes, but it wasonly after the 1980s that it started to receive moreattention and to be classified as a specific researchsubject. The basic idea of FGM is to improve theproperties of the part by varying the quantity of aningredient in specific regions in order to achieve dif-ferential proper
27、ties. An ingredient could be a basicelement such as carbon being used to increase thehardness of a steel part only at the surface. Anotherexample is the porosity variation from the outside tothe inside of the mammals bones. The low porosityfrom the outside increases the stiffness of the bonebut prov
28、ides interconnectivity to the inside. Thecore of the bone is porous, thereby allowing weightefficiency. By using this variation from one materialto another, optimized components can be obtained.Reduced number of joints and fasteners, weightreduction, structural enhancement, differential heatextracti
29、on, thermal barriers, embedded sensors, andbiocompatible implants are some of the potentialadvantages of using FGM 68. FGMs also couldgradually join dissimilar materials with differentproperties in the same component. The principle issimilar to composite materials. The difference isthat composites h
30、ave distinctive phases and do notvary their composition in the volume of the compo-nent. Despite the idea of FGMs being very simple,most of the potential FGM applications are restric-ted to technological limitations and high cost. Diffi-culties in controlling and depositing the gradientcomposition a
31、nd producing complex shapes withcomputer-aided design (CAD), computer-aided man-ufacturing (CAM), and finite element analysis (FEA)integration are some of the causes for restrictionsof use.The use of RP technologies to produce FGM partshas been investigated by many researchers 9. SinceRP technologie
32、s can produce free-form parts and canhandle different materials, it is possible to use them946V E Beal, P Erasenthiran, C H Ahrens, and P DickensProc. IMechE Vol. 221 Part B: J. Engineering ManufactureJEM764? IMechE 2007to produce FGM components. Most of the research-ers investigating the fabricatio
33、n of FGM by LMTs pro-cess the materials with the heat source delivered by alaser beam. As lasers can be easily automated andcan deliver high-energy densities with precision andspeed, they can process almost any material 10.Another aspect of FGM and RP is the frequent useof materials in the form of p
34、owder to be fused orpre-sintered under a laser spot. The main issue forusing rapid processing and manufacturing technolo-gies for producing FGM parts is the local compositioncontrol (LCC). This regards the principle for addingand joining the materials by controlling their percen-tages on each region
35、 of the part or layers. Someresearchers 11, 12 used miniature hopper-nozzlesand capillary tubes to control the deposition of pow-ders in the layer. Ensz et al. 13 studied the optimiza-tion of two powder flows in the laser engineering net-shaping (LENS?) process to build gradients fromH13 to M300 ste
36、el alloys. In addition, computationalmethods to represent the graded geometry havebeen the subject of study. Cho et al. 14 investigatedthe LLC for the 3D printing process after obtaininggeometry and material data from finite element andvoxel space geometries. By this method, it was possi-ble digital
37、ly to represent the 3D part with differentvolumetric gradients.The idea of adding an extra functional material toa base material to produce a FGM injection mouldhas been researched in previous work 15. A multi-compartment hopper was used to produce gradedstructures of H13 tool steel and Cu. The H13
38、iscommonly used as material for injection moulds asit has dimensional stability, toughness, and wearresistance at high temperature. Nevertheless theheat conduction of this material is low comparedwith Cu (kH13: 24.3W/m K; kCu: 385W/mK 16, 17).Elemental Cu powder was mixed with H13 in propor-tions of
39、 12.5, 25, 37.5, and 50%wt to produce FGMbars. The method for producing these bars was theselective laser fusion/melting (SLF or SLM) using ahigh-powder Nd:YAG pulsed laser. The laser pro-cessed the multi-composition powder bed that waspreviously loaded with powders from the multi-compartment hopper
40、. As the laser scanned thepowder bed, the powder was fused and bonded tothe previous added layers. After processing a layer,the powder bed was lowered and the powders werespread over the previous layer and the laser was setto fuse the powder to form a new layer. This processwas numerically controlle
