Method to Determine Hot Permeability and Strength of Ceramic Shell Moulds

of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
   Journal of Materials Processing Technology 211 (2011) 1336–1340 Contents lists available at ScienceDirect  JournalofMaterialsProcessingTechnology  journal homepage: Method to determine hot permeability and strength of ceramic shell moulds S. Amira a , D. Dubé b , ∗ , R. Tremblay b a  Aluminum Technology Centre, NRC-ATC, Saguenay, QC, Canada G7H 8C3 b Department of Mining, Metallurgical and Materials Engineering, Laval University, Quebec, QC, Canada G1V 0A6 a r t i c l e i n f o  Article history: Received 23 November 2010Received in revised form 1 March 2011Accepted 3 March 2011 Available online 10 March 2011 Keywords: Hot permeabilityHot strengthCeramic shellInvestment molding a b s t r a c t Thispaperreportsonanimprovedmethodtoevaluateboththestrengthandthepermeabilityofceramicshell specimens under high temperature conditions. In order to maintain safe testing conditions anduse lower testing pressure, a spherical wax model, larger than the standard ping-pong ball, was usedto prepare ceramic shells. Compressed air was introduced into ceramic shell specimen held at 900 ◦ C.Air flow and pressure drop across the shell wall were measured and hot permeability was calculatedaccording to Darcy’s law. Air pressure was subsequently increased and recorded up to bursting point. Ahoop stress formula was used to calculate the hot strength from the bursting pressure. This very simplemethod is easy to implement in foundries. © 2011 Elsevier B.V. All rights reserved. 1. Introduction Hot permeability and hot strength, two properties closelyrelated to the performance of in-service investment casting shells,areinfacttheirmostreliableperformanceindicators(Mills,1995).Hot strength refers to the mould capacity to resist metallostaticpressure and preserve the dimensional integrity of castings, whilehot permeability is required to promptly evacuate air and otherentrappedgasesduringmetalfilling.Thepermeabilityandstrengthofshellmaterialsaremostoftendeterminedindependentlyatroomtemperature using two different set-ups; however, extrapolationof their properties is unreliable and characterization at room tem-perature does not illustrate shell mould behavior at high servicetemperatures.Hot permeability can be evaluated by a method well describedby Lang et al. (1988) and recommended by the Investment Casting Institute (1979) using a ceramic shell built on the ping-pong ballmodel.Afterdryingandfiring,compressedairisintroducedintotheshell test specimen under high temperature. Pressure drop evolu-tion with flow across the shell walls subsequently leads either to apermeability factor or to absolute permeability using Darcy’s law(InvestmentCastingInstitute,1979).Permeabilityindex(similarto the AFS number) may also be obtained from the seconds requiredfor 1000ml of air compressed at 980Pa to flow through this shell(Beeley, 1972).Hot strength is usually determined by the four-point bendingtest which requires a tensile machine equipped with compo- ∗ Corresponding author. Tel.: +1 418 656 3533; fax: +1 418 656 5343. E-mail address: (D. Dubé). nents that are capable of withstanding high temperatures. Propertest bars alignment, regular surface finish, and constant thick-ness are crucial for sound results. This testing method is oftendifficult to conduct, as it is strongly influenced by experimentalconditions and is rather inaccessible for small- and medium-sizedfoundries.Shell materials strength has also been evaluated using vari-ous types of burst tests. Babu et al. (1990) have used sphericalshells pressurized with compressed air to test overall permeabil-ityatroomtemperatureandcomparedbendingstrengthandhoopstrength.ChennakesavaReddyetal.(2000)usedbursttestingwith pressurized air to study the permeability and strength of spheri-cal shells at room temperature and found that the burst test wassuperior to the bend test in simulating the casting conditions andfailureratesobservedininvestmentcasting.Inamorerecentstudy,the bursting resistance of square and cylindrical prismatic shellshas also been tested using pressurized water (Kline et al., 2009).The authors’ predictions with regard to the finite element methodoverestimated the resistance of the shells because of the presenceof natural defects that were difficult to anticipate.The burst test was also found to better simulate the propertiesof shells used in investment casting (Babu et al., 1990), contrary to the bend test which determines the strength using