ASM Handbook, Volume 4: Heat Treating Copyright © 1991 ASM International® ASM Handbook Committee, p 841-879 All rights reserved. DOI: 10.1361/asmhba0001205 www.asminternational.org Heat Treating of Aluminum Alloys

HEAT TREATING in its broadest sense, • Aluminum-copper-magnesium systems The mechanism of strengthening from refers to any of the heating and cooling (magnesium intensifies precipitation) precipitation involves the formation of co- operations that are performed for the pur- • Aluminum-magnesium-silicon systems herent clusters of solute atoms (that is, the pose of changing the mechanical properties, with strengthening from Mg2Si solute atoms have collected into a cluster the metallurgical structure, or the residual • Aluminum-zinc-magnesium systems with but still have the same crystal structure as stress state of a metal product. When the strengthening from MgZn2 the solvent phase). This causes a great deal term is applied to aluminum alloys, howev- • Aluminum-zinc-magnesium-copper sys- of strain because of mismatch in size be- er, its use frequently is restricted to the tems tween the solvent and solute atoms. Conse- specific operations' employed to increase quently, the presence of the precipitate par- strength and hardness of the precipitation- The general requirement for precipitation ticles, and even more importantly the strain hardenable wrought and cast alloys. These strengthening of supersaturated solid solu- fields in the matrix surrounding the coher- usually are referred to as the "heat-treat- tions involves the formation of finely dis- ent particles, provide higher strength by able" alloys to distinguish them from those persed precipitates during aging heat treat- obstructing and retarding the movement of alloys in which no significant strengthening ments (which may include either natural aging dislocations. The characteristic that deter- can be achieved by heating and cooling. The or artificial aging). The aging must be accom- mines whether a precipitate phase is coher- latter, generally referred to as "non-heat- plished not only below the equilibrium solvus ent or noncoherent is the closeness of treatable" alloys, depend primarily on cold temperature, but below a metastable miscibil- match or degree of disregistry between work to increase strength. Heating to de- ity gap called the Guinier-Preston (GP) zone atomic spacings on the lattice of the matrix crease strength and increase ductility (an- solvus line. The supersaturation of vacancies and on that of the precipitate. These nealing) is used with alloys of both types; allows diffusion, and thus zone formation, to changes in properties result from the forma- metallurgical reactions may vary with type occur much faster than expected from equi- tion of solute-rich microstructural domains, of alloy and with degree of softening desired. librium diffusion coefficients. In the precipi- or GP zones. Except for the low-temperature stabilization tation process, the saturated solid solution The exact size, shape, and distribution of treatment sometimes given for 5xxx series first develops solute clusters, which then be- GP zones depend on the alloy in which they alloys (which is a mill treatment and not come involved in the formation of transitional form and on the thermal and mechanical discussed in this article), complete or partial (nonequilibrium) precipitates. history of the specimen. Their shape can annealing treatments are the only ones used for non-heat-treatable alloys. A general overview of these heat treatments is covered in the article "Principles of Heat Treating of Nonferrous Alloys" in this Volume. 8°°/ L ,400

Precipitation from Solid Solution One essential attribute of a precipitation- hardening alloy system is a temperature- ////I////I/.'////////I/I///Jf..z,~/7/////x/~ - 1000 dependent equilibrium solid solubility char- 0 / I I I Temperaturerange for u_ acterized by increasing solubility with ! AI ]~ J i i solution heat treating ~_ increasing temperature (see, for example, the phase diagrams in Fig 1 and 2). Al- though this condition is met by most of the ~~~'~ L Temper;turerange E binary aluminum alloy systems, many ex- hibit very little precipitation hardening, and ~~'~~"~ J ~i~ iiai[ilran efor -- 600 1- these alloys ordinarily are not considered heat treatable. Alloys of the binary alumi- 200 precipitationheat num-silicon and aluminum-manganese sys- ~ J treating I tems, for example, exhibit relatively insig- nificant changes in mechanical properties as A,+ooA,2 I I i 200 a result of heat treatments that produce considerable precipitation. The major alu- ol I l t I minum alloy systems with precipitation 0 2 8 10 12 hardening include: Copper, % Portion of aluminum-copper binary phasediagram. Temperature rangesfor annealing, precipitation heat • Aluminum-copper systems with strength- Fig 1 treating, and solution heat treating are indicated. The range for solution treating is below the eutectic ening from CuAI 2 melting point of 548 °C (1018 °F) at 5.65 wt% Cu. 842 / Heat Treating of Nonferrous Alloys

700 The GP zones are characteristically meta- nar aggregates (GP zones), which form on 1200 stable and thus dissolve in the presence of a particular crystallographic planes of the alu- 600 more stable precipitate. This dissolution minum matrix. These aggregates create co- Solidus j~__ 1000 causes a precipitate-free, visibly denuded herency strain fields that increase resis- region to form around the stable precipitate tance to deformation, and their formation is 500 / 595 °C u_ particles. The final structure consists of responsible for the changes in mechanical ? J at 1.85% 800 oa; equilibrium precipitates, which do not con- properties that occur during natural aging. 40o ~.~ Mg2Si _ tribute as significantly to hardening. More At higher temperatures, transition forms of Solvus detailed information about preprecipitation approximate composition AI2Cu develop ~- 300 6oo & E E phenomena can be found in the article and further increase strength. In the highest / "Structures Resulting From Precipitation strength condition, both the 0" and 0' tran- 2OO 400 From Solid Solution" in Volume 9 of the sition precipitates may be present. When I 9th Edition of Metals Handbook. time and temperature are increased suffi- Precipitation in Aluminum-Copper Alloys. ciently to form high proportions of the equi- 100 I 200 I Figure l, which illustrates the required sol- librium 0, the alloy softens and is said to be Mg-Si ratio of 1.73:1 ubility-temperature relationship needed in "overaged." I I I 0.5 1.0 1.5 2.0 precipitation strengthening, shows the tem- The commercial heat-treatable aluminum Mg2Si, % perature ranges required for solution treat- alloys are, with few exceptions, based on (a) ment and subsequent precipitate hardening ternary or quaternary systems with respect in the aluminum-copper system. The equi- to the solutes involved in developing Temperature, °F librium solid solubility of copper in alumi- strength by precipitation. Commercial al- 570 660 750 840 930 10201110 num increases as temperature increases-- loys whose strength and hardness can be 1.0 1.4 from about 0.20% at 250 °C (480 °F) to a significantly increased by heat treatment maximum of 5.65% at the eutectic melting include 2xxx, 6xxx, and 7xxx series wrought 0.8 Solvus with silicon _ ~ 1.2 o~ and Mg2Si present /0_ 1.0 temperature of 548 °C (1018 °F). (It is con- alloys (except 7072) and 2xx.0, 3xx.0, and siderably lower than 0.20% at temperatures 7xx.O series casting alloys. Some of these o.6 0.8 below 250 °C.) For aluminum-copper alloys contain only copper, or copper and silicon, 0.6 § containing from 0.2 to 5.6% Cu, two distinct as the primary strengthening alloy addi- == 0.4 equilibrium solid states are possible. At tion(s). Most of the heat-treatable alloys, 0.4 ~ temperatures above the lower curve in Fig 1 however, contain combinations of magne- 0.2 .o.~ 0.2 (solvus), the copper is completely soluble, sium with one or more of the elements and when the alloy is held at such temper- copper, silicon, and zinc. Characteristical- o 0 " atures for sufficient time to permit needed ly, even small amounts of magnesium in 300 350 400 450 500 550 600 diffusion, the copper will be taken com- concert with these elements accelerate and Temperature, °C pletely into solid solution. At temperatures accentuate precipitation hardening, while (b) below the solvus, the equilibrium state con- alloys in the 6xxx series contain silicon and Equilibrium solubility as function of tempera- sists of two solid phases: solid solution, ct, magnesium approximately in the propor- Fig 2 ture for (a) Mg2Si in aluminum with an Mg-Si plus an intermetallic-compound phase 0 tions required for formulation of magnesium ratio of 1.73-to-1 and (b) magnesium and silicon in solid (AIECU). When such an alloy is converted to silicide (MgESi). Although not as strong as aluminum when both Mg2Si and silicon are present all solid solution by holding above the sol- most 2xxx and 7xxx alloys, 6xxx series al- vus temperature and then the temperature is loys have good formability, weldability, ma- sometimes be deduced by refined studies of decreased to below the solvus, the solid chinability, and corrosion resistance, with diffuse x-ray scattering. Under favorable solution becomes supersaturated and the medium strength. conditions, GP zones can be seen in trans- alloy seeks the equilibrium two-phase con- In the heat-treatable wrought alloys, with mission electron micrographs. Spherical dition; the second phase tends to form by some notable exceptions (2024, 2219, and solute-rich zones usually form when the solid-state precipitation. 7178), such solute elements are present in sizes of the solvent and solute atoms are The preceding description is a gross over- amounts that are within the limits of mutual nearly equal, as in the aluminum-silver and simplification of the actual changes that solid solubility at temperatures below the aluminum-zinc systems. If there is a large occur under different conditions even in eutectic temperature (lowest melting tem- difference in atom sizes, as in the alumi- simple binary aluminum-copper alloys. A perature). In contrast, some of the casting num-copper system, the GP zones usually variety of different nonequilibrium precipi- alloys of the 2xx.O series and all of the 3xx.O form as disks whose planes lie parallel with tate structures is formed at temperatures series alloys contain amounts of soluble some low-index plane of the matrix lattice. below solvus. In alloys of the aluminum- elements that far exceed solid-solubility Sometimes, the solute atoms occupy pre- copper system, a succession of precipitates limits. In these alloys, the phase formed by ferred lattice sites within the GP zone, and is developed from a rapidly cooled super- combination of the excess soluble elements thus form a small region of lattice order. saturated solid solution (SSS). These pre- with the aluminum will never be dissolved, The GP zones are of the size range of tens cipitates develop sequentially either with although the shapes of the undissolved par- of angstroms in diameter. They are essential- increasing temperature or with increasing ticles may be changed by partial solution. ly distorted regions of the matrix lattice, rath- time at temperature between room temper- Most of the heat-treatable aluminum alloy er than discrete particles of a new phase ature and the solvus. The several stages are systems exhibit multistage precipitation and having a different lattice. As such, they are identified by the following notation: undergo accompanying strength changes completely coherent with the matrix, impos- analogous to those of the aluminum-copper SSS ~ GP zones ~ 0" ing local but often large strains on it. These system. Multiple alloying additions of both 0"-~0 (A12Cu) mechanical strains, as well as the presence of major solute elements and supplementary a locally solute-rich, sometimes ordered lat- At temperatures in the natural aging elements employed in commercial alloys are tice, can account for large changes in me- range (about -20 to 60 °C, or 0 to 140 °F), strictly functional and serve with different chanical properties of the alloy before any the distribution of copper atoms changes heat treatments to provide the many differ- long-range microstructural changes occur. with time from random to the disklike pla- ent combinations of properties---physical, Heat Treating of Aluminum Alloys / 843 mechanical, and electrochemical--that are minum-lithium alloys aimed at optimizing magnesium alloys. These alloys were cho- required for different applications. Some al- mechanical properties after artificial aging. sen to superimpose the precipitation- loys, particularly those for foundry produc- The age hardening of aluminum-lithium hardening characteristics of aluminum- tion of castings, contain amounts of silicon alloys involves the continuous precipitation copper-, aluminum-copper-magnesium-, far in excess of the amount that is soluble or of 8' (AlaLi) from a supersaturated solid and aluminum-magnesium-base precipitates needed for strengthening alone. The function solution. The aluminum and lithium in the 8' to the hardening of lithium-containing pre- here is chiefly to improve casting soundness precipitates are positioned at specific loca- cipitates. Proceeding in this manner, alloys and freedom from cracking, but the excess tions. The eight shared corner sites are 2020 (A1-Cu-Li-Cd), 01429 (AI-Mg-Li), 2090 silicon also serves to increase wear resis- occupied by lithium, and the six shared (AI-Cu-Li), and 2091 and 8090 (A1-Cu-Mg- tance, as do other microstructural constitu- faces are occupied by aluminum. This gives Li) evolved. Besides these registered al- ents formed by manganese, nickel, and iron. rise to the aluminum-lithium composition of loys, other commercial aluminum-lithium Parts made of such alloys are commonly 8' precipitates. The geometrical similarity alloys include Weldalite 049 and CP276. used in gasoline and diesel engines (pistons, between the lattice of the precipitates and Properties and applications of these alloys cylinder blocks, and so forth). the face-centered cubic lattices of the solid are discussed in the article "Aluminum- Alloys containing the elements silver, solution facilitates the observed cube/cube Lithium Alloys" in Volume 2 of the 10th lithium, and germanium are also capable of orientation. The lattice parameters of the Edition of Metals Handbook. providing high strength with heat treatment, precipitate are also closely matched to In terms of 8' precipitation, the only and in the case of lithium, both increased those of the matrix. Consequently, the mi- effect of magnesium appears to be a reduc- elastic modulus and lower density, which crostructure of an aluminum-lithium alloy tion in the solubility of lithium. The micro- are highly advantageous--particularly for solution heat treated and aged for short structure of an aluminum-magnesium-lithi- aerospace applications (see the following times below the 8' solvus is characterized um alloy in the early stages of aging is section "Aluminum-Lithium Alloys" in this by a homogeneous distribution of coherent, similar to that of an aluminum-lithium alloy. article). Commercial use of alloys contain- spherical 8' precipitates. Precipitation in the aluminum-copper-lithi- ing these elements has been restricted either Aluminum-lithium-base alloys are micro- um system is more complicated than that in by cost or by difficulties encountered in structurally unique. They differ from most either the aluminum-lithium or aluminum- producing them. Such alloys are used to of the aluminum alloys in that once the magnesium-lithium systems. some extent, however, and research is be- major strengthening precipitate (8') is ho- Effects on Physical and Electrochemical ing directed toward overcoming their disad- mogeneously precipitated, it remains coher- Properties. The above description of the vantages. ent even after extensive aging. In addition, precipitation processes in commercial heat- In the case of alloys having copper as the extensive aging at high temperatures (> 190 treatable aluminum alloys (as well as the principal alloying ingredient and no magne- °C, or 375 °F) can result in the precipitation heat-treatable binary alloys, none of which sium, strengthening by precipitation can be of icosahedral grain-boundary precipitates. is used commercially in wrought form) af- greatly increased by adding small fractional with five-fold symmetry. Although the qua- fect not only mechanical properties but also percentages of tin, cadmium, or indium, or si-crystalline structure and the composition physical properties (density and electrical combinations of these elements. Alloys of these grain-boundary precipitates are not and thermal conductivities) and electro- based on these effects have been produced yet exactly known, it has been suggested chemical properties (solution potential). On commercially but not in large volumes be- that both the precipitates and the precipi- the microstructural and submicroscopic cause of costly special practices and limita- tate-free zones (PFZs) near the grain bound- scales, the electrochemical properties de- tions required in processing, and in the case aries might play a major role in the fracture velop point-to-point nonuniformities that of cadmium, the need for special facilities to process. account for changes in corrosion resistance. avoid health hazards from formation and The low ductility and toughness of binary Measurements of changes in physical and release of cadmium vapor during alloying. aluminum-lithium alloys can be traced, at electrochemical properties have played an Such alloys, as well as those containing least in part, to the inhomogeneous nature important role in completely describing pre- silver, lithium, or other particle-forming el- of their slip, resulting from coherent-parti- cipitation reactions and are very useful in ements, may be used on a selective basis in cle hardening of spherical 8' precipitates. analyzing or diagnosing whether heat-treat- the future. The presence of equilibrium ~ (aluminum- able products have been properly or im- Aluminum-Lithium Alloys. Like other age- lithium) precipitates at grain boundaries can properly heat treated. Although they may hardened aluminum alloys, aluminum-lithi- also cause PFZs, which can induce further be indicative of the strength levels of prod- um alloys achieve precipitation strengthen- strain localization and promote intergranu- ucts, they cannot be relied upon to deter- ing by thermal aging after a solution heat lar failure. Consequently, for the develop- mine whether or not the product meets treatment. The precipitate structure is sen- ment of commercial alloys, slip has been specified mechanical-property limits. Since sitive to a number of processing variables, homogenized by introducing dispersoids elements in solid solution are always more including, but not limited to, the quenching (manganese, zirconium) and semicoherent/ harmful to electrical conductivity than the rate following the solution heat treatment, incoherent precipitates, such as T1 same elements combined with others as the degree of cold deformation prior to (AIECuLi), 0' (AIECu), or S (AIELiMg), intermetallic compounds, thermal treat- aging, and the aging time and temperature. through copper or magnesium additions. ments are applied to ingots used for fabri- Minor alloying elements can also have a Magnesium and copper improve the cation of electrical conductor parts. These significant effect on the aging process by strength of aluminum-lithium alloys through thermal treatments are intended to precipi- changing the interface energy of the precip- solid-solution and precipitate strengthening, tate as much as possible of the dissolved itate, by increasing the vacancy concentra- and they can minimize the formation of impurities. Iron is the principal element tion, and/or by raising the critical tempera- PFZs near grain boundaries. Zirconium, involved, and although the amount precip- ture for homogeneous precipitation. Like which forms the cubic AlaZr coherent dis- itated is only a few hundredths of a percent, some other age-hardened 2xxx aluminum persoid, stabilizes the subgrain structure the effect on electrical conductivity of the alloys, aluminum-lithium-base alloys also and suppresses recrystallization. wire, cable, or other product made from the gain increased strength and toughness from Development of commercially available ingot is of considerable practical impor- deformation prior to aging. This unusual aluminum-lithium-base alloys was started tance. These alloys may or may not be heat phenomenon has given rise to a number of by adding lithium to aluminum-copper, alu- treatable with respect to mechanical prop- thermomechanical processing steps for alu- minum-magnesium, and aluminum-copper- erties. Electrical conductor alloys 6101 and 844 / Heat Treating of Nonferrous Alloys

6201 are heat treatable. These alloys are should, when possible, be above the temper- aluminum and particles of AI2Cu. When this used in tempers in which their strengthening ature at which complete solution occurs (sol- product is heated slowly, the AI2Cu begins tO precipitate, the transition form of Mg2Si, is vus). In the alloy represented by line (a) in Fig dissolve, and if heating is slow enough, all of largely out of solid solution to optimize both 1, these temperatures would be about 575 and the Al2Cu is dissolved when temperatures strength and conductivity. 515 °C (1065 and 960 °F), respectively. How- above the solvus (500 °C, or 932 °F) are ever, under production conditions, the tem- reached. When the heating rate is high, how- Strengthening by Heat Treatment perature interval for solution treatment ever, much of the A12Cu remains undis- (shown in Fig 1 for typical 2xxx or 2xx.x) solved. If a material with this microstructure Heat treatment to increase strength of alloys provides a margin to safeguard against is heated at or above the eutectic temperature aluminum alloys is a three-step process: eutectic melting and a cushion on the low side of 548 °C (1018 °F), melting will begin at the for increased solution and diffusion rates. interface between the AIECU and the matrix. • Solution heat treatment: dissolution of For alloys containing more than 5.65% With sufficient time above the eutectic tem- soluble phases Cu, complete solution can never occur. For perature, this metastable liquid will dissolve • Quenching: development of supersatura- these alloys, such as alloy 2219 (which has to form a solid solution and will leave no trace tion 5.8 to 6.8% Cu), the minimum solution heat- provided that hydrogen gas has not con- • Age hardening: precipitation of solute at- treating temperature is established so that it densed at the interface to form a void. If the oms either at room temperature (natural is as close as practical to the eutectic tem- product is quenched before the liquid has aging) or elevated temperature (artificial perature while providing a margin of safety time to equilibrate, however, it will solidify aging or precipitation heat treatment) commensurate with the capability of the and form fine eutectic rosettes. This nonequi- Each of these steps and the use of quench- equipment. Line (b) in Fig 1 is another librium melting should not be confused with factor analysis are described in the follow- example of a composition above 5.65% Cu true equilibrium melting, which would occur ing four sections. Typical solution and pre- that does not allow complete dissolution of in any alloy containing more than 5.65% Cu. cipitation heat treatments for mill products aluminum-copper precipitates. In such an alloy, eutectic melting is equilibri- are given in Tables l(a, b, and c) and 2, and For more complex ternary and quaterna- um melting. No matter how long such an treatments for castings are given in Table 3. ry systems, solution treatments are modi- alloy is held above the eutectic temperature, Temper designations are defined at the end fied according to the effect of new elements the liquid will never solidify. In commercial of this article. on the solid solubility and/or the eutectic alloys, which usually are temaries or quater- melting points of the basic binary system. In naries of the major alloying elements, the Solution Heat Treating aluminum-lithium alloys, for example, mag- situation is more complex. Different phases To take advantage of the precipitation- nesium reduces the solubility of lithium in have different solvus temperatures, and non- hardening reaction, it is necessary first to aluminum. In the aluminum-copper system, equilibrium melting may occur at different produce a solid solution. The process by magnesium also lowers the eutectic melting temperatures depending on composition, size which this is accomplished is called solution point. The proximity of typical solution- of precipitates, and rate of heating. When heat treating, and its objective is to take into treating temperature ranges to eutectic new solution heat-treating equipment (which solid solution the maximum practical melting temperatures for three common alu- provides higher heating rates) is employed, amounts of the soluble hardening elements minum-copper-magnesium alloys is shown careful examination of alloy microstructures in the alloy. The process consists of soaking in the following table: should be included as part of the certification the alloy at a temperature sufficiently high process. and for a time long enough to achieve a Solution-treating Eutectlc melting Underheating. When the temperatures at- temperature temperature nearly homogeneous solid solution. tained by the parts or pieces being heat treat- Nominal commercial solution heat-treat- Alloy *C *F *C *F ed are appreciably below the normal range, ing temperature is determined by the com- 2014 496-507 925-945 510 950 solution is incomplete, and strength some- position limits of the alloy and an allowance 2017 496-507 925-945 513 955 what lower than normal is expected. In the for unintentional temperature variations. 2024 488--499 910-930 502 935 aluminum-copper system (Fig l), the shallow Although ranges normally listed allow vari- slope of the solvus at its intersection with the ations of -+6 °C (-+ l0 °F) from the nominal, Similar considerations apply to other age- composition line indicates that a slight de- some highly alloyed, controlled-toughness, hardenable alloy systems such as aluminum- crease in temperature will result in a large high-strength alloys require that tempera- magnesium-silicon alloys. For example, ac- reduction in the concentration of the solid ture be controlled within more restrictive cording to Fig 2(a), a 1.08% Mg2Si alloy solution and a correspondingly significant de- limits. Broader ranges may be allowable for would be soaked at a temperature in excess of crease in final strength. The effect of solution- alloys with greater intervals of temperature 500 °C (930 °F) but below the solidus of 595 °C treating temperature on the strength of two between their solvus and eutectic melting (1100 °F) to avoid incipient melting. Howev- aluminum alloys is illustrated by the following temperatures. er, because some alloy constituents may form data: Overheating. Care must be exercised to complex eutectics that melt at temperatures Solution- avoid exceeding the initial eutectic melting below the equilibrium eutectic temperature, treating temperature. If appreciable eutectic melting the upper limit for solution treatment of alu- temperature Tensile strength Yield strength occurs as a result of overheating, properties minum-magnesium-silicon alloys is in the °C °F MPa ksi MPa ksi such as tensile strength, ductility, and frac- range of 515 to 540 °C (960 to 1000 °F). At 540 ture toughness may be degraded. Materials °C (1000 °F), about 0.6% Mg can be placed in 6061-T6 sheet 1.6mm(0.064in.)thick that exhibit microstructural evidence of solution (Fig 2b). 493 920 301 43.7 272 39.4 overheating are generally categorized as NonequUibrium Melting. When high heat- 504 940 316 45.8 288 41.7 unacceptable by specification. Evidence of ing rates are employed, the phenomenon of 516 960 333 48.3 305 44.3 grain-boundary melting that occurs above nonequilibrium melting must be considered. 527 980 348 50.5 315 45.7 the eutectic melting temperature of the alloy This phenomenon can also be explained with 2024-T4 sheet 0.8mm(0.032 in.)thick usually is not detectable by either visual the help of the aluminum-copper phase dia- 488 910 419 60.8 255 37.0 examination or nondestructive testing. gram (Fig 1). The room-temperature micro- 491 915 422 61.2 259 37.5 493 920 433 62.8 269 39.0 Although maximum temperature must be structure of an F-temper product containing 496 925 441 63.9 271 39.3 restricted to avoid melting, the lower limit 4% Cu consists of a solid solution of copper in Heat Treating of Aluminum Alloys / 845

Table l(a) Typical solution and precipitation heat treatments for commercial heat-treatable aluminum alloy mill products with copper alloying Solution heat treatment(a) Precipitation heat treatment Metal temperature(b) Metal temperature(b) Temper Time(c), Temper Alloy Product form °C °F designation °C °F h designation