41、d and continued until thecompletion of the part. The fusion process, usingthis laser, left a rough superficial aspect and requiredsome post processing including the removal of thesubstrate that was used to bond the first layers ofthe part to the powder bed platform. At the end ofthe process, graded
42、parts of H13 and Cu could beobtained. Therefore, FGMs could be used on injec-tion moulds to create high heat conductivity regionsto improve heat extraction. As the cooling/heatingchannels can be limited by manufacturability andthe ejector system, some regions of the impressioncould be over heated. T
43、his differential heat extractionfrom the part might cause warpage, sink and coldwelding marks, and poor surface quality, and couldreduce the production rate of the part. Anotherapplication of FGM on moulds is to build the cavityedges with gradients of tool steel and tungsten car-bide. This could imp
44、rove the part quality by red-ucing defects such as flashing caused by wear in themould edges 1.The use of FGM to obtain performance injectionmoulds was one of the stimuli for this research.Despite the limitations of the laser and layer deposi-tion systems used in this work, these experimentswere pla
45、nned to evaluate the influence of the Cuaddition to the H13 matrix. The effect of the additionof Cu on the mould temperature and on the injectedpolymer part crystallinity degree, compared with theH13 base material, was analysed. In theory, the addi-tion of Cu would increase the thermal conduct-ivity
46、 of the mould. The work was divided in twoparts: numerical modelling of the heat transfer andexperimental injection moulding. The first part pre-sents the numerical model of the heat transfer fromthe injected part to the mould and the metallicinserts. The model evaluated the temperature timestamp, s
47、imulating mould inserts in different materi-als: H13, Cu, and H1350%Cu. In the injectionmoulding experiment, FGM bars (mould inserts)were manufactured by laser fusion and placed in astereolithography (SL) mould. Polypropylene (PP)parts were produced by injecting the polymer overthese metallic insert
48、s. Two outputs were analysedfrom this experiment: temperature of the mould sur-face and crystallinity degree of the PP parts. Thetemperature of the mould was measured by thermo-couples in the exact same position taken in thenumerical model. The degree of crystallinity of theparts, moulded with diffe
49、rent inserts, was analysedby differential scanning calorimetry (DSC). The DSCtest was performed to identify if the parts mouldedover different inserts had different cooling rates. As aconsequence, the degree of crystallinity of the partscould be different too. The lower the cooling rate,the greater
50、is the organization of the polymer chainsreflecting in the crystallinity degree of the PP. A rapidcooling rate helps the polymer to hold an amorphousstructure. When heating a plastic, more heat will benecessary to dissolve the crystals (more stable andlower energy state) until the plastic is complet
51、elymelted. This phenomenon can be seen in the DSCcurves measuring the energy absorbed by the samplebefore melting 18.Evaluating the use of functionally graded materials inserts947JEM764? IMechE 2007Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture2METHODOLOGY2.1Numerical modelThe injection mo
52、ulding cycle is a transient phenom-ena and thermal conductivity is not the only materialproperty that counts when analysing the heat trans-fer. Density and specific heat capacity also determinethe capability of the material to store or to transportenergy 19. Considering volume control, the energysta
53、te is obtained by the balance of the energy that isabsorbed, generated, and lost. This variation of theenergy accumulated by the mass inside the volumecan be modelled by equation (1). The energy thatenters (_Ein) plus the energy generated (_Egen) insidethe volume minus the energy lost (_Eout) to the
54、 sur-roundings is equal to the variation of energy (E) ofthe mass inside the volume with respect to time (t)._Ein?_Eout_EgendEdt?vc1In the case of a mould, in the moments after themelted material fills the mould impression, there isno heat generated in defined volume control. Consid-ering the heat t
55、ransfer in one direction, equation (1)becomes further simplified for the heat flux throughan area A, generating equation (2). Simplifying thearea, equation (3) is generated. In these equations, q00represents the heat fluxes, r is the material density,cpsymbolizes the specific heat, T is the temperat
56、ure,and x is the axis of the direction of the heat flux.qin00A ? qout00A 0 ZxrcpTtAdx2qin00? qout00ZxrcpTtdx3The temperature in the mould, away from theimpression, could be considered constant. Takingthis into consideration, in a very short period theheat fluxes can be considered constant and can be
