A New Criterion for Prediction of Hot Tearing Susceptibility of Cast Alloys

A new criterion for prediction of hot tearing susceptibility of cast alloys is suggested which takes into account the effects of both important mechanical and metallurgical factors and is believed to be less sensitive to the presence of volume
of 4
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
  Communication A New Criterion for Prediction of HotTearing Susceptibility of Cast Alloys MOHAMAD REZA NASRESFAHANI andBEHZAD NIROUMANDA new criterion for prediction of hot tearing suscepti-bility of cast alloys is suggested which takes into accountthe effects of both important mechanical and metallur-gical factors and is believed to be less sensitive to thepresence of volume defects such as bifilms and inclu-sions. The criterion was validated by studying the hottearing tendency of Al-Cu alloy. In conformity with theexperimental results, the new criterion predicted reduc-tion of hot tearing tendency with increasing the coppercontent.DOI: 10.1007/s11661-014-2341-y   The Minerals, Metals & Materials Society and ASMInternational 2014Hot tearing is a common defect in cast alloys,especially in those with wide freezing ranges. [1 – 3] Mostcommonly used methods to study the hot tearing of castalloys have been traditionally based on visual observa-tion of the castings and measuring the dimensions of the tears incurred. [4 – 6] However, due to the possible errors indetection and measurement of the tears, as well as theirinability to distinguish the influence of different param-eters on the occurrence of hot tearing, they are not veryaccurate and reliable.Newer methods have been developed that use loadmeasurement equipments, ultrasonic waves, or X-raymicrotomographic observations to evaluate the hottearing tendency. [7 – 10] However, most of these methodsare expensive and their output data are complex.One of the established hot tearing criteria is that of Feurer. [11,12] Feurer considers two terms,  i.e. , SRG andSPV, which are abbreviations for  Schrumpfungs Ges-chwindigkeit  (solidification rate) and  Speisungsvermo ¨  gen (supply assets) in German, respectively. SRG and SPVfunctions, shown in Eqs. [1] and [2], denote the volumetric rate of solidification shrinkage and themaximum rate of volumetric feeding through a dendriticnetwork for a given mushy zone structure and hydro-static pressure, respectivelySRG  ¼  1 V  d V  d t  ¼  1 q d q d t  ½ 1  SPV  ¼  f  2L k 22  p s 24 p c 3 g L 2  ;  ½ 2  where d V   and d q  are the changes in volume and den-sity of the alloy by solidification, and  f  L ,  k 2  (m),  P s (Pa),  L  (m),  c , and  g  (Pa s) are liquid volume fraction,secondary dendrite arm spacing (or grain size), effec-tive feeding pressure, length of porous network, tortu-osity constant of the dendritic network, and viscosityof the liquid phase, respectively.According to this criterion, hot tearing occurs whentransfer of the molten metal between the dendritenetworks is unable to balance the volume reductiondue to solidification shrinkage. In other words, hottearing becomes likely if SRG > SPV. Feurer’s criterionuses only metallurgical parameters of the structuralelements in the mushy zone to predict the hot tearing. Itdisregards the contraction-induced tensile stresses devel-oped by the elements both inside and outside the mushyzone. These stresses are the main cause of hot tearingand significantly increase its occurrence.Some researchers have measured the contractionloads developed during solidification and tried toanalyze the features of the recorded load– time graphs to predict the occurrence of hot tearing. [8,9,13 – 15] How-ever, these methods also suffer from not considering themetallurgical parameters affecting the hot tearingoccurrence.Another well-known problem in hot tearing is therandom nature of its tests results. The scatter in theresults has been associated with the random presence of defects such as bifilms and inclusions in the hot spots of castings. [16] Provided the right conditions, these defectscan act as excellent initiation sites for subsequentgrowth of hot tears under contraction-induced tensilestresses. [16,17] The purpose of this study is to propose a hot