—-’- l I I Vii M" MW! HI 2 HI MI. W“ N M! f ,l H! .Hmd ION I moo—s SOLUTION RARE. 34 AN Mummy,“ 30-95%! ALLOY 02’ 1421334 MREW Week in: i313 Dag-2‘9: 3'} M. 3. MéCHifiéN sz‘fia CCLEGB " '1 t. :5 J" = 4. t'i ‘ 1" H “ ‘4. \r a ;-,;:.zszt~: Sammie :3 as: we: *7?! 3% .kaasus, This is to certify that the thesis entitled ‘\ : SOLUTION RATES IN AN ALUMINUM t COPPER ALLOY OF HIGH PURITY . presented by " William Edmonds Pearson L has been accepted towards fulfillment of the requirements for M. S. degree in Metallurgical Engineering _—7~——- )d/ professor 1'" 'T—."' SOLUTION RATES IN IN ALUMINUM COPPERMLLLOY'GI HIGH PUBITY IILLIAHLEDMONDS PEARSON 1.!EESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIMCE Department of Metallurgical Engineering. 1950 ACKNOWLEDGMENT The author wishes to express his appreciation to Dr. Austen J; Smith for his guidance, assistance and advice throughout this investigation. He also wishes to thank the other members or the faculty or the Metallurgical Engineering Department for their aid in carrying out this work. at go J 5.9-- .9 11 ‘0‘ V‘! Acknouledgement . Introduction . . Hunter: . . . Procedure . . e Tabulated Data . . Photomicrographs . Dilouseion of Results Conclusions . . Selected Bibliography Supplemental Bibliography 0 8o 4. 5o 9. 15. 26o 32. 42. 44. 48. INTRODUCTION Age hardening alloys of aluminum and other metals are among the most important engineering materials amp ployed in industry and construction today. The solu- tion heat treatment of such alloys has a marked bear- ing on the physical characteristics which may be ob- tained on aging. In order to reap the maximum.benefits of the hardening elements in the alloy, these elements should be in complete solid solution before aging begins. At the present time few data have been published concerning the rate of solution of many of these ele- ments during the solution heat treatment. For this inves- tigation an alloy of aluminum and copper was selected, the objective being to determine the rate of solution or the theta (Calla) phase in the aluminum. HISTORY Aluminumscopper alloys have been the basis for many heat treatable alloys since Alfred lulm.developed 'dural- uminF‘ at Durener'uetallserke, Duran, Germany, betseen 1903 and 1911.(5l Campbell did much of the early work on the aluminum, copper equilibriun.diagram.in the years 1903-1904.u'8 f Twenty years later-the limits of the solid solution field at the aluminumrrich end of the diagram.had still not been definitely determined and workers were questioning another Calla sas actually a compound.(°) In 1926 Di: and Rich- ardson, using very careful methods, finally determined the '7) They found that seven to eleven days at temperature sere diagram shich is accepted today with.miaor changes.( needed to determine points near the lbmits of the solid solution.field mentioned above. ‘Lccording to these inves— tigators, small amounts of iron and silicon caused much difficulty by obscuring the presence or absence of other microecnstitucnts. Stoekdale, using Xkray methods, checked the cork of Dix and Richardson in 1953. His results agreed very close- ly with theirs, but his sork.indicated that Guile could not be considered an actual compound even though the ratio of atoms Ias nearly that of a compound.(9) However, other English workers at this time were still confused mneern- ing the aluminumrrich.end of the diagrams(1°’ The presently accepted aluminum-rich end of the equi- librium diagram.is shown in Figure 1. It is very much the same as that determined by Di: and Richardson. Gayler discovered in same early work on the age hard- ening of aluminnmscopper alloys that prolonged times are required to bring Guile into solution. In one case, after holding a two per cent copper alloy at 9189?. for 1? days, a small amount of Gunlz was still present as a precipitate after quenching.e4’5l In more recent years, Japanese metallurgists have done some investigating of solution rates in aluminumecopper alloys. Hisatsune, in a study of aluminumrrieh aluminum! copper-silicon alloys, concluded that five hours at 9380!. was sufficient to put the soluble constituents into solu~ tion.(11) Obinata and rebate made an Ibray study of solu- tion velocities in duralumtn Du 1940, while looking into the effect of solution.rates on aging;propertiee. They found 80% of Gull: in solution after three minutes at 1004°I. and 90% after ten minutes. Their recommendation was to soak duralumin at least 20 minutes at 9380-95001. for all precipitates to be taken into solution.‘21) In a volume published in.1949, Dix said that solution heat treatment time may vary from.lO-30 minutes for sheet to several hours for massive wrought products.(36l Rhinos has investigated the process of homogenization z in.copper-nickel powder compacts.( 4) He states that the process in an aluminumrcopper alloy follows the Arrhenius equation and that the rate also varies as the square of the average distance between OuAlz particles.(3l) The electrical resistance method used in one part of this investigation has previously been employed hn other cork of a similar nature. Fink and Filey used electrical resistivity and microscopic methods to determine the slump inumpzinc equilibrium diagram. They used a type K poten- tiometer to measure the voltage drop across one by 23 on. strips.(13) The solubility of iron in pure aluminum.has been de- termined by Edgar with electrical resistivity measurements. Ihe samples were held at 86°F. in a thermostatically con- trolled bath.wh11e measurements were being made.(28) Glaser attempted to measure the degree of homogeni- zation of ironpsilicon powder compacts by the electrical resistivity method. He found that resistivity levels off as solution is completed.