57、described as in equation (4), where k is the heatconduction coefficient.q00 kTx?x4There is no easy solution for solving equations (3)and (4) and a numerical model is usually necessaryto solve them for complex shapes. A two-dimensional(2D) model of an injection moulded part inside a SLmould in contac
58、t with a metallic insert is shown inFig. 1. This model considered no contact resistancebetween the parts, moulds, and insert surfaces. Theinitial conditions were that the temperature in thenodes inside the area that represents the hot PPinjected part was 195C and the temperature for allother nodes,
59、including the connected nodes of thepart with other areas, was 20C. Temperature was cal-culated by employing a quadrangular mesh formedby planar four-node elements. The nodes chosen tobe analysed are indicated in Fig. 1. Thermocoupletemperature, Ttc, matched the same position in theexperimental work
60、 and insert surface temperature,Tis, matched the region from where DSC sampleswere taken in the PP part. The model and analysiswere performed using Ansys software.For inputting the material properties (density,specific heat, and thermal conductivity) in the numer-ical model, tabled values were used
61、for the H13 16.However, the H1350 per cent material propertiesvalues were estimated based on the Voight and Reussrules of mixtures 5, 6. The basic rule of mixtures(Voight) is presented in equation (5). An equivalentproperty () of the mixture formed by a and b phasesis calculated by the summation of
62、the property ofeach phase and the volume fraction (V) of the phasesin the mixture, resulting in a linear variation betweeneach phase value property. The second rule expressedin equation (6) is more elaborate, but neither rulecounts the phase interaction, phase geometry, spacedistribution, and other
63、factors that affect the finalproperty of the mixture. Nevertheless, the secondrule is more conservative than the first one. The mate-rial properties used in the numerical model arepresented in Table 1.Voight Vaa Vbb5ReussbaVab Vba6Intotal,sixsimulationsofthetimeversustemperature in the nodes Ttcand
64、Tis(refer to Fig. 1)were made using different insert materials specifiedin Table 1. The first four simulations were performedFig. 12D model for the heat transfer of an injected part incontact with a metallic insert in a SL mould (initialtemperature indicated for each area of the model)948V E Beal, P
65、 Erasenthiran, C H Ahrens, and P DickensProc. IMechE Vol. 221 Part B: J. Engineering ManufactureJEM764? IMechE 2007with all node temperature being time dependent.The other two simulations were calculated consider-ing that the nodes in the back surface of the insertwere kept at a constant temperature
66、 of 20C, withoutchanges with respect to time. This was made tosimulate the case of using a cooling channel in theback of the insert. The numerical simulations arelisted in Table 2.2.2Injection moulding experimentTo investigate the effect that FGM had on injectionmoulded parts, graded inserts of H13
67、and Cu weremanufactured using SLM and a multi compartmenthopper. The laser scanned tracks over a graded pow-der bed. This graded powder bed was spread by thehopper with different blends of powders. The laserenergy melted the powder building a graded struc-ture layer-by-layer. A basic sketch of the p
68、rocess isshown in Fig. 2.To produce the inserts, bimodal-sized mixturesof H13 and Cu powders were used. The powderswith H1312.5%Cu, H1325%Cu, H1337.5%Cu,and H1350%Cu (weight fractions) and pure H13powder were placed on ceramic ball mills for pro-per mixing of the powders. The average ratio bet-ween
69、the diameter of large and fine particles ofpowders was 6.6:1. This is close to 7:1, which is sug-gested in the literature for increasing the apparentdensity (packing) of powder consisting of sphericalparticles 22. The large particles of Cu and H13were within the range 105150mm. The small particlesof
70、 Cu were less than 22mm in diameter size and theH13 particles less than 38mm. The AISI H13 tool steelalloy (BS EN ISO 4957:2000 XV40CrMoV5-1) composi-tion was, in percentage weight: Fe 90.8%, C 0.320.42%, Cr 4.755.25%, Mn max. 0.4%, Mo 1.251.75%,Si 0.851.15%, and V 0.91.1% and the copper powderwas 9
71、9.99% Cu oxygen free (OHFC) 23.The different compositions of powder were used tofill the multi-compartment hopper and spread over amild steel substrate. The computer numerically con-trol (CNC)-laser-guided system melted the powder,layer-by-layer, using a specific scan strategy patterndeveloped for t