Al-Cu alloys without magnesium alloying 2011 Rolled or cold finished rod and bar 525 975 T3(d) 160 320 14 T8(d) T4 • • T45 l(e) • • 2025 Die forgings 515 960 T4 170 340 l0 T6 2219(0 Flat sheet 535 995 T31(d) 175 350 18 T81(d) T37(d) 165 325 24 T87(d) T42 190 375 36 T62 Plate 535 995 T3 l(d) 175 350 18 T81(d) T37(d) 175 350 18 T87(d) T351(e) 175 350 18 T851(e) T42 190 375 36 T62 Rolled or cold finished wire, rod, and bar 535 995 T351(e) 190 375 18 T851 (e) 2219(D Extruded rod, bar, shapes, and tube 535 995 T31(d) 190 375 18 T81 (d) T3510(e) 190 375 18 T8510(e) T351 l(e) 190 375 18 T851 l(e) T42 190 375 36 T62 Die forgings and rolled rings 535 995 T4 190 375 26 T6 Hand forgings 535 995 T4 190 375 26 T6 T352(f) 175 350 18 T8520 AI-Cu-Mg alloys 2018 Die forgings 51O(g) 950(g) T4 170 340 10 T61 2024(h) Flat sheet 495 920 T3(d) 190 375 12 T81(d) T361(d) 190 375 8 T861(d) T42 190 375 9 T62 190 375 16 T72 2024(h) Coiled sheet 495 920 T4 - • T42 190 375 9 T62 190 375 16 T72 Plate 495 920 T351 (e) 190 375 12 T851 (e) T361(d) 190 375 8 T861(d) T42 190 375 9 T62 Rolled or cold finished wire, rod, and bar 495 920 T4 190 375 12 T6 T351(e) 190 375 12 T851 (e) T36(d) 190 375 8 T86(d) T42 190 375 16 T62 Extruded rod, bar, shapes, and tube 495 920 T3 190 375 12 TSI T3510(e) 190 375 12 T8510(e) T3511 (e) 190 375 12 T851 l(e) T42 190 375 16 T62 Drawn tube 495 920 T3(d) • • T42 .. 2036 Sheet 5OO 930 T4 • • 2038 Sheet 54O 10O0 T4 205 400 2 T6 2218 Die forgings 510(g) 950(g) T4 170 340 10 T61 510(i) 950(i) T41 240 460 6 T72 AI-Cu-Mg-Si alloys 2008 Sheet 510 950 T4(d)(j) 205 400 1 T62(e) 2014(h) Flat sheet 50O 935 T3(d) 160 320 18 T62 T42 160 320 18 T6 Coiled sheet 500 935 T4 160 320 18 T6 T42 160 320 18 T62 Plate 500 935 T42 160 320 18 T62 T451(e) 160 320 18 T651(e) Rolled or cold finished wire, rod, and bar 500 935 T4 160(k) 320(k) 18 T6 T42 160(k) 320(k) 18 T62 T451 (e) 160(k) 320(k) 18 T651 (e) Extruded rod, bar, shapes, and tube 500 935 T4 160(k) 320(k) 18 T6 T42 160(k) 320(k) 18 T62 T4510(e) 160(k) 320(k) 18 T6510(e) Drawn tube 5OO 935 T4 160(k) 320(k) 18 T6 T42 160(k) 320(k) 18 T62 Die forgings 500(I) 935(1) T4 170 340 10 T6 2017 Rolled or cold finished wire, rod, and bar 5OO 935 T4 ...... T42 ...... 2117 Rolled or cold finished wire and rod 5OO 935 T4 ...... T42 ...... 2618 Forgings and rolled rings 530 985 T4 200 390 20 T61 4032 Die forgings 510(h) 950(h) T4 170 340 10 T6 AI-Cu-Li alloys 2090 Sheet 540 I00O T3(d) 165 325 24 T83(d) 2091 Sheet 530 990 T3(d) 120 250 24 T84(d) Extruded bar 530 990 T3(d) 190 375 12 Peak aged(d) 8090 Extruded bar 530 990 T3(d) 190 375 12 Peak aged(d) CP276 Extruded bar 540 10O0 T3(d) 190 375 12-15 Peak aged(d) Ca) Material should be quenched from the solution-treating temperature as rapidly as possible and with minimum delay after removal from the furnace. When material is quenched by total immersion in water; unless otherwise indicated, the water should be at room temperature, and should be suitably cooled so that it remains below 38 °C (100 °F) during the quenching cycle. Use of high-velocity, high-volume jets of cold water also is effective for some materials. (b) The nominal temperatures listed should be attained as rapidly as possible and maintained within -+6 °(2 (-+ 10 °F) of nominal during the time at temperature. (el Approximate time at temperature. The specific time will depend on the time required for the load to reach temperature. The times shown are based on rapid heating, with soak time measured from the time the load reaches a temperature within 6 °C (10 °FI of the applicable temperature. (d) Cold working subsequent to solution heat treatment and prior to any precipitation heat treatment is necessary to attain the specified properties for this temper. (e) Stress relieved by stretching to produce a specified amount of permanent set subsequent to solution heat treatment and prior to any precipitation heat treatment. (f) Stress relieved by I to 5% cold reduction after solution treatment and prior to precipitation heat treatment. (g) Quenched in water at 100 *{2 (212 °F). (h) These heat treatments also apply to alclad sheet and plate of these alloys. (i) Quenched with room-temperature air blast. (j) See U.S. Patent 4,840,852. (k) An alternative heat treatment of 8 h at 177 °C (350 *F) may also be used. (0 Quenched in water at 60 to 80 °C (140 to 180 °F). 846 / Heat Treating of Nonferrous Alloys

Table l(b) Typical solution and precipitation heat treatments for Mg-Si aluminum alloys (6xxx series alloys) Solution heat treatment(a) Precipitation heat treatment Metal temperature(b) Temper Metal temperature(b) Temper Alloy Product form °C *F designation *C *F Time(c), h designation 6005 Extruded rod, bar, shapes, 530(d) 985(d) T1 175 350 8 T5 and tube 6009(e) Sheet 555 1030 T4 205 400 1 T6(e) 6010 Sheet 565 1050 T4 205 400 1 T6(e) 6053 Die forgings 520 970 T4 170 340 10 T6 6061 (f) Sheet 530 985 T4 160 320 18 T6 T42 160 320 18 T62 Plate 530 985 T4(g) 160 320 18 T6(g) T42 160 320 18 T62 T451 (h) 160 320 18 T651(h) Rolled or cold finished 530 985 T4 160(i) 320(i) 18 T6 wire, rod, and bar 160(i) 320(i) 18 T89(j) 160(i) 320(i) 18 T93(k) 160(i) 320(i) 18 T913(k) 160(i) 320(i) 18 T94(k) T42 160(i) 3200) 18 T62 T451 (h) 160(i) 320(i) 18 T65 I(h) Extruded rod, bar, shapes, 530(d) 985(d) T4 175 350 8 T6 and tube T4510(h) 175 350 8 T6510(h) T451 l(h) 175 350 8 T651 l(h) 530 985 T42 175 350 8 T62 6061(0 Drawn tube 530 985 T4 160(i) 320(i) 18 T6 T42 160(i) 320(i) 18 T62 Die and hand forgings 530 985 T4 175 350 8 T6 Rolled rings 530 985 T4 175 350 8 T6 T452(I) 175 350 8 T652(1) 6063 Extruded rod, bar, shapes, (d) (d) T 1 205(m) 400(m) 1 T5 and tube 520(d) 970(d) T4 175(n) 350(n) 8 T6 520 970 T42 175(n) 350(n) 8 T62 Drawn tube 520 970 T4 175 350 8 T6 175 350 8 T83(j)(d) 175 350 8 T8310)(d) 175 350 8 T832(j)(d) T42 175 350 8 T62 6013(o) Sheet 570 1055 W(p) 190 375 4 T6 Plate 570 1055 W(p) 190 375 4 T651 6066 Extruded rod, bar, shapes, 530 990 T4 175 350 8 T6 and tube T42 175 350 8 T62 T4510(h) 175 350 8 T6510(h) T451 l(h) 175 350 8 T651 l(h) Drawn tube 530 990 T4 175 350 8 T6 T42 175 350 8 T62 Die forgings 530 990 T4 175 350 8 T6 6070 Extruded rod, bar, shapes, 545(d) 1015(d) T4 160 320 18 T6 and tube T42 160 320 18 T62 6111 Sheet 560 1040 T4 175 350 8 T6(q) 6151 Die forgings 515 960 T4 170 340 10 T6 Rolled rings 515 960 T4 170 340 10 T6 T452(I) 170 340 10 T652(1) 6262 Rolled or cold finished 540 1000 T4 170 340 8 T6 wire, rod, and bar 170 340 12 T9(k) T451 170 340 8 T651 (h) T42 170 340 8 T62 6262 Extruded rod, bar, shapes, 540(d) 1000(d) T4 175 350 12 T6 and tube T4510(h) 175 350 12 T6510(h) 540 1000 T42 175 350 12 T62 Drawn tube 540 1000 T4 170 340 8 T6 170 340 8 T9(k) T42 170 340 8 T62 6463 Extruded rod, bar, shapes, (d) (d) TI 205(m) 400(m) 1 T5 and tube 520(d) 970(d) T4 175(n) 350(n) 8 T6 520 970 T42 175(n) 350(n) 8 T62 6951 Sheet 53O 985 T4 160 320 18 T6 T42 160 320 18 T62 (a) Material should be quenched from the solution-treating temperature as rapidly as possible and with minimum delay after removal from the furnace. When material is quenched by total immersion in water, unless otherwise indicated, the water should be at room temperature, and should be suitably cooled so that it remains below 38 °C (100 °F) during the quenching cycle. Use of high-velocity, high-volume jets of cold water also is effective for some materials. (b) The nominal temperatures listed should be attained as rapidly as possible and maintained within -+6 °C (+- 10 °F) of nominal during the time at temperature. (c) Approximate time at temperature. The specific time will depend on the time required for the load to reach temperature. The times shown are based on rapid heating, with soak time measured from the time the load reaches a temperature within 6 °C (10 °F) of the applicable temperature. (d) By suitable control of extrusion temperature, product may be quenched directly from extrusion press to provide specified properties for this temper. Some products may be adequately quenched in room-temperature air blast. (e) Alternate heat treatments of 4 h at 190 °C (375 °F) or 8 h at 175 °C (350 °F) may also be used. See U.S. Patent 4,082,578. (f) These heat treatments also apply to alclad sheet and plate in these alloys. (g) Applicable to tread plate only. (h) Stress relieved by stretching to produce a specified amount of permanent set prior to precipitation heat treatment. (i) An alternative heat treatment of 8 h at 170 °C (340 °F) also may be used. (j) Cold working after solution treatment is necessary to attain specified properties during precipitation heat treatments. (k) Cold working after precipitation heat treatment is necessary to attain specified properties. (I) Stress relieved by 1 to 5% cold reduction subsequent to solution heat treatment and prior to precipitation heat treatment. (m) An alternative treatment of 3 h at 182 °C (360 °F) also may be used. (u) An alternative treatment of 6 h at 182 °C (360 °F) also may be used. (o) See U.S. Patent 4,589,932. (p) Two weeks of natural aging to a T4 condition. (q) Artificially aged in laboratory from T4 to T6. Heat Treating of Aluminum Alloys / 847

Table 1(c) Typical solution and precipitation heat treatments for heat-treatable Zn-Mg aluminum alloys from the 7xxx series Solution heat treatment(a) Precipitation heat treatment Metal temperature(b) Metal temperature(b) Temper Time(c), Temper Alloy Product form °C °F designation *C *F h designation 7001 Extruded rod, bar, shapes, 465 870 W 120 250 24 T6 and tube 120 250 24 T62 W510(d) 120 250 24 T6510(d) W511 (d) 120 250 24 T6511 (d)

7005 Extruded rod, bar, and ...... T53(e) shapes 7050 Plate 475 890 W51 (d) (f) (f) (f) T7651 (g) (h) (h) (h) T7451 (g) Extrusions 475 890 W510(d) (f) (f) (f) T76510(g) W51 l(d) (fl (I3 (f) T765 l l(g) Die and hand forgings 475 890 W (h) (h) (h) T74(g) W52(d) (h) (h) (h) T7452(g) 7075(i) Sheet 480 900 W 1200) 2500) 24 T6 1200) 2500) 24 T62 (f) (O (f) T76(g) (h)(k) (h)(k) (h)(k) T73(g) Plate 480 900 W 1200) 2500) 24 T62 W51(d) (h)(k) (h)(k) (h)(k) T7351(d)(g) 1200) 2500) 24 T651(d) (f) (f) (f) T7651(g) 7075(i) Rolled or cold finished 490 915 W 120 250 24 T6 wire, rod, and bar 120 250 24 T62 (h)(k) (h)(k) (h)(k) T73(g) W51 (d) 120 250 24 T651 (d) (h)(k) (h)(k) (h)(k) T7351 (d)(g) Extruded rod, bar, shapes, 465 870 W 120(1) 250(1) 24 T6 and tube 120(1) 250(1) 24 T62 (h)(k) (h)(k) (h)(k) T73(g) (f) (f) (f) T76(g) W510(d) 120(1) 250(1) 24 T6510(d) (h)(k) (h)(k) (h)(k) T73510(d)(g) (f) (0 (f) T76510(g) W511 (d) 120(1) 250(1) 24 T6511 (d) (h)(k) (h)(k) (h)(k) T73511 (d)(g) (f) if) (f) T76511 (g) Drawn tube 465 870 W 120 250 24 T6 120 250 24 T62 (h)(k) (h)(k) (h)(k) T73(g) Die forgings 470(m) 880(h) W 120 250 24 T6 (h) (h) (h) T73(g) W52(n) (h) (h) (h) T7352(n)(g) Hand forgings 470(m) 880(h) W 120 250 24 T6 (h) (h) (h) T73(g) W52(n) 120 250 24 T652(n) (h) (h) (h) T7352(n)(g) Rolled rings 470 880 W 120 250 24 T6 7175 Die forgings (o) (o) W (o) (o) (o) T66(o) (o) (o) W (o) (o) (o) T74(g)(o) (o) (o) W52(n) (o) (o) (o) T7452(n)(g)(o) Hand forgings (o) (o) W (o) (o) (o) T74(g)(o) (o) (o) W52(n) (o) (o) (o) T7452(n)(g)(o) 7475 Sheet 515(p) 960(p) W 120 250 3 plus 155 315 3 T61(p) (f) (f) (f) T761(g)(p) Plate 510(p) 950(p) W51(d) 120 250 24 T651(p) (f) (f) (f) T7651 (g)(p) (h) (h) (h) T7351(g)(p) Alclad Sheet 495 920 W 120 250 3 7475 plus 155 315 3 T61(p) (f) (f) (f) T761(g)(p) (a) Material should be quenched from the solution-treating temperature as rapidly as possible and with minimum delay after removal from the furnace. When material is quenched by total immersion in water, unless otherwise indicated, the water should be at room temperature, and should be suitably cooled so that it remains below 38 °C (100 °F) during the quenching cycle. Use of high-velocity, high-volume jets of cold water also is effective for some materials. (b) The nominal temperatures listed should be attained as rapidly as possible and maintained within -+6 °C (-+ 10 °F) of nominal during the time at temperature. (c) Approximate time at temperature. The specific time will depend on the time required for the load to reach temperature. The times shown are based on rapid heating, with soak time measured from the time the load reaches a temperature within 6 °C (10 °F) of the applicable temperature. (d) Stress relieved by stretching to produce a specified amount of permanent set after solution treatment and prior to precipitation heat treatment. (e) No solution heat treatment 72 h at room temperature following press quench followed by two-stage precipitation heat treatment comprised of 8 h at 107 °C (225 °F) plus 16 h at 149 °C (300 °F). (f) Aging practice varies with product, size, nature of equipment loading procedures and furnace-control capabilities. The optimum practice for a specific item can be ascertained only by actual trial treatment of the item under specific conditions. Typical procedures involve a two-stage treatment comprised of 3 to 30 h at 121 °C (250 °F) followed by 15 to 18 h at 163 °C (325 °F) for extrusions. An alternative two-stage treatment of 8 h at 99 °C (210 °F) followed by 24 to 28 h at 163 °C (325 °F) also may be used. (g) Aging of aluminum alloys 7050, 7075. 7175, and 7475 from any temper to the T73 or 176 temper series requires closer-than-normal controls on aging variables such as time, temperature, heatup rate, and so forth for any given item. In addition, when material in a T6-type temper is reaged to a T73- or T76-type temper, the specific condition of the T6 material (such as property levels and other effects of processing variables) is extremely important and will affect the capability of the reaged material to conform to the requirements specified for the applicable T73- or T76-type temper. (h) Two-stage treatment comprised of 6 to 8 h at 107 °C (225 °F) followed by: 24 to 30 h at 163 °c (325 °F) for sheet and plate; 8 to 10 h at 177 °C (350 °F) for rolled or cold finished rod and bar; 6 to 8 h at 177 °C (350 °F) for extrusions and tube; 8 to 10 h at 177 °C (350 °F) for forgings in the T73 temper; and 6 to 8 h at 177 °C (350 °F) for forgings in the T7352 temper. (i) These heat treatments also apply to alclad sheet and plate of these alloys. (j) An alternative two-stage treatment comprised of 4 h at 96 °C (205 °F) followed by 8 h at 157 °C (315 °F) also may be used. (k) For sheet, plate, tube, and extrusions, an alternative two-stage treatment comprised for 6 to 8 h at 107 °C (225 °F) followed by 14 to 18 h at 168 °C (335 °F) may be used, provided that a heatup rate of approximately 14 °C/h (25 OF/h) is employed. For rolled or cold finished rod and bar, the alternative treatment is 10 h at 177 °C (350 °F). (1) An alternative three-stage treatment comprised of 5 h at 99 °C (210 °F), 4 h at 121 °C (250 °F), and then 4 h at 149 °C (300 °F) may also be used. (m) Quenched in water at 60 to 80 °C (140 to 180 °F). (n) Stress relieved by I to 5% cold reduction after solution treatment and prior to precipitation heat treatments. (o) 7175-T74 and -T7452 heat treatments are directed to specific results, may vary from supplier to supplier and are either proprietary or patented. (p) Must be preceded by soak at 466 to 477 °C (870 to 890 °F). See U.S, Patent 3,79t,880, 848 / Heat Treating of Nonferrous Alloys

Table 2 Soak times and maximum quench delays for solution treatment of wrought more economical in the long run than heavi- aluminum alloys er loading, because with lighter loads heat- See Table 1 for solution-treating temperatures. ing rates are higher and fewer rejections and Soak time, minutes service failures are encountered. High-Temperature Oxidation. There is a Air furnace(b) Salt bath(c) Maximum quench condition, commonly but erroneously Thickness(a), mm (in.) min max(d) rain max(d) delay, s known as HTO or high-temperature oxida- --<0.41 (0.016) 20 25 10 15 5 tion, which can lead to deterioration of 0.51 (0.020) 20 30 10 20 7 properties in aluminum alloys. High-tem- 0.64 (0.025) 25 35 15 25 7 0.81 (0.032) 25 35 15 25 7 perature oxidation is a misnamed condition 1.02 (0.040) 30 40 20 30 10 of hydrogen diffusion that affects surface 1.27 (0.050) 30 40 20 30 10 layers during elevated-temperature treat- 1.35 (0.053) 30 40 20 30 10 ment. This condition can result from mois- 1.80 (0.071) 35 45 25 35 10 2.03 (0.080) 35 45 25 35 10 ture contamination in the furnace atmo- 2.29 (0.090) 35 45 25 35 10 sphere and is sometimes aggravated by 2.54 (0.100) 40 55 30 45 15 sulfur (as in heat-treatment furnaces also 3.18 (0.125) 40 55 30 45 15 used for magnesium alloy castings) or other 4.06 (0.160) 50 60 35 45 15 4.57 (0.180) 50 60 35 45 15 furnace refractory contamination. 6.35 (0.250) 55 65 35 45 15 Moisture in contact with aluminum at >6.35 (0.250)-12.7 (0.500) 65 75 45 55 15 high temperatures serves as a source of For each additional 12.7 0/2) or fraction +30 +30 +20 +20 (e) nascent hydrogen, which diffuses into the Rivets (all) 60 • • • 30 . • - 5 metal. Foreign materials, such as sulfur (a) Minimum dimension of thickest section. (b) Soak time begins when all pyrometer instruments recover to original operating temperature. (c) Soak time begins at time of immersion except when a heavy charge causes bath temperature to drop below specified compounds, function as decomposers of the minimum, in which case soak time begins when bath regains minimum temperature. (d) Applicable to alclad materials only. (e) Increases natural oxide surface film, eliminating it as a in thickness above 12.7 mm (½ in.) do not affect maximum quench delay, which remains constant at 15 s. barrier either between the moisture and the aluminum or between the nascent hydrogen and the aluminum. The most common man- In the tabulation above, note especially the time to be counted as soak time unless the ifestation of high-temperature oxidation is effects of small increments of temperature, bath temperature drops below the minimum surface blistering, but occasionally the only within the normal range, on the properties of the range. Even then, soak time begins as manifestations are internal discontinuities of 0.8 mm (0.032 in.) 2024-T4 sheet. soon as the bath temperature returns to the or voids, which can be detected only by Solution-Treating Time. The time at the minimum. In air furnaces, soak time does careful ultrasonic inspection or by metallo- nominal solution heat-treating temperature not begin until all furnace instruments re- graphic techniques. (soak time) required to effect a satisfactory turn to their original set temperature--that It is important to recognize that the symp- degree of solution of the undissolved or is, the temperature reading before insertion toms of high-temperature oxidation are precipitated soluble phase constituents and of the load. identical to those of unsoundness or high to achieve good homogeneity of the solid In air furnaces, thermocouples may also gas content in the original ingot or of other solution is a function of microstructure be- be attracted to, or buried in, parts located in improper mill practice. Blisters resulting fore heat treatment. This time requirement the load in such a manner as to represent from ingot defects, improper extrusion or can vary from less than a minute for thin the hottest and coldest temperatures in each improper rolling may be lined up in the sheet to as much as 20 h for large sand or zone. In this way, it is possible to ensure direction of working. However, it usually is plaster-mold castings. Guideline informa- that adequate soaking is obtained. impossible to distinguish among defect tion for soak times required for wrought Special consideration is given also to es- sources, and therefore the possibility that a products of various section thicknesses is tablishing soak times for hand and die forg- contaminated atmosphere is the cause of given in Table 2. Similar guidelines for ings; soak time in some specifications is the defects must be checked. castings are presented in Table 3. The time extended to complete solution and homog- Not all alloys and product forms are required to heat a load to the treatment enization in areas that received marginal equally vulnerable to this type of attack. temperature in furnace heat treatment also reduction during forging. Considerable vari- The 7xxx series alloys are most susceptible, increases with section thickness and fur- ation exists in the amount of soak time followed by the 2xxx alloys. Extrusions nace loading, and thus total cycle time added; some specifications call for an arbi- undoubtedly are the most susceptible form; increases with these factors. trary addition, such as one hour, and others forgings are probably second. Low-strength Soak time for alclad sheet and for parts require one hour per inch of thickness of the alloys and alclad sheet and plate are rela- made from alclad sheet must be held to a original forging. tively immune to high-temperature oxida- minimum, because excessive diffusion of In air furnaces, careful attention should tion. (Blistering of alclad material as a result alloying elements from the core into the be given to arrangement of the load. Air of inadequate bonding is not the same as the cladding reduces corrosion protection. For flow and natural temperature distribution blistering caused by high-temperature oxi- the same reason, reheat treatment of alclad within the furnace should be arranged to: dation.) sheet less than 0.75 mm (0.030 in.) thick If the protective oxide film formed during • Offer minimum resistance to air flow generally is prohibited, and the number of mill operations is removed from the mill • Produce the least disturbance in the natu- reheat treatments permitted for thicker al- product by a subsequent mechanical condi- ral temperature distribution clad sheet is limited. tioning operation such as sanding, the con- • Afford constant replenishment of the en- The soak times for wrought alloys take ditioned surface will be more susceptible to velope of air around each part into account the normal thermal lag be- high-temperature oxidation than those from tween furnace and part and the difference It is common practice to specify a minimum which the film was not removed. between surface and center temperatures spacing of 50 mm (2 in.) between parts, but Moisture can be minimized by thoroughly for commercial equipment qualified to the large complex shapes may require consider- drying parts and racks before they are standards of MIL-H-6088. The rapid heat- ably greater spacing. Many operators have charged. Drain holes often are needed in ing rates of salt baths permit all immersion found conservative loading practices to be racks of tubular construction to avoid en- Heat Treating of Aluminum Alloys / 849

Table 3 Typical heat treatments for aluminum alloy sand and permanent mold castings Solution heat treatment(b) Aging treatment Type of Temperature(c) Temperature(c) Alloy Temper casting(a) "12 *F Time, h *C *F Time, h

201.0(d) T4 S or P 490-500(e) 910-930(e) 2 ...... +525-530 +980-990 14-20 Minimum of 5 days at room temperature T6 S 51 0-515(el 950-960(e) 2 ...... +525-530 +980-990 14-20 155 3 I0 20 T7 S 51 0-515(el 950-960(e) 2 ...... +525--530 +980-990 14-20 190 370 5 T43(f) • • • 525 980 20 24 h at room temperature + l/~ to 1 h at 160 °C T71 • • - 490-500(e) 910-930(e) 2 ...... + 525--530 +980-990 14-20 200 390 4 204.0(d) T4 S or P 530 985 12 Minimum of 5 days at room temperature T4 S or P 520 970 10 ...... T6(g) S or P 530 985 12 (g) (g) • • • 206.0(d) T4 S or P 490-500(e) 910-930(e) 2 ...... +525-530 +980-990 14-20 Minimum of 5 days at room temperature T6 S or P 490-500(e) 910-930(e) 2 ...... + 525-530 +980-990 14-20 155 310 12-24 T7 S or P 490-500(e) 910-930(e) 2 ...... + 525-530 + 980-990 14-20 200 390 4 T72 S or P 490-500(e) 910-930(e) 2 ...... +525-530 +980-990 14-20 243-248 470--480 208.0 T55 S ...... 155 310 16 222.0 O(h) S ...... 315 600 3 T61 S 510 950 12 155 310 11 T551 P ...... 170 340 16--22 T65 • • • 510 950 4-12 170 340 7-9 242.0 O(i) S ...... 345 650 3 T571 S ...... 205 400 8 P ...... 165-170 330-340 22-26 T77 S 515 960 50) 330-355 625-675 2 (minimum) T61 S or P 515 960 4-120) 205-230 4(11)--450 3-5 295.0 T4 S 515 960 12 ...... T6 S 515 960 12 155 310 3--6 T62 S 515 960 12 155 310 12-24 T7 S 515 960 12 260 500 4--6 296.0 T4 P 510 950 8 ...... T6 P 510 950 8 155 310 1-8 T7 P 510 950 8 260 500 4--6 319.0 T5 S ...... 205 400 8 T6 S 505 940 12 155 310 2-5 P 505 940 4-12 155 310 2-5 328.0 T6 S 515 960 12 155 310 2-5 332.0 T5 P ...... 205 400 7-9 333.0 T5 P ...... 205 400 7-9 T6 P 505 950 6--12 155 310 2-5 T7 P 505 940 6--12 260 500 4--6 336.0 T551 P ...... 205 400 7-9 T65 P 515 960 8 205 400 7-9 354.0 • • • (k) 525-535 980-995 10-12 (h) (h) (1) 355.0 T51 S or P ...... 225 440 7-9 T6 S 525 980 12 155 310 3-5 P 525 98(~ 4-12 155 310 2-5 T62 P 525 980 4-12 170 340 14-18 T7 S 525 980 12 225 440 3-5 P 525 980 4-12 225 440 3-9 T71 S 525 980 12 245 475 4--6 P 525 980 4-12 245 475 3-.-6 C355.0 T6 S 525 980 12 155 310 3-5 T61 P 525 980 6--12 Room temperature 8 (minimum) 155 310 10-12 356.0 T51 S or P ...... 225 440 7-9 T6 S 540 1000 12 155 310 3-5 P 540 1000 4-12 155 310 2-5 T7 S 540 1000 12 205 400 3-5 P 540 1000 4-12 225 440 7-9 T71 S 540 1000 10-12 245 475 3 P 540 1000 4-12 245 475 3--6 A356.0 T6 S 540 1000 12 155 310 3-5 T61 P 540 1000 6-12 Room temperature 8 (minimum) 155 310 6--12 (continued) (a) S, sand; P, permanent mold. (b) Unless otherwise indicated, solution treating is followed by quenching in water at 65-100 °C (150-212 °F). (c) Except where ranges are given, listed temperatures are -+6 °C or -+ I0 °F. (d) Casting wall thickness, solidification rate, and grain refinement affect the solution heat-treatment cycle in alloys 201.0.204.0, and 206.0, and care must be taken in approaching the final solution temperature. Too rapid an approach can result in the occurrence of incipient melting. (el For castings with thick or other slowly solidified sections, a pre-solution heat treatment ran$ing from about 490 to 515 °C (910 to 960 *F) may be needed to avoid too rapid a temperature rise to the solution temperature and the melting of CuAI2. (f) Temper T43 for 201.0 was developed for improved impact resistance with some decrease in other mechanical properties. Typical Charpy value is 20 J (15 fl • lb). (g) The French precipitation treatment technology for the heat treatment of 204.0 alloy requires 12 h at temperature. The aging temperatures of 140, 160, or 180 °C (285, 320. or 355 OF) are selected to meet the required combination of properties. (h) Stress relieve for dimensional stability as follows: hold 5 h at 413 -+ 14 °C (775 -+ 25 °F); furnace cool to 345 °C (650 °F) over a period of 2 h or more; furnace cool to 230 *C (4.50 *F) over a period of not more than t/2 h, furnace cool to 120 °C (250 °F) over a period of approximately 2 h: cool to room temperature in still air outside the furnace. 6) No quench required: cool in still air outside the furnace. Ij) Air-blast quench from solution-treating temperature. (k) Casting process varies (sand, permanent mold, or composite) depending on desired mechanical properties. 0) Solution heat treat as indicated then artificially age by heating, uniformly at the temperature and for the time necessary to develop the desired mechanical properties. (m) Quench in water at 65-100 °C (150-212 °F) for 10-20 s only. (n) Cool to room temperature in stdl air outside the furnace. 850 / Heat Treating of Nonferrous Alloys