(27) The solidus of the lead-tin system has been determined by Enltgren and Lever with electrical resistivity measure- men”. (29 i gure l. The aluminum-rich on of tbed'a'luminum-copper equilibrium diagram. 35 PROCEDURE A. Microscopic Investigation. The high purity aluminum-copper alloy used in this investigation was obtained from the research laboratory of the Aluminum.company of America. The alloy was re- ceived in the form of a rod three-eighths of an inch in diameter and five feet long. the chemical analysis received with the rod was as follows: Aluminum 96.095 copper 4.89% Iron 0.017é Silicon 0.01% Several eight inch lengths were cut from the rod and given a 24 hour annealing treatment at lOlo°t. Ithe alloy was cooled in the furnace at the rate of 100%. per hour. a microscopic examination indicated that a longer anneal and a slower cooling rate were necessary for the desired aicrostructure. Therefore, the material was left at tem- perature for 48 hours and slowly cooled in the furnace to 600°I. in 58 hours. the furnace was then turned off and the rods allowed to cool to room temperature. licroscopic examination showed the alloy to be in the desired condition for the following steps in the investigation. 10 .s calibrated Filer eyepiece was employed to determine the average diameter of large, roughly circular particles of Guile in a sample after the second annealing treatment. Samples about one-fourth inch thick were cut from.the annealed lengths of rod. .s.length of rod, several samples and the coil of wire used in the second part of this in- vestigation are shown in Figure 8. Figure 2. Length.of rod, coil and ampla‘ e The faces of these small samples were made plans on a mill bastard file, and ground on #180, 1/0 and 3/0 paper lubricated with a solution of parafin in kerosene. The samples were marked for identification.and wrapped with wire for handling. 11 Solution treatment was carried out in a salt bath in order to insure rapid and uniform.heat flow. .A controller and a check.thermocouple were employed to hold the bath at 1018°r., the eutectic temperature for the aluminumrcuilz system. The maximum temperature variation was less than plus or minus five degrees. Samples were suspended in the salt bath for periods ranging from one minute to 48 hours, and then quickly removed and quenched in water at about eo°r. The arrangement of equipment for solution treatment is illustrated in Figure 3. Figure 3. Solution treatment. The quenched samples having been previously ground, were prepared for microscopic examination after quenching. 12 Heller’s Etch.was used to reveal the microstructure. (The composition of this etch is: 1.0 ml. cone. HP, 1.5 ml. cone. sci, 21.5 ml. cone. Bros and 05.0 ml. ago.) The polished and etched samples were examined at about 1000 diameters. ‘Using the Filer eyepiece some 30 of the larger, roughly circular particles of precipitated Guile were measured on each sample. Measurements of the annealed sample had shown that these larger particles predominated in the frequency distribution. The decrease in size could also be more easily determined in the round precipitates than.in.those of acicular fonm. ,L.srmilar'method involv- ing the use of only the larger particles for measuring purposes was employed in a study of nucleation and growth.(2°) luom.sueh particle size data a curve was plotted to show the rate of solution of Guile in.the aluminum.matrix at the eutectic temperature. Photomicrographs were made of eight representative samples to illustrate the microstruc- tare after varying degrees of solution.had taken place. B. Electrical Resistaneelmeasurements. In a second part of the investigation it was proposed to attempt to measure the change in electrical resistance of the alloy at the eutectic temperature as the Bull: went into solution. In order to do this it was necessary to put the alloy in wire forms A.six inch length of the three- eighthe inch diameter rod was cut lengthwise into quarters. One of the resulting strips, after having its edges ground down sufficiently, was then cold drawn through dies of #l 13 to 40 gauge site. After every few passes through the die block the wire was given a 80 minute stress relieving cancel at about 400°r. 1310 product of this work was a five foot length of cold drawn sire of #40 gauge size. The sire was given an annealing treatment correspond- ing to that given the material used in part A. Next it was wound on a cylindrical ceramic fora about two inches in diameter. The winding necessitated another stress reliev- ing anneal at 4000!. after which the wire was cooled to recs temperature in the furnace. A.Leeds and florthrup type K98 potentiometer was used to measure the voltage drops across the wire at the 101891. eutectic temperature. A.sensitive mirror type gelvenometer employing a moving light beam.snd glass scale provided an accurate indication of the null point. the electrical circuit consisted of a six volt auto- mobile battery, the sample, a standard one ohs resistance and a rheostat connected in series. Potential leads sere connected from.the sample and the standard resistance to a selector switch so that the voltage drop across either could be quickly measured. the rheostat ass adjusted to lhsit the current to such a value that the drop across the standard resistance see about 204 :10“6 volts. A.photograph of the arrangement for this part of the investigation is shown in Figure 4. 