72、his technique and described in detailin reference 15. This strategy was divided in to twostages. In the first step, the laser was set to fusespaced lines of powder leaving a gap between thelines. A refill of powder between these solidifiedlines was executed with the hopper sliding the pow-der bed (w
73、ithout moving the platform). In the secondstep of the scanning strategy, the laser was set tomelt the new powder deposited in the gaps. The laserprocessing parameters were: energy pulse 10J, pulsewidth 20ms, repetition rate 2Hz, and scan speedTable 2Conditions of the numerical simulationsCondition o
74、fthe insertInitial temperatureof the injected partInitial temperature ofthe moulds and insertsTemperature at theback surface of the insertH13195C20CInitial 20CCuH1350Cu VoightH1350Cu ReussH13195C20CConstant 20C(cooling system)H1350Cu ReussTable 1Material properties used in the numerical modelDensity
75、HeatcapacityHeatconductionModel area Material(g/cm3) (J/g.C)(W/mK)SL mouldRenShape 7580resin 201.222.00*0.2*PartPolypropylene 211.00y2.000.13InsertH13 167.800.46024.3Cu 178.960.385385.0H1350%Cu Voight 8.340.425192.2H1350%Cu Reuss8.300.42243.1*estimatedyaverageFig. 2The selective laser fusion process
76、 and the multi-compartmenthopperforspreadingx-gradedpowder bedEvaluating the use of functionally graded materials inserts949JEM764? IMechE 2007Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture5mm/s.Thelayerthicknessusedwas250mmand argon (Ar) was used to reduce oxidation. Afterbuilding the spe
77、cimens, they were post processedby cutting off the substrate and grinding its surfaces.The final dimensions of the graded insert and the dis-tribution of the gradient are presented in Fig. 3. Eachdifferent mixture of Cu and H13 was 3.5mm wide.A pure H13 insert was made from stock annealedH13 by cutt
78、ing and grinding and had the samedimensions as the FGM specimen.An injection mould was designed specifically tohold and swap the inserts. The design of the impres-sion consisted of a simple slab 30 60 2.5mm(Fig. 4). The SL process was used to build the mould.This choice was taken because SL resins h
79、ave lowthermal conductivity and can work as an insulator.Therefore, the influence of the metallic inserts wouldbe more distinctive. The mould design consideredalso the placement in the mould of two thermo-couples to read the mould surface temperatureduring the injection cycles. The positions of thet
80、hermocouples are shown in Fig. 4.The mould was built using a 3D Systems SLA7000stereolithography machine in Huntsman RenShape7580 resin. The K-type thermocouples were positionedand glued to the moulds surface using a commercialepoxy bi-compound resin (Araldite). After the epoxyglue hardened, it was
81、sanded until the thermocoupletips were exposed. Figure 5 shows in detail the mouldfinished and assembled in the bolsters ready to bemounted in the injection moulding machine. Theinjection moulding machine used was a BattenfeldTM750/210 and thermocouples were connected toa CR10X data logger from Camp
82、bell Scientific. Datawere collected each 5s and the room temperaturewas 20C. The injection moulding material was theSolvay Eltex-P HY202 PP.Theinjectionmouldingprocedurestartedbyplacing the FGM insert in the first slot and the H13in the second slot. After 20 injection cycles, the insertswere swapped
83、 and the FGM insert was placed inthe second slot, the H13 in the first slot, and a further20 cycles were performed. Table 3 shows the injectionmoulding cycle and the insert position on each slotcorresponding to Fig. 4.Fig. 5Injection mould (above) and the interchangeableinserts in detail (below)Fig.
84、 3FGM insert and the disposition of the gradientsFig. 4Position of the thermocouples in the cavity950V E Beal, P Erasenthiran, C H Ahrens, and P DickensProc. IMechE Vol. 221 Part B: J. Engineering ManufactureJEM764? IMechE 2007Between each mould opening, a delay of 120s wastaken to allow the inserts
85、 to cool down. A releaseagent was not necessary as the geometry was simpleand an ejection draft angle was used in the mouldimpression surfaces to make part ejection easier(1.5).Theinjectionmouldingparameterswerekept constant and they are presented in Table 4.The low injection and holding pressures a
86、re typicallyused in SL moulds. Most of the SL resins becamesoft above 80C and high pressures and temperatureshad to be avoided in order to increase the mouldslife.Samples from PP parts obtained in the second andsixteenth cycles of the A and B injection moulding ser-ies were taken from the DSC test.