Table 3 (continued)

Solution heat treatment(b) Aging treatment Temperature(c) Temperature(c) Type of Alloy Temper casting(a) *C *F Time, h °C *F Time, h

357.0 T6 P 540 1000 8 175 350 6 T61 S 540 1000 10-12 155 310 10-12 A357.0 • • • (k) 540 1000 8-12 (h) (h) (h) 359.O " " " (k) 540 lO00 10-14 (h) (h) (h) A444.0 T4 P 540 1000 8-12 ...... 520.0 T4 S 430 810 18(m) ...... 535.0 T5(h) S 400 750 5 ...... 705.0 T5 S ...... Room temperature 21 days 100 210 8 ...... Room temperature 21 days 100 210 10 707.0 T5 S ...... 155 310 3-5 P ...... Room temperature, or 21 days 100 210 8 T7 S 530 990 8-16 175 350 4-10 P 530 990 4--8 175 350 4-10 710.0 T5 S • • • ...... Room temperature 21 days 711.0 T1 P • • • ...... Room temperature 21 days 712.0 T5 S • • • ...... Room temperature, or 21 days 155 315 6--8 713.0 T5 S or P . • • ...... Room temperature, or 21 days 120 250 16 771.0 T53(h) S 415(n) 775(n) 5(n) 180(n) 360(n) 4(n) T5 S ...... 180(n) 355(n) 3-5(n) T51 S • • ...... 205 405 6 T52 S - • ' ...... (h) (h) (h) T6 S 590(n) 1090(n) 6(n) 130 265 3 T71 S 5900) 10900) 60) 140 285 15 850.0 T5 S or P • - • ...... 220 430 7-9 851.0 T5 S or P • • • ...... 220 430 7-9 T6 P 480 900 6 220 430 4 852.0 T5 S or P • • • ...... 220 430 7-9 (a) S, sand; P, permanent mold. (b) Unless otherwise indicated, solution treating is followed by quenching in water at 65-100 *C (150-212 °F). (c) Except where ranges are given, listed temperatures are -+6 °C or -+ 10 °F. (d) Casting wall thickness, solidification rate, and grain refinement affect the solution heat-treatment cycle in alloys 201.0, 204.0, and 206.0, and care must be taken in approaching the final solution temperature. Too rapid an approach can result in the occurrence of incipient melting. (el For castings with thick or other slowly solidified sections, a pre-solutioo heat treatment ranl~ing from about 490 to 515 °C (910 to 960 °F) may be needed to avoid too rapid a temperature rise to the solution temperature and the melting of CuAI2. (f) Temper T43 for 201.0 was developed for improved impact resistance with some decrease in other mechanical properties. Typical Charpy value is 20 J (15 ft • lb). (g) The French precipitation treatment technology for the heat treatment of 204.0 alloy requires 12 h at temperature. The aging temperatures of 140, 160, or 180 °C (285,320, or 355 °F) are selected to meet the required combination of properties. (h) Stress relieve for dimensional stability as follows: hold 5 h at 413 -+ 14 *C (775 ± 25 °F); furnace cool to 345 °C (650 °F) over a period of 2 h or more; furnace cool to 230 °C (450 *F) over a period of not more than ½ h; furnace cool to 120 °C (250 °F) over a period of approximately 2 h; cool to room temperature in still air outside the furnace. (i) No quench required; cool in still air outside the furnace. (j) Air-blast quench from solution-treating temperature. (kl Casting process varies (sand, permanent mold, or composite) depending on desired mechanical properties, tj) Solution heat treat as indicated, then artificially age by heating uniformly at the temperature and for the time necessary to develop the desired mechanical properties. (m) Quench in water at 65-100 *C (150-212 °F) for 10-20 s only. (n) Cool to room temperature in still air outside the furnace.

trapment of water. Another common re- ods is use of a protective fluoborate com- that are not susceptible to high-temperature quirement is adjustment of the position of pound in the furnace. Such a compound oxidation, may be detrimentally affected. the quench tank with respect to furnace usually is effective in minimizing the harm- Successful use of fluoborate protective doors and air intake. Because it is unlikely ful effects of moisture and other undesirable compounds appears to depend on specify- that all moisture can be eliminated from the contaminants because it forms a barrier ing the right amount for each furnace; this atmosphere in a production heat-treating layer or film on the aluminum surface. The must be established on a trial-and-error furnace, it is extremely important to elimi- additive is not a universal solution; in some basis. One aircraft manufacturer adds 4 nate all traces of other contaminants from applications, high-temperature oxidation g/m 3 (0.004 oz/ft 3) of furnace chamber to both the parts and the furnace atmosphere. has occurred even though a fluoborate com- each load. Another adds 0.45 kg (I lb) per The most virulent contaminants in attack- pound was employed. Also, the use of such shift to a metal container hung on the fur- ing aluminum are sulfur compounds. Resi- compounds, particularly ammonium fluo- nace chamber wall, thus avoiding loss of the dues from forming or machining lubricants, or borate, may present a hazard to personnel if compound during quenching. from a sulfur dioxide protective atmosphere used in poorly sealed furnaces or in furnac- A second method of combating high-tem- used in prior heat treatment of magnesium, es that discharge their atmospheres into perature oxidation is to anodize the work are potential sources of sulfur contamination. enclosed areas. before it is heat treated. The resultant alu- In one plant, surface contamination resulted Protective fluoborate compounds accen- minum oxide film prevents attack by con- from sulfur-containing materials in tote boxes tuate staining or darkening of the parts taminants in the furnace atmosphere. The used to transport parts. In another, an epi- being treated. (At times, this attack, partic- only deterrents to the use of anodizing are demic of blistering was cured by rectifying a ularly on parts located near the protective- its cost (in money and time) and the slight "sour" degreaser. In a third instance, it was compound container during heat treatment, surface frostiness which results from the found that a vapor-degreasing operation was has been severe enough to be termed "cor- subsequent stripping operation. not completely removing a thin, hard waxy rosion.") Although this minor nuisance The usual objection to the blistered sur- residue, and an alkaline cleaning operation might be considered a small price to pay for face produced by high-temperature oxida- was added. solution of a problem of high-temperature tion is its unsightly appearance. This often Very often, the source of contamination oxidation, the residual compound in the can be improved (for salvage purposes) by is obscure and difficult to detect, and the furnace dissipates slowly. Therefore, subse- applying local pressure to flatten each blis- problem must be combated in another way. quent loads of alloys and product forms ter and then finishing by a mechanical pro- The most common of the alternative meth- whose end uses require bright surfaces, and cess such as polishing, buffing, sanding, or Heat Treating of Aluminum Alloys / 851 abrasive blasting. In general, the effect of primary fabricating mills, by progressive ficially aged tempers, and in particular the HTO on static properties and fatigue flooding or high-velocity spraying with cold copper-free 7xxx alloys, are exceptions to strength is slight. However, if a void result- water. However, parts of complex shape, this rule. The effect of quench rate on ing from HTO is located close to another often with both thin and thick sections (such mechanical properties may also depend on stress concentration, such as a hole, much as die forgings, most castings, impact extru- the desired temper. In the underaged con- greater degradation of fatigue strength is sions, and components formed from sheet) dition, for example, a slow quench rate is likely. In critical aluminum alloy forgings, are commonly quenched in a medium that more detrimental on ductility and fracture any blistering must be evaluated carefully provides somewhat slower cooling. This toughness. Strength would be more affected for its effect on the integrity of the part. Any medium may be water at 65 to 80 °C (150 to after near-to-peak aging. "cosmetic" salvage should be performed 180 °F), boiling water, an aqueous solution Because of these effects, much work has only after it has been established that the of polyalkylene glycol, or some other fluid been done over the years to understand and blisters are superficial and will not remain in medium such as forced air or mist. predict how quenching conditions and prod- the finished product. If appreciable precipitation during cool- uct form influence properties. The relative Precipitation Heat Treating without Prior ing is to be avoided, two requirements must effects of quench methods can be compared Solution Heat Treatment. Certain alloys that be satisfied. First, the time required for in terms of average quench rates. In Fig 3, are relatively insensitive to cooling rate transfer of the load from the furnace to the for example, the effects of quenching on the during quenching can be either air cooled or quenching medium must be short enough to yield strength of four alloys are compared in water quenched directly from a final hot- preclude slow precooling into the tempera- terms of average quenching rates through working operation. In either condition, ture range where very rapid precipitation the range from 400 to 290 °C (750 to 550 °F). these alloys respond strongly to precipita- takes place. For alloy 7075, this range was For alloys relatively high in sensitivity to tion heat treatment. This practice is widely determined to be 400 to 290 °C (750 to 550 quenching rate, such as 7075, rates of about used in producing thin extruded shapes of °F), and some sources quote this range (or a 300 °C/s (540 °F/s) or higher are required in alloys 6061, 6063, 6463, and 7005. Upon slightly different range) as the most critical order to obtain near-maximum strength af- precipitation heat treating after quenching range for quenching of any aluminum alloy. ter precipitation heat treatment. The other at the extrusion press, these alloys develop Later work has shown that the most critical alloys in Fig 3 maintain their strengths at strengths nearly equal to those obtained by range is alloy-dependent, and as will be cooling rates as low as about 100 °C/s (180 adding a separate solution heat treating op- discussed in detail under "Quench-Factor °F/s). Similar comparisons in terms of aver- eration. Changes in properties occurring Analysis," significant errors can result from age quench rates are shown in Tables 4 and during the precipitation treatment follow the assumption that precipitation is negligi- 5. the principles outlined in the discussion of ble outside of a so-called "critical range." Average quench rates are useful in com- solution heat-treated alloys. The second requirement for avoidance of paring experimental results from various appreciable precipitation during quenching quench methods. In Table 4, for example, a Quenching is that the volume, heat-absorption capaci- severe reduction in strength occurred at the Quenching is in many ways the most ty, and rate of flow of the quenching medi- average quench rate of 36 °C/s (65 °F/s). critical step in the sequence of heat-treating um be such that little or no precipitation However, average quench rates only com- operations. The objective of quenching is to occurs during cooling. Any interruption of pare results in a "critical" temperature preserve the solid solution formed at the the quench that might allow reheating into a range, where precipitation is most likely to solution heat-treating temperature, by rap- temperature range where rapid precipitation occur. This method is not entirely accurate, idly cooling to some lower temperature, can occur must be prohibited. because significant precipitation can also usually near room temperature. From the For maximum dimensional stability, occur outside the specified critical temper- preceding general discussion, this statement some forgings and castings are fan cooled or ature range of average quench rates. More- applies not only to retaining solute atoms in still-air cooled. In such instances, precipita- over, for high-strength alloys, toughness solution, but also to maintaining a certain tion-hardening response is limited, but sat- and corrosion resistance may be impaired minimum number of vacant lattice sites to isfactory values of strength and hardness without significant loss of tensile strength. assist in promoting the low-temperature dif- are obtained. Extrusions produced without Therefore, a more sophisticated compar- fusion required for zone formation. The separate solution heat treatment can be air ison, known as quench-factor analysis, is solute atoms that precipitate either on grain or mist quenched, but thicker sections may needed for quantitative property prediction boundaries, dispersoids, or other particles, require water quenching by immersion or or property optimization. Quench-factor as well as the vacancies that migrate (with spraying. Alloys that are relatively dilute, analysis, as discussed in a later section, is extreme rapidity) to disordered regions, are such as 6063 and 7005, are particularly well useful when cooling rates are nonuniform. irretrievably lost for practical purposes and suited to air quenching, and their mechani- Delay in Quenching. Whether the transfer fail to contribute to the subsequent cal properties are not greatly affected by its of parts from the furnace to the quench is strengthening. low cooling rate. Lower quenching rates are performed manually or mechanically, it In most instances, to avoid those types of also employed for forgings, castings, and must be completed in less than the specified precipitation that are detrimental to me- complex shapes to minimize warpage or maximum time. The maximum allowable chanical properties or to corrosion resis- other distortion and the magnitude of resid- transfer time or "quench delay" varies with tance, the solid solution formed during so- ual stresses developed as a consequence of the temperature and velocity of the ambient lution heat treatment must be quenched temperature nonuniformity from surface to air and the mass and emissivity of the parts. rapidly enough (and without interruption) to interior. From cooling curves such as those illustrat- produce a supersaturated solution at room Effect of Quench Rate on Properties. As a ed in Fig 4, maximum quench delays (see temperature---the optimum condition for broad generalization, the highest strengths table accompanying Fig 4) can be deter- precipitation hardening. The resistance to attainable and the best combinations of mined that will ensure complete immersion stress-corrosion cracking of certain copper- strength and toughness are those associated before the parts cool below 400 °C (750 °F). free aluminum-zinc-magnesium alloys, with the most rapid quenching rates. Resis- MIL-H-6088 specifies maximum quench de- however, is improved by slow quenching. tance to corrosion and stress-corrosion lays for high-strength alloys of 5, 7, 10, and Most frequently, parts are quenched by cracking are other characteristics that are 15 s for thickness ranges of up to 0.016 in. immersion in cold water or, in continuous generally improved by maximum rapidity of (0.41 mm), 0.017 to 0.031 in. (0.43 to 0.79 heat treating of sheet, plate, or extrusions in quenching. Some of the alloys used in arti- mm), 0.032 to 0.090 in. (0.81 to 2.29 mm), 852 / Heat Treating of Nonferrous Alloys

Average cooling rate stopwatch or, if necessary, by attaching Average quenching rate from750-550 °E °F/s from750-550 o~ °F/s thermocouples to parts. However, although 10 102 103 104 10 s 10 102 103 104 the cooling rate between 400 and 260 °C (750 700 I I I I I I- 100 ~. 600 I I I |7~50_T 6 I_ 80 and 500 °F) is most critical and must be , 500 --7050-T736 ~r"/~-- .E I~. ~7178-T6 600 ~ 707'5-T6 extremely high for many high-strength al- .~-----"~'~7050-T73 ' loys, it cannot be directly measured in pro- /~,~- ~ ~ 7050-T73 500 ~ "7075-T73 80 ~3OO - " /~ "2014-T6 duction operations. It is usual to rely on ~'~- 2024-T4 •'a 200 ,J6061-T6 2024-T4 "o standardized practices, augmented by re- 400 607'0-T6 60 sults of tension tests and tests of suscepti- ~. 100 I I 20 1 10 102 103 104 "~ 300 ,,'/~ 6061-T6 bility to intergranular corrosion. Average quenching rate ~" 40 Water-immersion quenching normally is from400-290 °C,°C/s 2O0 controlled in practice by stipulating maxi- (a) 1 10 102 103 104 105 mum quench-delay time and maximum wa- Average cooling rate ter temperature. The first requirement con- from 400-290 °C, °C/s trols the cooling rate during transfer and, (b) for high-strength alloys, often is~ased on the criterion of complete immersion before 100 the metal cools below 415 °C (775 °F). This o~ Furnace cooling I I specification 0f415 °C (775 °F) is based on a \ s0 / J critical temperature for alloy 7075, which == Forced air cooling == 0"-" has one of the more severe C-curves (Fig 5). Therefore, the criterion for complete im- ~ 6o mersion of other alloys might be based on a d J temperature lower than the 415 °C (775 °F) 0 N 40 7 specification, depending on the characteris- Alloy and condition (Source: Ref 1) tics of the particular C-curve. g O 8090, peak aged The second requirement controls the -~ 20 • 2090, peak aged cooling rate during immersion. MIL-H-6088 z~ 7150, aged 24 h at 120 °C • 7475, aged 24 h at 120 °C specifies that for water-immersion quench- ing, except quenching of forgings and cast- o l II 0.01 0.1 10 WO ings, the temperature of the water shall not Average cooling rate, °C/s exceed 38 °C (100 °F) upon completion of quenching. This requirement controls both It) the temperature of the quench water prior Quench sensitivity of various aluminum alloys as a function of average quench rates. (a) Yield strength to immersion and the ratio of the combined Fig 3 after aging of four wrought alloys. (b) Tensile strength after aging of eight wrought alloys. (c) Relative mass of load and rack to the volume of quench sensitivity of two aluminum-lithium alloys (2090 and 8090, both solution treated for 1 h at 520 °C, or 970 °F) and two Zn-Mg-Cu aluminum alloys (7150 and 7475, both solution treated for 40 min at 480 °C, or 895 °F) water. However, to ensure adequate quenching effectiveness, it is necessary also that the cooling fluid flow past all surfaces and over 0.090 in., respectively. Quench de- maximum delay time is permitted if tempera- of each part during the first few seconds lay is conservatively defined as commencing ture measurements of the load prove that all after immersion. Before parts enter the fur- "when the furnace door begins to open or the parts are above 415 °C (775 °F) when nace, their placement in racks or baskets first corner of a load emerges from a salt quenched. The C-curves used in quench-fac- should be compatible with this requirement. bath" and ending "when the last corner of the tor analysis can also assist in determining a During the first few seconds of quenching, load is immersed in the water quench tank." maximum allowable delay. agitation of the parts or the water should be Recommended maximum quench-delay times It is relatively easy to control quench sufficient to prevent local increases in tem- are listed in Table 2. However, exceeding the delay in day-to-day operations by using a perature due to the formation of steam pockets. In one application, it was found that Table 4 Effect of average quench rate on tensile properties of aluminum-lithium alloy 2024-T4 plates 13 by 760 by 760 mm 0/2 by 2090 30 by 30 in.), quenched singly into a large Yield Tensile volume of still water, were quite susceptible strength(a) strength(a) to intergranular corrosion. This susceptibil- Average quench rate at center of plate Condition MPa ksl MPa ksi Elongation(a), % ity disappeared completely when the 0.5 °C/s (13 mm plate, air As-quenched 162 334 2 quenching practice was modified by adding cooled) 6% stretch + aged 448 513 5 sufficient agitation to break up the insulat- 8 h at 190 °C 36 °C/s (38 mm plate, quenched As-quenched 128 312 12 ing blanket of steam that formed on the in room-temperature water) 6% stretch + aged 338 surface of the hot metal. Quenching prac- 476 6 8 h at 190 °C tices for small parts such as fasteners and 46 °C/s (13 mm plate, quenched As-quenched 138 331 16 hydraulic fittings have been modified for the in boiling water) 6% stretch + aged 530 570 9 8 h at 190 °C same reason. Dumping in bulk from baskets 48 °C/s (13 mm plate, quenched As-quenched 139 331 17 has been replaced by methods, such as the in room-temperature water) 6% stretch + aged 526 570 7 use of shaker hearth furnaces or special 8 h at 190 °C racking, which permit parts to be quenched 85 °C/s (13 mm plate, quenched As-quenched 135 349 19 in ice brine) 6% stretch + aged singly. 8 h at 190 °C 535 575 7 Spray Quenching. For spray quenching, (a) Data are averages from 4 specimens. the quench rate is controlled by the velocity of the water and by volume of water per unit Heat Treating of Aluminum Alloys / 853

Table 5 The effect of quench rate on the mechanical properties of age-hardened aluminum-lithium alloy 8090 Ultimate tensile Cooling from Yield strength(b) strength(b) Elongation in !;0 mm Alloy composition solution treatment(a) Stretch, % Aging treatment MPa ksi MPa ksi (2 in.)(b), % AI-2.28Li-0.86Cu- Air cool 2 190 °C for 16 h 380 55 446 64.5 7.7 0.90Mg-0.13Zr- (-0.25 °C/s) 0.13Fe-0.06Si 4 170 °C for 24 h 401 58 465 67.5 6.0 Polymer quench 2 190 °C for 16 h 415 60 481 70 8.0 (-18 °C/s) 4 170 °C for 24 h 415 60 481 70 7.2 Water quench 2 190 °C for 16 h 428 62 492 71.4 8.1 (- 120 °C) 4 170 °C for 24 h 417 60 483 70 7.5 AI-2.58Li-1.36Cu- Air cool 2 190 °C for 16 h 417 60 485 70.3 6.5 0.89Mg-0.13Zr- (-0.25 °C/s) 0.17Fe-0.04Si 4 170 °C for 24 h 442 64 503 73 4.5 Polymer quench 2 190 °C for 16 h 448 65 524 76 6.8 (-18 °C/s) 4 170 °C for 24 h 448 65 519 75 5.0 Water quench 2 190 °C for 16 h 464 67 535 77.5 8.2 (--120 °C/s) 4 170 °C for 24 h 448 65 517 75 6.3 (a) Solution treatment of 550 °C (1020 °F) for 1 h. (b) Data are averages from two specimens. area per unit time of impingement of the tween the quenchant and the part. These ly the heat transfer coefficient (C), which is water on the workpiece. Rate of travel of quantities are related by the equation H = related to the Grossmann number (H). The the workpiece through the sprays is an C/2k, where the coefficient of heat transfer application of polymer quenchants is cov- important variable. (C) is affected by the quenchant velocity at ered in AMS specifications 3025 and 2770, Local increases in temperature that occur the surface of the part and several inherent although many aluminum and aerospace within the first few seconds of quenching, characteristics of the quenchant (such as companies have developed internal specifi- caused by a phenomenon such as plugged quenchant boiling point, viscosity, density, cations that differ from AMS-2770. Typical spray nozzles, are particularly deleterious. thermal conductivity, and specific heat). parameters for quenching wrought products The remaining "internal heat" may be suf- Water, which is the most widely used and (other than forgings) in glycol-water solu- ficient to reheat the surface region. When effective quenching medium, can obtain tions are presented in Table 7. this happens, a large loss in strength occurs cooling rates up to about 200 °C/s (400 °F/s) OtherFactors Affecting Quench Rate. at the previously quenched surface. The at the midplane of 25 mm (1 in.) thick Quenching rates are very sensitive to the loss of strength in the affected area of a aluminum alloy plate (see the dashed line in surface condition of the parts. Lowest rates heavy part is much more severe than that Fig 7). No rates higher than those defined are observed with products having freshly caused by an inadequate quenching rate by this line have been observed, although machined or bright-etched, clean surfaces, alone. This is illustrated for 75 mm (3 in.) rates approaching them were measured with or products that have been coated with thick 7075-T62 plate in Fig 6, which com- impinging spray quenches. Lower cooling materials that decrease heat transfer. The pares, at various depths, the properties of a rates are achieved by immersion in heated presence of oxide films or stains increases plate for which quenching was interrupted water (Fig 7) or by reducing the velocity of cooling rates. Further marked changes can on one side after 3 s with those of a plate the quenchant around the part (Table 6). be effected through the application of non- that was quenched from one side only. Cooling rates can also be reduced by low- reflective coatings, which also accelerate Quench Severity and Quenchant Selec- ering surface tension or by increasing the heating (Fig 8). Surface roughness exerts a tion. Quench severity is commonly ex- stability of the vapor film around the part. similar effect; this appears related to vapor pressed in terms of an H-value (or Gross- Polymer quenchants, which retard cool- film stability. The manner in which complex mann number), where the H-value is related ing rates by the formation of films around products, such as engineered castings and to the thermal conductivity (k) of the part(s) the part, are compared with water in Table die forgings, enter the quenching medium and the coefficient of heat transfer (C) be- 6. The effective film coefficient is essential- can significantly alter the relative cooling

Time per unit thickness, s/in 0 1000 2000 3000 600 I I I I000 Thickness Maximum quench delay, s LL o mm in. Alclad Nonclad 400 ~ 800 ~Alclad *~ 600 0.41 0.016 6.4 4.4 0.51 0.020 8.0 5.5 0.64 0.025 10.0 6.8 0.81 0.032 12.8 8.8 2o0 Nooc,adJ--- i2oo 1.02 0.040 20.0 11.0

0 50 1O0 150 Time per unit thickness, s/ram Fig 4 Cooling curves for alclad and nonclad aluminum products cooled from 495 °C (920 °F) in forced air. Air temperature, 25 °C (80 °F); air velocity, 2.3 m/s (450 ft/min). Tabulated values of quench delay (maximum delay before the material being quenched has cooled below 400 °C, or 7,50 °F) were determined from cooling curves shown. 854 / Heat Treating of Nonferrous Alloys