'Ihen the current in the circuit had been adjusted as described above, the potentiometer ass zeroed against a 14 standard cell. The sample was then placed in an electric muffle furnace at the eutectic temperature. The tempera- ture of the furnace was held constant by a controller and also checked vith another thermocouple connected to a portable indicator. Figure 4. Electrical resistance set-up. veltage drop measurements were made across the sample and the standard resistance at intervals after the sample had been placed in.the furnace, The potentiometer was bel- anced against the standard cell at frequent intervals dur— ing this time. The voltage drop readings were plotted against time to give a second indication of the rate of solution of the Guile precipitate in the aluminum. 15 TABULATED DATA Table 1. Particle size data for alloy after annealing 48 hours at lOOO°F. and slow cooling to room temperature. Diameters of roughly circular particles of Guile in Filer eyepiece micrometer units. 88 93 92 102 79 100 106 76 78 126 144 79 76 82 102 82 104 102 88 124 85 106 83 79 118 78 82 113 95 100 88 86 94 100 77 108 87 90 101 85 Average diameter 3 92.5 units or 92.5 1:8.00 x 10-5 millimeters/Filer eyepiece unit: 0.0074lmm. Table 2. Particle size data for annealed alloy after one minute of solution treatment at 1018°F. Diameters of roughly circular particles of Guile in Filer eyepiece micrometer units. 88 83 71 83 90 91 101 87 91 78 81 93 80 88 85 108 80 91 94 94 77 81 88 94 100 75 101 81 95 88 Average diameter : 87.9 units or 87.9 r 8.00 x 10-5: 0.00703 mm. 16 Table 3. Particle sine data for alloy after annealing and solution treatment at 101801. for one and one-half minutes. Monsters of roughly circular particles of Gull; in rilar eyepiece micrometer units. 85 88 59 100 73 74 89 106 98 96 9O 90 66 70 66 94 79 73 96 76 7 3 7B 66 70 188 89 77 94 75 78 Average disaster I 86.1 units or 33.1 x 0.00 x 10-5 : 0.00005 as. Table 4. Particle size data for alloy after annealing and solution treatment at 101891. for two minutes. Diameters of roughly circular particles of M13 in l'ilar eyepiece micrometer units. 68 64 95 79 85 81 81 88 79 76 73 78 69 79 94 77 ' 76 105 a1 80 74 70 71 71 78 70 77 78 70 76 Average diameter 6 ".7 units. or 77.? x 8.00 x 10" : 0.00688 mm. 17 Table 5. Particle sise dag; for alloy after annealing and solution treatment at 1018 . for three minutes. Diameters of roughly circular particles of GuAlz in lilar eyepiece micrometer units. 67 78 68 56 68 66 57 68 75 86 71 65 86 64 67 69 79 74 74 56 78 65 70 80 7 8 58 60 78 71 66 Average diameter I 69 units or as x 0.00 x 1.0-5 = 0.0055 mm. Table I. Particle sise data for alloy after annealing and solution treatment at 1018°!'. for four minutes. Diameters of roughly circular particles of Bull, in Filer eyepiece micrometer units. 55 68 55 78 66 68 75 54 55 56 57 59 65 61 85 61 56 55 71 85 67 68 r 59 55 e0 64 78 56 68 55 Average diameter 8 64 units or 54 x 0.00 x 10"!5 a 0.0051 as. 18 Table 7. Particle size data for alloy after annealing and solution treatment at 1018°F. for five minutes. Diameter: of roughly circular particles or Calla in Filer eyepiece micrometer unite. 55 45 67 58 78 6O 45 69 77 50 65 48 45 58 45 60 48 45 45 65 78 67 58 69 58 55 77 48 5O 56 Average diameter = 59 units or 50 x 3.00 x 10'5 = 0.0047 mm. Table 8. Particle cine data for alloy after annealing and solution treatment at lOlBOF. for seven minutes. Enameters or roughly circular particles or GuAlg in Filer eyepiece micrometer units. 41 46 45 48 55 42 52 51 78 61 56 54 47 52 58 54 57 79 52 55 66 47 69 58 54 55 56 46 45 55 Arerage diameter : 54 units or 54 x 8.00 x 10-5 = 0.0043 mm. 19 Table 9. Particle size data for alloy after annealing and solution treatment at 101803. for ten minutes. Diameters of roughly circular particles of Cuhlg in Filer eyepiece micrometer units. 40 45 50 29 56 54 56 59 42 58 42 55 28 50 47 41 55 55 52 42 54 54 65 58 57 52 57 59 46 46 Axerage diameter s 39 units or 59 x 8.00 x 10-5 : 0.0031 mm. Table 10. Particle size data for alloy after annealing and solution treatment at 1018°F. for 15 minutes. Diameters of roughly circular particles or CuAlg in Filer eyepiece.micrometer units. 26 57 55 51 40 41 42 29 46 4O 52 5O 55 55 50 52 55 20 55 49 40 57 55 25 55 52 54 58 50 25 Awerage diameter : 34 units or 04 x 8.00 x 10“5 : 0.0027 mm. 20 Table 1].. Particle size data for alloy after annealing and solution treatment at 101803. for 20 minutes. Diameters of roughly circular particles or Cullg in Filer eyepiece micrometer units. 27 54 55 58 50 21 27 27 27 25 22 24 25 24 29 55 46 54 40 56 48 29 24 51 5O 24 5O 51 50 28 Average diameter 2 30 units or 30 x 8.00 x 10‘5 : 0.0024 m. Table 12. Particle size data for alloy after annealing and solution treatment at 101801. for 30 minutes. Diameters of roughly circular particles 01‘ 0:11.13 in Pilar eyepiece micrometer units. 25 16 2O 20 15 25 4O 21 15 21 15 15 18 16 50 55 27 51 25 16 25 25 24 16 57 25 25 5O 17 14 Average diameter : 23 units or 23 x 8.00 x 10-5 : 0.0018 mm. 21 Table 16. Particle size data for alloy after annealing and solution treatment at 10180F. for one hour. Diameters or roughly circular particles or GuAlz in Filar eyepiece micrometer units. 10 25 29 29 51 19 14 9 14 12 11 12 15 17 14 17 18 10 15 2O 15 16 14 12 15 15 10 2O 12 14 Average diameter : 16 units or 10 x 8.00 x10"5 : 0.0010 mm. fable 14. Particle size data for alloy after annealing and solution treatment at 1018°E. for two hours. Diameters or roughly circular particles or Guile in Filer eyepiece micrometer units. 19 8 9 10 10 18 8 11 16 16 12 16 17 12 15 12 11 11 12 15 14 18 12 14 11 11 12 11 10 9 Ayerage diameter : 13 units or 15 x 8.00 x 10-5 : 0.0010 mm. 22 Table 15. Particle size data for alloy after annealing and solution treatment at 1018°F. for four hours. Diameters or roughly circular particles or Cuslg in Filer eyepiece micrometer units. 8 9 9 11 8 12 14 9 9 10 9 10 10 9 9 10 10 11 11 8 15 10 11 11 9 ‘9 14 15 18 11 Average diameter : 11 units or 11 x 8.00 x 10*5 : 0.00088 ms. Table 16. Particle size data for alloy after annealing and solution treatment at 1018°F. for six hours. Diameters or roughly circular particles of Gull; in Filer eyepiece micrometer units. 