Project D3: Surface coatings and cladding
D3.1: Kinetic metallisation
•	Key researchers: K. Spencer, M.X. Zhang
Summary
Kinetic metallisation is being used to deposit metallic or composite coatings onto metal substrates at low temperatures. This technique is a low-pressure variation of cold spray designed to minimise the use of helium gas, which is the most effective (and expensive) process gas. The process temperature is well below the recrystallisation temperature of the coating and substrate, making it suitable for thermally-sensitive materials such as light metals and for amorphous materials. Results to date suggest that the compacted powder coatings form a strong metallurgical bond with the substrate, similar to that formed in explosive welding. There are three key research directions in the project.
An investigation of the effect of various Al-based metal matrix composite (MMC) coatings on corrosion and wear resistance is in progress. To date, pure Al coatings on Mg substrates have been studied and some preliminary results for corrosion and wear resistance have been obtained. In addition, fundamental aspects such as the coating formation mechanism and bond strength have been studied.
Preliminary work on corrosion resistance has shown that pure Al coatings on Mg alloy substrates do not exhibit good immersion corrosion resistance in the as-sprayed condition. Heat treating the Al coatings leads to substantially better corrosion resistance, approaching that of bulk Al (Figure D6).
Figure D6 - Pitted surface of pure Al coatings on a pure Mg substrate in a 5% NaCl solution, after a 24 hour immersion.
The bond strength of pure Al coatings on pure Mg has been studied, in addition to the bonding mechanism. In the as-sprayed condition the coating/substrate interface is stronger than the cohesive strength of the coating, and failure occurs inside the coating (Table D1). The effect of heat treatment on coating bond strength is complex, as there are different processes that occur simultaneously: annealing and softening of the heavily deformed coating material, sintering of the coating particles and the formation of a diffusion bond at the coating/substrate interface. Table D1 shows that heat treating a coating of pure Al on an Mg substrate will shift failure to the interface, though this may not be true for all material/substrate combinations.
Condition Bond Strength [MPa] Failure Mode As-Sprayed 29.8 Cohesive – within coating Heat Treated - 2 hours at 400°C 28.4 Adhesive – at coating/substrate interface Table D1: Bond strength measurements of pure Al powder KM sprayed on a pure Mg substrate
Mg-Al intermetallic layers have also been generated on AZ91 substrates (Figure D7). The layers have a high bond strength with the substrate, and have shown very good pitting corrosion resistance in 48 hour ASTM B117 salt spray tests as compared to exposed AZ91 (Figure D8). The sample on the right shows the exposed substrate, while the left and centre samples show the intermetallic layer; the surface is largely undamaged, showing only minor pitting.
Figure D7 - Mg-Al intermetallic layers on an AZ91 Mg substrate.				Figure D8 - Intermetallic layer face after 48 hour salt spray test, 5%NaCl (left and centre); exposed AZ91 substrate (right).
Key items of infrastructure acquired to facilitate this work include:
•	Usage of the Kinetic Metallisation coating development system, which was commissioned in December 2006.
•	M-2000 sliding wear and abrasion testing machine.
D3.2: Cold spray
•	Key researchers: S. Li, J. Soria
Summary
This project deals with the numerical modelling of the Cold Gas Dynamics Spray process. The present numerical investigation of supersonic impinging jets is being carried out with two objectives in mind.
1.	The mean flowfield properties of varying degree of underexpanded jets impinging on a flat plate for different impingement distances is established.
2.	The interaction between the injected particles and shock structures within the flow is studied for particles of different sizes and mass loading.
The method used is the conservation form of the non-dimensionalised Navier-Stokes equations being used to model the single phase impinging jet with additional particle-gas interaction terms included for the two phase jet.
The results for a round supersonic impinging jet have been generated. The supersonic jet is issued from a converging-diverging nozzle with an exit angle of 15 degrees at a pressure ratio (PR = Pexit/Pambient) of 1.2 and Mach number of 2.2. The impingement plate is placed at a distance of 2 diameters from the nozzle exit. All flow variables plotted have been normalised with respect to exit conditions. The length is normalised by the nozzle exit diameter, the density by ρe, velocity by the speed off sound ae, temperature by Te pressure by aeρe2.
The current computational domains are defined by the nozzle exit (bottom left), the jet centreline (bottom) and impingement plate (right). The velocity contour plot in Figure D9 is overlayed with the velocity vector field. A vertical plate shock is evident from the discontinuity in vector length and contour magnitude. The shock spans across most of the jet, causing an outward deflection of jet flow and a reduction in flow velocity. In order to locate the position of the plate shock, the variation of flow properties along the jet centerline is plotted.				Figure D9 - Velocity contour at PR = 1.2, Ma = 2.2, z=2
Figure D10 shows the gradual expansion of jet followed by a sudden decrease in the u component velocity at approximately x/D = 0.4 which is an indication of the plate shock. This corresponds with sudden increases in both pressure and temperature. After the flow has been re-directed along the plate the velocity increases again inside the wall jet region where a series of expansion and compression cells is observed.
The flow within the impingement region appears stagnant (i.e no recirculation bubble) and the pressure distribution along the impingement plate supports this.
Figure D10 - Centreline profile at PR = 1.2, Ma = 2.2, z=2