Traditionally, hot water is often selected as an alternative to cold-water quenching of heat-treatable aluminum alloys for distortion reduction. The levels of reduction are often modest, however, and more effective means are required.

Over the years, Type-I polymer quenchants – as defined by SAE AMS 3025 – have been used increasingly because the levels of distortion reduction typically are dramatically lower than that achievable with hot water while still meeting Mil Handbook 5 design minimums. Distortion and residual-stress reduction achievable with Type-I quenchants are discussed here, and reasons for this behavioral advantage are discussed.

Introduction

Aluminum is solution treated at temperatures generally in the range of 400-540°C (750-1000°F). During solution treatment, some alloying elements are re-dissolved to produce a solute-rich solid solution. The objective of this process is to maximize the concentration of hardening elements, including copper, zinc, magnesium and/or silicon in the solid solution. The concentration and rate of dissolution of these elements increases with temperature. Therefore, solutionizing temperatures are usually near the liquidus temperature of the alloy.[1,2]

If an aluminum alloy is slowly cooled from an elevated temperature, alloying elements precipitate and diffuse from solid solution to concentrate at the grain boundaries, at small voids, on undissolved particles, at dislocations and at other imperfections in the aluminum lattice.[2] For optimal properties, it is desirable to retard this diffusion process and maintain the alloying elements in solid solution. This is done by quenching from the solution temperature.

For quench-hardenable wrought alloys 2xxx, 6xxx and 7xxx and casting alloys such as 356, this is accomplished by the quenching process. The objective is to quench sufficiently fast to avoid an undesirable concentration of the alloying elements in the defect and grain-boundary structure. After quenching, aluminum alloys are aged, and a fine dispersion of elements and compounds are precipitated that significantly increase material strength. The diffusion process and precipitation kinetics varies with the alloy chemistry.

The cooling process of age-hardenable aluminum alloys not only affects properties such as strength and ductility, but it also affects thermal stresses. Thermal stresses are typically minimized by reducing the cooling rate from the solutionizing temperature. If the cooling rate is too slow, however, undesirable grain-boundary precipitation will result. If the cooling rate is too fast, there is an increased propensity for distortion. Therefore, one of the primary challenges in quench-process design is to select quenching conditions that optimize strength while minimizing distortion and avoiding undesirable properties such as intergranular corrosion, which is also cooling-rate dependent.[3]

Various quenchants and quenching processes are utilized in the heat treatment of aluminum, and many are reviewed in reference 4. Of these potential quenchants, however, cold water (10-32°C) is used most often in order to maximize mechanical properties.[5] When distortion or cracking is encountered, either hot water (60-71°C)[4] or a polymer quenchant is used.[6-9]

Currently, the most common polymer quenchant used throughout the aluminum heat-treating industry is an aqueous solution of a specific poly(alkylene glycol) copolymer (PAG).[10] This article will focus only on water and the PAG-based quenchant defined by AMS 3025 as Type I.[11,12] The objective of this article is to provide an overview of aluminum quenching and minimization of distortion and cracking.

Discussion

Cold water, especially when highly agitated, is an excellent quench medium for maximizing the strength of a quenched aluminum-alloy part. This is illustrated in Figure 1, where it is shown that the tensile strength of AA7075 decreases when the water temperature is greater than approximately 130-140°F (55-60°C). However, the relatively high cooling rates associated with cold-water quenching produce large temperature differentials between thick and thin sections that often lead to high residual stresses, distortion and, occasionally, cracking. Raising the water temperature reduces these temperature differentials and therefore may result in a corresponding reduction in residual stress and distortion.

The PAG copolymer that is perhaps most commonly specified for aluminum quenching applications is designated as “Type I” by AMS 3025.[11] A Type-I quenchant is a solution of water, a PAG copolymer and additives (such as corrosion inhibitors) and is defined by the physical properties shown in Table 1 for the as-delivered aqueous concentrate. In addition to the physical properties of the quenchant, there are specific performance requirements that are also required, including: cooling rates from aluminum-alloy probe cooling-curve data, tensile properties and intergranular corrosion resistance. It is wise for the user to examine all data required and specifically outlined in AMS 3025,[11] not just quenchant physical properties, to properly qualify the Type-I quenchant being used.

The physical property values in Table 1, taken together with the required performance properties, reasonably specify the composition of a Type-I quenchant. For example:

• Viscosity, at constant water content, is indicative of the copolymer molecular weight (size).
• Refractive index, while a very old characterization method, has traditionally been related to the molar refractance values (structure) of chemical compounds and their concentrations.
• Specific gravity, while not a critical parameter, is often used in chemical engineering processes as an indicator of composition and concentration.
• Diluted solution viscosity is a validation that the correct polymer molecular weight and concentration is being used.
• Separation temperature, as referred to here, is indicative of the chemical composition of the PAG copolymer being used to formulate the quenchant.

