Machining of Metal-Composite Stacks and Hybrid Aerospace Materials through Milling and Abrasive Water Jet

Thesis (Ph.D.)--University of Washington, 2019Recent trends in aerospace industry towards reducing emissions, weight and buy-to-fly ratio has led to extensive use of high strength hybrid materials and structures. These materials range from brittle carbon fiber reinforced plastic (CFRP) to ductile Titanium, and their hybrid combinations. Often, secondary machining such as edge trimming and drilling is required to meet the geometric tolerances and functional requirements. Owing to their inhomogeneous nature, the machining of materials such as metal-composite stacks and cellular carbon foam is challenging. Conventional machining often results in defects and damages such as delamination, fiber pullouts and fuzzing in CFRP, burr formation and thermal distortion in titanium and chipping in carbon foam. In addition, significant tool wear especially at high machining speeds often questions the feasibility of conventional machining. Abrasive Water jet is a high speed alternative to traditional machining but is limited by kerf striations, width variation and delamination. In this study, Abrasive Water Jet (AWJ) is developed and demonstrated as a versatile tool for a wide range of materials. Three material systems were undertaken - discontinuous fiber composites (DFC), open-cell carbon foam, and hybrid metal-composite alliance (Titanium-CFRP (continuous fiber) stacks and Titanium-Graphite fiber metal laminates). The machinability of these materials was primarily evaluated in terms of machining forces and kerf characteristics - surface roughness and kerf geometry. Throughout the study, two process variables were used - hydraulic pressure and jet traverse speed. A simplified lumped parameter- Jet power-to-speed was effectively used to understand the kerf responses in terms of kerf geometry and surface quality. Random chopped discontinuous fiber composites (DFC) is rife with resin rich hot spots due to the random distribution of prepreg chips. Upon comparison with conventional trimming, a conspicuous surface topology was resulted in AWJ machining. The sub-surface damage which was concealed beneath the smeared matrix, was occasionally visible at the top and bottom side. Next, machinability of carbon foam was studied with parametric evaluation and surface characterization. Carbon foam is a composite tooling material and consists of thin walled, open-cell structure which is susceptible to breaking, chipping and tearing during the machining process. The open cell structure also provides an opportunity to the jet to expand and intrude in the free spaces resulting in divergent and uneven kerf geometry. Morphological inspection revealed some common material removal mechanisms - erosion and cell shape distortion, cell and pore engulfment, ligament fractures, debris entrapment with cell wall cracking. Erosion and shape distortion was caused at low jet energy while cell wall damage along with cracks originating from pores were common at high pressure conditions. Besides, the conventional techniques in assessing and characterizing the surface roughness were incompetent due to the large cell size (average 1 mm diameter). A novel algorithm was proposed in which discrete wavelet transform was employed on the surface profile map. The wavelet parameters (decomposition level and mother wavelet) were identified based on the strong relationship of the filtered roughness profile with the process parameters. The proposed method can be easily extended to any cellular material like ceramic an metallic foams. Next, machining of Titanium-CFRP was studied in three phases. First phase was the machining of thin stacks and second phase was partial depth penetration of stacked Titanium and Ti-CFRP. A simplistic model using the jet energy and principle of superposition was developed to predict the kerf geometry. In the third phase a full depth penetration was studied with Ti-CFRP and CFRP-Ti stacking sequence. An energy based semi-analytical model was developed to predict the kerf shape. A major development from this research is an approach to monitor the AWJ process performance. A multi-sensor study revealed the suitability of Acoustic Emission signals. A novel algorithm was developed which used Wavelet packet analysis to decompose the signals and characteristic features were calculated. Features such as energy-entropy coefficient and standard deviation were identified as best classifiers and a strong correlation was identified with both process parameters and surface quality. A comparative assessment with several time domain and conventional spectrogram monitoring techniques showed the superiority of the proposed analysis with coefficient of determination as high as 97%. Similar approach was developed to monitor conventional milling process in milling unidirectional, multidirectional and discontinuous fiber composites. Upon selection of optimal process parameters through aforementioned studies, the effect of edge trimming condition on the surface integrity was studied through low velocity impact of DFC and fiber-metal-laminates. The sensitivity of energy absorption and damage evolution on the edge trimming condition was revealed. In general up to 42% degradation in energy absorption was resulted due to the marginal surface quality, including those machined conventionally, thus meriting the importance of machinability studies and process monitoring. Overall AWJ was demonstrated as a single tool to machine hard-to-cut, hybrid and inhomogeneous materials and structures. The study further integrated jetting technology with metal Additive Manufacturing, and explored the versatility of AWJ through erosion and surface treatment of 3D printed titanium