600 1110 flow between parts, jets should not impinge directly and cause rapid localized cooling. Quenching to Minimize Residual Stress 500 930 and Warpage. Although cold-water immer- sion or flushing is most common, because it produces the most effective quench (and 400j "~~.a .I/- t "~-" ," ,"" 750 has been required by MIL-H-6088 for 2014, ~- -~" ~B ,c 2017, 2024, 2117, 7075, and 7178 alloys except forgings), it presents problems in- e 300 ~ ~~~ "' "-7. 570 volving residual stress and warpage. Residual stresses in heavy sections of E aluminum alloys originate from differential 200 ""-'- --'~7 -.. 390 thermal expansion during quenching --that is, the still-warm central material contracts, A: 7075 pulling in the already cooled outer shell. B: 2017 100 C: 6061 212 The magnitude of stresses increases with D: 6063 section size, as shown in Fig 9. The distribution pattern of residual 0 32 stresses in as-quenched parts (compression 10 100 103 in the outer layers and tension in the central "[]me, s portion) is usually desirable in service. Time-temperature-property curves at 95% of maximum tensile stress for various alloys. See the section Compressive stresses inhibit failure by fa- Fig 5 "Quench-Factor Analysis" for discussion. Source: Ref 2 tigue and stress corrosion--two mecha- nisms that initiate in the outer fibers. Un- fortunately, metal-removal operations rates at various points, thereby affecting tions, placement and spacing of parts on the required after heat treating often expose mechanical properties and residual stresses racks can be a major factor in determining material that is stressed in tension. Also, established during quenching. Similarly, the quenching rates. In immersion quench- metal-removal operations that are asym- quenching complex extruded shapes whose ing, adequate volumes of the quenching metrical (with respect to residual stresses) wall thicknesses differ widely poses special medium must be provided to prevent an cause distortion by redistributing residual problems if distortion and stresses are to be excessive temperature rise in the medium. stresses. When close-tolerance parts are minimized. In batch heat-treating opera- When jet agitation is used to induce water being fabricated, the resulting warpage can be costly and difficult to correct. Although service performance is some- Table 6 Grossmann numbers and heat transfer coefficients (C) of quenchant-to-part films times a factor, the major incentive for re- Quenchant ducing residual stress differentials has been a reduction in warpage during machining or Effective film heat transfer Temperature Velocity Grossmann coefficient (C) an improvement in shape before machining. Number Type *C *F m/s ft/min (H = CI2k) W/cm2 • K Btu/ft 2 • h • OF One approach to reducing the cooling- rate differential between surface and center Water 27 80 0.00 0 1.07 3.55 2460 is the use of a milder quenching medium-- 0.25 50 1.35 4.78 3105 0.50 100 1.55 5.14 3565 water that is hotter than that normally used Water 38 100 0.00 0 0.99 3.28 2275 or water-glycol solutions. Boiling water, 0.25 50 1.21 4.01 2785 which is the slowest quenching medium 0.50 100 1.48 4.91 3400 Water 49 120 0.00 0 1.10 3.65 2530 used for thick sections, is sometimes em- 0.25 50 1.29 4.29 2970 ployed for quenching wrought products 0.50 100 1.60 5.31 3680 even though it lowers mechanical properties Water 60 140 0.00 0 0.86 2.85 1980 and corrosion resistance. Quenching of 0.25 50 1.09 3.62 2510 0.50 100 1.33 4.41 3060 castings in boiling water, however, is stan- Water 71 160 0.00 0 0.21 0.70 485 dard practice, and is reflected in design 0.25 50 0.57 1.89 1310 allowables. 0.50 100 0.79 2.62 1815 Another approach to the minimization of Water 82 180 0.00 0 0.11 0.36 255 0.25 50 0.21 0.69 485 residual stresses that is generally successful 0.50 100 0.27 0.89 620 consists of rough machining to within 3.2 Water 93 200 0.00 0 0.06 0.20 138 mm (0.125 in.) or less of finish dimensions, 0.25 50 0.08 0.27 184 heat treating, and then finish machining. 0.50 100 0.09 0.30 207 Water 100 212 0.00 0 0.04 0J3 92 This procedure is intended to reduce the 0.25 50 0.04 0.13 92 cooling-rate differential between surface 0.50 100 0.04 0.13 92 and center by reducing thickness; other Polyalkylene glycol benefits that accrue if this technique is used (UCON A)(a) 30 85 0.00 0 0.19 0.63 429 0.25 50 0.21 0.70 475 to reduce or reverse surface tension stress- 0.50 100 0.23 0.77 529 es in finished parts are improvements in Polyvinyl strength, fatigue life, corrosion resistance, pyrrolidone and reduced probability of stress-corrosion (PVP90)(a) 30 85 0.00 0 0.44 1.49 1012 0.25 50 0.40 1.34 912 cracking. 0.50 100 0.42 1.41 966 Several factors (especially quenching (a) Polymer quenchants with concentrations of 25%. K is equal to the thermal conductivity of the aluminumalloy (7075). Source: Ref 4 warpage) sometimes preclude general use of this procedure. The thinner and less sym- Heat Treating of Aluminum Alloys / 855

Depth, in. Because of the difficulties encountered 0.5 1.0 1.5 2.0 2.5 3.0 ioo ! with quenching in cold water, milder quen- f I I I I chants have been employed. Indiscriminate use of milder quenchants can have cata- J strophic effects; however, when their use is based on sound engineering judgment and a metallurgical knowledge of the effects on the specific alloy, significant cost savings or performance improvements can be realized. { 70-- The most frequent advantage is the re- duction in costly straightening operations | and in resultant uncontrolled residual 60L © ~trol specimen stresses. For example, one aircraft manu- • Q~ inched from side A only facturer utilizes water-spray and air-blast +=nched from side B, interrupted eCter 3 s quenching for weldments and complex 5o L I I I l \ formed parts made from 6061, an alloy 0l 10 20 30 40 50 60 70 80 whose corrosion resistance is insensitive to Side A Depth, mm Side B quenching rate. Straightening requirements Depth, in. are negligible and, through careful control 0 0.5 1.0 1.5 2.0 2.5 3.0 of racking and coolant flow, the decrease in ~E 6oo mechanical properties is minimized, as 80 = shown by the data in Fig 10. 70 _~ Another development for reducing straight- ening costs is quenching in water-polymer e0 ~ solutions. Quenching of formed sheet-metal 4OO parts in aqueous solutions of polyalkylene ? 50 ~ glycol or in similar inversely soluble media 30o has significantly reduced the cost of straight- 0 10 20 30 40 50 60 70 8(1 ening these parts after quenching. The SAE Side A Depth, mm Side B heat-treatment specification AMS-2770 rec- Depth, in. ommends, for several alloys, maximum thick- O 0.5 1.0 1.5 2.0 2.5 3.0 nesses that can be quenched in solutions of 550 I i I I I 1 specific concentrations while maintaining ac- ceptable property levels. Typical parameters 450 --- ~" for quenching wrought products (other than forgings) in glycol-water solutions are pre- I sented in Table 7. Additional information on z~ 350 -- ! " 50 "~ polymer quenchants for aluminum alloys can ; ' y___. be found in Ref 5. 4o ~ Forming and Straightening after Quench- 250 ing. Immediately after being quenched, 3 ~ ,3o 3 most aluminum alloys are nearly as ductile as they are in the annealed condition. Con- 150 0 10 20 30 40 50 60 70 80 sequently, it is often advantageous to form Side A Depth, mm Side B or straighten parts in this temper. More- over, at the mill level, controlled mechani- Fig 6 Through-thickness75mm (3 in.) thick property variations due to quench rate and temperature-rise effects in 7075-T62 plate cal deformation is the most common meth- od of reducing residual quenching stresses. Because precipitation hardening will occur at room temperature, forming or straighten- metrical a section, the more it will warp resultant inconsistent warpage usually re- ing usually follows as soon after quenching during quenching, and the residual stresses quires costly hand straightening. Conse- as possible. In addition, maximum effec- resulting from straightening of warped parts quently, a significant amount of effort has tiveness in stress relief is obtained by work- (plus straightening costs) often are less de- been devoted to reducing or eliminating ing the metal immediately after quenching. sirable than the quenching stresses. Holding warpage by changing racking positions to Forming and straightening operations fixtures and die quenching may be helpful, achieve symmetry of cooling. vary in degree from minor corrections of but precautions must be taken to ensure For sheet-metal parts, one manufacturer warpage to complete forming of complex that they do not retard quenching rates uses a double screen floor in the quenching parts from solution-treated flat blanks. Par- excessively. Other factors that must be con- rack to reduce the force of initial contact ticular value is gained when enough forming sidered are the availability of heat-treating between water and parts. Others allow parts can be done at this stage of processing to facilities and whether or not the advantages to "free fall" from rack to quench tank. eliminate the distortion caused by quench- of such a manufacturing sequence offset the Spacing and positioning on the rack are ing. However, production operations must delay and cost entailed in a double-machin- carefully controlled so that parts will enter be adjusted so that most of the plastic ing setup. the water with minimum impact. With this deformation is accomplished before an ap- Warpage of thin sections during quench- technique, water turbulences must be preciable amount of precipitation hardening ing is also a problem. Even in the same avoided, because it will often cause parts to takes place. load, symmetry of cooling usually varies float for a few seconds, greatly reducing Although the most severe forming opera- significantly among identical parts and the their cooling rate. tions may have to be arranged to avoid 856 / Heat Treating of Nonferrous Alloys

Average cooling rate at 400-290 °C, °C/s precipitation rate is low despite the high 0.1 1 10 100 103 degree of supersaturation. At intermediate , , 10 temperatures, precipitation rate is highest. Consequently, times to produce equal ~ Computed maximum amounts of precipitation follow a C-shape (assumes instantaneous "~ cooling of surface from pattern. Using isothermal quenching techniques, Fink and Willey pioneered the attempts to \ \ describe the effects of quench rates with the use of C-curves (Ref 6). The C-curves plot E " the time required at different temperatures E to precipitate a sufficient amount of solute to: reduce strength by a certain amount (Fig 7.5 ~" 5); cause a change in the corrosion behavior from pitting to intergranular (Fig 12); pro- t-- duce a given electrical conductivity (Fig 0.1 13); or relate other properties, such as frac- 2.5 o ~ ture toughness, to isothermal quench con- ditions. The nose of the C-curves identifies 150 °F 200 °F the critical temperature range (the region of 0.75 highest precipitation rates). Investigators Immersion in water at use critical temperature ranges in conjunc- indicated temperature tion with properties of samples quenched 0.25 I 0.01 continuously from the solution temperature 0.1 1 10 100 103 104 to compare relative sensitivities of alloys to Average cooling rate at 750-550 °F, °F/s quenching condition. Although average quench rates through a Effects of thickness and quenching medium on average cooling rates at midplane of aluminum alloy critical temperature range can provide rea- Fig 7 sheet and plate quenched from solution temperatures. The dashed line delineates the maximum cooling sonable property predictions if cooling rates rates theoretically obtainable at the midplane of plate, assuming an infinite heat transfer coefficient (c") and a diffusivity factor of 1400 cm2/s. Source: Ref 3 are fairly uniform, average quench rates can- not provide quantitative predictions when cooling rates vary considerably during the natural aging, it often is desirable to allow ciously for parts that are critical in fatigue quench. For such instances, a procedure some natural aging to occur and thus avoid (Fig 11) or stress corrosion. known as "quench-factor analysis" uses in- formation of Ltiders lines. This condition of Re-solution heat treatment of parts formation from the entire C-curve to predict nonuniform deformation is most likely to formed after quenching often causes exces- how any quench curve affects properties. occur shortly after quenching and diminish- sive grain growth in critically strained re- Quench-factor analysis is useful in designing es significantly after a few hours of natural gions and thus is not recommended. suitable limits for quench delays, or when it is aging. Complete freedom from Ltiders not sufficient just to ensure that the cooling lines, however, may require one or two Quench-Factor Analysis curve misses the nose of the C-curve. days of natural aging prior to forming. Thus, During the quenching of alloys from a The method of quench-factor analysis, as the forming operation may have to be timed solid-solution temperature condition, the outlined by Evancho and Staley (Ref 7), is so as to obtain the most appropriate trade- rate of precipitation during quenching is based on the determination of a quench off of these characteristics for the specific maximized in a so-called "critical" temper- factor (r), which is the major variable in the parts involved. Ltiders lines also can be ature range, because the diffusion of dis- following equation for precipitation kinetics reduced by employing low strain rates or by solved species and the subsequent nucle- during continuous cooling: forming at temperatures of 150 to 175 °C ation of precipitates exhibit opposite ~=1- exp (k-r) (Eq 1) (300 to 350 °F). behavior as a function of temperature. At Residual stresses in sheet-metal parts high temperatures, nucleation rates are where ~ is the fraction transformed and k is formed in the quenched condition are higher small because of the low degree of super- a constant related to the transformation than those in parts formed in the annealed saturation, and so precipitation rates are fraction of a given C-curve. The quench condition. Consequently, forming in the low despite the high diffusion rates. At low factor (r) is defined as: quenched condition should be selected judi- temperatures, diffusion rate is low, and thus "r= (Eq 2)

Table 7 Limits for quenching in glycol-water solutions where t is time and Ct is critical time as a Data are for wrought aluminum alloy products other than forgings. function of temperature to transform a spec- ified fraction (x). The locus of critical times Glycol Maximum thickness concentration, for a given transformation fraction x (or a vol % Alloys in. percentage of mechanical properties from 12-16 2014, 2017, 2117, 2024, 2219 2.03 0.080 precipitation) is the C-curve, and the value 7075, 7175 25.4 1.000 ofk is related to x as follows: k = ln(l - x), 17-22 2014, 2017, 2117, 2024, 2219 1.80 0.071 7075, 7079, 7175, 7178,6061 12.7 0.500 or e k = I - x. Therefore, when "r = 1, the 23-28 2014, 2017, 2117, 2024, 2219 1.60 0.063 fraction transformed, 4, equals the fraction 7075, 7079, 7175, 7178,6061 9.53 0.375 value designated by the C-curve. Equation 29-34 2014, 2017, 2117, 2024, 2219 1.02 0.040 2 is based on the assumption that the reac- 7075, 7079, 7175, 7178,6061 6.35 0.250 35-40 7075, 7079, 7175, 7178,6061 2.03 0.080 tion rate is a function only of the amount transformed and temperature. Heat Treating of Aluminum Alloys / 857

500 The numerical evaluation of the quench factor involves the integration of Eq 2. This 20 °C water quench 800 integral can be graphically integrated using ? 400 the method illustrated in Fig 14. Examples of the way to use the quench factor (r) in the analysis of quench methods are described 600 below. Neither the average quenching rate 300 I through a critical temperature range nor quench-factor analysis can predict strength when the temperature increases during 200 400 E quenching after it is cooled below some critical temperature. Under this condition, strength in the affected areas can be signif- 100 icantly lower than in other areas of the 200 material. The most likely way for this phe- Black oxide As-rolled Sanded etched coating surface surface nomenon to occur is during spray quench- 0 ing, when the surface cools rapidly by the 0 2 3 4 6 impinging spray, but reheats by heat flow Time, s from the hotter interior when the spray is (a} interrupted. Predicting Strengths of Thick Products. 500 Effects of the quenching rate on alloy I I strengths can be represented on a general- Boiling water quench ized graph of the type shown in Fig 3, and -- 800 the expected quenching rates of products ,oo having various dimensions can be deter- mined from Fig 7. Nevertheless, combining these two kinds of information to predict ? I 300 / ~ ~~ 600 mechanical properties must be done with etched caution. Inconsistencies were encountered, rface for example, in correlating properties of E 200 400 E thick sections quenched in high-cooling-rate media with properties of thinner sections etchedcoating quenched in media affording milder quench- ~ ing action. One of the reasons for the incon- 100 \_ 200 sistencies is believed to be the different Powdered oxide As-rolled Sanded sprayed coating shapes of the cooling curves. This difficulty surface surface can be overcome by using quench-factor 0 I ,,, I I analysis. The other reason is that the degree 10 20 30 40 50 60 70 80 of recrystallization and texture of the thick Time, s and thin sections may be different. (b) Predicting Corrosion Behavior. Alloy 2024-T4, for example, is susceptible to in- Effect of surface conditions on the midplane cooling of a 13 mm (0.5 in.) thick plate of 7075 from Fig 8 quenching in (a) 20 °C (70 °F) water and (b) boiling water. Source: Ref 5 tergranular corrosion when a critical amount of solute is precipitated during quenching, but will corrode in the less se- vere pitting mode when lesser amounts are precipitated. For predicting the effects of proposed quenching conditions on the cor- Cross section of solid cylindrical Cross section of solid cylindrical rosion characteristics of 2024-T4, the postu- specimen, in. 2 specimen, in. 2 lated quench curve is drawn and the quench O 10 20 30 4Q 0 10 20 30 40 100 I I I l factor is calculated using the C-curve in Fig 12. Corrosion characteristics are predicted t ' Quenched - 1o g ._ g Quenched - 10 tg from the plot in Fig 15. When the quench 50 i:ab:iling -- - ~ 50 -- in cold __ factor (r) is less than 1.0, continuously water ~_ 5 quenched 2024-T4 will corrode by pitting. These relationships are applied to studies 0 "~ i"~ - o ~ ~ o \ 0 of effects of proposed changes in quench practice on design of new quenching sys- tems. For example, consider that the goal of 3 o 5O =s ~3 3 ~s0 o o J a proposed quenching system for 2024-T4 sheet products is to minimize warpage while I00 I '°i i,oo preventing susceptibility to intergranular 0 10 20 30 20 30 corrosion. Warpage occurs when the stress- Cross section of solid cylindrical Cross section of solid cylindrical es imposed by temperature differences specimen, 10 3 mm 2 specimen 10 3 mm 2 across the parts exceed the flow stress. As quenching rate decreases, the tendency for Fig 9 Effect of quenching from 540 °C (1000 °F) on residual stresses in solid cylinders of alloy 6151 large differences in temperature to occur 858 / Heat Treating of Nonferrous Alloys

Thickness, 0.001 in. 400 100 150 200 250 t00 Curve Bend radius Condition during flattening

Water \ 1 Not bent Not applicable -- 50 7spray •x\ ~\\ 2 3.2 mm Annealed 9O \ ~\ 3 3.2 mm As-quenched + 3 days storage \ 300 ~ 4 3.2 mm As-quenched + 14 days storage o ~o 5 1.6 mm As-quenched + 3 days storage 80 ,~nm~ ~~ -- 40 ~u;

\ last E "~ 70 E "5 ~ 200 - 30 6O

6061-T6sheet Stress ratio, O.1 ~.~ ~"~'~ ~, "~ ~~ 1

I ~_ "~'~'- ,-- ~ -- ~2 -- 20 1250 2500 3750 5000 6250 7500 Thickness, #m 100 Thickness, 0.001 in. 0.01 0.1 10 100 150 200 250 Millions of cycles to failure 100 I Fatigue characteristics of 1 mm (0.04 in.) alclad 2024-T4 sheet after 90 ° bending in the annealed ~Water spray Fig 11 condition and subsequent flattening as indicated. Flattening (unbending) was done either in the 90 annealed condition (curve 2), or in the solution-treated and quenched condition (curves 3, 4, 5) with indicated ~ storage times at -18 to -12 °C (0 to 10 °F). 80 ir blast able corrosion behavior in 2024-T4 sheet Other curves could be drawn, of course, E ~ (quench factor, 0.99) were calculated. Some but the important points are that air-blast E of these curves are plotted in Fig 16. This quenching cannot be continued to more "S 70 \ illustration shows that 2024 can be than a few degrees below 395 °C (740 °F) 6061-T6 sheet quenched at a rate of 470 °C/s (850 °F/s) or and cannot be initiated at more than a few 60 1 higher and still develop acceptable corro- degrees above 270 °C (520 °F) even if infi- 1250 2500 3750 5000 6250 7500 sion characteristics if the quenching rate is nite quenching rates are attained from 395 Thickness, /~m linear from the solution temperature to 150 to 270 °C (740 to 520 °F). °C (300 °F). If sheet 3.2 mm (0.125 in.) thick Predicting yield strength is more complex Effect of quenching medium on strength of is air-blast quenched (rate of heat removal, than predicting corrosion behavior and re- Fig 10 6061-T6 sheet. Water-immersion quench equals 100%. Control of coolant flow will minimize 5.68 W/m 2. °C) to 395 °C (740 °F), however, quires some knowledge of the relationship decrease in mechanical properties. the quenching rate from 395 to 150 °C must between extent of precipitation and loss in be at least 945 °C/s (1700 °F/s) to maintain ability to develop property. Because attain- the acceptable corrosion behavior. It may able strength of precipitation-hardening alu- decreases but the tendency for intergranular also be air-blast quenched to 395 °C (740 minum alloys is a function of the amount of corrosion to occur increases. °F), spray quenched at 3300 °C/s (6000 °F/s) solute remaining in solid solution after The C-curve in Fig 12 indicates that to 250 °C (480 °F), then air-blast quenched quenching, relationships between strength quenching rate can be decreased near the to 150 °C (300 °F). (,rx) attainable after continuous cooling and solution heat-treating temperature and near room temperature without greatly sacrific- ing corrosion characteristics, but this infor- 700 425 mation does not provide a quantitative an- IACS swer. Simple calculations, however, can 18% 1 17% reveal a multitude of hypothetical cooling 650 375 curves that provide slow quenching during a v ? large portion of the quench cycle but suffi- I 19% ciently rapid quenching where critical times 600 20% 325 are short so that desirable corrosion char- f f acteristics are obtained. E E As an example, one-, two-, and three-step 21% quench curves that would ensure accept- == 550 275 == -5 -1- 500 ....._.._~.._.~ ~, --,900 ~. 500 225 400 ~ Predominantly_ 700 17% 21% intergranular 22% 300 __co.o ,o soo / 7 2y / ~- Predom ~ 450 175 E 200 pitting --300 19.~- 10 100 103 104 105 106 100 0.1 1 10 100 103 Holding time, s Critical time, s Change in electrical conductivity of an AI-2.5% Li binary alloy after the following: solution treated at 540 C-curve indicating type of corrosion attack Fig 1 3 °c (1000 °F) for 1/2 h, immersed into an adjacent salt or oil bath for the appropriate isothermal holding Fig 1 2 on 2024-T4 sheet temperature and time, then quenched into water. Source: Ref 1 Heat Treating of Aluminum Alloys / 859

Quench curve C-curve 500 -- AirbT~t0°Jch q ~00 \~/3.2-ram (0.125-in.) sheet I 11 ~on both sides ~ 800 12 (11 + "1"2)12 L) 400 ~]470°C/s(850°F/s)~700 u_ (T2 + 7"3)12 o o T~ 1~.~2780 °C/s (5000 °F/s} 2 | I Ab blast quench - 600 I I ~ 300 I II 3.2-mm(0.125-in.) t f E I"~_ /1~ sheet /1 500 E E E I I TF1 I--- I 200 I/L":l'-'-- 400 TF I I- - T (TF_ 1 + TF)/2 {330 °C/s ~. 945 °C/s I I I (6000 OF/s) (w00OF/s) - 300 I1%--AtA 2 I--q-z&tF_l1--4- I p I 100 ~ I Ill I I ..... I I l 0 2 4 tlt2t 3 tF_l tF C 2 C 1 OF_ 1 ~lme, s Elapsed time Critical time Quench curves for 2024-T4 sheet, to elimi- At1 At2 AtF 1 Fig 1 6 nate susceptibility to intergranular corrosion ~=~+~+'"+ c~,

Fig 14 Method of determining quench factor, r, using a cooling curve and a C-curve The advantage of using the quench factor for predicting yield strength from cooling 0.20 curves is apparent. Cooling curves that Quench have long holding times either above or o Air-water-air below the critical temperature range from • Water.air 0.15 - 0006 400 to 290 °C (750 to 550 °F) cannot be used E ~Air-water E • Water n~ to predict yield strength from average E] Air o~,. .oo" o~ t~ ~ quenching rate. In such instances, predic- -~ 0.10 0.004 tion of yield strength on the basis of quench factor is particularly advantageous. o_ ,~ < With the use of finite-element analysis, 005 i'~ o 0.002 quench factors can also be plotted as a function of Grossmann quench severity val- ues (H) or the heat transfer coefficients (C) I J_ I I I I I I 0.1 0.2 0.4 0.6 0.8 1.0 2.0 4.0 6.0 8.0 10.0 20.0 between the quenchant and a particular part (Fig 20). However, an underlying assump- Quench factor, T tion of both quench-factor analysis and av- erage-cooling-rate estimation is that the p[ .... I P Pitting -- ' I .... -~ o,ll~wT,~l-F~,~ P + SI - Pitting and slight intergranular t only effect of temperature is on the kinetics of precipitation. This assumption is not val- P~S , - }nterg .... }ar '~ ii ~ iiiiii.li I _ _~& P + I- Pitting and interg .... lar t id, however, when portions of the metal are quenched locally but reheated significantly ~- I olo before quenching is complete. I Age Hardening 0.1 0.2 0.4 0.6 0.8 1.0 2.0 4.0 6.0 8.0 10.0 20.0 Quench factor i' After solution treatment and quenching, hardening is achieved either at room tem- Fig 15 Type and depth of attack on 2024-T4 sheet versus quench factor perature (natural aging) or with a precipita- tion heat treatment (artificial aging). In some alloys, sufficient precipitation occurs quench factor (r) can be expressed as fol- following comparison. Four specimens of in a few days at room temperature to yield lows: alloy 7075-T6 quenched by various means stable products with properties that are (see Fig 17) were selected. Yield strengths adequate for many applications. These al- (rx = ~rmax exp (ka'r) (Eq 3) were predicted both from average quench- loys sometimes are precipitation heat treat- where O'max is the strength attainable with ing rate between 400 and 290 °C (750 and ed to provide increased strength and hard- an infinite quenching rate and: 550 °F) and from quench factor. Quench ness in wrought or cast products. Other dt factor was calculated using the C-curve for alloys with slow precipitation reactions at T= J ~xx (Eq 4) 99.5% maximum yield strength for 7075-T6 room temperature are always precipitation (Fig 18), and yield strength was estimated heat treated before being used. where t is time and Cx is the C-curve for from the above equation defining the In some alloys, notably those of the 2xxx (r~----that is, critical time as a function of quench factor (~) (see Fig 19). series, cold working of freshly quenched temperature to reduce attainable strength to A comparison of predicted yield strength material greatly increases its response to x of (rm,x. The constant k, is related to the with actual yield strength is given in Table later precipitation heat treatment. Mills take natural logarithm of x. For example, if ~ is 8. Yield strengths predicted from quench advantage of this phenomenon by applying based on the C-curve for 99.5% of maxi- factor agree very well with measured yield a controlled amount of rolling (sheet and mum yield strength, then kl = -0.005013 = strengths for all specimens, the maximum plate) or stretching (extrusion, bar, and In (0.995). error being 19.3 MPa (2.8 ksi). Yield plate) to produce higher mechanical proper- The advantage of predicting yield strengths predicted from average quenching ties. However, if the higher properties are strength from quench factor instead of from rates, however, differ from measured val- used in design, reheat treatment must be average quenching rate is illustrated by the ues by as much as 226 MPa (32.8 ksi). avoided. 860 / Heat Treating of Nonferrous Alloys