a relativelylimitedvolumeofmaterial,lesslikelytocontaindefectscapableof rupturing during tests. These authors also found burst strength tobe significantly more sensitive to defects present in ceramic shellscompared to bending strength, although no test was performed tofind a more representative strength at high temperature.The present study therefore explored an innovative method torapidly assess the hot permeability and hot strength of ceramicshells using the same test specimen. Of interest is that this novel 0924-0136/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jmatprotec.2011.03.002  S. Amira et al. / Journal of Materials Processing Technology  211 (2011) 1336–1340 1337 Fig. 1.  Spherical aluminum mould and hollow wax ball used as model (a); wax ball and Vycor glass tube assembly (b); and shell mould test specimens after dipping andstuccoing (c). approach requires no expensive equipment and can thus be easilyimplemented in small foundries. 2. Experimental  2.1. Materials A colloidal silica binder was used to prepare the shell speci-mens. REMASOL  ® SP-30 was used as a binder for both the primaryand back-up slurries. A single primary zircon layer was applied.REMET ® milledzirconGF( − 325mesh)wasusedasthefillermate-rial in the slurry. REMET ® Zircon Sand A (+70–200mesh) was usedas primary stuccoing sand. For the back-up layers, aluminosilicateREMASIL  ® 48RP325(48%Al 2 O 3 ,52%SiO 2 )andREMASIL  ® 60RP325(60% Al 2 O 3 , 40% SiO 2 ) ( − 325 mesh) were used as filler materialsin the slurry. Aluminosilicate REMASIL  ® 48 RG30 (+20–40 mesh)wasusedasback-upstuccoingsand.Dryingtookplaceforapprox-imately 3h between each coat and a total of three layers werenecessary for each shell.  2.2. Shell specimen preparation Hollow wax balls (outer diameter 80mm) were used as themodel for shell mould test specimens. Each wax ball was preparedmanuallybypouringmoltenwaxintothesphericalcavityofanalu-minummould(Fig.1a).Themouldwasrotatedtoevenlyspreadthe liquid wax on the walls of the mould. Once solid, the hollow waxball was extracted and connected to the end of a Vycor glass tubewithstickywax(Fig.1b).Theendofthewaxball-Vycortubeassem- bly was first dipped in the zircon slurry and stuccoed with zirconsandtoformtheprimarylayer.Theshellmouldwasthendippedinthe aluminosilicate slurry and stuccoed with aluminosilicate sandtoproducethesecondarylayer.Finally,theshelltestspecimenwasdipped in the aluminosilicate slurry without stuccoing to form thesealing layer (Fig. 1c). Following a final drying period, the wax ball was removed by flash firing which produced a thin ceramic shellwith thickness of approximately 2mm ( ± 0.2mm).  2.3. Hot permeability and hot strength measurements Fig. 2 shows the device used to measure the hot permeabilityand hot strength of the shell test specimen. Valves a–d were usedtoputthedeviceineitherthehotpermeabilitymeasurementmode(Fig. 2a), or the hot strength measurement mode (Fig. 2b).  2.4. Hot permeability measurement  The hot permeability measurement was performed prior to thehot strength measurement. In hot permeability mode (Fig. 2a), the shell test specimen (5) was placed inside a horizontal tubu-lar furnace with an inner diameter of 100mm and temperaturewas gradually increased. The shell was held at 900 ◦ C for 30minto obtain an even temperature distribution. The Vycor glass tubewas then connected to the testing device, as shown in Fig. 2a, and compressedairwasallowedtoflowintotheshell.Theairflowrate( F  ) was varied using the appropriate flowmeter (2) and the cor-responding pressure drops (  P  = P  1 − P  0 ) were measured using a Fig.2.  Set-upsusedforthemeasurementof(a)hotpermeabilityand(b)hotstrengthofshelltestspecimen.(1)pressureregulator;(2)flowmeters;(3)watermanometer;(4) tube furnace; (5) shell test specimen; (6) pressure sensor; (7) data acquisitioncard. (a–d represent the valves).  1338  S. Amira et al. / Journal of Materials Processing Technology  211 (2011) 1336–1340 water manometer (3). Darcy’s law was used to calculate the hotpermeability of shell according to the flow rates values and corre-sponding pressure drop. An example of permeability calculation isgiven further.  2.5. Hot strength measurement  Thehotstrengthoftheshelltestspecimen(5)wasdeterminedat900 ◦ Cusingcompressedairappliedintothetestspecimen(Fig.2b). Air pressure was then gradually increased at a rate of close to1.4 × 10 − 3 MPas − 1 bymeansofapressureregulator(1)untilburst-ingoftheshellwasachieved.Thecorrespondingburstingpressure( P  b ) was then measured with the pressure sensor (6) connected toa data acquisition card (7). Pressure variation over time within theshell specimen was recorded by a computer. The data acquisitionratewas10Hz.Safetyprecautionswereobservedduringthetesttoprotect both the operators and the equipment from debris thrownoutbytheburstingshells.Forthispurpose,asteelgridcoveredthetube furnace. 