tearingcriterion which considers both mechanical and metal-lurgical conditions of the castings during solidificationand is independent of the presence or absence of hot tearencouraging factors such as bifilms and inclusions.The proposed criterion for determination of hottearing tendency of cast alloys,  i.e. , NNC (short forNasresfahani and Niroumand’s criterion), complementsthe Instrumented Constrained T-shaped Casting (ICTC)hot tearing test developed and evaluated by the authorspreviously. [13 – 15] The schematic of the ICTC apparatusis shown in Figure 1. The apparatus provides thenecessary constraints to contraction by two boltsinserted from two sides of the mold cavity. It alsoprovides a real time load–time ( F  - t ) graph duringsolidification of the casting (Figure 2). The tempera-ture–time ( T  - t ) graphs are recorded by two K-typethermocouples positioned 10 mm apart in the T-junc-tion (hot spot) of the casting. [13,14] In order to validate the proposed criterion, the effectof the copper content (in the range of 4.4 to 5.1 wt pct) MOHAMAD REZA NASRESFAHANI, Ph.D. Student, andBEHZAD NIROUMAND, Associate Professor, are with the Depart-ment of Materials Engineering, Isfahan University of Technology,84156-83111 Isfahan, Iran. Contact e-mail: submitted October 25, 2013.Article published online May 21, 2014 METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 45A, AUGUST 2014—3699  on hot tearing tendency of Al-Cu alloys was evaluated.In each experiment, Al-Cu alloy was poured into the T-shape mold at pouring temperature of 973 K (700   C).Load and temperature recordings continued until thehot spot temperature reached to about 523 K (250   C).After each test, hot tearing severity at T-junction wasvisually investigated. Grain size measurements of the T- junctions of castings were consequently carried outaccording to standard metallographic techniques [18,19] using CLEMEX image analysis software.The new criterion considers two effects as the maincauses of hot tearing. The higher rate of development of contraction loads during solidification would result inmore rapid strain concentration in the hot spots of thecasting and earlier opening up of tears. On the otherhand, low flowability of the melt in the mushy zone doesnot allow for healing of the tears by displacement of theremaining liquid. Therefore, NNC is proposed as Eq.[3]. It consists of two terms,  i.e. , modified inverse of SPVpart of Feurer’s criterion (as a measure of melt flowdifficulties) and the rate of increase in the contractionload developed during solidification. Parameters of thefirst term of Eq. [3] are computed by Eqs. [4] to [7], where  c sl  (N m  1 ) is the solid–liquid interface energy,  g (m s  2 ) is the gravity constant,  h  (m) is the distance tothe melt surface, and  f  s  is the solid fraction.  P 0 ,  P M , and P C  (Pa) are atmospheric, metallostatic, and capillarypressures, and  q L  and  q s  (kg m  3 ) are the densities of theliquid and solid, respectively [11,12] NNC  ¼  g L 2  f  2L k 22 P S   r min t min    t 0 ½ 3  P S  ¼  P 0  þ  P M  þ  P C  ½ 4  P M  ¼  q  gh  ½ 5  q  ¼  q S  f  S  þ q L  f  L  ½ 6  P C  ¼  4 c sl = k 2  ½ 7  r min  (Pa),  t min  (s), and  t 0  (s) in the second term of theequation are extracted from ICTC tests and are theminimum recorded stress in the negative (pressure) sec-tion of the tensile load–time curve, the time corre-sponding to  r min , and the time when the negative loadreaches to zero, respectively.  r min  is obtained from Eq.[8], where  F  min  ( N  ) is the minimum recorded load and A  (m 2 ) is the cross section area of the casting r min  ¼  F  min = A :  ½ 8  Figure 2 shows an example of the load–time graphalong with its first derivative for an Al-4.6 pctCu alloy.The coherency and solidus temperatures of the castinghave been detected by thermal analysis of the readings Fig. 1—Schematics of the instrumented constrained T-shaped casting(ICTC) apparatus used for the hot tearing test. [11] Fig. 2—Load–time graph and its derivative for Al-4.6 pct Cu alloy cast at 973 K (700   C). [11] 3700—VOLUME 45A, AUGUST 2014 METALLURGICAL AND MATERIALS TRANSACTIONS A  of the thermocouples positioned in the hot spot of thecastings. [20,21] It should be