500 ...... I I I I I I °F) + cold-water quench -- 900 500 900 800

o 700 LL 400 o 400 I ~NN~ ~n!boiling walter to ~ o c 700 ~ 600 = 300 2 500 ~_ 500 E I- | ~ F6 - Quenched in denatured alcohol to 290 °C 200 t- 40O 200 ~ I [ [ (550 °F) 4t cold-water quench 300 / 300 100 Cold water que ch 0.1 10 100 1000 10 000

,00 i CHtical time, s 0 5 10 15 20 25 30 35 Time, s

Fig 17 Cooling curves for 7075-T6 sheet Fig 18 C-curve for 99.5% maximum yield strength of 7075-T6 sheet

Natural Aging. The more highly alloyed duration of changes in tensile yield strength Unanticipated difficulties may arise as a members of the 6xxx wrought series, the of representative alloys of the three types result of failure to control refrigerator or copper-containing alloys of the 7xxx group, are illustrated in Fig 21. Because of the part temperature closely enough. If opening and all of the 2xxx alloys are almost always relative instability of the 7xxx alloys, the of the cold box to insert or remove parts is solution heat treated and quenched. For naturally aged temper (after solution heat done too frequently, the cooling capacity of some of these alloys--particularly the 2xxx treatment and quenching) is designated by the refrigerator may be exceeded. At times, alloys--the precipitation hardening that re- the suffix letter W. For a specific descrip- the rate at which heavy-gage parts can be suits from natural aging alone produces tion of this condition, the time of natural cooled in a still-air cold box has been found useful tempers (T3 and T4 types) that are aging should be included (example: 7075-W, to be insufficient. This problem has been characterized by high ratios of tensile to 1 month). solved in one plant by immersing parts in a yield strength and high fracture toughness Aging characteristics vary from alloy to solvent at -40 °C (-40 °F) before placing and resistance to fatigue. For the alloys that alloy with respect to both time to initial them in the refrigerator. are used in these tempers, the relatively change in mechanical properties and rate of The T3-type tempers are distinguished high supersaturation of atoms and vacan- change, but aging effects always are less- from T4-type tempers by significant me- cies retained by rapid quenching causes ened by reductions in aging temperature chanical-property differences resulting from rapid formation of GP zones, and strength (see Fig 21). With some alloys, aging can be cold work strain hardening associated with increases rapidly, attaining nearly maxi- suppressed or delayed for several days by certain mechanical operations performed mum stable values in four or five days. holding at a temperature of -18 °C (0 °F) or after quenching. Roller or stretcher leveling Tensile-property specifications for products lower. It is usual practice to complete form- to achieve flatness or straightness introduc- in T3- and T4-type tempers are based on a ing and straightening before aging changes es modest strains (on the order of ! to 4%) nominal natural aging time of four days. In mechanical properties appreciably. When that cause changes in mechanical properties alloys for which T3- or T4-type tempers are scheduling makes this impractical, aging (primarily, increases in strength). Further standard, the changes that occur on further may be avoided in some alloys by refriger- increases in strength can be obtained by natural aging are of relatively minor magni- ating prior to forming. It is conventional cold rolling, additional stretching, combina- tude, and products of these combinations of practice to refrigerate alloy 2024-T4 rivets tions of these operations, or for products alloy and temper are regarded as essentially to maintain good driving characteristics. such as hand forgings, compressive defor- stable after about one week. Full-size wing plates for current-generation mation. The tempers produced by these In contrast to the relatively stable condi- jet aircraft have been solution heat treated operations followed by natural aging alone tion reached in a few days by 2xxx alloys and quenched at the primary fabricating (no precipitation heat treatment) are classi- that are used in T3- or T4-type tempers, the mill, packed in dry ice in specially designed fied as T3-type tempers, and an additional 6xxx alloys and to an even greater degree insulated shipping containers and transport- digit is used to indicate a variation in strain the 7xxx alloys are considerably less stable ed by rail about 2000 miles to the aircraft hardening that results in significant changes at room temperature and continue to exhibit manufacturer's plant for forming. in properties. In the most recently intro- significant changes in mechanical properties for many years. The differences in rate and Table 8 Yield-strength values for 7075-T6 sheet predicted from cooling curves using average quench rate and quench factor Average quench Yield strength rate from 400 to predicted from Yield strength 290 °C (750 to Measured yield average quench predicted from 100 550 OF) Quench strength rate quench factor factor, 80 \ Quench °C/s °F/s T MPa ksi MPa ksi MPa ksi '~- 60 -- 100 ~/~ma×" exp(-O.O05013 "~) - Cold water 935 1680 0.464 506 73.4 499 72.4 498 72.3 E_ 40 Denatured alcohol to E 290 °C (550 °F), then cold water 50 90 8.539 476 69.1 463 67.2 478 69.4 ~ 20 Boiling water to ~ o 315 °C (600 °F), 0.01 0.1 1 10 100 103 then cold water 30 55 15.327 458 66.4 443 64.2 463 67.1 Quench factor, Still air to 370 °C (700 °F), then cold water 5 9 21.334 468 67.9 242 35.1 449 65.1 Fig 19 Yield strength versus quench factor Heat Treating of Aluminum Alloys / 861

Sheet thickness, in. Plate thickness, in. 0 0.05 0.1 0.15 0.2 0.25 0.3 0 0.5 1 1.5 2 2.5 3 50 50

,5 2- ,5 -- 2-- ~- 40 t- 40 / //,~v~//~~

=~25 ~ =~25

0 'l ~.b. 0 0 1,25 2.5 3.75 5 6.25 7.5 0 12.5 25 37.5 50 62,5 75 Sheet thickness, mm Plate thickness, mm (a) (b) Plot of quench factors derived from finite element analysis with given product sizes and film (heat transfer) coefficients (C). Heat transfer coefficients between Fig 20 the quenchant and part are expressed in W/cm 2 • K. Source: Ref 4 duced 2xxx aircraft alloy, 2324, high wrought alloys and 2xx.O series casting al- tures of the precipitation-hardening process strength is achieved by cold rolling plate to loys. This contrast is noticeably decreased reduce the probability of obtaining the re- a T39 temper. by precipitation heat treatment. quired properties. Precipitation heat treatments generally Differences in type, volume fraction, T6 and T7 Tempers. Precipitation heat are low-temperature, long-term processes. size, and distribution of the precipitated treatment following solution heat treatment Temperatures range from 115 to 190 °C (240 particles govern properties as well as the and quenching produces T6- and T7-type to 375 °F); times vary from 5 to 48 h. changes observed with time and tempera- tempers. Alloys in T6-type tempers gener- Choice of time-temperature cycles for ture, and these are all affected by the initial ally have the highest strengths practical precipitation heat treatment should receive state of the structure. The initial structure without sacrifice of the minimum levels of careful consideration. Larger particles of may vary in wrought products from unre- other properties and characteristics found precipitate result from longer times and crystallized to recrystallized and may ex- by experience to be satisfactory and useful higher temperatures; however, the larger hibit only modest strain from quenching or for engineering applications. Alloys in T7 particles must, of necessity, be fewer in additional strain from cold working after tempers are overaged, which means that number with greater distances between solution heat treatment. These conditions, some degree of strength has been sacrificed them. The objective is to select the cycle as well as the time and temperature of or "traded off" to improve one or more that produces optimum precipitate size and precipitation heat treatment, affect the final other characteristics. Strength may be sac- distribution pattern. Unfortunately, the cy- structure and the resulting mechanical prop- rificed to improve dimensional stability, cle required to maximize one property, such erties. particularly in products intended for service as tensile strength, is usually different from Because mechanical properties and other at elevated temperatures, or to lower resid- that required to maximize others, such as characteristics change continuously with ual stresses in order to reduce warpage or yield strength and corrosion resistance. time and with temperature, as shown in Fig distortion in machining. T7-type tempers Consequently, the cycles used represent 22(a), (b), and (c) by typical curves for three frequently are specified for cast or forged compromises that provide the best combi- wrought alloys, treatment to produce a engine parts. Precipitation heat-treating nations of properties. combination of properties corresponding to temperatures used to produce these tem- Production of material in T5- through a specific alloy-temper combination re- pers generally are higher than those used to Tl0-type tempers (see the section on tem- quires one or more rather specific and co- produce T6-type tempers in the same al- per designations near the end of this article) ordinated combinations of time and temper- loys. necessitates precipitation heat treating at ature, with both parameters being subject to Two important groups of T7-type elevated temperatures (artificial aging). Al- practical limitations. Recommended com- tempers---the T73 and T76 types--have though the hardening precipitate developed mercial treatments often are compromises been developed for the wrought alloys of by this operation is submicroscopic, struc- between time and cost factors and the prob- the 7xxx series, which contain more than tures before and after precipitation heat ability of obtaining the intended properties, about 1.25% copper. These tempers are treatment often can be distinguished by with consideration of allowances for varia- intended to improve resistance to exfolia- etching metallographic specimens. In alumi- bles such as composition within specified tion corrosion and stress-corrosion crack- num alloys in the solution heat treated and range and temperature variations within the ing, but as a result of overaging, they also quenched condition, coloration contrast be- furnace and load. Use of higher tempera- increase fracture toughness and, under tween grains of differing orientation is rela- tures may reduce treatment time; but if the some conditions, reduce rates of fatigue- tively high, particularly in 2xxx series temperature is too high, characteristic fea- crack propagation. The T73-type temper 862 / Heat Treating of Nonferrous Alloys

2014 2024 600 600 sition. If first-step aging time is too short, 80 if first-step aging temperature is too far below the GP-zone solvus, or if heating 500 500 rates are too high, the GP zones will 7O RT 7O 2 RT.....---- ~= ~ , f dissolve above 150 °C (300 °F), and the 60 #/ 0 °C (32 F) 60 ~' resultant coarse and widely distributed 400 I # 0 °C (32 °F)-- - 400 / / == i / ..--- -18 ~C (0 ~F) precipitate will provide lower strength. 50 50 The T76-type treatments have the same .... J -18°C(0°F) 300 300 operational sequence but employ second- stage heating only long enough to develop a resistance to exfoliation corrosion higher 200 30 200 30 than that provided by the T6-type tempers. I year 1 veal Materials in the T73-type temper also have 30 min day 1 week 2 months- 20 30 mir 1 day 1 week 2 months- 20 high resistance to exfoliation corrosion. 100 I I I [I 100 I I I iI 0.1 1 10 102 103 104 0.1 1 10 102 103 104 Recommended treatments to produce TS- Elapsed time after quenching, h Elapsed tTme aftra quenching, h and T6-type tempers, and those of the T7- type employed for dimensional and proper- 500 7O 500 ,.~ 7o ty stabilization, provide adequate tolerance for normal variations encountered with 160 60 400 400 good operating practices. On the other hand, the T73, T74 (formerly T736), and 5O -- 50 T76 tempers for alloys 7049, 7050, 7075, 300 • h 3o0 7175, and 7475 involve changes in strength ~o 40 } RT f that occur significantly more rapidly at the / ~/0 °C (32 °F) i temperatures employed in the second stage 200 / / 3o -~ 200 ii f F) 30 -~ >- >- >- of the T7x precipitation heat-treatment cy- J // J ¢" ~---18 C (0 F) ...... s 20 2O cle compared to the changes occurring at 100 -18 °C (0 °F) 100 the temperatures employed to produce the I I yeal- 10 I yea,-- 10 T6 temper. 30 min 1 day 1 week 2 months 30 mini ldaVl 11 week 21 m°nd's/ I1 As illustrated in Fig 23, variations in soak 1 I I II o 0 0 0 time of several hours, and variations in soak 0.1 1 lO 102 10 3 104 102 103 10 4 0.1 10 temperature of up to I l °C (20 °F) from the Elapsed time after quenching, Elapsed time after quPnching, h II nominal aging practice of 24 h at 120 °C (250 4O °F) affect the strength of 7075-T6 by as d 6 much as 28 MPa (4 ksi). In contrast, similar c~ variations in second-step soak time and 5 30 3O E E temperature for 7075-T73--that is, varia- E E '~ -- - 18

7050 7075 6061 600 , 600 BBO t- 80 80 / / I 500 500 500 7O 7O 7O 2_ ST z oi , 2o RT / ~o / 6O 6O ~" 400 ~:£ 400 -S" 400 f / / 5o ~- 5o .~ -18 °C (0 ~F) - 5o .~

300 -- 300 300 c RT 4o 40 p- / ~ O~C (32°F) J 2O0 30 2OO 200 30 --"~ ~J-18 "~C (0 ~FI 1 yeal [ I / lye~, 20 30 mm l ooy, ,wrok r,onth,_l 30 rain 1 day 1 week 2 months ' 30 rain 1 day 1 week 2 months 100 I I 100 ~ 100 I i I II JI 20 01 1 10 102 103 104 0.1 10 102 103 104 0.1 1 10 102 103 104 Elapsed time after quenching, h Elapsed time quenching, h Elapsed time after quenching, h

500 500 500 ~70 70 70

7 60 400 400 / 400

/ _ 50 - -- 5(I E / 50 - / RT / - 300 . 300 .c / RT 3O0 ~j 40 =- -40 ~, E,, /O~C (32 °F) 4O = / ~ 30 ~ -e 20EI 30 / _~ 200 ~E z: 200 / >- >- >- RT ~-- 20 100 100 -18 °C (O °F) 0 °C (32 "F) 100 jl i/'l*~°C (~j °'F) 1 year" l0 1 year 10 1 day 1 week 2 months I I 1 yeaT 30 min I davl l~e;k 2ni~nths 10 30 rain O -- I ], 0 1 l I I iI II I 0 3omin,, 1lay l w:ok2 ooths I0 0 0.1 1 10 102 103 104 0.1 1 10 102 103 1G4 0.1 1 10 102 103 104 Elapsed time after quenching, h Elapsed time after quenching, h Elapsed time after quenching, h

4O 4O 40 d o 30 30 E S E E E E o RT E _18°C iO°F) 20 c 20 20 ~ ---.-_~... j ------__.. I 0 °C(32 °F) RT m iI 1 yea Lu uJ 30 min 1 day 1 week 2 months ,inT/ (Jay week 2 rqonths w 0 t- J~ I ~l__ iJ .- o ~__L~ U LJ~ 30 min 1 day week 2 cnonths 01 I 10 102 103 104 0.1 I 10 102 103 104 0 Elapsed time aftel quenching, h Elapsed time after quenching, h 0.1 1 10 102 103 104 Elapsed time after quenching, h Fig 21 (continued)

18 090 0= /e d__~t (Eq 6) Fys=I.45x10 -j6 exp ( ~K ) (Eq 7b) The effects of neglecting to compensate J Fvs for soaking at temperatures other than the where t is time during heating. where TK is temperature in K. nominal can be large (Fig 25). For exam- Equation 5 provides the basis for selec- In one experiment, lengths of 7050-W (4 ple, the calculated difference in strength tion of a nominal aging time that will result days) extrusions were aged at 24 h at 120 between alloy 7050 extrusions soaked 29 h in the desired yield strength and gives the °C (250 °F) plus the equivalent of 3 to 42 h at 160 °C (320 °F) and at 165 °C (325 °F) is furnace operator a method of compensating at 165 °C (325 °F). For the second step, a about 50 MPa (7 ksi), and the calculated for heating rate and for differences between logarithmic heatup was used in which l0 h difference in strength between 7050 extru- desired and attained soak temperatures. were required for the load to reach 155 °C sions soaked 29 h at 155 °C (315 °F) and at Specifics will be illustrated using data for (315 °F), and nominal soak temperature 170 °C (335 °F) is about 100 MPa (14 ksi). alloy 7050. The value of Frs (in units of was 165 °C (325 °F). Figure 24 indicates Neglecting to compensate for time spent hours) for 7050 can be calculated by the that yield strength generally agreed with heating the work to the soak temperature following equation: values predicted using Eq 5. The deviation will increase the variability. Strength loss of the curve for short-transverse strength attributed to heatup was 14 MPa (2 ksi). 32 562 at the short aging times indicates that the These kinetic relationships also can as- Fvs=I.45× 10-16 exp ( ~ ) (Eq 7a) method is inadequate for predicting sist in selection of equivalent aging times strength on the underaging side of the for alternate second-step aging tempera- where TF is temperature in °F, or aging curve. tures. Equations 5 and 7 can be rearranged 864 / Heat Treating of Nonferrous Alloys

55O 550 175 'C .~ 1305°F) C (350 'F)

500 500 105 "C -- RT .~"" ,~ \ ~- 1oo

450 - 150 'C 450 % i3oo,,, 60 -~ ~E ~ ~ 135 rC 60 400 =- £ 400 g 190 'C

\150 rC 350 350 50 ~' 1300[ "F} 50 c

175 "C 300 (350 "F) 300 190 "C 205 C 40 260 ~C (400 '~F~" ,?j p~5O0 'F) 250 250 ~(4230 r'c (400 "F) 50 "F) I 30 min 1 day 1 week 2 months 1 veal 30 min 1 day 1 week 2 months 1 year I I I I [I " 1[ 30 [ I I I II ~1 30 200 20O 105 0 0.01 0,1 1 lO 100 103 104 105 0 0.01 0.1 1 10 100 103 104 Duration of precipitation heat treatment, h Aging time, h 500 500 I 70 175 °C 130 °C'(265 °F) 150 °C - 70 105 °C (/3300~F) -~<'~....~ f'l~soFI - 60 400 F 60 400 ~. "f, \\ J 135°c RT 100'~C - 50 x"x/ \ \ (2,5 oF, - 50 1212 'F) ~E 300 , 160 "C £ - 300 4O 190 °C ~) 40 p~ (375 "F)

3O 200 30 -~ 200 205 'C )- >- \ 1400 F) . . 2 5° .0ioF, 2O 260 °C (500 °F)~ 20 100 100 260 'C 10 ~500 "F- 30 mini 1 (lay 1 week 2 months 1 yeal 30 rain day 1 week 2 months 1 yeal 0 a I 0 I I iI [I 11 0 0.01 0.1 1 10 100 103 104 105 0 0.01 0.1 1 10 100 103 104 105 Duration of precipitation heat treatmenL h Aging tin](!, h

4O 20 E RT 100 "C c~ 30 _205 °C ~175 °C _150 "C _135 °C ~~-1400°F)~ ~ (350 °F) (300 °F) (275 OF) ~~~__ E -130 ~C 205 'C E 6~ (400 'F) o 20 :- 10 260 °C 15o ._o 0. 13oo % ~ s o o w 1375, (/350 (11320 w (500~F) (450~F) 30rain day 1week 2months 1year 30 rain 1 clay 1 week 2 months 1 year 0 0 i, I I [I 1 IJ O 0.01 0.1 1 10 100 103 104 105 0 0.01 01 I lO 100 103 104 105 Duration of precipitation heat treatment, h Aging time, h

Fig 22(a) Aging characteristics of alloy 2014 sheet Fig 22(b) Aging characteristics of alloy 2024 sheet (see also Fig 22d) to yield the following equation: t350=29/exp (1.28)=29/3.6=8 h quenching, whereas other alloys show little or no added strengthening when treated by 32 562 32 562 Thermomechanical effects on aging occur this sequence of operations. t2=tl exp -- (Eq 8) T1+460 T2+460 from deformation after solution heat treat- Alloys of the 2xxx series such as 2014, ment. The deformation step may be warm 2124, and 2219 are particularly responsive where t 1 is aging time at temperature T1, t 2 is or cold and before, after, or during aging. to cold work between quenching and aging, aging time at temperature T2 that will provide The simplest thermomechanical practices and this characteristic is the basis for the equivalent yield strength, and T~ and T2 are in are those of the conventional T3, T8, or T9 higher-strength T8 tempers. The strength °F. For example, the time at 175 °C (350 °F) tempers. The rate and extent of precipita- improvement accruing from the combina- equivalent to aging alloy 7050 for 29 h at 165 tion strengthening are distinctly increased tion of cold working and precipitation heat °C (325 °F) is calculated as follows: in some alloys by cold working after treating is a result of nucleation of addition- Heat Treating of Aluminum Alloys / 865

350 50 type of approach results in strengths sim- ilar to those obtained with T8 processing 120 °C but with the better toughness and fatigue 325 f (250 °F) characteristics of T3 products. Alloys

-45 2024, 2124, and 2219 in T8-type tempers are particularly well suited for supersonic 300 and military aircraft; alloy 2219 in such ~150 °C tempers, and alloy 2014-T65, were the (300 °F) ~ principal materials for the fuel and oxidiz- - 275 - 40 x: ~~ "~ 170 °C er tanks (which also served as the primary (340 °F) structure) of the Saturn V space vehicles. 25C Re-solution heat treatment of mill prod- ---%, ucts supplied in these tempers can result in \ \ 35 ~ 205 °C F- grain growth and in substantially lower (400 °F) 225 strength than is normal for the original temper. Such reheat treatment is not rec- 230 °C (450 °F) -30 ommended. 200 Alloys of the 7xxx series do not respond ~/260 °C ~1500 °F) favorably to the sequence of operations 301mini /1 day\/\ llwleek 2 monthsll l yearll used to produce T8-type tempers, and no 175 0.01 0.1 1 10 100 103 104 105 such tempers are standard for these alloys. Duration of precipitation heat treatment, h The strains associated with stretching or compressing of 7xxx alloys have relatively 300 I little effect on the mechanical properties of 120 °C 40 material precipitation heat treated to T6- ~ /(250°FI type tempers. On the other hand, these 250 -- ~ 150 °C "~ ~ (300 °F) -- 35 operations have measurable detrimental 2 /I 170 °C effects on final strength when T73-, T736, ~ J (340 °C) or T76-type tempers are produced, partic- - 200 _!J 30 ~ ~ ~...205°~ ularly in the direction opposite the direc- (400 °F) 2s ~ tion of cold work. Accordingly, specifica- ~_ 15~ f tion properties are somewhat lower for ..~-..-..-. z...~ % 230 °C >- (450 °F) 20 ~ the stress-relieved versions of these tem- pers. Decreasing the overaging time to 100 compensate for the loss in strength is not advisable, because this would impair de- 30 rain 1 day 1 week 2 months 1 year 10 velopment of the desired corrosion char- I I I 11 II I[ acteristics. 0.1 1 10 100 103 104 105 0 0,01 Temperature control and uniformity pre- Duration of precipitation heat treatment, h sent essentially the same problems in pre- 4O cipitation heat treating as they do in solu- d tion heat treating. o4 = Good temperature control and uniformity o 30' 230 °C _205 °C 170 °C 150 °C E (450 OF) (400 OF) (340 OF) (300 OF) throughout the furnace and load are re- E 260 °C / (500 °F) quired for all precipitation heat treating. Recommended temperatures are generally .E 20 ~~ 120°C ~'- 1250 °F) those that are least critical and that can be used with practical time cycles. Except for ~, lO 7xxx alloys in T7x tempers, these tempera- o ,v, tures generally allow some latitude and 30 min 1 day 1 week 2 months 1 year should have a high probability of meeting property specification requirements. Fur- 0 0.01 o.1 1 lo loo lO3 lO4 105 Duration of precipitation heat treatment, h nace radiation effects seldom are trouble- some except in those few furnaces that are Fig 22(c) Aging characteristics of alloy 6061 sheet used for both solution and precipitation heat treating. Generally, such situations should be avoided, because the high heat capacity al precipitate particles by the increased strength T8-type tempers of alloys 2011, needed for the higher temperatures may be strain. In some alloys of the 2xxx series, 2024, 2124, 2219, and 2419, which are difficult to control at normal aging temper- strain introduced by cold working after so- produced by applying controlled amounts atures. lution heat treatment and quenching also of cold rolling, stretching, or combinations Soak time in precipitation heat treating is induces nucleation of a finer precipitate of these operations. Normally, cold work not difficult to control; the specified times dispersion that increases strength. Depend- is introduced by stretching; however, oth- carry rather broad tolerances. Heavier loads ing on the aging temper, however, tough- er methods such as cold rolling can be with parts racked closer together, and even ness may be adversely affected, as illustrat- used. Recently, 2324-T39 was developed. nested, are not abnormal. The principal haz- ed in Fig 26 for 2024 sheet. The T39 temper is obtained by cold rolling ard is undersoaking due to gross excesses in Strengthening from thermomechanicai approximately 10% after quenching fol- loading practices. Some regions of the load processing is the basis for the higher- lowed by stretching to stress relieve. This may reach soak temperature long after soak 866 / Heat Treating of Nonferrous Alloys

2024-T4 2024-T3 2024-T36 (not cold worked) (cold worked 1 to 2%) (cold worked 5 to 6%) 600 600 600 80 80 190 °C (375 °F)[ 175 °C 1350 °F) B0 .- =3 /190 °C (375 °F) - 500 .~190 °C (375 °F)_ ~" 500 70 ~ ~ 500 "I--"~ -~° ~"~ ~--~--- 205 °C (400 °F) 70 ~ 205 (400 o~> \ °c 220 °C 1425 OF) ~" 60 220 °C (425 °F) 60 "~ "~ 60 = = 400 ,~ 400, -~ ~ 400 F- ~''~220°C(425°F) -50 ~ 5o~- ~- 50 ~- 30C 300 3OO 4 8 12 16 4 12 16 4 8 12 16 Aging time, h Aging time, h Aging time, h

600 600 600 I 80 80 190 °C (375 °F) 175 °C (350 °F) - 80

500 500 5OO ~o ,190 °C (375 °F) -- 70 70 ~ =- --"'./'-I g 400 190 °C (375 °F)--- 400 r\ ,. ~/ " ~220 °C (425'°F) 50 ~///~'20 °C-2" (425 °F) - ~o ~ >- >- >" 300 300 ~.~ 205 °C (400 °F)-~ 3OO 40 -- 40 40 220 °C (425 °F) 200 30 200 --30 200 30 0 8 12 16 0 4 8 12 16 4 12 16 Aging time, h Aging time, h Aging time, h