9' 10 9 9 8 8 7 7 10 9 9 9 10 11 10 9 10 8 9 9 15 8 11 9 11 10 9 9 8 11 Average diameter : 9 units or 0 x 8.00 x 10'5 : 0.0007 mm. 25 Table 1?. Particle size data for alloy after annealing and solution treatment at 1018°F. for 24 hours. Diameters of roughly circular particles of 60.112 in Filer eyepiece mite. 0"..0‘i.0~l 'Ifl'lvlfil'l...‘ pa “0060....QO Average diameter 8 7 units or 7 x 8.00 x 10‘”5 a 0.0006 In. table 18. Particle sise ta tor alloy after annealing and solution trcement at 1018 . for 48 hours. Diameters or roughly circular particles of 0m; in Filer eyepiece micrometer units. 4Q.‘Gfl.‘6im 0......00. 9....0‘9.Q. Average diameter a 6 units or c x 8.00 x 10-5 : .0005 en. table 19. Smry of data on average sizes or larger cir- cular particles or Guile and per cent or original amount or precipitate in solution after heat treatment. time of Average Average Average Per cent or solution diameter diameter diameters original 00113 arm ".035. °* use. 78:5 3.21:8? nealing mar milli- x 10“) treatment (minutes) eyepiece meters) . (per out ) units) 92. 5 0. 00741 54. 9 Zero 87.9 0.00705 49.4 9.81 1'} 85.1 0.00665 44.2 19.5 2 77.7 0.00622 58.7 29.5 5 69 0 . 0055 50 45 4 04 0.0001 - as as 5 59 0 . 0047 25 56 7 54 0. 0045 18 67 10 59 0 . 0051 9.6 85 1‘. 34 0.0027 7.: av 20 50 0. 0024 5.6 69 50 25 0 . 0016 5. 2 ' 94 60 16 0. 0016 1. 7 97 120 15 0. 0010 1.0 96 340 11 0.00088 0.7? 98.6 560 9 0.0007 0.5 - 99.1 1440 7 0.0006 0.6 99.6 2680 6 0.0005 0. 2 ’ 99 . 6 figure 6. lioroatrnoture or aluminum-copper alloy after annealing treatment. 10001 Keller'a Itch. I .9 3 I l '. l - W figure 7. lieroetrnoture or alminnn-oopper alloy after annealing and solution treatment 1} ninntea at 101801. 10001 Ioller'a ltoh. 27 Figure 8. ZMioroatruoture or aluminumsoopper alloy after annealing and aolution treatment 3 minutes at 101801. 10001 Keller‘e Etoh. . ’ C) . ‘ ' ‘5 I Figure 9. Microetruoture of aluminumeeopper alloy after annealing and solution treatment 5 minutes at 1018°F. 10001 Keller'e Etoh. aluminum-comer Keller'a Btoh. of alloy after annealing and aolution treatment 10 minutes at 1018 lieroatruoture 10001 It“. 10. III: Or Keller'e Itoh.’ Hieroatruoture of aluminum-copper alloy after annealing and solution treatment 20 minutes at 1018 10001 Figure 11. alloy after annealing and solution treatment one hour at lOlBOF. Miorontrueture or aluminum-copper lOOOX Figure 12. Keller'e Etch. 01803. and solution Keller’a Etoh. of aluminum-copper hours at l annealing treatment four 10001 Hieroetruoture alloy after Figure 13. 50 Table No. 20. Relative resistance or aluminunpooppar alloy vire vs. tine at 101801. Relative Resistance Tine (voltage drop x 10-5) (minutes) 62.0 0 95.2 1.25 93.8 5.25 97.2 4 101.5 5 102.7 6 104.0 7 140.5 9 l41.8~ 15 150.2 15 149.1 20 150.2 25 162.5 50 156.5 45 145.0 50 142.5 60 155.0 120 157.0 , 185 205.5 211 216.0 240 252.7 275 505.0 282 525.5 288 526.0 554 577.8 , 1520 577.1 1565 560.5 1412 561.0 1560 578.2 1688 386.5 5994 588.0 6165 380.2 7147 . .. I. '0 '5‘- 52 DISCUSSION OF RESULTS The optimum.properties which may be realized in the age hardening type alloys depend to a large extent on the heat treatments to which they are subjected. According to generally accepted theories, the strength and hardness of such alloys is due to the precipitation of some hard phase during the aging process. Both the size and number of precipitate particles are important. A large number or very small particles is usually more effective than a few large ones. ' In the alloy in question it is the Gullz or theta phase shieh precipitates trom.the supersaturated solution and causes an increase in strength and hardness. The theta prime transitional phase is also formed in aluminum? copper alloys under certain conditions. However, in the procedure used in this work it is probable that the theta phase is formed directly and remains stable.(lz1 the aging process in these alloys has received can» siderahle attention fron.many investigators. (1" 8’ " 17’ 15' 22, 25’ 36’ 50) Before aging can take place the hardening constituent must be in solid solution. We are coneerned here with the problem.ot placing the Cull: phase he that condition.and more specifically vith its rate of solution. 35 A.high.purity binary alloy was used to eliminate as much as possible the confusing effects of impurity ela- nents upon the investigation. Since the accuracy of the results sould depend on the amount of copper present in the alloy, it was desirable to include as high a percent~ age of this element as possible sithout exceeding its solubility limit in aluminum. The alloy selected con- tains 4.91% copper. This approaches the solubility limit but allovs a broad enough temperature range for ease of sorking.in the solid solution field. Reference to Figure 1., the equilibrium diagram, indicates that the solution field for this percentage of copper is from.about 985° to 1040°I. The temperature for solution treatment should be as close as possible to the Inuit of solubility of the hardening constituent.