NASA recommends homogenizing large aluminum 2219 ingots before forging them

Aluminum 2219 is an age-hardenable, over-saturated aluminum-copper alloy, developed by Alcoa in 1954 for service up to 600 degrees Fahrenheit ^[4]. In a badly processed ingot the copper stays concentrated in the boundary zones between dendrites ^[8]. Homogenization holds the ingot at temperature so that diffusion, following Fick's laws, spreads those copper atoms evenly through the aluminum grains ^[8][10]. According to the bulletin, residual phases dissolve into the matrix and segregation of every element drops sharply ^[8].

The order of steps matters because the damage can be permanent. The bulletin accepts that forging, solution treatment, quenching and aging can partly mitigate as-cast defects ^[7]. Once an ingot has interdendritic segregation, banded clusters of large copper intermetallics and mixed grain sizes, it says, no later thermomechanical processing will fix the resulting properties, and larger ingots are the worst case ^[7].

The bulletin's firmest numbers describe the metal's structure. After upset forging and rolling, followed by solution treatment and aging, the area fraction of coarse Al2Cu particles fell from 5.5% to 1.0% as they dissolved into the matrix ^[14]. That is a drop of about 82% ^[1]. Grain size fell from 230 to 58.6 micrometers ^[15], roughly a fourfold refinement ^[2]. The bulletin credits the finer grains to stored energy and nucleation during the lower-temperature rolling ^[15].

The comparison runs from the starting state to the finished one across four processing steps ^[14], so it measures the whole route. It cannot tell you how much any single step contributed. In one of the studies the bulletin cites, superior properties came from multidirectional forging at 510 degrees Celsius followed by warm rolling at 240 C ^[16].

The available text of the bulletin does not include anodizing or corrosion test results. It also omits the values in the Wang et al. table comparing forged 2219 in the T6 temper with and without homogenization, though the bulletin calls the homogenized material "clearly superior" ^[13]. Link by link, the evidence shown supports the microstructural claims most directly. The link from those changes to poorer anodizing and more corrosion is stated in NASA's summary ^[1].

Scaling up from lab samples to plant ingots brings in section thickness. Wang et al. homogenized at 535 C for 10 hours ^[10]. The bulletin says the parameters must be tuned to each ingot's cross-section thickness so the whole section responds evenly to heat ^[9]. Its list of variables includes melting point, the amount and dissolution rate of the eutectic phase, ingot size, grain size and copper content ^[11]. I would treat the Wang schedule as a starting point for supplier trials on thick sections. The bulletin's concern is greatest for exactly those larger castings ^[3].

The listed contact for the bulletin is Donald S. Parker at Kennedy Space Center ^[17].