30 30 30

c E ~Z 2o ~o ~ 20 ~ 20 k ~ 190 °c (375 °F) 205 °C (400 °F) 190 °C (375 OF) ~ E E 190 °C (375 OF) ~E c ~/Z/ 175 °C (350 OF) oo lO oo 10 -<. -

220 °C 1425 °F) p 220 °C (425 OF) 205 °C (400 °F) I I 220 °C 425 °F) ] 4 8 12 16 4 8 12 16 4 8 12 16 Aging time, h Aging time, h Aging time, h

Fig 22(d) Effects of cold work after quenching and before aging on tensile properties of alloy 2024 sheet time has been called. Placement of load ther- to the T73 and T76 tempers. As discussed Hardening of Cast Alloys mocouples is critical, and limiting the size and above, soak time is not as critical for peak- spacing of a load may be necessary for aging aged (T6 and T8) tempers. In general, the principles and procedures for heat treating wrought and cast alloys are similar. The major differences between so- 225 lution-treating conditions for castings and ~" ,~ 360 those for wrought products are found in ~--~.. 385 400 soak times and quenching media. Solution 200 of the relatively large microconstituents ~~ 410 MPa P o present in castings requires longer soaking

175 350 = ~ 700 I i lO0 E kongit inal 90 ~ ,550 ~ 150 \ 300 ~ 600 64 / °i 8o 68 ~ ~ 500 125 t I I F \ - 70 -o Predicted from Eq 3 ~ -$ ksi 250 O • Actual 4O0 I L I ] I 60 100 2 4 6 8 10 20 30 40 60 0.01 0.1 1 10 100 1000 Equivalent aging time Aging time, h at 165 °C (325 °F), h

Fig 24 Actual versus predicted yield strengths for Fig 23 Iso-yield-strength curves for alloy 7075 alloy 7050 extrusions Heat Treating of Aluminum Alloys / 867

Aging temperature, °F Transverse yield strength, ksi not by the heat treater. This effect of casting 310 320 330 340 40 45 50 55 60 65 70 methods on property development is shown 1.8 100 I I I I I I I I II 5000 in Fig 27. Because of the finer cast structure ~. lO T4~ Cu Mg Mn Fe Si 4.2 1.4 0.6 0.34 0.14 and higher supersaturation of the more rap- 50 ~" 1.6 4500 idly solidified permanent mold castings, ~ , I -5) x7' Stretched Io5Yo their tensile properties are superior to those 4000 0 0 73 1.4 n of sand castings of the same composition \ ° similarly heat treated. .~. 3500 ~ Tempers. Cast products of heat-treatable •y. -50 [ -10 # 1.2 aluminum alloys have the highest combina- Compared for standard ~ '5 / 1-6 \ 3000 tions of strength, ductility, and toughness -100 conditions of 29 h . x~ -15 Not stretched at 165 °C w when produced in T6-type tempers. Devel- ~ 1.0 N 2500 uJ oping T6-type tempers in cast products re- -150 I I I -20 N-- quires the same sequence of operations 150 155 160 165 170 175 T7 ~K,T8~ 0.8 2000 employed in developing tempers of the Aging temperature, °C 250 300 350 400 450 500 same type in wrought products--solution Effect of aging temperature on yield strength Transverse yield strength, MPa heat treating, quenching, and precipitation Fig 25 of alloy 7050-T736 Fig 26 Effect of stretching and aging on the tough- heat treating. Premium-quality casting spec- ness and yield strength of 2024 sheet ifications such as MIL-A-21180 can require different strengths and ductility levels in the periods than those used for wrought prod- same casting. ucts (Table 3). When heat treatment of ified castings. The microsegregation of sili- Among precipitation treatments unique to castings must be repeated, solution times con and magnesium is not severe in the castings are those resulting in the T5 and T7 become similar to those for wrought prod- aluminum-silicon-magnesium casting al- tempers. The T5 temper is produced merely ucts, because the gross solution and homog- loys, and hence it takes only a short time to by applying a precipitation treatment to the enization has been accomplished and is homogenize the alloy and to place the as-cast casting, without previous solution irreversible under normal conditions. Re- Mg2Si into solution. treatment. A moderate increase in strength duction of stresses and distortion from Quenchants. Quenching of aluminum is achieved without warpage and subse- quenching are also important, because cast- castings is often done in boiling water or a quent straightening. High hardness and di- ings generally are complex shapes with vari- milder medium to reduce quenching stress- mensional and strength stability at elevated ations in section thickness. es in complex shapes. A commercially im- temperatures account for the almost univer- Different casting processes and foundry portant variety is a mixture of polyalkylene sal use of materials in T5 tempers for pis- practices also result in microstructural dif- glycol and water, which has no detrimental tons and other engine parts. Some applica- ferences with relevance to heat-treatment effect on properties for thicknesses under tions demand combinations of strength, practice, because the coarser microstruc- approximately 3.2 mm (0.125 in.). Quen- toughness, and dimensional stability that tures associated with slow solidification chant additions can be made for the follow- cannot be met by heat treating to T5-, T6-, rates require a longer solution heat treat- ing purposes: or T8-type tempers. For these applications, ment exposure. Therefore, the time re- T7-type tempers are developed by solution • To promote stable vapor film boiling by quired at temperature to achieve solution is heat treating, quenching in a medium that the deposition of compounds on the sur- progressively shorter for investment, sand, provides a moderate cooling rate, and then face of parts as they are submerged in the and permanent mold castings. Foundry precipitation heat treating at a temperature quench solution practice (chills, gating, type of mold) also higher than those used to develop T5-, T6-, • To suppress variations in heat flux by plays an important role in the response of a and T8-type tempers. Heat treating to T7- increasing vapor film boiling stability casting, or a portion of a casting, to heat type tempers results in lower strength than through chemically decreased quench so- treatment. For example, thin-wall sand that of material in T6- or T8-type tempers, lution surface tension castings produced with extensive use of develops high ductility and toughness, and • To moderate quench rate for a given water chills can often display finer microstruc- carries precipitation far enough to minimize temperature tures than heavy-section permanent mold further precipitation during service. parts produced in such a way that process The key to the compromise between advantages are not exploited. goals involving property development and Stress Relief For these reasons, solution heat-treat- the physical consequences of quenching is ment practices can be optimized for any uniformity of heat extraction, which is in Immediately after being quenched, most specific part to achieve solution with the turn a complex function of the operable heat aluminum alloys are nearly as ductile as shortest reasonable cycle once production extraction mechanism. Nucleate, vapor they are in the annealed condition. Conse- practice is finalized, even though most film, and convective boiling occur with dra- quently, it is often advantageous to stress foundries and heat treaters will standardize matically different heat extraction rates at relieve parts by working the metal immedi- a practice with a large margin of safety. different intervals. Differences in section ately after quenching. Numerous attempts There also exists a fundamental difference thickness, load density, positioning, racking also have been made to develop a thermal between unmodified and modified alloys in methods, surface condition, and casting ge- treatment that will remove, or appreciably which the size and shape of silicon crystals ometry also influence the results. reduce, quenching stresses. Normal precip- are modified with additions of elements Property Development. Yield strength is itation heat-treating temperatures are gener- such as calcium, sodium, strontium, or an- largely controlled by the limiting hardening- ally too low to provide appreciable stress timony. Modified alloys undergo rapid element level, and tensile strength (in a relief. Exposure to higher temperatures (at spheroidization while complete sphe- general sense) is related to the ductility at a which stresses are relieved more effective- roidization is not achieved in unmodified given yield strength. Ductility, however, is ly) results in lower properties. However, alloys even after very long times. The prac- controlled for a given yield strength by such treatments are sometimes utilized tical implication is that shorter solution heat soundness and microstructural fineness, when even moderate reduction of residual treatment could be employed in fully mod- and is thus determined in the foundry and stress levels is important enough so that 868 / Heat Treating of Nonferrous Alloys

Permanent mold castings Sand castings 7079 350 -- 50 350 -- 50 Spar forging Planes of A saw cuts 300 ~\ c~ 300 150 °C (300 °F) -160 mm ~/~ 1501C (300 °F) _ 40 ~ - 40~ ~ 250 205 °C (400 °F) .... =" ~ 250 g ~--~- ..~5 °C (400 °F) ~ ~30 200 \ _~ ~ 200 \ ~o °~ (5oo °~t ~ ~ \~ 260 °C (500 OF) 150 150 ~ ~ ~ 20 20 Direction of 100 100 compressive deformation 10 20 30 0 10 20 30 Aging time, h Aging time, h 7500 250 .~ E 200 oo 250 - 4500 250 150 °C (300 °F) 150 d (-- _ --, 3°L~ ~ 2oo - 30~ 100 2OO 150 °C (300 °F) ~ 1500 .c" 50 ~ 0 ~ t150 -~205 °C 1400 °F)-- -1500 ]rallel T652 normal -50 ~ 205 °C (400 °F) -~05 i -o 260 °C (500 °F) '%. 260 °C (500 °F) < Length,- 1.8m >1 100 >-,~- >,.~ loo >- 2500 I0 --10 80 "- 50 5O E 10 20 3O 10 20 30 60 Aging time, h Aging time, h E 1500 c~ o 4O E o = 500 20 "~ 6 20 ~ ~ 260 pc (500 °F) . / 205 °C (400 °F) 0 N -500 -20 ~ '~ "'°E 10 '°°°c'°°°°L' 1 0 (300 L '° t--" °c (4oo"i < Length,-1.8m > ,T,oo m I 150 °C (300 °F) .__ 0 ,7,~ 0 t Effect of 3% permanent deformation in corn- 0 10 20 30 "- 0 10 20 30 Fig 28 pression (T652 treatment) on distribution of Aging time, h Aging time, h stress in a large forging. Parallel and normal refer to Comparison of the precipitation-hardening characteristics of 356.0-T4 sand and permanent mold warpage directions with respect to the plane of the Fig 27 castings saw cut. some sacrifice in mechanical properties can induced during the quenching operation. iization, involve cycling of parts above and be accepted. The T7 temper for castings is a For products stress relieved by stretching, below room temperature. The temperatures typical example of this kind of treatment. the digits 51 follow the basic Tx designation chosen are those that can be readily ob- Mechanical Stress Relief. Deformation (T451, for example). For products stress tained with boiling water and mixtures of consists of stretching (bar, extrusions, and relieved by compressive deformation, the dry ice and alcohol--namely, 100 and -73 plate) or compressing (forgings) the product supplementary digits are 52. °C 1212 and -100 °F)--and the number of sufficiently to achieve a small but controlled An additional digit is added to designa- cycles ranges from one to five. The maxi- amount (1 to 3%) of plastic deformation. If tions for extrusions: an added zero specifies mum reduction in residual stress that can be the benefits of mechanical stress relieving that the product has not been straightened effected by these techniques is about 25%. are needed, the user should refrain from after final stretching; an added one indicates The maximum effect can be obtained only if reheat treating. that straightening may have been performed the subzero step is performed first, and Figure 28 illustrates the beneficial effect after final stretching. immediately after quenching from the solu- of 3% permanent deformation in compres- Effect of Precipitation Heat Treating on tion-treating temperature while yield sion on a large forging. Residual Stress. The stresses developed dur- strength is low. No benefit is gained from These methods are most readily adapt- ing quenching from solution heat treatment more than one cycle. able to mill and forge shop products and are reduced during subsequent precipitation A 25% reduction in residual stress is require equipment of greater capacity than heat treatment. The degree of relaxation of sometimes sufficient to permit fabrication that found in most manufacturing plants. stresses is highly dependent upon the time of a part that could not be made without this Application of these methods to die forgings and temperature of the precipitation treat- reduction. However, if a general reduction and extrusions usually requires construc- ment and the alloy composition. In general, is needed, as much as 83% relief of residual tio,n of special dies and jaws. Stretching the precipitation treatments used to obtain stress is possible by increasing the severity generally is limited to material of uniform the T6 tempers provide only modest reduc- of the uphill quench--that is, more closely cross section; however, it has been applied tion in stresses, ranging from about 10 to approximating the reverse of the cooling- successfully to stepped extrusions and to a 35%. To achieve a substantial lowering of rate differential during the original quench. 3 by 14 m (10 by 47 ft) aircraft wing skin quenching stresses by thermal stress relax- This may be accomplished by a patented roll-tapered to a thickness range of 7.1 to ation, higher-temperature treatments of the process that involves extending the subzero 3.2 mm (0.280 to 0.125 in.). T7 type are required. These treatments are step to -195 °C (-320 °F) and then very Specific combinations of the supplemen- used when the lower strengths resulting rapidly uphill quenching in a blast of live tal digits are used to denote the tempers from overaging are acceptable. steam (Fig 29). The rate of reheating is produced when mechanical deformation is Other thermal stress-relief treatments, extremely critical, and therefore, to ensure used primarily to relieve residual stresses known as subzero treatment and cold stabi- proper application of the steam blast, a Heat Treating of Aluminum Alloys / 869

Plate Table 9 Reheating schedules for wrought aluminum alloys thickness The schedules given in this table normally will not decrease strength more than 5%. ~-- 50 mm ~ Reheating time at a temperature of: 150 *C 165 *C 175 *C 190 *C 205 *C 220 *C 230 *C Alloy and temper (300 *F) (325 *F) (350 OF) (375 *F) (400 *F) (425 *F) (450 *F) I Range of g - A B C D residual 2014-T4 (a) (a) (a) (a) (a) (a) (a) stress 2014-T6 2-50 h 8-10 h 2-4 h V2-1 h 5-15 rain (b) (b) -j MPa ksi 2024-T3, 2024-T4 (a) (a) (a) (a) (a) (a) (a) E A 28 4.0 2024-T81, 2024-T86 20-40 h • • • 2-4 h 1 h V2 h 15 min 5 min o B 86 12.4 6061-T6, 6062-T6, C 130 19.0 6063-T6 100-200 h 50-100 h 8-10 h 1-2 h l/z h 15 min 5 min D 165 24.0 7075-T6, 7178-T6 10-12 h 1-2 h 1-2 h 1/2--1 h 5-10 min (b) (a) (a) Reheating not recommended. (b) Bring to temperature Treatment A: Cooled to -195 °C, then uphill quenched in a steam blast B: Cooled to -75 o C, then uphill temper without first carefully testing the For both heat-treatable and non-heat-treat- quenched in a steam blast effects of such reheating. In one such test, able aluminum alloys, reduction or elimina- C: Cooled to -75 or -195 o C, then uphill quenched in boiling water 2024-T4 sheet was found to be very suscep- tion of the strengthening effects of cold work- D: Standard specimen, quenched and tible to intergranular corrosion when sub- ing is accomplished by heating at a aged to T6 temper in conventional jected to a 15-rain drying operation at 150 °C temperature from about 260 to about 440 °C manner with no further treatment (300 °F) during the first 8 h after quenching; (500 to 825 °F). The rate of softening is no susceptibility was evident when the Fig 29 Effectiveness of various uphill quenching strongly temperature-dependent; the time re- treatments in reducing residual quenching same drying operation was performed more quired to soften a given material by a given stresses in 2014 plate. Note: uphill quenching treat- than 16 h after quenching. In another test, amount can vary from hours at low tempera- ments (single-cycle only) were applied from 1/2 to 11/2 h 7075-W (0.2 to 600 h) bar and plate were tures to seconds at high temperatures. after quenching from the recommended solution- treating temperature. All specimens were aged to the reheated for hot forming at 175 °C (350 °F) If the purpose of annealing is merely to ]-6 temper after uphill quenching. for 20 rain. Strengths after aging to the T6 remove the effects of strain hardening, heat- temper were 10 to 15% lower than those for ing to about 345 °C (650 °F) will usually standard 7075-T6. In contrast, similar re- suffice. If it is necessary to remove the special fixture usually is required for each heating ofT6 material for up to 1 h at 175 °C hardening effects of a heat treatment or of part. (350 °F) produced no detrimental effect. cooling from hot-working temperatures, a This process will not solve all problems of If reheating is performed on material in treatment designed to produce a coarse, warpage in machining. It may reduce the W or T4 condition, its effect can be widely spaced precipitate is employed. This warpage internally but increase warpage of estimated from families of precipitation usually consists of soaking at 415 to 440 °C the extreme outer layers, although in the heat-treating curves such as those present- (775 to 825 °F) followed by slow cooling (28 opposite direction (Fig 30). Also, the effect ed in Fig 22. Such curves can also be used °C/h, or 50 °F/h, max) to about 260 °C (500 of the altered residual-stress pattern on per- for reheating of precipitation heat-treated °F). The high diffusion rates that exist dur- formance must be evaluated carefully for material at the precipitation heat-treating ing soaking and slow cooling permit maxi- each part. This is particularly important for temperature. For reheating at other temper- mum coalescence of precipitate particles parts subjected to cyclic loading or exposed atures, other data may be needed (Fig 31). and result in minimum hardness. to corrosive environments such as marine The heat-treating and reheating curves may As a result of this treatment, only partial atmospheres, especially if the process is be used as the bases for limitations on precipitation occurs in 7xxx alloys, and a introduced after the start of production and reheating (Table 9). second treatment (soaking at 230 +-- 6 °C, or original performance tests are not repeated. 450 --- 10 °F, for 2 h) is required. When the Further disadvantages are the cost and haz- Annealing need arises for small additional improve- ard involved in handling liquid nitrogen and ments in formability, cooling at 28 °C/h (50 live steam. Annealing treatments employed for alu- °F/h) should be extended to 230 °C (450 °F), minum alloys are of several types that differ and the material should be soaked at 230 °C Effects of Reheating in objective. Annealing times and tempera- for 6 h. The effects of eliminating or pro- tures depend on alloy type as well as on longing the 230 °C second step on the duc- The precipitation characteristics of alumi- initial structure and temper. tility of 7075-0 sheet are compared with the num alloys must be considered frequently Full Annealing. The softest, most ductile, standard treatment in Table 10. during evaluation of the effects of reheating and most workable condition of both non- In annealing, it is important to ensure that on mechanical properties and corrosion re- heat-treatable and heat-treatable wrought the proper temperature is reached in all sistance. Such evaluations are necessary for alloys is produced by full annealing to the portions of the load; therefore, it is common determining standard practices for manu- temper designated "O." Strain-hardened to specify a soaking period of at least 1 h. facturing operations, such as hot forming products in this temper normally become The maximum annealing temperature is and straightening, adhesive bonding, and recrystallized, but hot-worked products moderately critical; it is advisable not to paint and dry-film lubricant curing, and for may remain unrecrystallized. In the case of exceed 415 °C (775 °F), because of oxidation evaluating the effects of both short-term heat-treatable alloys, the solutes are suffi- and grain growth. The heating rate can be and long-term exposure in elevated temper- ciently thoroughly precipitated to prevent critical, especially for alloy 3003, which atures in service. natural age hardening. A higher maximum usually requires rapid heating for preven- The stage of precipitation that exists in an temperature than that used for stress-relief tion of grain growth. Relatively slow cool- alloy at the time of reheating plays a signif- annealing, controlled cooling to a lower ing, in still air or in the furnace, is recom- icant role in the effects of reheating. Con- temperature, and additional holding time at mended for all alloys to minimize distortion. sequently, it is extremely dangerous to re- the lower temperature generally are em- Typical annealing conditions used for some heat material in a solution heat-treated ployed. alloys in common use are listed in Table 11. 870 / Heat Treating of Nonferrous Alloys

2600 +100 25 bv-25-'mm (1-by-'l-in.) bar 475

+1300 425 © 120 °C ~ 60 ~ 0~ 0~ 400 -• 135 °C neheatln n 0 0 ~ A 150°C - - "' ~ '~ g 375 temperature - 55 ~

1300 § 425 g ~ - 60~ O o c" ~=- '°° Fv-~~~"°,% =- o -26oo -1oo ~ D c~ ,~ 350 "o O 120°C -- 50 _"~ -3900 "~ 325 -e 135°C .-Q 300 A 150°C - 45 > o Control specimen • Quenched from - 75 °£ -200 A -5200 ---- (-100 °F) to steam /k Quenched from liquid o'- nitrogen to steam WE -6500 2 3 o E 1 10 100 103 Tine number Duration of reheating, days

+50 Fig 31 Effects of reheating on tensile properties of alclad 2024-T81 sheet

34 0 tions, this ductility often is slightly lower (2 by 2-in.) bar than that of material that has not been d -1300 -50 subjected to prior heat treatment--that is, § material annealed at the producing source. 0 ,,,,H, Therefore, when maximum ductility is re- c~ -2600 -t00 ~ quired, annealing of a previously heat-treat- % IIIIIII ed product is sometimes unsuccessful. tm Partial Annealing. Annealing of cold- 7075 \_ -3900 -150 worked non-heat-treatable wrought alloys to o Control specimen • Quenched from - 75 °C (-100 °F) to steam obtain intermediate mechanical properties Z~ Quenched flom liquid nitiogen to steam (H2-type tempers) is referred to as partial -5200 I I t J -200 2 3 4 5 annealing or recovery annealing. Tempera- Tine numbel tures used are below those that produce ex- tensive recrystallization, and incomplete soft- 2600 I I 100 ening is accomplished by substructural 75-by-75 mm (3-by-3-in.) bar changes in dislocation density and rearrange- ment into cellular patterns (polygonization). +1300 +50 Bendability and formability of an alloy an- E nealed to an H2-type temper generally are E ~ 0~ o significantly higher than those of the same o alloy in which an equal strength level is de- veloped by a final cold-working operation -1300 -50 (H l-type temper). Treatments to produce H2- type tempers require close control of temper- c~ ature to achieve uniform and consistent me- 7075 -2600 -100 chanical properties. o Control specimen Figure 32 shows changes in yield strength 1 • Quenched from -75 °C t-100 °F) /k Quenched from liquid nitrogen to as functions of temperature and time for -3900 I i J J -150 sheet of two non-heat-treatable alloys (1100 1 2 3 4 5 6 7 and 5052) initially in the highly cold-worked Tine number condition (H18 temper). From these curves, Effect of uphill quenching on deflection of tines. Six-tine specimen was machined from 50 by 50 mm it is apparent that, by selection of appropri- Fig 30 (2 by 2 in.) bar. Similar specimens machined from 25 by 25 mm (1 by I in.) and 75 by 75 mm (3 by 3 in.) ate combinations of time and temperature, bars had four and eight tines, respectively. mechanical properties intermediate to those of cold-worked and fully annealed material can be obtained. It is also evident that yield Products that can be heated and cooled few minutes. For these extremely rapid strength depends much more strongly on very rapidly, such as wire, are annealed by operations, maximum temperature may ex- temperature than on time of heating. continuous processes that require a total ceed 440 °C (825 °F). Stress-Relief Annealing. For cold-worked heating and cooling time of only a few Although material annealed from the pre- wrought alloys, annealing merely to remove seconds. Continuous annealing of coiled cipitation-hardened condition usually has the effects of strain hardening is referred to sheet is accomplished in a total time of a sufficient ductility for most forming opera- as stress-relief annealing. Such treatments Heat Treating of Aluminum Alloys / 871