(lgl 1018°F. las chosen since it is the eutectic horizontal and the temperature of maximum.solution of capper in pure aluminum. The results of the microscopic section of this inves— tigation are portrayed in Figure 5. This curve vas con- structed from the data summarized in Table 19. The per. cent of Gudla in solution is plotted against the logarithm. of the solution tune in minutes. It will be observed that the amount of precipitate going into solution varies lin- early with the logarithm of the soaking time until the ~ latter reaches the nine minute point. At this time about 75% is in solution. Beyond this time the rate of solution 34 decreases rapidly, the last few per cent of the precipi- tated phase going into solution very slowly. These results are in good agreement with those ob- tained by the Japanese intestigators, referred to in.the History section.of this paper, eho vorked with related a1- loys. For example, as previously mentioned, Hisatune found five hours were required for the solution of the soluble phases in an aluminumscopper-silicon.alley. la find that after five hours at temperature about 99.5% of the precipitate has dissolved. Secondly, Obinata and re- bata, in a study of duralumin, stated that 90% of the Guilt sea in solution after 10 minutes at 1004°E. The re- sult as found here is that 85% is in solution at this time. .Ls an.attempted check.on.the microscopic investigation the relative variation of electrical resistance of the alloy vith time at the solution.tempereture vas studied. The data obtained from the electrical measurements are con- tained in Table 20. 1 curve plotted from this data, rela- tive resistivity versus logarithm of time at solution temp pereture is shown in Figure 14. Considerable scatter is apparent in.the points on the curve, however, its shape at longer solution treatment times is similar to that plotted from.microscopic data. Both shoe that the rate of solution is very slow after the first 100 minutes at temperature. Figure 14. also gives some indication of the and point “for the solution of Gull: in aluminum. complete solution is indicated after about 74 hours of solution treatment. 55 Since we wished to study the rate of solution.of the euilz in the aluminum.matrix, it was first necessary to have all this phase precipitated except the three or four- tenths of a per cent of copper which is soluble at room temperature. In order to do this lengths of the three- eighths inch rod were held at about 101003. for 24 hours and cooled at the rate of 100°F. per hour in the furnace. This resulted in a iidmannstatten type structure. As it see desirable that the precipitate be in the form of par- ticles having a roughly circular cross-section, the meter- ial sea given a second treatment. The time at temperature was increased to 48 hours, and 52 hours were alloved for cooling to 600%. Since four-fifths of the sun, had pre- cipitated at this temperature, the alloy was then furnace cooled to room temperature. After this second treatment the appearance of the al- ley was as shown in.Figure 6. of the photomicrographs. Some nodular particles of Guilz are still present, repre— senting perhaps the edges of platelets or longitudinal sections of rods. This results from.the fact that some traces of a Widmsnnstattcn pattern still remained. Such particles probably had an adverse effect upon the accuracy of the results of this investigation. A.majority of the precipitate, however, was in a roughly circular form which permitted easy measurement of particles diameters. The determination of the mean diameter was made by averaging the diameters of 40 particles. This data is contained in Table 1.’ 1,111ar micrometer eyepiece which had been 56 calibrated with a stage micrometer was used as a measuring device. Two different determinations indicated that the aver- age diameter of the larger Cuilz particles in this sample was 92.5 eyepiece units or 0.00741 mm. “With.this value as a starting point, the decrease in the average size of the larger particles of Guilz after solution treatment could be closely estimated. Samples were suspended in a molten salt bath held at 101803. They were left in the salt pot for periods ranging from one minute to 48 hours and then quenched in water at about 6091. The temperature variation of the bath during this time use not more than plus or minus five degrees. The average Calla diameter in a sample after one min- ute in the salt bath was 0.00703 mm. Assuming that the amount of precipitate varies with the surface area on the face of the sample, we can compute the percentage of this phase in solution at any time. For example: the average diameter in the above sample is squared and divided by the square of the value for the annealed case. This amount is multiplied by 100 and subtracted from 100%. numerically: (o.oovos)3/(o.oov4l)3 : 0.9019 0.9019 1.100 2 90.19% 100% - 90.19% : 9.18%. Thus, after one minute at 101801. the data indicates that 9.81% of the Gulls originally precipitated is in solur tion. 3? The microstructure of the sample given a minute and one-half solution treatment is shown in.Figure 7. The average diameter of the larger particles in this case is 0.00665.mn., and the amount in solution is 19.5%. Photomicrographs of several other representative sam- phes shew continued decrease in precipitate size as the length of solution treatment is increased. After three minutes, 45% of the precipitate has gone into solid solu- tion, the average diameter at this time being 0.0055ll. Five minutes of soaking at 10180E. reduces the average Gudlz diameter to 0.004? mm. and the percentage in solar tion increases to 58%. Figures 10. and 11. illustrate the appearance of same plan which have received 10 and 20 minute solution treat- ments. By this time the amount in solution has reached 83% in the first case and 89% tn the second. lhen these photcnicrographs are compared with Figure 6.. the marked decrease in the average size of the larger particles of precipitate is very apparent. uncrostructures of two other samples are shown in Figures 12, and 13. One small particle of Cullg is clear- 1y visible in.the lower center of Figure 12., near the grain boundary.. This sample was in the salt bath for one hour at which time 97% of the precipitate was in solution and the particle diameter was 0.0013 mm. In Figure 13. the etchant reveals a very small; particle