3. Results and discussion  3.1. Hot permeability determination Hotpermeability k isdeterminedbyDarcy’slaw(initsdifferen-tial form) as applied to flows through spherical shells: F  4 r  2  =− k d P  d r   (1)where    is the dynamic viscosity of the flowing fluid(4.78 × 10 − 5 Nsm − 2 for air at 900 ◦ C),  k  is the permeability (m 2 ),and  F   is the volume flow rate (m 3 s − 1 ). Under these conditions, airis compressible and considered an ideal gas: k d P  d r   =− ˚ mass 4 r  2   =− ˚ mass 4 r  2 ( M  mole P/RT  ) (2)where  ˚ mass  is the mass flow rate,    is the hot air density,  M  mole is the molar mass of air,  T   is the temperature ( K  ) and  R  the gasconstant.Rearrangement of Eq. (2) thus led to: ˚ mass 4 r  2  d r   =− kM  mole RT  P  d P   (3)Integration between  r  o  and  r  i  for the first term (outer and innerradius of the shell specimen, respectively), and between  P  1  and  P  0 for the second term (inner and outer pressures of shell test speci-men, respectively) led to the following equation: ˚ mass 4    1 r  1 − 1 r  0  =− kM  mole 2 RT   ( P  12 − P  02 ) (4)where, after rearrangement, Eq. (4) becomes: k  ˚ mass   ( r  1 − r  0 )4 r  0 r  1  = ( P  )( P  1 + P  0 )2 P   (5)where  P   represents the pressure prevailing at the location of flowmeasurement.Assumingthatforsmallpressuredropthroughshellwall,  (P  1 + P  0 )/2 P  ≈ 1, the following approximation was deemedvalid: P  1 − P  0  = P   ∼= k ( r  o − r  i )4 r  o r  i F   (6)The method used to calculate the hot permeability for a setof   n -tested shell specimens is presented hereafter. As previouslydescribed, the hot permeability assessment was based on therecording of air flow  F   evolution with the corresponding pressuredrop   P   through the shell specimen walls. Fig. 3 shows data fitted Fig.3.  Pressuredropvariation(  P  )withtheflowrate( F  )throughthewallsofthreeshellspecimenstestedunderidenticalconditions.Linearregressionwasdeterminedusing the least squares method. by a linear regression in the case of three shell specimens tested at900 ◦ C.The slope  s  of the linear regression curve, as determined byleast squares method, was the proportionality coefficient betweenpressure drop   P   and volume flow rate  F  , giving: s = k ( r  o − r  i )4 r  o r  i (7)and hot permeability was expressed as: k = s ( r  o − r  i )4 r  o r  i (8)From this example (Fig. 3), the calculated hot permeability was k =3.4 × 10 − 15 m 2 witharelativeerrorofapproximately10%,whichwas mainly attributed to uncertainty as to shell thickness.  3.2. Hot strength determination Assuming that the shell specimen was a thin-walled sphericalshell, the hot strength was equivalent to the hoop stress at burst-ing point. The formula used to determine the hot strength wastherefore:    = P  b r  i 2 t   (9)where  P  b  is the bursting pressure (MPa),  d  is the internal radius of the spherical shell specimen (m) and  t   is its thickness (m).Usingtheburstingpressure P  b  expressedinFig.4andEq.(9),the hot strength of the shell was estimated to be 2.1MPa, with a rela-tiveerrorofapproximately10%whichwasessentiallyattributedtouncertaintyastoshellthicknessandvariationsofburstingpressure. Fig. 4.  Air pressure variation inside the shell test specimen and determination of the bursting pressure ( P  b ) at  T  =900 ◦ C.  S. Amira et al. / Journal of Materials Processing Technology  211 (2011) 1336–1340 1339  Table 1 Comparison of various ceramic shell permeability and strength determination studies.Method Test specimens Testing temp ReferencesModel Shape/sizePermeability (absolute) Ping-pong ball (6–11 layers) Spherical (38mm dia. × 5–6mm)High temp. Lang et al. (1988)Bend test (mod. of rupture) Wax strips (N/A) Rectangular (dimensions notgiven)Permeability (index) Tennis ball (6 layers) Spherical (75mm dia.)Room temp. Babu et al. (1990)Burst test (hoop strength)Bend test (mod. of rupture) Perspex strips (6 layers) Rectangular(25mm × 30mm × 4mm)Permeability (index) Ping-pong ball (6layers)Spherical (38mmdia. × 5mm) Room temp. Chennakesava Reddyet al. (2000)Burst test (hoop strength)Bend test (mod. of rupture) Wax strips (6 layers) Rectangular(25mm × 32mm × 5mm)Permeability (AFS index and Darcy) Foam (5–8 layers) Cylindrical (30.5mmdia. × 76.2mm