noted that shortly after the pouring, theload cell (and the casting) starts to experience somepressure due to the initial expansion of the mold andother equipments. Therefore, for a certain period of time, the recorded load shows negative values. However,solidification contraction of the casting gradually over-comes the initial expansion and as a result the graphpasses from a minimum at certain coordinate ( F  min ,  t min )and then moves toward the positive (tension) part (inlayin Figure 2).In conventional hot tearing tests, hot tearing suscep-tibility is measure under contraction-induced tensilestresses. Therefore, the results are very sensitive to thepresence of volume defects such as bifilms, inclusions,and porosity. A significant point of NNC is the secondterm of Eq. [1]. This term is the slope of the stress–timegraph (stress rate) between the lowest point and wherethe curve intercepts the horizontal coordinate axis. Infact, the proposed criterion examines the rate of development of tensile stresses in the casting when thecasting is still under compression (between  t min  and  t 0 ).Since during this period, the casting is not under tension,the occurrence of hot tearing is very unlikely. In otherwords, NNC studies the susceptibility to hot tearing notthe actual occurrence of hot tearing. Therefore, NNC isbelieved to be less sensitive to the random presence of volume defects in the hot spot and primarily influencedby the intrinsic alloy characteristics.At times longer than  t 0 , the load experienced by thecasting becomes tensional and any hot tearing that mayoccur changes the slope of the curve. Therefore, usingthe slop of the curve at times longer than  t 0  may causewrong estimation of the hot tearing tendency.In order to use NNC,  P s  and  L  must be determined. P s  can be readily obtained from Eq. [4] using q L  = 2480 kg m  3 ,  g  = 0.0013 Pa s, and  c sl  = 0.84N m  1[11,12] and the real measured density of the samples. L  is the distance between the positions of coherencyand solidus temperature isotherms. In other words,  L  isthe average distance between the points in the mushyzone, where a coherent dendrite network has beenformed and the points of the casting where the last dropof molten metal is being solidified (Figure 3). Thefollowing procedure was adapted to estimate  L  usingEqs. [9] to [11] T   ¼  G    R  ½ 9  R  ¼  D h th = D t sol  ½ 10  G  ¼ ð T  coh    T  sol Þ = L ;  ½ 11  where  T    (K s  1 ) is the cooling rate between the coher-ency and solidus temperatures and is extracted fromthe cooling curves.  R  (m s  1 ) is the rate of advance-ment of the solidification front.  D h th  is the distance be-tween the two thermocouples positioned in the hotspot (10 mm).  D t sol  (s) is the time it takes for the solid-ification front to move from one thermocouple to theother thermocouple which is derived from the corre-sponding cooling curves of the two thermocouples. T  coh  and  T  sol  (K) are coherency and solidus tempera-tures, respectively, which are also derived from thecooling curves.  G  (K m  1 ) is the temperature gradientalong  L. G  is obtained from Eq. [9] and inserted inEq. [11] to yield  L .The metallurgical and mechanical parameters for eachalloy and the corresponding values for NNC calculatedbased on the described procedure are presented inTable I. It is evident that according to NNC, themaximum hot tearing tendency is predicted to occur at4.4 pct Cu and the hot tearing tendency is expected todecrease with increasing the copper content. Thisconforms to the results of visual observation of the castings [13, 14] as well as the findings of other researcherswho have shown that the maximum hot tearing ten-dency for Al-Cu-Mg alloys occurs at about 0.8 pctcopper and the hot tearing tendency decreases withincreasing the copper content [for example Ref  1, 22]. It is thought that copper has a complex effect. On theone hand, it changes the volume fraction of the eutecticmelt. On the other hand, it affects the growth rate of thedendrites, the grain size, the heat transfer, and thecooling rate of the alloy. Therefore, as the resultspresented in Table I suggests, it may have a substantialnon-linear effect. Fig. 3—Schematic definition of   