of Guila in the upper part of‘the picture to the right of the grain boundary. This shows the microstructure after four hours of soaking by which time 98.6% of the precipitated phase is again in solution. The maximum solution heat treatment time investigated was 48 hours. At this time a few tiny particles of Guile could still be measured with the Filer eyepiece, however, beyond this time the measurement of smaller particles would have been impracticable. The average size of these particles after 48 hours was 0.0005 mm. Only about four- tenths of one per cent of the Guile originally precipitat» ed was still out of solution by this time. As previously mentioned, it has been found in the past that many days may be required for complete solution of Cuilg even though the equilibrium diagram indicates that the temperature is sufficiently high. Due to this and because of the diffi- culty of making accurate measurements of smaller particles the investigation.was suspended after 48 hours at the. solution temperature. Prom.the data collected above calculations of the dif- fusion coefficient of copper in aluminum were made. The method used followed that of Brick and Phillips‘ls) and was explained by Meh1(15). The values found for D (cm.z per second) ranged from.8.87 x 10’13 for one minute of solution treatment to 2.20 x 10'10 for 30 minutes. The values found by Brick and Phillips were in order of 1.40 x 10”. The discrepancy in these values indicated that the equations used were not strictly applicable to the 59 conditions under which this investigation was made. It would be necessary to derive a special equation.to fit these conditions in order to make accurate calculations of D. Hewever, the process of solid solution is essenti- ally one of diffusion. is such this work is closely re- lated to the heat treatment of both non-ferrous and far- rous alloys(83). The circuit used in the measurement of relative elec- trical resistance has been described under Procedure. The sample used for this part of the study was a length of No. 40 gauge wire drawn from a part of the original alloy rod. The voltage drop across the sample at room tempera- ture was 62.0 microvolts. This value was the average of a number of readings taken over a period of more than 48 hours. Under the same conditions the drop across a stan- dard one ohm.reaistance averaged 203.8 microvolts, the largest variation from this value being less than 0.2%. This indicated that the current remained at nearly con- stant value. When the ceramic form.on which the wire was wound was placed in the muffle furnace its temperature was steady near the eutectic value of 101801. The potentiometer was zeroed against the standard cell Just previous to this in order to expedite the first few readings. The first volt- age drop was measured one and one-quarter minutes after placing the sample in the furnace and other readings were taken intermittently for the next 119 hours. 40 The resistivity of a solid solution is in almost all cases greater than that of a pure metal. This is because the solute atoms destroy the periodicity of the field within the solvent lattice thus increasing the opposition (32) This variation in.re- to the movement of electrons. sistivity should be roughly linear as the amount of solute in the solid solution increases. However, if a superlat- tics is formed the more ordered structure will result in a (34) Moreover, it has been found decrease in resistivity. in work on precipitation from solid solution that the in- formation gained from.changes in.this property may be anomalous.(33) The scatter in the data from.the voltage drop meas- urements was probably partially due to the fact that at the time the sample was placed in the furnace there was a considerable drop in temperature. This was caused at least in part by the relatively large mass of the ceramic form.on which the sample wire was wound and also due to the cool air which entered the furnace at the time. The shape of the curve determined from.the microscop- ic data is better supported by theory. Based on diffusion theory the linear part of this curve at short times is to be expected since the diffusion coefficient will remain fairly constant until the concentration gradient has changed radically. When most of the Guile has gone into solution the driving force will be relatively small since the concentration gradient is then quite flat. This explains why the last few tenths of a per cent of the pre- cipitate go into solution very slowly. Because there was insufficient time to carry it fur- ther, this investigation was terminated after gathering the data discussed above. However, there obviously re- mains a large amount of work, both experimental and theo- retical, to be conducted in.the field of solution rates in solid alloys. 42 CONCLUSIONS Although the data collected in this investigation is insufficient to draw definite conclusions which might serve to alter present industrial practice, some general state- ments concerning the problem of solution rates in pure bi- nary aluminumecopper alloys can be made. ‘ The microscOpic study indicates that about 80% of the precipitated Ouilz phase will be in solution after the al- loy has been at the eutectic temperature, 1018°r, for ten minutes. The results also lead to the conclusion that the solution rate will be quite constant during this period when time at temperature is plotted on a logarithmic basis. According to the data collected by microscOpic study, nearly all the precipitate will have dissolved in the ma- trix.after one hour at 1018°F. This investigation also demonstrates that the last few per cent of Guile can be expected to go into solution very slowly. We conclude from the electrical resistance data that the