87、The apparatus usedfor these tests was a Shimadzu DSC-60. The sampleswere taken from the same position for all parts andmatched the face in contact with the inserts (FGMand H13) on the same region where the thermo-couples were placed. Figure 4 indicates these regionsfrom sampling. The PP samples, hav
88、ing masses bet-ween 5 and 7mg, were inserted on aluminium pansfor DSC, covered, and placed in the apparatus. Thetests were performed from room temperature (1921C) to 300C with a heating rate of 10C/min. Theresults were analysed using the Shimadzu ta60 version1.51 analysis software. As explained earl
89、ier, the mainobjective of the DSC tests was to compare the degreeofcrystallinity ofa partobtained using different insertsin the injection moulding impression. The percentageof crystallinity can be calculated by equation (7). DH isthe heat of fusion for the polymer tested and theDH100%is the heat of
90、fusion for the same polymerwith 100 per cent crystallinity 24.x% DHDH100%73RESULTS3.1Numerical analysis resultsFrom the numerical simulations it was possible toplot the temperature against the time for the areas(nodes) that represented the thermocouples and sur-face of the inserts. The calculated te
91、mperature in thenode that represents a thermocouple (Ttc) is shownin Fig. 6. For the simulations without cooling system(H13; Cu; H1350Cu Voight; H1350Cu Reuss)the curves of Fig. 6 showed that there was no differ-ence between the cooling rate, although the materialof the insert was different. For the
92、 simulations withthe cooling channel (temperature of the back surfaceof the insert constant at 20C), it showed that thecurves for the two tested insert materials (H13 andH1350Cu) had a more accelerated cooling effectcompared with the inserts without cooling. Neverthe-less, it showed no difference in
93、 the cooling ratebetween them.The results from the other node that representedthe insert surface temperature were relevantly differ-ent. Figure 7 shows the temperature plotted againsttime for node Tis. The heating and cooling rates ofthe insert surface without simulation of the coolingsystemshowedth
94、attherewerenosignificantchanges in the cooling rate although the materialsproperties were different. For the other two analysesthat simulated the cooling system in the back surfaceof the insert, the results showed that the materialaffected the cooling rate. In Fig. 7 the curves of H13and H1350Cu, es
95、timated by the Reuss equation,had higher cooling rates than the other simulatedconditions and it demonstrated that the insert simu-lating the FGM (H1350Cu Reuss) would have afaster cooling rate than the insert of pure H13.Table 4Injection moulding parametersParameterSettingClamping force64kNInjectio
96、n pressure160barInjection speed10ccm/sInjection temperature195C (nozzle)Holding pressure60bar (during 1s)Cooling time30sTime before next injection (open mould)120sTable 3Injection moulding order and position of theinsertsFirst slotSecond slotInjection moulding A series (IMA)FGM insertH13 insertInjec
97、tion moulding B series (IMB)H13 insertFGM insertFig. 6Estimated temperature of the thermocouple (Ttc,mould surface) plotted against the time simulatingdifferent material properties for the insertEvaluating the use of functionally graded materials inserts951JEM764? IMechE 2007Proc. IMechE Vol. 221 Pa
98、rt B: J. Engineering Manufacture3.2Injection moulding experiment resultsThe temperatures measured from the thermocouplesduring typical injection moulding cycles are plottedin Fig. 8. This figure shows that the temperature forthe FGM inserts was higher than the temperaturefor inserts made of only H13
99、. This behaviour wasindependent from the order of the slot as shown inFig. 8. Also, the temperature of the mould surfacefor the position of the FGM insert was considerablyhigher than the H13 insert. In addition, the tempera-ture for H13 inserts was similar to the numericalmodel. On the other hand, t
100、he FGM temperaturewas considerably higher than the numerical resultsobtained.The DSC results reflected the behaviour of thetemperature measured with the thermocouples. TheFGM insert presented higher temperature and lowercapacity to absorb the heat from the part. This wasidentifiable by the higher en
101、ergy necessary to dis-solve the more crystalline samples moulded overthe FGM insert. The curves of the DSC shown inFig. 9 also reveal that with a hotter mould, sixteenthcycle, a higher crystallinity could be observed in thesamples. Table 5 shows the difference between theenergy per gram that was nec
