The machinability of TiNi shape memory alloy

Download Statistics

Downloads

Downloads per month over past year

Beck, Robert James (2011). The machinability of TiNi shape memory alloy. University of Birmingham. Ph.D.

Beck11PhD.pdf
Text - Accepted Version
Restricted to Repository staff only until 31 December 2026.
Available under License All rights reserved.
Download (27MB) | Request a copy

Abstract

The research involves an in-depth investigation into the machinability of a titanium-nickel (TiNi) shape memory alloy (SMA) which can potentially be applied to fabricate novel aerospace components and actuation systems. SMA materials posses the unique ability to undergo a dramatic change in properties due to a reversible microstructural phase transformation that may be induced by a temperature change or applied stress. Such functional properties however severely hinder manufacturing process operations, which is the principal reason that use of TiNi has hitherto been largely limited to wire products. The current work therefore concentrates on the machining of larger scale workpieces than previously reported and was supplied in the form of 3.5mm and 12mm thick rolled sections measuring ~65mm wide and ~0.8m long. Following a literature review focusing primarily on the material properties / characteristics and machinability of NiTi / TiNi shape memory alloys, five phases of experimental work were performed to assess the influence of key process variables on workpiece integrity. This comprised four different cutting processes together with subsequent fatigue performance testing of selected machined surfaces. The machining processes evaluated were electrical discharge wire machining (EDWM), twist drilling, slot milling and reciprocating surface grinding (RSG). These were selected due to their potential for forming part of the production route for a selected target component specified by Rolls-Royce. Phase 1 trials relating to EDWM of TiNi was carried out on two different machine tools (Agie Vertex and Charmilles FI240CC) utilising multiple pass (roughing and several trim operations) technology parameters predefined in the machine controller database. Experiments on the smaller Agie Vertex involved workpiece sections of 3mm and 60mm high and entailed the real time recording of voltage and current pulse profiles for analysis of the discharge energy. Microhardness evaluation indicated that the depth of damage after roughing passes was 90μm and 60 μm for the 3mm high and 60mm high workpieces respectively. This was reduced to 30μm for the 3mm specimen after 4 trim passes however there was no appreciable decrease in damage with the thicker workpiece sample following finishing passes. Tests on the Charmilles FI240CC machine employed workpieces 60mm and 150mm tall. Here, heat affected zones (HAZ) extending down to 60μm and 70μm beneath the surface respectively was observed following roughing however both decreased 30μm thick after the use of trim passes. As a result of the properties of the TiNi material, a novel method for the management of residual stress when applying multiple pass EDWM was developed to improve surface quality and part geometry when machining compliant parts by this process.

In Phase 2 work, drilling of Ø4mm through holes was investigated. Initial feasibility experiments resulted in random catastrophic failure of drills by fracture with tool life as short as 5mm drilled length. The mechanism for this failure was shown to be due to the excessive deflection of the workpiece material due to high cutting forces and low Young‟s modulus. A method for drilling through holes in TiNi was developed using a backing material to prevent the workpiece deflecting elastically under the action of the drill, this was shown to improve tool life and reduce burr formation. Subsequent mainstream tests evaluated the effects of tool geometry and resulted in the development of a bespoke drill suitable for TiNi. Tool life of up to 600 holes (~1500mm drilled length) was demonstrated when drilling at 30m/min 0.075mm/rev. Hole quality (in terms of burr formation and hole size) and surface integrity was evaluated. The slot milling process was assessed using Ø12mm corner radius end mills in Phase 3 testing. The influence of tool types (solid and insert tungsten carbide) and a range of operating parameters (cutting speed 15-45m/min, feed rate 0.05-0.1mm/tooth, depth of cut 0.5-1mm) were investigated and correlated against results from analysis of tool life, surface integrity damage and surface quality. A maximum tool life of 19200mm3 material removed was achieved when operating at a cutting speed of 30m/min, 0.05mm/tooth feed rate, and 1mm depth of cut, which was a 100 fold improvement over results reported in the literature. Machined surfaces showed evidence of extensive sub surface deformation although under certain conditions this was limited to ~40μm.

Phase 4 research assessed the performance of RSG on TiNi in a plunge grinding mode for the production of low aspect slot features. A range of conventional abrasive wheels (GC and WA wheels 46 – 80 grit) were initially evaluated to identify a preferred wheel type, which was then used in subsequent trials to evaluate the effect of grinding parameters. White alumina abrasive was shown to be ineffective, and a silicon carbide wheel (80 grit) was found to provide superior results. The machined surfaces were shown to have been affected predominantly by the thermal load on the workpiece, while microhardness evaluation showed damage up to 400μm from the surface under certain conditions although damage was reduced to 120μm through development of the process.

Phase 5 detailed high cycle fatigue testing to evaluate and compare the performance of machined surfaces produced using RSG, fine EDWM and rough EDWM processes. The testing was carried out at elevated temperature (150°C) to be representative of the target R-R application. RSG samples demonstrated the best performance (run out to 1.2x107 cycles at ~390MPa), while the fatigue strength of fine EDWM surfaces was ~70% of those prepared by RSG. Similarly the fatigue strength of rough EDWM surfaces was only ~43% of RSG surfaces. The stress for run-out to 1.2x107 cycles achieved for the rough EDWM surface represented a significant proportion of the „yield‟ stress for the material.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):