long) (internaldimensions)Room temp.Kline et al. (2009)Burst test (with pressurized water) Square prisms(30.5mm × 10.2mm × 127mm)(internal dimensions)Bend test (mod. of rupture) Foam strips (N/A) Rectangular (25.4mm widespecimens)Room/high temp.Permeability (absolute) Hollow wax ball (3layers)Spherical (80mmdia. × 2mm) High temp. Present workBurst test (hoop strength)  3.3. Discussion The sequential of hot permeability and hot strength measure-ments at high temperature on the same specimen were simple toexecute, thereby enabling to easily replace the four-point bendtest with a burst test. Compared to the conventional hot per-meability test described by Lang et al. (1988) and recommendedby the Investment Casting Institute (1979), the advantage of a larger wax model lies in the greater ceramic shell surface exposedto the conditions encountered in investment mould castingpreparation.Another benefit of using a burst test is that the defects gen-erated within the ceramic shell during the manufacturing process(dryingandde-waxingsteps),moreresemblethoseencounteredinfoundryoperation.Inaddition,thebursttestrequiresnomachiningto obtain a smooth surface finish which is mandatory in the four-pointbendtestmethod.Besides,comparedtofour-pointbendtest,tensilestressesinthesphericalshellaremoreuniformlydistributedover the shell volume during testing.Machined specimens used in the four-point bend test are sub- ject to both tensile and compression stresses which is far from thestress distribution in service conditions and during the manufac-turing process. Furthermore, this test may lead to errors causedby unequal moments at the inner loading points, twisting result-ing from skewed contact lines, wedging stresses at the contactpoints, and counter moments produced by friction at the loadingpoints/specimen interface (Hoagland et al., 1976). Calculating the strength from the maximal bending torque located in the mid-dle of the sample may also lead to an over-estimation of thestrength. Indeed, in the case of brittle materials, the failure maynot necessarily be initiated at the point of the highest nominalstress, but rather at the greatest surface or volume critical flaw(Quinn, 1991).Moreimportant,usingthefour-pointbendtesttodeterminehotstrength is not only technically difficult but also requires sophisti-catedequipments,whichmaylimititsaccessformanyinvestmentcasting professionals. It is interesting to note that the four-pointbendtest,asdescribedbytheInvestmentCastingInstitute,pertainsto green or fired samples at room temperature; high temperaturestests, on the other hand, are never mentioned (Quinn, 1991).The burst test enables operators to avoid all of the above-mentioned drawbacks (unequal moments, twisting, wedging, andfriction), as there is no direct contact between the shell specimenand a loading device (typically knife edges or cylindrical rollersin the case of four-point bend test). In contrast, during the bursttest, stress concentration sources are minimized because (1) thepressure is applied uniformly over the inner surface of the shellspherical specimen, (2) the shell specimen has no edges, and (3)the thickness of the shell is relatively uniform.Contrary to the four-point bend test, the burst test at high tem-perature requires additional safety precautions. In this study, thethickness of the tested shell specimens was limited to approx-imately 3mm so as to safely maintain the maximum burstingpressure ( P  b ) below 0.35MPa. The heating elements of the fur-nace were also covered with a ceramic shield to protect againstthe bursting fragments. In this regard, a strong wire mesh enclo-surewasplacedinfrontofthefurnaceduringtestingtocollectanyfragment while allowing for the evacuation of the compressed airreleasedduringtheburstingprocess.Aventilationsystemwasalsoemployed to collect the dust.MostoftheleadingstudiesonpermeabilityandstrengthtestingofceramicshellsusedininvestmentcastingarelistedinTable1and represent the wide variety of methods and test specimens used indifferentlaboratories,thusmakingitdifficulttopresentanaccuratecomparison of shell performances. 4. Conclusions ã  Themethodproposedinthisstudyallowsforthesequentialeval-uation of both the hot permeability and hot strength of ceramicshells used for investment casting. ã  Hot permeability and hot strength can be determined using thesame device and with the same specimen, contrary to conven-tional methods where permeability and strength are assessedusingdifferentspecimenswithvariousmodels,shapesandsizes.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks