The Science of Forming

A year or two ago, print media praised the forming of a very complex automotive trunk lid with only a single die. The secret ingredient was a superplastic aluminum that stretched into very sharp features without necking or other failure modes. The aluminum stretched to strain levels way beyond any other material. Much superplastic forming research is not new but occurred during the 1960s at the Massachusetts Institute of Technology. Two laboratory demonstrations showed the fascinating capabilities of this very special aluminum alloy. In one demonstration, a standard round tensile test specimen (0.505-in. dia.) reached total elongations ranging from 2000 to 3000 percent before failure. In some tests, the specimen approached the diameter of a human hair without failure.

The other demonstration began with a box having 1-ft. square sides. The box contained one of the old spherical typewriter balls set in the center of the bottom of the box. This ball had raised letters and numbers around the surface. In the typewriter, the ball rotated to imprint each letter or number into the inked tape. The forming process began with clamping the sheet of superplastic aluminum over the top of the box to ensure a tight air seal. Then a vacuum pump removed the air from the sealed box. The aluminum sheet gradually sank into the box and took the shape of the inside of the box by forming all inside zero-radius edges and corners of the box. In the process, the aluminum sheet also wrapped itself around the type ball and embossed every letter and number through the sheet to create a duplicate of the original ball. Those two demonstrations made complete believers of everyone who witnessed the tests.

Why didn’t the capability of this alloy

Three stress-strain curves illustrate the increase in strain after the onset of a diffuse (width) neck, marked by downward arrow plus N. This additional strain is proportional to the m-value.

immediately spread throughout the entire forming world? Mother Nature does not provide something for nothing. Almost all new capabilities come with a price. This time, creating the superplastic behavior incurs several different costs. First, the aluminum requires very special composition, microstructure and processing. Second, the sheet of aluminum is heated to a specific temperature. Third, the deformation must be very slow.

Once a new material capability is discovered, product designers immediately want to incorporate the new features into future part designs. Potential application by a huge industry (such as automotive) provides the driving force for additional research to further focus the material characteristics to end-user needs. In this case, elongations of 2000 to 3000 percent are far greater than target parts required. Therefore, metallurgists tweaked the chemistry and processing to reduce cost. New microstructures allowed an increase in forming speed. Ultimately, the total part cost equaled the value of the benefit received and the process went into production.

One fundamental measure of superplastic behavior is the strain-rate hardening exponent or m-value.

The strain-rate hardening equation is similar to the strain hardening equation containing the n-value. When a strain gradient, local neck, or other localization of strain begins to terminate useful deformation in sheetmetal, the strain rate in the localization (?´ in the equation) must increase relative to the surrounding material. A positive m-value means that the material becomes stronger as the local forming speed increases and resists or delays the localization. A zero m-value means that the strain rate has no effect. A negative m-value means that the material becomes weaker than surrounding material and accelerates the localization.

The m-value effect is shown in the illustration. For all three alloys, the onset of the diffuse or width neck occurs at the load maximum marked by the downward arrow and letter N. The zero m-value for the 20-30 brass means no change in strength and the neck forms rapidly (little additional strain) between the load maximum (arrow + N) and the forming limit curve (angular slash near the end of the curve). For aluminum-killed draw-quality (AKDQ) steel, the 0.012 m-value shows an additional 16 percent strain before reaching the forming limit curve (FLC). For the Zn-Ti (zinc-titanium) alloy, the 0.06 m-value adds 42 percent more stretch before the FLC terminates useful deformation.

In the illustration, the n-value is proportional to the amount of strain at the load maximum. Looking at the stress-strain curves, one may incorrectly conclude that the m-value increases as the n-value decreases. Instead, these three curves illustrate that the forming limit strains in the tensile test relate to the total effect of the n-value and the m-value. Why then are m-values generally ignored for steels? Additional research showed that the m- and n-values are proportional for steel. Therefore, the traditional n-value obtained automatically during computerized tensile testing provides sufficient information without additional tests utilizing speed changes.

Returning to superplastic forming, one can now understand why this process generates such tremendously high strains. The Zn-Ti alloy in the graph has an m-value of only 0.06. The early work at MIT required a minimum m-value of 0.5 for superplastic forming. Since the m-value is an exponent, the effect becomes even more powerful with increasing m-value.

The search for high positive m-values goes on. Recent work presented at the SAE International Congress in Detroit showed that warm forming of another aluminum alloy raised the m-value from negative to positive and thereby increased the total elongation from 25 to 125 percent. Competition among producers of different alloys continues to drive many improvements to target customer needs. MF