Atomistic studies of dislocation-interface interactions in lamellar Ti-Al

Lead Research Organisation: Queen's University of Belfast
Department Name: Sch of Mathematics and Physics


The metallurgy of high temperature alloys presents challenges both at a technical and at a fundamental level. Alloys that are used in turbines at high temperature have to resist high stresses over prolonged timescales and high flow rates of noxious and corrosive gases. Moreover they are often found in critical applications where failure may lead to human distasters. This is because as well as finding applications in electricity generation, high temperature alloys are now found in the high temperature zones of jet engines, both civil and military.The superalloy nickel aluminide was invented and first produced by the Mond Nickel Company in 1941. It is still the workhorse of the industry despite extensive searches for lighter alloys having the same mechanical and chemical properties. The difficulty clearly resides in the need to meet the most stringent criteria simultaneously: high temperature oxidation and sulphidation resistance; high temperature strength and creep resistance; and finally low temperature ductility. The latter requirement is essentially a manufacturing, not a service condition. Unless a metal has at least a few percent room temperature ductility it's just too difficult to work with in manufacture and installation. A component may not even survive being dropped on a concrete floor!Titanium aluminides have recently emerged as promising new candidates to replace the excellent, but weighty gamma / gamma-prime nickel based superalloys. Our proposed work is concerned with their plasticity and ductility at room temperature. Although monolithic TiAl intermetallic is very brittle, a remarkable alloy produced by certain heat treatments shows ductility at least at the few percent level. This alloy, called (polysynthetically twinned) PST-TiAl contains grains (crystals) with a microstructure of layers ( lamellae ) of two phases, one of these phases (gamma-TiAl) being itself separated into layers having different crystal orientations. This sandwich structure is thermodynamically stable and is now the object of intense study in academia and industry worldwide.Our remit is atomistic simulation. We take large collections of atoms in a computer, bonded together by some model of interatomic forces (this may be fully quantum mechanical, or a classical ball and spring, or anything in between). Then we observe their behaviour as they relax under the influence of interatomic forces (molecular statics) or move at some temperature following Newton's equation of motion (molecular dynamics). Here we do particularly difficult simulations, namely of crystal dislocations. These are two dimensional crystal defects, the carriers of slip (plastic deformation) in metals. Such simulations are demanding because of the complexity of these defects, greatly amplified in alloys compared to pure metals; and because of the nature of the associated long ranged stresses. The rewards of a successful simulation are very great. We may view directly in the computer the motions of dislocations. We may follow their progress through perfect crystal and as they interact and possibly penetrate the lamellar boundaries. We will learn how slip is transmitted within and between lamellae and uncover the secret of why the PST titanium aluminide is ductile. We propose to use results of atomistic simulations to construct a multiple pile up model to predict behaviour at a length scale on the order of the size of a grain. This is called bridging the length scale gap. Armed with this model one should then be able to improve ductility and we shall be able to make such suggestions to the practical alloy designers.Further reading: Alloys by Design, Tony Paxton, Physics World, vol 5, No 11, p 35(Nov 1992) Electron Theory in Alloy Design, Edited by D G Pettifor andA H Cottrell, (Institute of Metals, 1992) Interatomic Forces in Condensed Matter, Mike Finnis, (OUP 2003)
Description All objectives have been met, except that a macroscopic pile up model is not yet complete. The quantum embedding problem is solved in principle and has been demonstrated in a computer programme written by Dr Katzarov and Dr Pashov. A paper is in preparation. Dr Katzarov is now employed under an FP7 project at QUB in which he plays a leading role. A new EPSRC proposal is in preparation. The student Dr Pashov has been award his PhD and is now completing a one year post working for the Numerical Algorithms Group (NAG) under the dCSE scheme. In July 2013 he will take up a post doctoral position at King's College London in the Physics Department.

We studied the dislocation (ds'n) core geometries that the lattice produces in proximity to lamellar boundaries (LB) and in which the interfaces affect the activation of ds'n slip inside the gamma lamellae and transfer of slip across lamellar boundaries. Core structures are modified by the elastic field and chemical presence of LBs. We arrive at a detailed understanding of the different glide mechanisms. We have discovered three new phenomena:

1. Two new roles for the interfaces emerge: dramatic reduction of the Peierls stress for both ordinary and superds'ns when they lie in the 60 and 120 degree boundaries and stabilisation of the glissile, planar core of superds'ns as a consequence of greatly reduced complex stacking fault (CSF) energy. Conversely, the CSF energy is increased with respect to bulk at the 180 degree lamellar boundary

2. Transformation of superds'ns within the 60 degree and 120 degree lamellar boundaries and subsequent detachment of an ordinary ds'n which is released into the lamella.

3. Emission of a twinning ds'n following the blocking of an ordinary ds'n after gliding in the 60 and 120 degree LBs.

When the loading axis is parallel to the LBs we would expect slip to occur on planes that intersect them. Instead one observes no macroscopic shape change in the direction perpendicular to the LBs. This unusual and curious phenomenon (channelled flow) has been supposed to result from the action of slip and twinning and we have been able to confirm this.

1. The interfacial step at a LB is cancelled by a single 60 degree ordinary ds'n.

2. Dipole nucleation and lateral twin growth is initiated at the same Peierls stress as the motion of ordinary ds'ns.

3. The Peierls stress for longitudinal growth of twins is significantly lower than that for ordinary ds'ns.

4. A 60 degree ordinary ds'n in monolithic gamma-TiAl has non planar core and a Peierls stress of 0.065. At stresses close to this it emits Shockley partial dipoles as it glides. The same ds'n inside a microtwin does not dissociate and has much lower Peierls stress.

5. 120 and 60 degree LBs act as obstacles to transmission of both twins and ds'ns. For the first time ever, we have modelled the lateral growth of twins.

In glide of the 60 degree ordinary ds'n in monolithic gamma-TiAl we find that modest stress causes the ds'n to emit a dipole of partials onto the cross slip plane after which a small increase in stress causes it to glide on its slip plane. As it glides one expects the stress to fall slightly and this will allow it to pause and emit a further dipole before it continues to glide. The picture emerges of a ds'n under a stress close to the Peierls stress moving between self-locking events at which it emits a Shockley partial dipole. This jerky flow is consistent with observation, but the atomistic mechanism that we advance is a new one.

We have examined the core structure of screw superds'n finding a complex set of core structure, whose cores may transform under stress and violate Schmid's law. Superds'n activity dominates at lower temperatures. We now suggest that at low temperatures ds'ns, which are sessile and hardly contribute to strain, are products of the decomposition of superds'ns.

Finally we have moved right up the length scale, using kinetic Monte Carlo to arrive at a new understanding of the pinning of ds'ns that results in the famous yield anomaly. We have shown that the effect is extrinsic.
Exploitation Route The design of high temperature superalloys
Sectors Energy,Transport

Description European Union Framework 7
Amount € 307,723 (EUR)
Funding ID 263335 
Organisation European Commission 
Department Seventh Framework Programme (FP7)
Sector Public
Country European Union (EU)
Start 05/2011 
End 04/2015