Table 10 Effectsof annealing treatments on ductility of 7075-O sheet ly controlled. Even allowing the load to cool Elongation in bending(c), % in the furnace may result in an excessively Elongation in tension(a), % in 50 mm (2 in.) Bend angle(b), degrees, for in 50 mm (2 in.) for high rate. Similarly, lowering the furnace- for thickness of: thickness of: thickness of: control instrument by 28 °C (50 °F) each Annealing 0.5 mm 1.6 mm 2.6 mm 1.6 mm 2.6 mm 1.6 mm 2.6 mm hour may produce stepped cooling, which is treatment (0.020 in.) (0.064 in.) (0.102 in.) (0.064 in.) (0.102 in.) (0.064 in.) (0.102 in.) not satisfactory for severe forming opera- Treatment l(d) 12 12 12 82 73 48 50 tions. For maximum softening, a continu- Treatment 2(e) 14 14 14 91 76 58 57 Treatment 3(0 16 16 • • - 92.5 84 56 60 ous cooling rate of not more than 28 °C/h (50 °F/h) is recommended. (a) Uniform elongation of gridded tension specimens~ Ib) Bend angle at first fracture. (c) Elongation in bend test for 1,3 mm (0.05 in.) gage spanning fracture. (d) Soak 2 h at 415 ± 14 °C (775 ± 25 °F); furnace cool to 260 °C (500 °F) at 30 °C/h (50 °F/h); air cool, (c) Soak Annealing of castings for 2 to 4 h at 2 h at 425 °C (800 °F), air cool; soak 2 b at 230 °C (450 °F), air cool. (f) Soak 1 h at 425 °C (800 °F); furnace cool to 230 ~C (450 °F) at 30 °C/h (50 °F/h); soak 6 h at 230 °C (450 °F), air cool temperatures from 315 to 345 °C (600 to 650 °F) provides the most complete relief of residual stresses and precipitation of the phases formed by the excess solute retained employ temperatures up to about 345 °C even very small amounts of magnesium will in solid solution in the as-cast condition. (650 °F), or up to 400 -+ 8 °C (750 -+ 15 °F) form a surface magnesium oxide unless the Such annealing treatments provide maxi- for 3003 alloy, and cooling to room temper- atmosphere in the annealing furnace is free mum dimensional stability for service at ature. No appreciable holding time is re- of moisture and oxygen. Examples include elevated temperatures. The annealed tem- quired. Such treatment may result in simple alloy 3004, which is used for cooking uten- per is designated "O." (This temper was recovery, partial recrystallization, or full sils, and alloys of the 5xxx series. designated "T2" prior to 1975.) recrystallization. Age hardening may follow Another problem that control of the an- stress-relief annealing of heat-treatable al- nealing atmosphere helps to overcome or Grain Growth loys, however, because a concentration of avoid is oil staining by oil-base roll lubri- soluble alloying elements sufficient to cause cants that do not burn off at lower annealing Many of the aluminum alloys in common natural aging remains in solid solution after temperatures. If the oxygen content of the use are subject to grain growth during solu- such treatments. furnace atmosphere is kept very low during tion treatment or annealing. This phenome- A special form of stress-relief temper is such annealing, the oil will not oxidize and non can occur during or after recrystallization used for heat-treatable alloy products that stain the work. of material that has been subjected to a small subsequently will be inspected ultrasonical- Temperature control for full and partial critical amount of prior cold work. It is usu- ly. The product is heated to its normal annealing is somewhat more critical than for ally manifested by surface roughening during solution heat-treating temperature, then stress-relief annealing; the temperatures subsequent fabrication operations and fre- cooled in still air to room temperature. This and times specified are selected to produce quently results in rejections for appearance or temper is referred to as the Ol temper. recrystallization and, in the case of heat- functional reasons. Less frequently, some de- Controlled-Atmosphere Annealing and treatable alloys, a precipitate of maximum terioration of mechanical properties is en- Stabilizing. Aluminum alloys that contain size; for this the cooling rate must be close- countered, and this is undesirable regardless of surface-roughening effects. Degree of susceptibility to grain growth Table 11 Typical full annealing treatments for some common wrought aluminum alloys varies with alloy, structure, and chemical- These treatments, which anneal the material to the O temper, are typical for various sizes and methods of composition variation, and from one prod- manufacture and may not exactly describe optimum treatments for specific items. uct form to another. The critical range of Metal temperature Approximate Metal temperature Approximate cold work is ordinarily about 5 to 15%. time at time at Alloy *C °F temperature, h Alloy °C °F temperature, h Usually, temperatures of 400 °C (750 °F) and above must be reached before grain 1060 345 650 (a) 5457 345 650 (a) 1100 345 650 (a) 5652 345 650 (a) growth occurs, but some growth has been 1350 345 650 (a) 6005 415(b) 775(b) 2-3 encountered at temperatures as low as 345 2014 415(b) 775(b) 2-3 6009 415(b) 775(b) 2-3 °C (650 °F). Grain growth that occurs during 2017 415(b) 775(b) 2-3 6010 415(b) 775(b) 2-3 initial recrystallization is more a function of 2024 415(b) 775(b) 2-3 6053 415(b) 775(b) 2-3 2036 385(b) 725(b) 2-3 6061 415(b) 775(b) 2-3 composition, structure, and degree of cold 2117 415(b) 775(b) 2-3 6063 415(b) 775(b) 2-3 work than of temperature per se; tempera- 2124 415(b) 775(b) 2-3 6066 415(b) 775(b) 2-3 tures in excess of 455 °C (850 °F) in common 2219 415(b) 775(b) 2-3 7001 415(c) 775(c) 2-3 alloys can lead to secondary-recrystalliza- 3003 415 775 (a) 7005 345(d) 650(d) 2-3 3004 345 650 (a) 7049 415(c) 775(c) 2-3 tion grain-growth problems. The common 3105 345 650 (a) 7050 415(c) 775(c) 2-3 symptom indicating moderately large-grain 5005 345 650 (a) 7075 415(c) 775(c) 2-3 material is roughening or "orange peel" on 5050 345 650 (a) 7079 415(c) 775(c) 2-3 the external surfaces of bends. Severe 5052 345 650 (a) 7178 415(c) 775(c) 2-3 5056 345 650 (a) 7475 415(c) 775(c) 2-3 growth of grains to fingernail size and larger 5083 345 650 (a) sometimes is evident in parts made from 5086 345 650 (a) Brazing sheet annealed (O temper) material by stretch 5154 345 650 (a) forming and then thermal treating or similar 5182 345 650 (a) 5254 345 650 (a) No. 11 and 12 345 650 (a) operations. This type of grain growth often 5454 345 650 (a) No. 21 and 22 345 650 (a) is detected during subsequent anodizing, 5456 345 650 (a) No. 23 and 24 345 650 (a) etching, and chemical milling operations. (a) Time in the furnace need not be longer than necessary to bring all parts of the load to appealing temperature. Cooling rate is Cracking during welding or brazing is unimportant. (b) These treatments are intended to remove the effects of solution treatment and include cooling at a rate of about 30 °C/h (50 °F/h) from the annealing temperature to 260 °C (500 °F). Rate of subsequent cooling is unimportant. Trealment at 345 °C (650 °F), another characteristic which may indicate followed by uncontrolled cooling, may be used to remove the effects of cold work or to partly remove the effects of heat treatment. (c) that severe grain growth has occurred. In These treatments are intended to remove the effects of solution treatment and include cooling at an uncontrolled rate to 205 °C (400 °F) or less, followed by reheating to 230 °C (450 °F) for 4 h. Treatment at 345 °C (650 ~F), followed by uncontrolled cooling, may be used such instances, cracks propagate along to remove the effects of cold work or to partly remove the effects of heat treatment, (d) Cooling rate to 205 °C (400 °F) or below is less than or equal to 30 °C/h (50 °F/hL grain boundaries that provide little obstruc- tion to their progress. 872 / Heat Treating of Nonferrous Alloys

200 reducing the size of furnace loads or by 1100-H18I changing from an air furnace to a salt bath. I In one application, severe grain growth was ; 150j /175 °C (350 °F) found during bending of alloy 1100 rectan- ,/ / 2o L~ g gular tubing. The roughening of the inside ~2~5°C (400 °F) surfaces of the parts, which occurred during 100 forming of the large-grain material, im- -o \I I 230 °C (450 °F) -- 10 .-~ paired their functioning as radar waveguides. Investigation disclosed that, to 50 I I ,/-260 °C 1500 °F) minimize handling marks, the material was ,,....,.,., J I procured in the strain-hardened (H14) tem- °and 316 °C 1560 and 600 °F) \ 88I 0 per and was stress-relief annealed at 345 °C 0 0.5 I 1.5 2 2'.5 ; 315 4 4.5 (650 °F) immediately prior to forming. Grain Time, h growth occurred during annealing as a re- sult of the moderate amount of cold work 350 50 introduced at the mill. The problem was 5052-H18I eliminated by changing the stress-relieving operation to a 5-min heating period in an air 300 ~__ furnace operating at 540 °C (1000 °F). The //175 °C (350 OF) / 205 °C (400 OF) 40 explanation advanced for the success of this treatment was that, due to the rapid heating 250 rate, the temperature of the material was ~'230 °C 1450 °F) raised through the recrystallization range 30 for the less severely cold-worked grains - 200 260 °C (500 °F)" before the critically cold-worked grains had time to grow appreciably. 150 >- 20 ~ ,.•0°C (550 °F) Heating Equipment and Accessories 100 "X315 o 1600 OF) The general methods for heat treating 10 aluminum alloys include the use of molten 50 salt baths, air-chamber furnaces, and induc- tion heaters. The choice of heating equip- ment depends largely on the alloy and the 0 0 0.5 1 1.5 2 2.5 3.5 4 4.5 configuration of the parts to be processed. Time. h The type of heat treatment can also influ- ence the choice of heating equipment. For Fig 32 Representative isothermal annealing curves for alloys 1100-H18 and 5052-H18 example, both molten salt baths and air- chamber furnaces are suitable for solution treating of aluminum alloys, while induction If the surface roughening is objectionable In other similar investigations, no detri- heating requires additional analysis to de- from either an appearance or a functional mental effects have been discovered, and in fine the proper temperature range for solu- aspect, the desirability of surface-smooth- many cases such parts have served satisfac- tion treatment. Induction methods can pro- ing operations, such as sanding or buffing, torily in critical applications. vide high heating rates, which affect must be evaluated. If reductions in mechan- When a grain-growth problem is discov- transformation behavior (see, for example, ical properties are suspected, these must be ered, it is too late to change the condition of the section "Nonequilibrium Melting" in established by test and evaluated in relation the parts in question, but several possible this article). to the anticipated service. methods are available for preventing recur- Molten salt baths and air-chamber furnac- In one application, a part that had been rence of the difficulty. The simplest of these es both have advantages and disadvantages made by stretch forming O-temper 2 mm is relieving the causative stress by interject- in solution heat treatments, as discussed (0.080 in.) sheet and heat treating exhibited ing a stress-relief anneal into the manufac- below. Oil- and gas-fired furnaces, in de- significantly lower tensile and yield turing sequence immediately prior to the signs that allow the products of combustion strengths in portions where severe grain solution-treating or full-annealing cycle in to come in contact with the work, are growth had occurred than in portions hav- which the grain growth occurred. This ap- usually unsatisfactory because they pro- ing normal grain size: proach is usually successful and practical. mote high-temperature oxidation. Another possibility is to adjust the amount Salt baths heat the work faster (see Table Tensile strength Yield strength Grain of stress present in the part immediately 2) than air furnaces, provided that the Test structure MPa ksi MPa ksi prior to the critical heat treatment so that amount of work introduced at any one time the stress level is outside the critical range. is controlled to prevent the temperature Transverse This may be done by adding a cold-working from falling below the desired range. If the 1 Coarse 265 38.5 247 35.8 operation before forming, such as pre- temperature is permitted to fall below the 2 Coarse 263 38.2 241 35.0 minimum limit, much of the advantage of 3 Fine 311 45.1 261 37.8 stretching of blanks, or by forming in mul- tiple stages with a stress-relief anneal before the salt bath is lost, because of the necessity Longitudinal each stage. for reheating the large mass of salt. 1 Coarse 259 37.6 243 35.3 A third method that is sometimes suc- Salt baths are also more readily adapted 2 Coarse 269 39.0 245 35.6 to the introduction, at any time, of small 3 Fine 305 44.2 270 39.1 cessful consists of increasing the heating rate during the critical heat treatment by amounts of work requiring different soaking Heat Treating of Aluminum Alloys / 873

Table 12 Frequency selection for induction heating with a longitudinal-flux coil and a Production rate, tons/h transverse-flux coil 0 1 2 3 4 5 6 1800 ii Minimum part thickness, mm (in.), for a frequency of: -- Solution heat treat at 890 °C Material 60 Hz 200 Hz I kHz 3 knz 10 kHz 1500 -- -- Full anneal at 425 °C .... Partial anneal at 315 °C Solenoid (longitudinal-flux) coil -~ 1200 J

- Steel below Curie temperature >38 (1.5) 13 (0.5) 5 (0.2) 2.3 (0.09) 1 (0.04) j Steel above Curie temperature >175 (7.0) 100 (4.0) 43 (1.7) 25 (1.0) 13 (0.5) 900 / . i Brass >50 (2.0) 28 (1.1) 13 (0.5) 7 (0.28) 4 (0.16) Aluminum :>38 (1.5) 22 (0.85) 9 (0.375) 5 (0.2) 3 (0.12) 800 Transverse-flux coil Aluminum >5 (0.2) 1.3 (0.05) 0.25 (0.01) 0.08 (0.003) 0.04 (0.0016) 300 ~ . - Brass :>10 (0.4) 2.5 (0.1) 0.5 (0.02) 0.15 (0.006) 0.08 (0.0032) Steel :>50 (2.0) 13 (0.5) 2.5 (0.1) 0.9 (0.035) 0.5 (0.020) 0 5 10 15 20 25 30 Line speed, m/rain periods. (Economical utilization of air fur- bottom-heated pots can also lead to local Power requirement for transverse-flux induc- naces usually dictates accumulation of a overheating. Overheat controls are essen- Fig 33 tion heating of aluminum strip 1 mm (0.04 in.) large load of parts of similar thickness be- tial to ensure against temperatures exceed- thick and 1270 mm (50 in.) wide. Source: Ref 9 fore charging.) Also, the buoyant effect of ing 595 °C (1100 °F). the salt reduces distortion during heating, It is vitally important that water be kept and the large reservoir of heat facilitates away from a nitrate tank. In controlling a temperature control and uniformity. nitrate fire, do not use water or any fire als. Therefore, transverse-flux coils are ide- Salt bath operation entails special house- extinguisher containing water. The best ex- ally suited for heating nonferrous materials, keeping requirements. Dragout is costly and tinguisher is dry sand, a supply of which because transverse-flux lines do not exhibit unsightly. Because residual salt on parts should be kept near the tank. the degree of current cancellation associat- may result in corrosion, all salt must be Extra sacks of salt should be stored in a ed with longitudinal flux lines. This aspect completely removed, including that from dry place, distant from the tank. If the fresh of transverse-flux coils improves efficiency crevices and blind holes. In addition, salt salt being added to the bath is even slightly and also permits the use of lower frequen- residue from the quench water must be kept damp, it should be added very slowly or cies (Table 12). This reduces the capital to a minimum by a constant water overflow when the bath is frozen. equipment costs, and where it shifts from or by providing a fresh-water rinse for all Air furnaces are used more widely than salt requiring RF frequencies, the power source parts after quenching. When these provi- baths because they permit greater flexibility conversion efficiency is also significantly sions are impractical, corrosion can be in- in operating temperature. When production improved. Aluminum, brass, copper, and hibited by adding 14 g (1/2 oz) of sodium or schedules and the variety of alloys requiring austenitic stainless steel strip lines are ide- potassium dichromate to each 45 kg (100 lb) heat treatment necessitate frequent changes ally suited for transverse-flux heating. Each of the molten salt. in temperature, the time and cost of adjusting of these materials often requires in-line pro- Precautions. Molten salt baths are poten- the temperature of a large mass of salt makes cesses like partial or full annealing or solu- tially hazardous and require special precau- the use of an air furnace almost mandatory. tion heat treating to provide necessary me- tions. Operators must be protected from However, waiting periods are often required chanical properties for subsequent finishing splashing and dripping of the hot salt. Be- to allow the walls of air furnaces to stabilize at operations. cause heated nitrates are powerful oxidizing the new temperature before parts are intro- Transverse-flux induction heating offers agents, they must never by allowed to come duced. Otherwise, parts may radiate heat to several benefits for in-line strip heating and in contact with combustibles and reducing colder walls or absorb radiant heat from hot- has been known for many years. However, agents, such as magnesium and cyanides. ter walls, and the temperature indicated by it requires specially designed iron-cored Most authorities advise against inserting the control instrument will not reflect actual laminated inductor coils and tighter control aluminum alloys containing more than a few metal temperature in the usual manner. Air of the power, strip handling, and process percent of magnesium into molten nitrate. furnaces are also more economical when the parameters. Frequency selection is dictated To avoid exposure of personnel to nitrous product mix includes a few rather large parts; by the resistivity and thickness of the ma- fumes produced during decomposition of holding the temperature of a large volume of terial, while power requirements depend on nitrates, good ventilation is essential. salt in readiness for an occasional large part is the production rates, the specific heat, and When molten nitrates are being used, the far more expensive than heating an equal the processing temperatures for a given possibilities of explosions resulting from volume of air. material. Figure 33 shows typical power both physical and chemical reactions must Induction heating with either solenoid source requirements for transverse-flux be avoided. The former result from rapid (longitudinal-flux) coils or transverse-flux heating of aluminum strip. expansion of gases entrapped beneath the coils provides an efficient method for in-line surface of the bath. Hence, parts entering heating of flat-rolled products (particularly Furnace Temperature Control the bath must be clean and dry; they must strip). Solenoid coils create a longitudinal also be free of pockets or cavities that flux, which can produce efficient heating for The importance of close temperature con- contain air or other gases. Chemical-reac- thicker and/or lower resistivity materials. trol in solution treating has been noted in tion explosions result from rapid break- Solenoid coils can also be used efficiently in the previous section on solution treating. down of the nitrates due to overheating or the heating of thinner magnetic material Each control zone of each furnace should reaction with the pot material. Stainless (see, for example, steel below the Curie contain at least two thermocouples. One steel pots (preferably of type 321 or 347) are temperature in Table 12). thermocouple, with its instrument, should more resistant to scaling than those made of For several nonferrous materials (alumi- act as a controller, regulating the heat input; carbon steel or cast iron and therefore pre- num, copper, brass), however, efficiency the other should act independently as a sent a lower probability of local overheat- and power factors with solenoid coils are safety cutoff, requiring manual reset if its ing. Sludge or sediment accumulations in significantly lower than for ferrous materi- set temperature (usually the maximum of 874 / Heat Treating of Nonferrous Alloys the specified range) is exceeded during the factors should be applied after each probe solution-treating cycle. check, but if the correction required ex- Safety cutoffs are mandatory for salt ceeds ---3 °C (---5 °F), the source of the baths to guard against explosions and often deviation should be corrected. MIL-H-6088 have paid for themselves in air furnaces by recommends that this check be made week- saving a load of parts or even the furnace ly, but many operators make the check as itself. It is important, however, that they be frequently as once each shift. tested periodically (by deliberately over- Temperature-Uniformity Surveys. In con- shooting the empty furnace) to guard trolling the temperature of parts that are against "frozen" corroded contacts result- being heat treated it must first be deter- Rectangular furnace ing from prolonged periods of idleness. mined that the temperature indicated by the At least one of the instruments for each furnace instruments truly represents the zone should be of the recording type, and temperature of the nearby air or salt. Sec- both instruments should have restricted ond, the uniformity of temperature within scales for instance, 400 to 600 °C (750 to the working zone must be shown to be 1110 °F), rather than 0 to 600 °C (32 to 1110 within a range of 11 °C, or 20 °F (6 °C, or 10 °F). This is required for maximum accuracy °F, for precipitation heat treatment of alloy because manufacturers' guarantees are 2024). This is accomplished by measuring specified in terms of percent of scale. the temperature at several test locations, In the placement of instruments, expo- using calibrated test thermocouples and a sure to extremes in ambient temperature, calibrated test potentiometer, and reading humidity, vibration, dust, and corrosive furnace instruments nearly simultaneously. fumes should be avoided. Ambient temper- MIL-H-6088 recommends monthly surveys Cylindrical salt bath Cylindrical air furnace atures between 5 and 50 °C (40 and 120 °F) with one test location per 1.1 m 3, or 40 ft 3 Location of thermocouples for surveying are satisfactory, but temperature changes of (0.7 m 3, or 25 ft 3, for air furnaces on initial Fig 34 temperature uniformity in the working 6 °C/h (10 °F/h) or more should be avoided. survey), but with a minimum of nine test zones of air furnaces and salt baths It is also essential that instruments and locations distributed as shown in Fig 34. thermocouple circuits be shielded from Despite the large size of some furnaces, electromagnetic fields commonly associat- rather surprising temperature uniformities ed with the leads of high-amperage furnace have been reported. In one instance the stabilized at the test temperature. Then a heating elements. initial survey of an air furnace measuring rack containing the test thermocouples is Temperature-sensing elements must be 12.5 by 1.2 by 3.0 m (41 by 4 by 10 ft) inserted into the furnace. By using multiple capable of responding more rapidly to tem- showed maximum temperature variations of switches or a multipoint recording instru- perature changes than the materials being +1.7, -1.1 °C (+3, -2 °F). When a parti- ment, all test thermocouples and furnace processed. Therefore, thermocouple wire tion 0.3 m (1 ft) thick was lowered, convert- instruments are read every 5 min. As the diameter should not exceed IVz times the ing the furnace to two chambers 6.1 by 1.2 temperature approaches the test range, it is thickness of the minimum-gage material to by 3.0 m (20 by 4 by 10 ft) each, the spread advisable to increase the frequency of read- be heat treated, and should in no case was +1.1, -0.6 °C (+2, -1 °F) in one ings to detect possible overshooting. After exceed 14 gage. Thermocouples for salt section and +0.6, -I.1 °C (+1, -2 °F) in thermal equilibrium is reached, readings baths should be enclosed in suitable protec- the other. should be continued until the recurrent tem- tion tubes. Air-furnace thermocouples For each furnace load, one thermocouple perature pattern is established. should be installed in open-end protection (the "cold" couple) should be placed in the Surveys of salt baths generally are con- tubes, with the thermocouple junction ex- coldest area of the furnace and another (the sidered acceptable whether they are made tending sufficiently beyond the tube to pre- "hot" couple) in the hottest area. In addi- while the bath is empty or filled with work. vent any loss in sensitivity. tion to these two thermocouples, a load It is controversial whether surveys of air Temperature-sensing elements should be thermocouple should be installed. The load furnaces should be made with or without a located in the furnace work chamber, not in couple should be of approximately the same load. Undoubtedly, recovery overshoots ducts and plenums, and should be as close gage as the sheet or other product being are most likely to occur with a very light as possible to the working zone. Specifica- heat treated. If heavy plate, forgings, or load and would not be detected if a heavier tion MIL-H-6088C restricts distance be- castings are being heat treated, a similar load were used. Certainly, if all loads are tween the sensing element and the working discarded item should be used at the con- essentially alike, surveys should be made zone to a maximum of 100 mm (4 in.). The trolling load couple. The thermocouple with typical loads. With widely varying safety-cutoff thermocouple should be locat- should be placed in a drilled hole and loads, the optimum approach is to make ed to reflect the highest temperature in the packed to hold it firmly in place during the several surveys initially, including one with working zone. The control thermocouple heat-treating cycle. In some instances, the an empty furnace, and then to make suc- should be located in a position where it will items being heat treated can be used as the ceeding surveys with an empty furnace to read a temperature approximately halfway load couples. The thermocouples can be ensure against changes in furnace charac- between the hottest and coldest tempera- placed in holes drilled io areas that will be teristics. If any changes are made in the tures. removed in making the finished article. furnace that might affect temperature distri- Probe Checks. After the temperature- It is important that items of different bution, such as repair of vanes or louvers, measurement equipment is properly in- thicknesses--1 mm (0.040 in.) sheet and 25 several surveys should be repeated. stalled, it must be checked frequently for mm (1 in.) plate, for example--not be heat Another aspect of the problem of temper- accuracy. This is accomplished by inserting treated in the same furnace load. ature control in air furnaces is the necessity a calibrated probe thermocouple into the In salt baths, uniformity surveys usually of ensuring that the temperature of the parts furnace adjacent to each furnace thermo- are made by holding a probe thermocouple is the same as that of the surrounding air. couple and comparing its reading on a cali- in each location until thermal equilibrium is Furnace components whose temperature brated test potentiometer with that indicat- reached; in air furnaces, a mock heat-treat- differs from the air temperature must be ed by the furnace instrument. Correction ing cycle is required. First, the air furnace is suitably shielded to prevent radiation to or Heat Treating of Aluminum Alloys / 875 from the parts being heat treated. In a as possible to the actual temperature. To Average cooling rate at center furnace used for solution heat treating of achieve this, it is necessary to apply correc- of cylinders, °F/s 10 100 rivets, unshielded heating elements have tion factors obtained during calibration to 500 I I been known to produce part temperatures the next lower echelon of accuracy. Even / as much as 20 °C (35 °F) higher than the g 400 6O g then, if all errors inherent in the chain are in c control temperature, resulting in eutectic the same direction, a considerable differ- 300 / melting and cracking. In two other instanc- ence will exist between the measured and 40 es, reradiation through inadequate shielding actual temperatures. Therefore, it is advis- 200 produced a radiation effect of as much as 11 able to operate as close to the mean of the 2D :5 °C (20 °F). One of these problems was desired range as possible. 100 solved by painting the shield with reflective g g -~ 0 aluminum paint and the other by adding a 13 J 10 100 Dimensional Changes during mm (1/2 in.) thick layer of asbestos to the 1.6 Average cooling rate at center mm 0/16 in.) stainless steel shield. Heat Treatment of cylinders, °C/s Furnace-wall temperatures that differ ap- In addition to the completely reversible Fig 35 Effect of quenching rate on longitudinal preciably from the temperature of the parts changes in dimensions that are simple func- stress ranges in alloy 2014-T4 cylinders also must be avoided. Consequently, when tions of temperature change and are caused quenched in various media. Cylinders were 75 mm (3 the operating temperature of an air furnace by thermal expansion and contraction, di- in.) in diameter by 230 mm (9 in.) long. Cooling rate was measured from 400 to 290 °C (750 to .555 °F). Stress is changed, waiting periods are required mensional changes of more permanent char- range is maximum tensile stress plus maximum com- after the furnace instrument indicates sta- acter are encountered during heat treat- pressive stress. bility, to allow the furnace walls to stabilize ment. These changes are of several types, at the new temperature. The magnitude of some of mechanical origin and others this limitation is directly proportional to the caused by changes in metallurgical struc- thin material and in thin sections of parts efficiency of the furnace as an insulated ture. Changes of mechanical origin include that contain variations in thickness. For chamber, but possibilities of such radiation those arising from stresses developed by thick-section products or parts, changes in should be recognized even in thin-wall fur- gravitational or other applied forces, from external shape may be small because of naces. thermally induced stresses or from relax- rigidity, but the interior-to-surface temper- Radiation effects are potentially danger- ation of residual stresses. Dimensional ature gradients that form with rapid cooling ous because they often cannot be detected changes also accompany recrystallization, create residual stresses; these stresses nor- by ordinary thermocouples. Specially pre- solution, and precipitation of alloying ele- mally are compressive at the surfaces and pared radiation panels with thermocouples ments. tensile in the interior. attached are used, and their readings are Solution Heat Treatment. Distortion as a As previously discussed, warpage or dis- compared with adjacent free thermocou- result of creep during solution heat treat- tortion of thin-section material can be re- pies. These panels normally are made of ment should be avoided by proper loading duced by using a quenching medium that material of the same gage as the thinnest of parts in baskets, racks, or fixtures, or by provides slower cooling; however, cooling parts to be heat treated and should have a provision of adequate support for long piec- must be sufficient to produce the required single surface area of about 650 cm 2 (100 es of plate, rod, bar, and extrusions heat properties. Slower quenching can also re- in.2). A thermocouple is attached to the treated in horizontal roller hearth furnaces. duce the magnitude of residual stresses in center of the panel by welding or peening. Sheet is provided with air-pressure support thicker parts or pieces, as shown in Fig 9 for In order to detect the maximum effect, in continuous heat-treating furnaces to cylindrical specimens of alloy 6151 panel surfaces should be darkened so that avoid scratching, gouging, and distortion. If quenched in cold or boiling water. Stress their emissivity is at least as high as that of parts are to be solution heat treated in range (maximum tensile stress plus maxi- any material to be processed. During the fixtures or racks made of materials (such as mum compressive stress) for a cylinder with test, the panel surfaces should be parallel to steel) with coefficients of thermal expansion a radius of 89 mm (3.5 in.) is about 205 MPa the suspected source or recipient of radia- lower than that of the aluminum being treat- (30 ksi) when the cylinder is quenched in tion. As an example of the number of panels ed, allowance should be made for this dif- cold water but less than 70 MPa (10 ksi) required, several aerospace companies ferential expansion to ensure that expansion when it is quenched in boiling water. The specify one panel for every 1.5 linear meters of the aluminum is not restricted. Straight- effects of average cooling rate through the (5 linear feet) of furnace wall. ening immediately after solution heat treat- temperature range from 400 to 290 °C (750 Instrument Calibration. All instruments ing may be preferable to fixturing. to 550 °F) on longitudinal stress ranges and thermocouples must be accurately cal- Solution of phases formed by major alloy- developed in alloy 2014 cylinders 75 mm (3 ibrated, and it is essential that the calibra- ing elements causes volumetric expansion in.) in diameter are shown in Fig 35. tions be traceable directly to the National or contraction, depending on the alloy sys- High stresses induced by rapid quenching Bureau of Standards. The chain of trace- tem, and this may have to be taken into generally are reduced only modestly by the ability should consist of not more than four account in heat treatment of long pieces. precipitation heat treatments used to pro- links for sensing elements and three links For example, solution heat treatment and duce T6- or T8-type tempers. Consequent- for measuring elements. To illustrate, if the quenching of alloy 2219 causes lengthwise ly, for the alloys that require rapid cooling article calibrated by the National Bureau of contraction of about 2 mm/m (0.002 in.fin.). to develop the properties of these tempers, Standards is called a primary standard, then Solution heat treatment and quenching of those incorporating mechanical stress relief the chain of traceability of measuring ele- alloys of the 7xxx series is accompanied by (Tx51, Tx52) usually are specified when ments should consist of primary standard, lengthwise expansion--about 0.6 mm/m substantial metal must be removed to pro- test potentiometer, and furnace instrument. (0.0006 in.fin.) for alloy 7075 rod or plate. duce final shapes. Other T8-type tempers, Similarly, the chain for sensing elements Quenching. The most troublesome such as T86 and T87, also have low residual should consist of primary standard, second- changes in dimensions and shape are those stress as a result of the stretching required ary standard, test thermocouple, and fur- that occur during quenching or that result to produce them. nace thermocouple. Every effort should be from stresses induced by quenching. Due to Heat Treatments for Precipitation and Sta- made to ensure that the temperature indi- its nonuniform cooling, quenching may pro- bilization. The most significant dimensional cated by the furnace instruments is as close duce warpage or distortion, particularly in changes associated with precipitation heat 876 / Heat Treating of Nonferrous Alloys treatments and stabilizing heat treatments The T3- and T4-type tempers are the least Yield strength, ksi arise from concurrent dilution of the solid stable dimensionally because of possible 60 65 70 75 50 I I I I solution (which changes lattice parameter) precipitation in service. Alloys 2024 and its i 7075-T6 and formation of precipitate. Changes in variants have the smallest dimensional Alclad sheet density and specific volume resulting from change in aging; the total change from the 40 these changes in metallurgical structure are quenched to the average state is of the order 180 specimens f the reverse of those caused by solution of of 0.06 mm/m (0.00006 in./in.), less than the from a single sheet~ the alloy phases. However, because the change due to a temperature variation of 3 3O strongest tempers are those in which the °C (5 °F). These alloys therefore can be used j4290 routine precipitate is present in nonequilibrium in the T3- and T4-type tempers, except for mill tests transition forms, the amount of change dur- precision equipment. For all other alloys, 2O J ing precipitation heat treatment does not T6- or T8-type tempers should be used, \ totally compensate for the previous (and because in these tempers all the alloys have opposite) change that occurred during solu- good dimensional stability. lo 1 I tion heat treatment. Most of the heat-treat- Stability of Precision Equipment. Proper able alloys expand (grow) during precipita- maintenance of high-precision devices, 0 tion heat treatment. Exceptions are alloys such as gyros, accelerometers, and optical 400 425 450 475 fi00 525 of the 7xxx wrought series and the 7xx.O systems, requires use of materials in which Yield strength, MPa casting series, which exhibit contraction. dimensional changes from metallurgical in- Comparison of distribution of yield strength In alloys of the 2xxx series, the amount of stability are limited from 10 i~m/m (10 ixin./ Fig 3(} in heat-treated 7075-T6 clad sheet product growth decreases with increasing magne- in.). Several laboratory investigations and with distribution in a single sheet. A is 95% probability sium content. Thus, growth of about 1.5 considerable practical experience have that not more than 1% of all material will fall below this mm/m (0.0015 in./in.) can be expected dur- shown that wrought alloys 2024 and 6061 value; B is 95% probability that not more than 10% of all material will fall below this value. (A and B refer ing precipitation heat treatment of alloy and casting alloy 356.0 are well suited and only to curve representing 4290 routine mill tests.) 2219-T87, about 0.5 mm/m (0.005 in./in.) for generally preferred for such applications. treatment of alloy 2014-T6 and less than 0.1 Dimensional changes were no greater than mm/m (0.0001 in./in.) for treatment of alloy l0 ~xm/m when alloys 2024-T851 and -T62, erties allow the use of tensile properties 2024-T851. Alloys 7050 and 7075, on the 6061-T651 and -T62, and 356.0-T51, -T6, alone as acceptance criteria. The minimum other hand, contract about 0.3 mm/m and -T7 were tested for more than a year at guaranteed strength is ordinarily that value (0.0003 in./in.) on precipitation heat treating room temperature and for several months at above which it has been statistically pre- from the W temper to the T6 temper and 70 °C (160 °F), and then the same alloys dicted with 95% probability that 99% or about 0.7 mm/m (0.0007 in./in.) on treating were tested with repeated thermal cycling more of the material will pass. The inherent from the W temper to the T73 temper. between 20 and -70 °C (68 and -94 °F). variability within lots and among specimens Stabilizing T7-type treatments cause great- Because stresses applied or induced by from a given piece is shown in Fig 36. er amounts of growth than the T5-, T6-, or acceleration in such devices generally are Testing provides a check for evidence of T8-type treatments for the same alloys. This not high, strength levels lower than those of conformance; process capability and pro- increased growth is associated either with the highest-strength tempers frequently are cess control are the foundations for guaran- formation of increased amounts of transition satisfactory. To increase precision of ma- teed values. precipitates or with transformation of transi- chining to intended dimensions, as well as Published minimum guaranteed values tion precipitates to equilibrium phases. to promote maximum stability, it is com- are applicable only to specimens cut from a mon practice to apply additional thermal specific location in the product, with their Dimensional Stability in Service treatments for stress relief and precipitation axes oriented at a specific angle to the of l to 2 h at temperatures of 175 to 205 °C direction of working as defined in the appli- Dimensional stability of heat-treated (350 to 400 °F) after rough machining. These cable procurement specification. In thick parts in service depends on alloy, temper, additional treatments sometimes are repeat- plate, for example, the guaranteed values and service conditions. Of the latter, ex- ed at successive stages of processing, and apply to specimens taken from a plane cluding mechanical conditions such as ap- even after final machining. In addition, it midway between the center and the surface, plied loads, the most important is service has been claimed that one or two cyclic and their axes parallel to the width dimen- temperature range relative to the range in treatments consisting of cooling to -100 °C sion (long transverse). Different properties which precipitation occurs. Residual stress- (-150 °F), holding for 2 h, heating to 232 to should be expected in specimens taken from es constitute another source of dimensional 240 °C (450 to 465 °F) and again holding for other locations, or in specimens whose axes changes. Stress relief minimizes changes 2 h can improve dimensional stability of were parallel to thickness dimension (short due to residual stresses, and most mill prod- 356-T6 castings. transverse). However, the specified "refer- ucts usually are supplied in tempers that ee" locations and orientations do provide a include stress relief. Potential dimensional Quality Assurance useful basis for lot-to-lot comparisons, and change as a result of further precipitation in constitute a valuable adjunct to other pro- parts that operate at elevated temperatures Quality-assurance criteria that heat-treat- cess-control measures. is minimized for wrought products by use of ed materials must meet always include min- Tensile tests can be used to evaluate the T7-type stabilizing treatments and for cast- imum tensile properties and, for certain effects of changes in the process, provided ings by use of T5-type treatments. Howev- alloys and tempers, adequate fracture specimens are carefully selected. A varia- er, components of high-precision equip- toughness and resistance to detrimental tion in process that produces above-mini- ment, such as instruments for aerospace forms of corrosion (such as intergranular or mum properties on test specimens, howev- guidance systems and optical and telescopic exfoliation attack) or to stress-corrosion er, is not necessarily satisfactory. Its devices, may require special supplementary cracking. All processing steps through heat acceptability can be judged only by compar- treatments during manufacture to further treatment must be carefully controlled to ing the resulting properties with those de- reduce stresses or subsequent precipitation. ensure high and reliable performance. veloped by the standard process on similar- (These treatments are discussed below, un- Tensile Tests. In general, the relatively ly located specimens. Finally, variations in der "Stability of Precision Equipment.") constant relationships among various prop- heat-treating procedure are likely to affect Heat Treating of Aluminum Alloys / 877