102、essary to melt the sam-ples. Also, the percentage crystallinity is calculatedbased on equation (7). The reference for 100 percent crystalline polypropylene was 209J/g 25.4DISCUSSION AND CONCLUSIONSThe measured temperature of the mould surfaceshowed different results from those obtained fromthe numer
103、ical experiments. The temperatures inthe mould surface were higher than the simulatedresults. In addition, analysis of the curves dropdownshowed that the cooling rate of the simulation washigher than the real readings. The difference mightbe from different causes. The model did not considerphase and
104、 property changes of the polymer during itssolidification. Additionally, the fittings of the insertsin the rough SL bolster do not provide a perfect con-tact surface, as considered in the numerical model.Another source affecting the results was that thecomputational model was based on estimations of
105、Fig. 9Calorimetric plot of the samples taken from PPparts of the second injection moulding cyclesover the first slot (H13 2nd and FGM 2nd) andover the second slot in the 16th cycles (H13 16thand FGM 16th)Fig. 7Temperature of the insert surface (Tis) plottedagainst the time simulating different mater
106、ial pro-perties for the insertFig. 8Temperature of the mould surface, Ttc, during theinjection mouldingTable 5DSC summary of the PP samples analysedSlotInjectionmouldingseriesInsertPeaktemperature(C)Heat(mJ)Heat/mass(J/g)Crystallinitydegree(%)1stIMAFGM167.22635.10 109.50 52.4IMBH13165.28548.0994.50
107、45.22nd IMBFGM166.69464.5776.16 36.4IMAH13165.99382.8563.81 30.5952V E Beal, P Erasenthiran, C H Ahrens, and P DickensProc. IMechE Vol. 221 Part B: J. Engineering ManufactureJEM764? IMechE 2007properties of the graded material. The FGM materialhas its porosity and internal cracks from the buildproce
108、ss. The rules of mixtures did not consider thesevoids in the material. The stock H13 insert presentedless difference between the injection moulding andnumerical results as it did not present these voidsinside its microstructure. Confirming this theory,the comparison between the results showed that t
109、hebehaviour of the cooling curves was equivalent, butthere was a small distinction between the H13 andFGM curves in the results from the numerical model.For the real data measurements, it was possible toidentify that the H13 was more efficient than theFGM inserts in absorbing and storing the heat. T
110、hedifference between the inserts performance showeda temperature peak around 7 degrees lower for theH13 insert. As the objective of the addition of Cu tothe H13 was to increase conductivity it was necessaryto simulate a condition to compare the heat beingtransferred through the inserts. Consequently
111、, thenumerical results simulating the cooling channelsin the back surface of the inserts showed that the50%CuH13 insert had better capacity to transportthe heat away from the part than the pure H13 insert.The DSC results showed no surprise as the tem-perature studies in the numerical and practicalex
112、periments had shown that the FGM mould wasabsorbing the heat slower than the pure H13 insert.Although the properties of the 50%CuH13 mixturewere unknown, it is reasonable to say that theseproperties are lower than the expectation comparingthe thermal performance of pure Cu. Some of the fac-tors that
113、 might be affecting the results have alreadybeen identified in previous work 15. Porosity andcracks in the microstructure of the different propor-tions of Cu and H13 affected the thermal properties.Optimization in the building process can reducethese voids and increase material properties.Although t
114、he results that the conductivity of theCuH13 mixtures was lower than expected, it showedthat it could be used for achieving different coolingrates in the mould using cooling channels. The com-bination of FGM with conformal cooling can bringclear benefits to improve heat extraction, especiallyin case
115、s of deep grooves, thin walls, and highly com-plex parts. Higher proportions of Cu in the gradientmight increase the thermal heat transfer capabilitiesof these regions. Even so, care must be taken as themechanical properties are not characterized and itwould be expected that the material would be so
116、fterand less tough.ACKNOWLEDGEMENTSThe authors would like to thanks to Capes andCNPq (Ministries of Education and Science andTechnology, Brazil) for funding support. Also a spe-cial thanks to Juliano Heidrich, Naguib Saleh, BobTemple, and Rod Springthorpe.REFERENCES1 Menges, G. and Mohren, P. How to
