Papers by Author: I.C. Howard

Abstract: This paper presents a modelling strategy that combines neuro-fuzzy methods to dene the material model with cellular automata representations of the microstructure, all embedded within a nite element solver that can deal with the large deformations of metal processing technology. We use the acronym nf-CAFE as a label for the method. The need for such an approach arises from the twin demands of computational speed for quick solutions for ecient material characterisation by incorporating metallurgical knowledge for material design models and subsequent process control. In this strategy, the cellular automata hold the microstructural features in terms of sub-grain size and dislocation density which are modelled by a neuro-fuzzy system that predicts the ow stress. The proposed methodology is validated on a two dimensional (2D) plane strain compression nite element simulation with Al1%Mg alloy. Results from the simulations show the potential of the model for incorporating the eects of the underlying microstructure on the evolving ow stress elds. In doing this, the paper highlights the importance of understanding the local transition rules that aect the global behaviour during deformation.

Abstract: Charpy upper shelf energy is widely used as a fracture controlling parameter to estimate the crack arrest/propagation performance of gas transportation pipeline steels. The measurement of this fracture criterion particularly for modern steels and its apportion into different components, i.e. fracture and non-related fracture energy, are of great importance for pipeline engineers. This paper presents the results of instrumented Charpy impact experiments on high-grade pipeline steel of grade X100. First, the instrumentation technique including the design and implementation of a strain gauge load-cell and the details of the data-recording scheme are reviewed. Next, the experimental data obtained from the Charpy impact machine so instrumented are presented and discussed. These include the test data from full and sub-sized Charpy V-notched specimens. The instrumented Charpy machine was able to capture the load history in full during the fracture process of the test specimens resulting in a smooth load-time response. This eliminated the need for filtering used in similar test techniques. From the recorded test data the hammer displacement, impact velocity and fracture energy were numerically calculated. The results showed that there was a significant drop in hammer velocity during the impact event. This resulted in a change in the fracture mode from dynamic to quasi-static which was more appreciable for full-size Charpy test samples. As a result, sub-sized specimens might be preferable for impact testing of this steel in order to guarantee the conditions of dynamic crack propagation in the specimen ligament. Accurate analysis of the instrumented impact test data showed that the ratio of crack initiation energy to propagation energy was around 30% for the X100 steel. It can be concluded that in impact testing of high-grade pipeline steel a significant portion of overall fracture energy is consumed in non-related fracture processes. This high fracture initiation energy should be accounted for if the current failure models are going to be used for toughness assessment of highstrength low-alloy gas pipeline steels.

Abstract: This paper reports recent results from a set of experimental and computational studies of ductile flat fracture in modern gas pipeline steel. Experimental data from plain and notched cylindrical tensile bars and standard C(T) specimens together with damage mechanics theories have been used to capture the flat fracture characteristics of a gas pipeline steel of grade X100. The modelling was via finite element analysis using the Gurson-Tvergaard modified model (GTN) of ductile damage development. The assumption of effective material damage isotropy was sufficiently accurate to allow the transfer of data from the notched bars to predict, in a 2D model, the crack growth behaviour of the C(T) specimen. This was in spite of the considerable ovalisation of the bars at the end of their deformation. However, it was not possible to obtain similar accuracy with a 3D model of the C(T)test, even after a large number of attempts to adjust the values of the GTN parameters. This, and the anisotropic shape change in the tensile bars, suggests very strongly that the damage behaviour is so anisotropic that conventional models are not good enough for a full engineering description of the flat fracture behaviour. Suitable averaging (of shape) in the modelling of the notched bar data, and the companion averaging associated with the 2D model of the C(T) data provide a relatively fast way of transferring engineering data in the tests. There is a discussion of potential ways in which to incorporate 3D effects into the modelling for those purposes where the considerable increase in computational time (due to the microstructurally-sized finite elements needed to capture the damage behaviour) is acceptable in order to include through-thickness effects.

Abstract: The grid technique is an experimental method for measuring the deformation in hot rolling. An AA3004 sample -fitted with an insert - was rolled in a single hot rolling pass at 400 oC. The insert was hand engraved with a 1x1 mm grid and the analysis of the image of the deformed grid enabled the calculation of the components of the deformation gradient tensor. In order to prevent relative motion between the insert and the work-piece, four steel pins were used; after the test no detachment was observed between insert and sample. The temperature was monitored during rolling using two embedded thermocouples, one close to the surface and the other on the centre-line of the slab. The commercial finite element (FE) code ABAQUS was used to build a threedimensional model of the rolling process. The recorded temperature was compared with the FE values evaluated after tuning the heat transfer coefficient. The FE model was run several times with different friction coefficients and the deformation gradient checked against the experimental measurement of the deformed grid in order to obtain the optimum friction coefficient. The experimentally determined deformation gradient and the measured temperature agreed well with the numerical values.

Abstract: A common feature that stimulates modelling efforts across the various physical sciences is that complex microscopic behaviour underlies apparently simple macroscopic effects. Mathematical formulations attempt to capture the initial and evolving microstructural entities either implicitly or explicitly and link their effects to measurable macroscopic variables such as load or stress by averaging out any microscopic fluctuations. The implicit formulations that ignore the inherent spatial heterogeneity in the deforming domain form the basis of constitutive models for input to finite element (FE) systems. On the other hand, explicit formulations to capture and link microstructural entities rely on narrowing down the size of each finite element, thereby increasing the number of finite elements in the deforming domain, an effect accompanied by a rapid growth in computational time. The model described here, Cellular Automata based Finite Elements (CAFE), utilises the Cellular Automata technique to represent initial and evolving microstructural features (e.g., dislocation densities, grain sizes, etc.) in C-Mn steels at an appropriate length scale by linking the macro-scale process variables obtained using an overlying finite element mesh. Differences will be illustrated between single and two-pass hot rolling experiments.