Therefore, these values reflect the chemical composition of the copolymer, molecular weight, concentration and additive content. While additional tests could be performed, they are probably not necessary for fresh, as-received quenchant from the supplier. In addition, the tensile properties, intergranular-corrosion resistance and alloy cooling rates must be determined for proper quenchant certification as already required by ASM 3025. The mechanical properties of the quenched-and-tempered aluminum alloys must at least meet the Mil Handbook 5 design minimums.[13] It is not sufficient to qualify a quenchant by determining the originating vendor. Recommended Type-I quenchant concentrations and heat-treating conditions for wrought and cast aluminum-alloy parts are provided in AMS 2770 and AMS 2771, respectively.[14,15]

The maximum attainable strength properties are dependent on the cooling rate between 750-550°F (400-290°C). Generally, faster cooling rates provide greater strengths, up to a limit. This is illustrated in Figure 2.[22] As Figure 3 shows, cooling rates decrease with increasing water temperature, and Figure 4 shows that cooling rates decrease with increasing polymer quenchant concentration.[5] However, increasing cooling rates produce increasing thermal gradients, which produce increasing thermal stresses and the potential for increased distortion. Thermal gradients are reduced by reducing the cooling rate. (This is why “hot” or boiling water is used.) There is a cooling rate “window” that must be identified to obtain both the design minimum optimal strength and minimum residual stresses and distortion. The position and width of the window is a function of the specific hardenability of the alloy and the geometry of the part.

Although hot-water quenching is commonly performed to reduce the thermal gradient between the surface and the core of aluminum alloys to reduce distortion, it is often insufficient to produce optimal minimization of distortion, particularly of thin-sheet and plate components. Figure 5 shows that hot-water (70°C) quenching (Figure 5B) produced essentially the same high total distortion measured on three 1.6-mm test panels of AA2024-T4 as cold-water (30°C) quenching (Figure 5A) although the variation of the distortion was somewhat less.

By comparing Figure 5C with Figure 5A and Figure 5B, dramatically lower distortion was obtained when a AA2024 test panel was quenched in a 12% solution of a Type-I quenchant at 40°C. This example, in addition to other published examples,[4,5,8] clearly shows the advantage of using a Type-I quenchant for distortion reduction when quenching aluminum alloys. In such cases, an aqueous PAG copolymer quenchant may be used. The distortion-reduction advantages of Type-I quenchants relative to both cold and hot water is illustrated in Figure 5.

In other studies, Suttie reported that a Type-I quenchant provided significant distortion reduction for 7075-T6 forgings relative to cold water while producing substantially higher tensile strengths than attainable with boiling water.[16] Similar results were reported by Collins and Maduell on the same alloy.[17]

Type-I polymer quenchants have been reported to offer significantly greater residual-stress reduction than hot water in a number of studies. The results of one of the earliest reported studies that was conducted using the Sach’s Bore-Out Method on aluminum A356 castings are shown in Figure 6.[18] The studies compare the residual stress obtained by cold-water quenching with varying concentrations of a Type-I quenchant. These data show the substantial reduction in the residual stress obtained as the concentration of the Type-I polymer quenchant is increased up to 30% by volume.

Torgerson and Kropp conducted extensive residual-stress and other mechanical-property comparisons between hot water and Type-I quenchant at various concentrations on 7050-T736 forgings and plate.[19] It was found that Type-I quenchants provided minimum distortion while still meeting the design minimums for forgings up to 5 inches thick.

Tensi, et.al compared cooling curves obtained by quenching an aluminum-alloy (AlMgSiCu) probe in a Type-I polymer-quenchant solution at 10% by volume (in distilled water at 25°C) and distilled water only, also at 25°C.[21] The temperature of both probe materials when quenched was 520°C (968°C). Both probes were cleaned with 600-grit abrasive paper before each test. The polymer film surrounding the probe surface ruptured simultaneously around the entire surface, also called “explosive” rewetting for the Type-I-polymer-quenched probe, whereas the rewetting process was slower and with three different boiling mechanisms often occurring simultaneously on the water-quenched probe surface. The slower rewetting process obtained with water leads to substantial thermal gradients during quenching and increased residual stresses and distortion relative to that obtained with a Type-I polymer quench. This non-uniform wetting process is the primary reason for the increased distortion observed with both the hot- and cold-water-quenched panels compared to the panel quenched in the Type-I polymer quenchant shown in Figure 5 and the increased residual stress shown in Figure 6.

Silver probes are also used to evaluate quench severity exhibited by different quenchants. (ASTM D7646 has been issued recently and provides a standard test method using silver probes to model aluminum quenching performance.)[20]

Conclusions

This discussion has shown that a Type-I PAG polymer quenchant, as defined in AMS 3025, will provide dramatically lower residual stress and distortion compared to that attainable with cold or hot water. The reason given for this is that the water-quenching mechanism involves an inherently non-uniform wetting process that produces substantially greater thermal gradients than the Type-I polymer quenchant. Although increasing the concentration of Type-I quenchants will produce correspondingly lower strength properties, the Mil Handbook 5 design minimums for the section size and alloy can be achieved by following the concentration use guidelines specified in AMS 2770 and AMS 2771. IH

For more information: Contact George E. Totten, Portland State University, G.E. Totten & Associates, LLC, PO Box 30108, Seattle, 98113: tel: 206-788-0188; fax: 815-461-7344; e-mail: getotten@getottenassociates.com. Other authors are Patricia Mariane Kavalco (patikavalco@gmail.com) and Lauralice C. F. Canale (lfcanale@sc.usp.br) of the University of São Paulo, Department of Materials, Aeronautical and Automobile Engineering, School of Engineering, São Carlos, SP, Brazil