117、 make injectionmolds, 2nd edition, 1993 (Hanser, Munich).2 Center for Epidemiologic Research. Former berylliumworkers medical surveillance program. Oak Ridge Insti-tute for Science and Education, Oak Ridge, Tennessee,USA,availableonlineathttp:/www.orau.gov/cer/BMSP_pro/be-home.htm (accessed December
118、 2005).3 Jacobs, P. F. Stereolithography and other RP&M tech-nologies, 1996 (RPA/ASME Press, New York).4 Ahrens, C. H., Ribeiro, A. S., and Beal, V. E. An alterna-tive cooling technique for stereolithography moulds.ProceedingsofTime-compressiontechnologies2001Conference, Manchester, UK, 2627 Septemb
119、er 2001.5 Dalgarno, K. and Stewart, T. Production tooling forpolymer moulding using the RapidSteel process. RapidPrototyping J., 2001, 7(3), 173179.6 Miyamoto, Y., Kaysser, W. A., Rabin, B. H., Kawasaki, A.,and Ford, R. G. Functionally graded materials: designprocessing and applications, 1999 (Kluwe
120、r AcademicPublishers, Dordrecht, The Netherlands).7 Suresh, S. and Mortensen, A. Fundamentals of function-allygraded materials, 1998(InstituteofMaterials,London).8 Calder, N. Rapid manufacturing of functional materials.Proceedings of Time-compression technologies 2001Conference, Manchester, UK, 2627
121、 September 2001.9 Shishkovsky, I. Synthesis of functionally gradient viaRP methods. Rapid Prototyping J., 2001, 7(4), 207211.10 Steen, W. M. Laser material processing, 1991 (Springer-Verlag, New York).11 Kumar, P., Santosa, J. K., Beck, E., and Das, S. Direct-write deposition of fine powders through
122、 miniaturehopper-nozzles for multi-material solid freeform fabri-cation. Rapid Prototyping J., 2004, 10(1), 1423.12 Yang, S. and Evans, J. R. G. Acoustic control of powderdispensing in open tubes. Powder Technol., 2004,139(6), 5560.13 Ensz, M. T., Griffith, M. L., and Reckaway, D. E. Criticalissues
123、for functionally graded material deposition bylaser engineered net shaping (LENS TM). Proceedingsof the 2002 International Conference on Metal powderdeposition for rapid manufacturing, San Antonio, Texas,USA, 2002, pp. 195202.14 Cho, W., Sachs, E. M., and Patrikalakis, N. M. Localcompositioncontroli
124、nsolidfreeformfabrication.Invited paper in Rapid Prototyping, quarterly of theRapid Prototyping Association of the Society of Manu-facturing Engineers, 7(2), 15.15 Beal,V.E.,Erasenthiran,P.,Hopkinson,N.,Dickens, P., and Ahrens, C. H. Fabrication of x-gradedH13 and Cu powder mix using high power puls
125、edNd:YAG laser. Proceedings of Solid freeform fabricationsymposium, Austin, Texas, USA, 2004.16 Matweb. AISI Type H13 Hot Work Tool Steel, air oroil quenched from 9951025C, available online athttp:/ (accessed February 2005).Evaluating the use of functionally graded materials inserts953JEM764? IMechE
126、 2007Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture17 Matweb. Copper, Cu; annealed, cold drawn and cold-worked, available online at http:/(accessed February 2005).18 Harris,R.A., Fouchal, F.,Hague,R.J.M.,andDickens, P.Quantifying part irregularities and subsequent morpho-logy manipulation
127、in stereolithography plastic injectionmoulding. Plastics Rubbers Composites, 2004, 33(2/3),9298.19 Incropera, F. P. and DeWitt, D. P. Fundamentalsof heat and mass transfer, 3rd edition, 1990 (Wiley,New York).20 Huntsman.RenShape:StereolithographyMaterials:SL 7580, available online at http:/ (accesse
128、d February 2005).21 Matweb. Overview polypropylene, molded, availableonline at http:/ (accessed February2005).22 German, R. M. Powder metallurgy science, 2nd edition,1994, p. 472 (MPIF, New Jersey).23 Osprey Metal Powders. Microfine powder H13 andOHFC Cu datasheets, available online at http:/www.osp
129、reymetals.co.uk (accessed November 2003).24 Potsch, G. and Michaeli, W. Injection moulding: anintroduction,1995(CarlHanserVerlag,Munich,Germany).25 Marinelli, A. L. and Bretas, R. E. S. Blends of poly-propylene resins with a liquid crystalline polymer:I-isothermal crystallization. J. Appl. Polymer S
130、ci., 2003,87(6), 916930.APPENDIXNotationAareacpspecific heatEvariation (balance) of energy_Egengenerated energy in the volume control_Einenergy flux entering the volume control_Eoutlost energy from the volume control?Hheat of fusion for PP sample?H100%heat of fusion for 100 per cent crystalline PPkh
131、eat conduction coefficientq00heat fluxesttimeTtemperatureTtcthermocouple temperatureTisinsert surface temperatureVvolume fraction or volumexaxis of the direction of the heat fluxX%degree of crystallinity?material phase?material phaseequivalent property?density954V E Beal, P Erasenthiran, C H Ahrens, and P DickensProc. IMechE Vol. 221 Part B: J. Engineering ManufactureJEM764? IMechE 2007