solution will be entirely complete after about 74 hours. The method used here, that of measuring the decrease in the average diameter of the larger particles of the precipitated phase, seems to provide reliable data. Since the problem of solution rates in solid alloys is important 43 in the heat treatment of both ferrous and non-ferrous al- loys, it is hoped that this method may prove useful in other work on the problem. Solution rates are closely related in theory to the process of diffusion in solid metals. This relationship is probably the key to mathematical expressions describ- ing solution rates. SILETBD BIBLI OGRAPHY ARTICLES (1) Campbell, I. and Mathews, 0., "The Alloys of Aluminum", Journal 9; the American Chemical 8 ciet , 24, 264, (1952). ""'"' (2) Campbell, W., "Aluminum Al1oys", Jo al of the Ameri- _._ag Chemig; Society, as, 129%“, 904T: "" "‘""' (3) Knerr, 3.0., "Duralumin, A Digest of Information", American Society for Stee;= IIreatin , 3, 13, (1922). (4) Gayler, M.L.V., ”The Constitution and Age Hardening of Alloys of Aluminum with Copper, Magnesium and Sili- con in the Solid State”, Journal 9; the ggtitgte QLMetals, 28, 213, (1922 . (5) Gayler, M.L.Y. and Hanson, D., "The Heat Treatmmt and Mechanical Properties of Alloys of Aluminum with Small Percentages of Copper", Journa g; the Insti- tute g; Metals, 89, 491, (less . (6) Jette, 3.3., Phragmen, 0., and Iestgren, A.l‘., 'I-Bay Studies on the Copper-Aluminum Alloys", Jogrna; 2f. thg Institute 91 Metals, 31,, 193, (1924). ('7) Dir, 13.3. Jr. and Richardson, 1.3., ”Iquilibrium Bela- tions in Aluminum-Copper Alloys of High Purity", Institute of Metals Division, America; Institute of E—igg Effie-Flip}? ca' 1' "fig‘ineera , 73 , '5'6‘0","'T"'19 2'57. (8) Mehl, 11.1., Barrett, 0.8. and Rhinos, I.N., "Studies upon the lidmannstatten Structure, III-The Alumi- num-Rich Al1oys of Aluminum with Copper, and of Alu- minium with Magnesium and Silicon“, Institute 9; fitals Division, American Institute _o_f_ m1 app. tallurgica; En ineers, 99, 203, (1932 . (9) Stockdale, D. "The Constitution of the Aluminum-Rich Aluminum-Copper Alloys above 400%.", Journal 9_f_ 3;; guns. gustag, 52, 111, (1933 . (10) Bradley, A.J. and Jones, P., ”An I-Ray Investigation of The Copper-Aluminum Alloys“, Journal 2; the La;- stitute QLMstals, 51, 131, (1933 . (ll) Eisatsune, C. , ”A Study of the Aluminum-Rich Aluminum- (12) (13) (14) (15) (16) (17) (18) (19) (so) (21) (38) 45 Copper Alloys”, Journal at; W Abstracts (Japan) 10, 48, (1934) and tallurgicg Abstracts London), 1, 284, (1934). “sagas, G. and Hearts, Metallwirtschaft, 14, 605, Pink, LL. and Wiley, _L.A., "Equilibrium Relations in Aluminum-Zinc Alloys of High Purity'. Institute 9; Motels Division, American Institute 9; mm §_n_d_ Etallurgical En ineers, 122, 544, (1936 . rm, l.L. and Smith, D.l., "Age Hardening of Aluminum Alloys", Institute 9_f_ Metals Division, American Institute of Minin and Etallurgica;= agineers IE?— , 284,- T1955"). ' Mehl, R.l'., ”Diffusion in Solid Metals”, Inst tute of Metals Division, American Institute of min Ln}:- Metallurgical mineers, 122, 11, (15-36 . Brick, 3.1!. and Phillips, A., ”Diffusion of Copper and Magnesium into Aluminum", Institute 93 Metals Di- vision, American Institute p; W and Eta-fur- gical Engineers, 124, 331, 1937 . Cayler, M.L.V., "Sane Characteristics of Copper-Alumi- num Alloys Made from Aluminum of Very High Purity", Journal 9; the Institute 9;; Metals, 63, 67, (1938). Pink, LL. and Smith, D.I., "Age Hardening of Aluminum Alloys, III-Double Aging Peaks", Institute of Metals Division American Institut of Mini and fitglur- "gi‘eai gg"’1n"'e"""‘ers, '1' '28“, are, (15558). Irmann, 11., "Solution Treatment of Aluminum Casting Alloys”, Metal Igdustry (London), 54, (25), 663, 1939,. Mehl, 3.1. and Johnson, I.A., ”Reaction Kinetics in Processes of Nucleation and Growth”, Institute 9_f_ Metals Division, Amegican Institute 9; Minin ;a_n_d_ Metallurgical Engineers, 136, 416, (1939 . Obinata, I. and Tabata, K., ”Studies on the Solution Heat Treatment of Duralumin" Institute 9_f_ Metals (Japan), 4( (10), 324, (1940) “and Metallurg—_ica]_. 1. Abstracts ondon), 8, 70, (1940). Banana, C.H., "The Process of Precipitation from Solid Solution, I-A Crystallographic Mechanism for the Aluminum-Copper Alloys", Institute _o_f_ Metals Divi- sion, America Institute .o_f_ Mg and Metallurgiog]i a. 46 m; ineers, 137, 85, (1940). (23) Iells, C. and Mehl, R.P., "Rate of Diffusion of Carbon in Austenite in Plain Carbon, in Michel and in Man- ganese Steels”, I_1_-___on an___d_ §_____tee1 Divisio , American Institute of Mini—_— and Metal—_iurgicg En ineers 14 , 27 9, '(‘1'94""o"")"£. ' (24) Hhinea,HF.N. and Colton, S.A., "Homogenization of Cop- per- ickel Powder A110 8", American Society gor Me___t__als, 30, 166, (.1942) (25) Dix, H.H. Jr., "New Developments in High-Strength Aluminum Alloy Products', American Society fo_;_ Meta___J_.__s, 35, 130, (1945). (26) Geisler, A.H. and Keller, F., "Precipitation in Age Hardened Aluminum Alloys", gtitute 9__f Metals Division Ameri____c_____an Institutg i a_n__ etg- ic urineers, 171, 192 ,‘Ti‘TEE—uv . (27) Glaser, l'.I.',‘ "Electrical Resistance Measurements on Iron-Silicon Compacts Prepared by the Powder Metal- lurgy Procedure", M_e____tals Branch, American Institute fiMini and Metallurgical En ineers, 185,475, 949 . (28) Edgar, J.K., "Solubility of Iron in Solid Aluminum”, Institute of MetalsD i;_i_sion, Amerigan Institute of Mini end-Metallurgical agineers, 180, 225, (1949;.