L . Table I. Calculated or Measured Parameters for Each Alloy and the Corresponding Values of NNC Alloy Stress Rate (N m  2 s)  L  (cm)  k 2  ( l m)  f  L  P s  (MPa) NNC (kPa)Al-4.4 pct Cu 1379 3.12 118 0.01 129,838 9.676Al-4.6 pct Cu 1199 2.59 154 0.01 123,955 3.619Al-5.1 pct Cu 1527 1.48 120 0.01 130,168 2.154 METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 45A, AUGUST 2014—3701  It is interesting, however, to note how the suggestedNNC simultaneously considers the effects of mechani-cal, thermal, and microstructural factors on the hottearing tendency. For example, it is believed that for5.1 pct Cu alloy the effects of finer grain ( k 2 ) and smallermushy zone ( L ) have neutralized the effect of higherstress rate which otherwise would have resulted in higherhot tearing tendency. Also for the casting with thelowest copper percent, although its stress rate is betweenthose of the other two alloys, larger mushy zone hasresulted in higher hot tearing tendency.For further validating the criterion, predictions of three criteria,  i.e. , NNC, loading rate, and Clyne andDavis [23] on the effect of super heat, copper content,grain refinements, and degassing on hot tearing suscep-tibility were compared. [24] These comparisons alsoshowed that NNC had a good agreement with othercriteria.In conclusion, it is believed that if the requiredmechanical and thermal parameters are correctly deter-mined through accurately performed ICTC hot tearingtests, NNC can be used as a good hot tearing indicator.This criterion is less sensitive to random volume defectsin the melt and can determine the hot tearing tendencyof cast alloys even when the hot tearing does not occur. REFERENCES 1. Z. Wang, Y. Huang, A. Srinivasan, Z. Liu, F. Beckermann, K.U.Kainer, and N. Hort:  Mater. Des. , 2013, vol. 47, pp. 90–100.2. J.B. Mitchell, S.L. Cockcroft, D. Viano, C. Davidson, and D.Stjohn:  Metall. Mater. Trans. A , 2007, vol. 38A, pp. 2503–12.3. D.G. Eskin:  Physical Metallurgy of Direct Chill Casting of Alu-minum Alloys , 1st ed., Taylor & Francis, Boca Raton, 2008, p. 45.4. P.R. Beeley:  Foundry Technology , 2nd ed., Butterworth-Heine-mann, Oxford, 2001, p. 281.5. J. Zhang:  Scripta Mater. , 2003, vol. 48, p. 677.6. H. Kamguo Kamga, D. Larouche, M. Bournane, and A. Rahem: Mater. Sci. Eng. A , 2010, vol. 52, pp. 7413–23.7. S. Terzi, L. Salvo, M. Sue ´ry, N. Limodin, J. Adrien, E. Maire, Y.Pannier, M. Bornert, D. Bernard, M. Felberbaum, M. Rappaz,and E. Boller:  Scripta Mater. , 2009, vol. 61, pp. 449–52.8. M.G. Pokorny, C.A. Monroe, C. Beckermann, Z. Zhen, and N.Hort:  Metall. Mater. Trans. A , 2010, vol. 41A, pp. 3196–3207.9. G. Cao and S. Kou:  Proc 111th Metal Casting Cong. , 2007,pp. 7–34.10. A.B. Phillion, S.L. Cockcroft, and P.D. Lee:  Scripta Mater. , 2006,vol. 55, pp. 489–92.11. U. Feurer: Report of Delft University of Technology, 1977.12. U. Feurer:  Gießerei-Forschung , 1976, vol. 28, pp. 75–80.13. M.R. Nasresfahani and B. Niroumand:  J. Mater. Charact. , 2010,vol. 61, pp. 318–24.14. M.R. Nasresfahani and B. Niroumand:  Proceedings of 2nd Inter-national Conference on Aluminium Casting , Moscow, 2009,pp. 26–30.15. M.R. Nasresfahani and B. Niroumand:  Metall. Mater. Int. , 2010,vol. 16, pp. 35–38.16. J. Campbell:  Complete Casting Handbook , 1st ed., Butterworth-Heinemann, Oxford, 2011, p. 465.17. J. Campbell and M. Tiryaioglu:  Metall. Mater. Trans. B , 2012, vol.43B, pp. 902–14.18. D.B.A. Granger:  ASM Metals Handbook , ASM, Metals Park,1999.19. ASTM Committee:  Standard Test Methods for DeterminingAverage Grain Size , ASTM, 1999.20. D. Emadi, L.V. Whiting, S. Nafisi, and R. Ghomashchi:  J. Therm.Anal. Calorim. , 2005, vol. 81, pp. 235–42.21. M.B. Djurdjevic, J.H. Sokolowski, and Z. Odanovic:  J. Therm.Anal. Calorim. , 2012, vol. 109, pp. 875–82.22. Y. Xinyan and L. Jenc:  Metall. Mater. Trans. B , 2006, vol. 37B,pp. 913–18.23. T.W. Clyne and G.J. Davies:  Proceedings of the Conference onSolidification and Casting of Metals , Metals Society,London, 1979.24. M.R. Nasresfahani and B. Niroumand: Isfahan University of Technology, Unpublished research, 2013. 3702—VOLUME 45A, AUGUST 2014 METALLURGICAL AND MATERIALS TRANSACTIONS A
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