Investigation into Manufacturing Complexities Intrinsic to Superplasticly Formed and Diffusion Bonded (SPF/DB) Components by Statistical Modeling, Simulation, and Experiments

Wednesday, March 13, 2024: 1:30 PM
E 216 C (Charlotte Convention Center)
Mr. Bryan Ferguson , University of Washington, Seattle, WA
Mr. Eric Bol , University of Washington, Seattle, WA
Mrs. Neha Kulkarni , University of Washington, Seattle, WA
Dr. Daniel G. Sanders, PhD Mech. Engr. , University of Washington, Lake Tapps, WA
Prof. M Ramulu , University of Washington, Seattle, WA
Superplastic forming is a sheet metal forming process that produces parts free of residual stresses, dimensionally accurate, and with strains otherwise unobtainable. Diffusion bonding is a solid-state welding operation which combines two relatively flat, clean surfaces at high temperature to create a near flawless weld over a large surface area. Combination of superplastic forming with diffusion bonding (SPF/DB) creates a phenomena, where under similar processing conditions, the material involved will produce a weld with itself, enabling a variety of reinforcing internal structures.

Most superplastic parts are blow formed to a die surface with small variations in thicknesses. This work, investigated a process unique to SPF/DB using four sheets of titanium. The two outer sheets are formed to the die surface, while the two unsupported inner sheets form a complex sandwich structure that is all diffusion bonded. Because superplasticity is stress-history dependent, and somewhat chaotic, the geometry of the inner sheets are free to shift and translate such that small variations in the initial geometry create large changes in the final geometry. Modelling of diffusion bonding has been challenging due to variations in voids formed at the mating surfaces.

A variety of techniques to measure the variances in forming were quantified using cell wall measurements, gas pathway measurements, and computer vision-based geometry analysis. Finite element simulations of the inner sheet forming process compared the experimental results with idealized geometry. A stochastic diffusion bonding model was implemented to estimate the probability of voids formed; it used a statistical version of Pilling’s model, combined with theoretical voids formed from interacting surfaces based on surface roughness initial conditions. The results of the model, and simulations, were compared with experimental results for three different titanium alloys at different process conditions. These findings provide insights into the complexities of manufacturing four sheet SPF/DB structures.