— (29) Hultgren, R. and Lever, S.A., “Use of Electrical Re- sistivity Measurements to Determine the Solidus of the Lead-Tin System”, Metals m, Am___e______rican In- stitute of Mining and Ego Iiur urgica; ggineefif's , 185, __‘T'ev, 194?). (30) Lemon, 3.0. and Hunsicker, H.Y., “lffects of Quenching Rate and Quench-Aging on the Tensile Properties of Aluminum Alloy 618*, American Society _f_o_1_-_ Metals, 42, 357, (1950). . (31) Hhines, F.N., Private Communication, February 4, 1950. BOOKB (32) Mott, NJ. and Jones, H., ”The Theor of the Pro or- ties of Metals and Alloys, _anra, EfigIan aren- 3011:1388, 1936, ID. 286‘502e (33) -----A_g9_ Hardenng p; Metals, Clegeland, Ohio, r1 1‘ 47 American Society for Metals, 1940, Pp. 9-11, 18, 22, 34-38, 40-42, 345. (34) Hume-Bothery, I., Atomic Theory for Stgdents g; mtg,- lurgy, London, Institute of Metals, 1946, Pp. 204- (35) --- a1s Handbook, Cleveland, Ohio, American Society for tals, 1948, Pp. 761-840. ' (36) ---Physical Metallurgy 9_f_ uminum Alloys, Cleveland, Ohio, American Society for tals, 1949, Pp. 48-50, 95-99 , 212 O SUPPLEAENTAL BIBLIOGRAPHY Edwards, J.D., ”Mechanism of Solidification of a Copper- Aluminum Alloy” Chemica and Metallurgical gagin- e rin , 24, 217: (19‘21' ). Portovin, w. and Chevenard, P., "A Diletometric Study of the Transformations and Thermal Treatment of Light Alloys of Aluminum" Joggnal 2;, Egg Institute 9; Me}.- a s, so, 329, (1923). -----eStruoture of Aluminum-Copper Alloys' £319; pro ress 18. 81, (October 1930). .’ ___s.__... Dorn, J.E. and Harder, 0.3., "A Theory of Diffusion in Solids”, Institute of Metals Division, American Insti- tute )1; Mi—Mgg an mat-'mi urgi'TFcc inee——-rs, 128775-15 , (1938 . Rhinos, LN. and Mehl, R.F., ”Rates of Diffusion in the Alpha Solid Solutions of Copper”, gistitute of Metals Division, American Institute of MiniLgT—TM Torin"? gIcEI En inoers, I28, I85, (1938). link, I.L. and Wiley, L.A., 'Equilibrium Relations in Alum- mum-Zirconium Alloys of High Purity“, Institute 9_f_ Metals Division, American Institute of Minin gag. Mo‘t'alluF-Wmm, 153, 69, (1939). anith, 0.8., ”Constitution and Microstructure of Copper- Rith Copper-Silicon Alloys”, Institute of Metals Divi- sion, American Institute of Mini_n_g and Motallurgical iii—g :_L_n oW,, W457. Mondolfo, L.r., Metallography 21: Aluminum Alloys, New York, John Wiley and ens, c., 1943, E. 16- 8, 168-169, 176-177, 298-302. Jetter, L.K. and Mehl, R.F., "Rate of Precipitation of Sil- icon from the Solid Solution of Silicon in Aluminum”, Institute of Metals Division, American Institute 9_f_ M_—_——ining ang'fic—tchrgm—icci En inee s, 152'“, (1"94"3'T‘. Gayler, M.L.V., "Diffusion in Relation to Changes in Micro- structure", Institute 2; Motels Division, American Institute a; Mining and Mofllurgical En ineors, 156, m . 4 ' t ( ¢ . ' C s . . t 4 L C . . _ . . ' . e - t Z - o . . ‘ e 0 g e . o I - O 5 - ' e ' 5' . . . . e . . . a O . ' - O O n — I ' g l , I I s 9 x V” o‘ - f I I y ‘ x . I ‘ . Q ' . ' V I ‘ 6 . a e e ' e . , . O a ' I t r e . . . . f e . v . I . . . . . ' . V, 9 a e ‘ . . t . C ‘ : s l O o ( . ’ O a . o . l u . . . ' ’ ' e s . I l l e \ . . . Q I - w - v‘ - ~. a t a s K s ' r e ‘ ( . I e 49 Tomes, C.H., "An I-Ray Study of the Time-Rate of Precipi- tation from a Solid Solution of Capper in Aluminum", American Society for Metals, :53, 281, (1944). Mehl, Rd" and Anderson, S.A., "Recrystallization of Alum- inum in Terms of Rate of Nucleation and Rate of Growth”, Institute of Metals Division, American Insti- tute of mumigg gpg Eat—"_ailurgi—‘E—cu ins—__ors, 1"e"1',"f‘d"4 , (1925)"? Mondolfo, L.P., ”Diffusion in R301 Alloy and Its Effect on the Corrosion Resistance“, Institute 2; Motels Diyi- sion, American Institute of Mining _a_n_q Metallurgical En ineers, 166,"2'2'_('19, 9461'. Brown, H., Aluminum and Its Applications, Nev York, Pitman Publishing Corp., 1948. Bowen, H.G. Jr. and Bernstein, H., "Effects of 16 Alloying Elements upon the Grain Size of Cast 4.5% Copper- Aluminum Alloy", Ameyican Society for Metals, 40, 209, 1948 . Dean, R.S., Potter, 3.7. and Huber, R.I., "‘nio Electrical Resistivity and Temperature Coefficient of Resistance of Copper-Manganese Alloys”, American Society ;_c_>_1_'_ Metals, 40, 381, (1948). ‘ Pink, I.L., Wiley, LA. and Stumpf, H.0., ”Equilibrium Re— lations in Aluminum-Sodium Alloys of High Purity", Institute of Metals Division, American Institute 9_f_ ming egg-Metallurgical Egineers, 1'75, 364, (1948). Harrington, R.H., "The Effect of Single Addition Metals on the Recrystallization, Electrical Conductivity and Rupture Strength of Pure Aluminum”, American Society LgyMotals, 41, 443, (1949). Barker, L.J., ”Revealing the Grain Structure of Common Aluminum Alloy Metallographic Specimens", American Society for Metals, 42, 347, (1950). ROOM usa (my f‘fi.- " -.a h- v-I- 2’. " r'l -' ‘_ I'Oe . I f- - . ' .7 . ‘ ’ 7"." "-. h . > ‘ I. V - 3". ‘AI ' >C ‘ It \LI.‘ I ' H ’ A, r W (w b; L. as» » a, u 2’ ’9): . ‘. .t I, ‘~.‘.- I I "‘1 v p \‘ ,_ I; ,‘v- t.‘ ,, \. ('r ',.. ’ .‘( ." -E’ *,.!p";:5;; i“ i ”H, f . A ya.“ ' ' 7;. ' ‘3"va cw We.) .1 l )1 "if ’7 ‘ P -“:L 1".“ '0.“ 1“, t. ‘\ .“‘.‘AII.'A ‘ b . -l -J"" fly.) ‘J! ('4 "Aj' .‘-'-—‘T~""\ "a 1, "y" 1' '1' “affix, t“ a a <‘Lt>;t'7~ t L .1 i . J‘ lv""’l=.":.\; r ‘- ‘ If“: 3" LUP‘A f anew =‘ .L 342.19.- 2. (M _1 t 4:1 I» '."‘.‘.'{~'- I . 3'19," 3 (“9". (‘ ' "r, 01‘»: ii...‘ (.1.’ ‘71“. . ,'.’ Lid . " a; RX“, 93;? I., :13?) ‘1 I‘- '..r a"? ' 7 “"‘q . . I l a. \ . . — Q ‘4‘ A a . . . . .. a I '1' I l - y A. . .> . . . , ... n . . . . 7.. . . . .. \. 4 .. .‘ . ... a . . y . .r. . in H., u a . . . . . L. 9)....” .... 4 . .. u . I . . V . .. c I a. . n H. 1 n ..1 . I . ...K . . . I \ ... y t“ . . .y t . . w. .... .it w ... . . .. i l a 1 A . .4. ..h. .... . y. a N i \. .... . i\. o: \ . . y . . .“WM . . .. I. . .v. ....a.#.‘..; a d. a. ...«L a n ..u. . .r n .....a . J.. 14. . .....3; .. . .m. ......u V. .... . . V . all ,. ai. ,., .. t. , .. w _ . . l.‘1 (I)- .‘t ( tit. . . ”TITI'IMI’IILEM)((JIIEIETI'IMHJFIZIHI'I‘ITFs