600 Table 13 Typical acceptable hardness values for wrought aluminum alloys 85 Acceptable hardness does not guarantee acceptable properties; acceptance should be based on acceptable hardness plus written evidence of compliance with specified heat-treating procedures• Hardness values 550 80 higher than the listed maximums are acceptable provided that the material is positively identified as the correct alloy• 75 Hardness 500 Alloy and temper Product form(a) HRB HRE HRH HRI5T 7(] 2014-T3, -T4, -T42 All 65-70 87-95 45(] 2014-T6, -T62, -T65 Sheet(b) 80-90 103-110 6s All others 81-90 104-110 2014-T61 All 100-109 (](] ~_ 2024-T3 Not clad(c) 69--83 97-106 111-118 82.5-87.5 I-- 400 Clad, -<1.60 mm (0.063 in.) 52-71 91-100 109-116 80--84.5 Clad, > 1.60 mm (0.063 350 in.) 52-71 93-102 109-116 50 2024-T36 All 76-90 100-110 85-90 2024-T4, -T42(d) Not clad 97-106 111-118 82.5-87.5 69--83 45 Clad, -<1.60 mm (0.063 3OO in.) 52-71 91-100 109-116 80-84.5 50 60 70 80 9(] 10(] Clad, >1.60 mm (0.063 Hardness, HRB in.) 52-71 93-102 109-116 Fig 37 Tensile strength versus hardness for various 2024-T6, -T62 All 74.5-83.5 99-106 84-88 aluminum alloys and tempers 2024-T81 Not clad 74.5-83.5 99-106 84-88 Clad 99-106 2024-T86 All 83-90 105-110 87.5-90 6053-T6 All 79-87 74.5-78.5 • Use a specimen that has at least 19 cm 2 (3 6061 -T4(d) Sheet 6O-75 88-100 64-75 in. 2) of surface area Extrusions; bar 70-81 82-103 67-78 Not clad, 0.41 mm • Remove any cladding by filing or etching 6061-T6 (0.016 in.) 75-84 • Clean the specimen by immersing it for l Not clad, ->0.51 mm min in a solution containing 5% concen- (0.020 in.) 47-72 85-97 78--84 trated nitric acid and 0.5% hydrofluoric Clad 84-96 6063-T5 All 55-70 89-97 62.5-70 acid at a temperature of 95 °C (200 °F); 6063 -T6 All 70-85 rinse in distilled water. Immerse for 1 rain 6151-T6 All 91-102 in concentrated nitric acid at room tem- 7075-T6, -T65 Not clad(e) 85-94 106-114 87.5-92 perature; rinse in distilled water Clad: -<0.91 mm (0.036 in.) 102-110 86-90 • Immerse the specimen for 6 h in a freshly >0.91 ~ 1.27 mm prepared solution containing 57 g of sodi- (>0.036 ~ 0.050 in.) 78-90 104-110 um chloride and l0 mL of 30% hydrogen >1.27 -< 1.57 mm peroxide per liter of water at a tempera- (>0.050 <- 0.062 in.) 76-90 104-110 >1.57 ~ 1.78 mm ture of 30 -+ 5 °C (86 +- 9 °F). More than (>0.062-< 0.070 in•) 76-90 102-110 one specimen may be corroded in the >1.78 mm (0.070 in.) 73-90 102-110 same container provided that at least 4.6 7079-T6, -T65 All(e) 81-93 104-114 87.5-92 mL of solution is used for each square 7178-T6 Not clad(t) 85 min 105 min 88 min Clad: centimeter (30 mL/in. 2) of specimen sur- -<0.91 mm (0.036 in.) 102 min 86 min face and that the specimens are electrical- >0.91 -< 1.57 mm ly insulated from each other (>0.036 <- 0.062 in.) 85 min • After the immersion period, wash the >1.57 mm (0.062 in.) 88 min specimen with a soft-bristle brush to re- (a) Minimum hardness values shown for clad products are valid for thicknesses up to and including 2.31 mm (0.091 in.); for heavier-gage material, cladding should be locally removed for hardness testing or test should be performed on edge of sheet. (b) 126 to 158 HB (10 move any loose corrosion product. Cut a mm ball, 500 kg load). (c) 100 to 130 HB (10 mm ball, 500 kg load). (d) Alloys 2024-T4, 2024-T42 and 6061-T4 should not be rejected for cross-sectional specimen at least 19 mm low hardness until they have remained at room temperature for at least three days following solution treatment. (e) 136 to 164 HB (10 mm ball, 500 kg load). (f) 136 HB min (10 mm ball, 500 kg load) (3/4 in.) long through the most severely corroded area; mount and metallographi- cally polish this specimen • Examine the cross-sectional specimen mi- the relationships among tensile properties influences the type of corrosion attack and croscopically at magnifications of 100x and other mechanical properties. In appli- the corrosion resistance. With thin-section and 500x both before and after etching cations where other properties are more products quenched at rates sufficiently rapid with Keller's reagent important than tensile properties, the other to prevent precipitation in the grain bound- • Describe the results of the microscopic properties should be checked also. aries during the quench, short periods of examination in terms of the five degrees Hardness tests are less valuable for ac- precipitation heat treating produce localized of severity of intergranular attack illus- ceptance and rejection of heat-treated alu- grain boundary precipitates adjacent to the trated in Fig 38 minum alloys than they are for steel. Nev- depleted areas, producing susceptibility to ertheless, hardness tests have some utility intergranular corrosion. Additional heating, Electrical Conductivity. For control of the for process control. Typical hardness val- however, induces extensive general precipi- corrosion and stress-corrosion characteris- ues for various alloys and tempers are given tation within the grains, lowering the corro- tics of certain tempers, notably the T73 and in Table 13. Figure 37 shows the general sion potential differences between the grains T76 types, the materials must meet combi- relationship between longitudinal tensile and the boundary areas, thus removing the nation criteria of yield strength plus electri- strength and hardness for aluminum alloys. cause of the selective corrosion. cal conductivity. Although these criteria are Intergranular-Corrosion Test. The extent The most common test for susceptibility based on indirect measurements of proper- of precipitation during elevated-temperature to intergranular corrosion is carried out as ties, their validity for ensuring the intended aging of alloys 2014, 2219, and 2024 markedly follows: corrosion and stress-corrosion resistance 878 / Heat Treating of Nonferrous Alloys

Alloy Condition Product form may differ from those employed to produce (a) ~--- 2048 T8 Sheet and plate the same temper in another alloy. 2124 T3, T8 Sheet and plate Designations for the common heat-treat- 2419 T8 Sheet, plate, extrusions, and ed tempers, and descriptions of the se- forgings quences of operations used to produce 7049 T7 Plate, forgiugs, and extrusions 7050 T7 Sheet, plate, forgings, and those tempers, are given in the following extrusions paragraphs. (For the entire aluminum alloy 7150 T6 Sheet and plate temper designation system, including desig- 7175 T6, T7 Sheet, plate, forgings, and nations for non-heat-treatable alloys, see 1¢) extrusions 7475 T6, T7 Sheet and plate Volume 2 of this Metals Handbook series.) Basic temper designationsfor heat-treated conditions include the codes O, W, and T. Other basic temper designations are F (as (e) The fracture toughness of these alloys and fabricated) and H (strain hardened). tempers range in measured Ktc values from O, annealed. Applies to wrought prod- Five degrees of severity of intergranular at- about 20 MPaV~ (18 ksiv'~.) upward. Con- ucts that are annealed to obtain lowest Fig 38 tack. Severity of intergranular attack (sche- trolled-toughness alloys are often derivatives strength temper and to cast products that matic), as observed microscopically in transverse sec- tions after test for susceptibility to intergranular of conventional alloys. For example, 7475 are annealed to improve ductility and di- corrosion. Top of each area shown in surface exposed alloy is a derivative of 7075 with maximum mensional stability. The O may be followed to corrosive solution compositional limits on some elements that by a digit other than zero. were found to decrease toughness. W, solution heat treated. An unstable In products of the newer controlled- temper applicable to any alloy that naturally has been firmly established by extensive toughness high-strength alloys 2090, 2091, ages (spontaneously ages at room tempera- correlation and testing. 2124, 2224, 2324, 7050, 7149, 7150, 7175, ture) after solution heat treatment. This Low tensile strengths may be accompa- 7475, and 8090, which provide guaranteed designation is specific only when the period nied by high levels of electrical conductiv- levels of fracture toughness, minimum val- of natural aging is indicated--for example, ity, so electrical conductivity is sometimes ues of the applicable indices, K~c or K¢, W I/z h. (See also the discussion of the Tx51, used as a quality-assurance diagnostic are established by accumulation of statis- Tx52, and Tx54 tempers, in the section tool. However, because the correlation tical data from production lots as a basis below on subdivision of the T temper.) between strength and electrical conductiv- for guaranteed minimum values. If the T, heat treated to produce stable tempers ity is strongly a function of chemical com- minimum specified fracture toughness val- other than O. Applies to products that are position and fabricating practice, use of ue is not attained, the material is not thermally treated, with or without supple- electrical conductivity is not recommend- acceptable. Some specifications allow use mentary strain hardening, to produce stable ed except for rough screening. This of less-expensive screening tests (such as tempers. The T is always followed by one or screening must be followed by hardness the notch tensile or chevron-notched short more digits, as discussed below. testing, and then by tensile testing if the bar) as a basis for release of high-tough- Major Subdivisionsof T Temper. In T-type hardness tests indicate that the heat treat- ness alloy products. In these instances, designations, the T is followed by a number ment was suspect. correlations between K~c and the screening from 1 to 10; each number denotes a specif- Fracture Toughness Indices. Fracture test result is used to establish the appro- ic sequence of basic treatments, as de- toughness is rarely, if ever, a design con- priate notch-yield ratio as a lot-release scribed below. sideration in the 1000, 3000, 4000, 5000, criterion. TI, cooled from an elevated-temperature and 6000 series alloys. The fracture tough- shaping process and naturally aged to a ness of these alloys is sufficiently high that Temper Designations for substantially stable condition. Applies to thicknesses beyond those commonly pro- products that are not cold worked after an duced would be required to obtain a valid Heat-Treatable Aluminum Alloys elevated-temperature shaping process such test. The temper designations used in the Unit- as casting or extrusion, and for which me- Fracture toughness is a meaningful de- ed States for heat-treatable aluminum alloys chanical properties have been stabilized by sign-related parameter for some conven- are part of the system that has been adopted room-temperature aging. If the products are tional high-strength alloys and all the con- as an American National Standard (ANSI flattened or straightened after cooling from trolled-toughness, high-strength alloys. H35.1). Used for all wrought and cast prod- the shaping process, the effects of the cold Conventional aerospace alloys for which uct forms except ingot, the system is based work imparted by flattening or straightening fracture toughness minimums may be use- on the sequences of mechanical or thermal are not recognized in specified property ful in design include 2014, 2024, 2219, treatments, or both, used to produce the limits. 7075, and 7079. These alloys have tough- various tempers. The temper designation T2, cooled from an elevated-temperature ness levels that are inferior to those of follows the alloy designation and is separat- shaping process, cold worked, and natural- their controlled-toughness counterparts. ed from it by a hyphen. Basic temper des- ly aged to a substantially stable condition. Consequently, these products are not used ignations consist of individual capital let- Applies to products that are cold worked in fracture-critical applications, although ters. Major subdivisions of basic tempers, specifically to improve strength after cool- fracture toughness can be a meaningful where required, are indicated by one or ing from a hot-working process such as design parameter. Fracture toughness is more digits following the letter. These digits rolling or extrusion, and for which mechan- not guaranteed in conventional high- designate specific sequences of treatments ical properties have been stabilized by strength alloys. that produce specific combinations of char- room-temperature aging. The effects of cold Fracture toughness quality control and acteristics in the product. Variations in work, including any cold work imparted by material procurement minimums are appro- treatment conditions within major subdivi- flattening or straightening, are recognized in priate for controlled-toughness, high- sions are identified by additional digits. The specified property limits. strength alloys. The alloys and tempers conditions during heat treatment (such as T3, solution heat treated, cold worked, currently identified as controlled-tough- time, temperature, and quenching rate) and naturally aged to a substantially stable ness, high-strength products include: used to produce a given temper in one alloy condition. Applies to products that are cold Heat Treating of Aluminum Alloys / 879 worked specifically to improve strength af- strength after they have been precipitation Temper designations T42 and T62 have ter solution heat treatment, and for which heat treated. been assigned to wrought products heat mechanical properties have been stabilized TIO, cooled from an elevated-tempera- treated from the O or the F temper to by room-temperature aging. The effects of ture shaping process, cold worked, and demonstrate response from the heat treat- cold work, including any cold work impart- artificially aged. Applies to products that ment described below. Temper designations ed by flattening or straightening, are recog- are cold worked specifically to improve T42 and T62 also may be applied to Wrought nized in specified property limits. strength after cooling from a hot-working products heat treated from any temper by T4, solution heat treated and naturally process such as rolling or extrusion, and for the user when such heat treatment results in aged to a substantially stable condition. which mechanical properties or dimensional the mechanical properties applicable to Applies to products that are not cold stability, or both, have been substantially these tempers. worked after solution heat treatment, and improved by precipitation heat treatment. • T42. Solution heat treated from the O or for which mechanical properties have been The effects of cold work, including any cold the F temper to demonstrate response to stabilized by room-temperature aging. If the work imparted by flattening or straighten- heat treatment and naturally aged to a products are flattened or straightened, the ing, are recognized in specified property substantially stable condition effects of the cold work imparted by flatten- limits. • T62. Solution heat treated from the O or ing or straightening are not recognized in Other Subdivisions of T Temper Codes for the F temper to demonstrate response to specified property limits. Stress-Relieved Products. When it is desir- heat treatment and artificially aged T5, cooled from an elevated-temperature able to identify a variation of one of the ten shaping process and artificially aged. Ap- major T tempers described above, addition- Subdivision of the O Temper. In temper plies to products that are not cold worked al digits, the first (x) of which cannot be designations for annealed products, a digit after an elevated-temperature shaping pro- zero, may be added to the designation. following the O indicates special character- cess such as casting or extrusion, and for The following specific sets of additional istics. For example, O1 denotes that a prod- which mechanical properties or dimensional digits have been assigned to stress-relieved uct has been heat treated according to a stability, or both, have been substantially wrought products. time/temperature schedule approximately improved by precipitation heat treatment. If Tx51, stress relieved by stretching. Ap- the same as that used for solution heat the products are flattened or straightened plies to the following products when treatment, and then air cooled to room after cooling from the shaping process, the stretched to the indicated amounts after temperature, to accentuate ultrasonic re- effects of the cold work imparted by flatten- solution heat treatment or after cooling sponse and provide dimensional stability; ing or straightening are not recognized in from an elevated-temperature shaping pro- this designation applies to products that are specified property limits. cess: to be machined prior to solution heat treat- T6, solution heat treated and artificially ment by the user. aged. Applies to products that are not cold Product form Permanent set, % worked after solution heat treatment, and Plate 11/2-3 REFERENCES for which mechanical properties or dimen- Rod, bar, shapes, extruded sional stability, or both, have been substan- tube 1-3 1. S. Hirano Quench Sensitivity in Drawn tube ~/2-3 et al., tially improved by precipitation heat treat- A1-Li Based Alloys, Proceedings of Con- ment. If the products are flattened or ference on Aluminum-Lithium Alloys straightened, the effects of the cold work Tx51 applies directly to plate and to rolled (Vol 1), Materials and Component Engi- imparted by flattening or straightening are or cold finished rod and bar. These products neering Publications, 1989, p 335-344 not recognized in specified property limits. receive no further straightening after 2. T. Sheppard, Mater. Sci. Technol., Vol T7, solution heat treated and stabilized. stretching. Tx51 also applies to extruded 4, July 1988, p 636 Applies to products that have been precip- rod, bar, shapes, and tubing, and to drawn 3. J.E. Hatch, in Aluminum Properties and itation heat treated to the extent that they tubing, when designated as follows: Physical Metallurgy, American Society are overaged. Stabilization heat treatment • Tx510. Products that receive no further for Metals, 1984, p 165-166 carries the mechanical properties beyond straightening after stretching 4. C.E. Bates, Selecting Quenchants to the point of maximum strength to provide • Tx511. Products that may receive minor Maximize Tensile Properties and Mini- some special characteristic, such as en- straightening after stretching to comply mize Distortion in Aluminum Parts, J. hanced resistance to stress-corrosion crack- with standard tolerances Heat Treat., Vol 5 (No. 1), 1987, p 27-40 ing or to exfoliation corrosion. • Tx52, Stress relieved by compressing. Ap- 5. T. Croucher, Critical Parameters for TS, solution heat treated, cold worked, plies to products that are stress relieved Evaluating Polymer Quenching of Alu- and artificially aged. Applies to products by compressing after solution heat treat- minum, Heat Treat., Vol 19 (No. 12), that are cold worked specifically to improve ment, or after cooling from a hot-working Dec 1987, p 21-25 strength after solution heat treatment, and process to produce a permanent set of 1 6. W.L. Fink and L.A. Willey, Quenching for which mechanical properties or dimen- to 5% of 75S Aluminum Alloy, Trans. AIME, sional stability, or both, have been substan- • Tx54. Stress relieved by combining Vol 175, 1948, p 414-427 tially improved by precipitation heat treat- stretching and compressing. Applies to 7. J.W. Evancho and J.T. Staley, Kinetics ment. The effects of cold work, including die forgings that are stress relieved by of Precipitation in Aluminum Alloys dur- any cold work imparted by flattening or restriking cold in the finish die. (These ing Continuous Cooling, Metall. Trans, straightening, are recognized in specified same digits--and 51, 52, and 54---may be A, Vol 5A, Jan 1974, p 43-47 property limits. added to the designation W to indicate 8. J.T. Staley, Industrial Heating XLIV, T9, solution heat treated, artificially unstable solution heat-treated and stress- Oct 1977, p 6-9 aged, and cold worked. Applies to products relieved tempers) 9. G.F. Bobart, J. Heat Treat., Vol 6 (No. that are cold worked specifically to improve 1), 1988, p 47-52