Ironmaking & Steelmaking: Processes,Products and Applications

The present paper describes the application of a neural network technique for the prediction of tensile strength (TS) test results for steels heat treated using a batch heat treatment process. Industrial process data often contain outlying points, some of which can be spurious for a number of reasons. A data cleaning technique has been used to ensure that spurious data points are not present in the final neural model, which would otherwise hinder the model's representation of the true process. The effectiveness of this technique is demonstrated by comparison of the TS model trained on cleaned and uncleaned data. A model trained on cleaned data is generated for the prediction of TS in the form of an ensemble network, which was found to provide more reliable predictions and give a better representation of the degree of uncertainty in the network predictions. The performance of the model is evaluated from a metallurgical perspective. Application areas for the model are examined with particular attention being drawn to the need for caution when entering inputs into the neural model. An assessment of the model's ability to generalise to new treatment sites and new steel compositions is made, together with experimentation to determine the effect of measurement tolerances on the predicted values from the model.

Cold bonded ore–coal composite pellets developed on a laboratory scale (Part 1 of this two part paper) were tested in a rotary kiln sponge iron plant. This plant had a 12 m length refractory lined rotary kiln of 8 t/day iron ore throughput capacity. Kiln operation was optimised to achieve more than 90% metallisation of sponge iron using ore–coal composite pellets. The kiln productivity increased and energy consumption decreased, compared with levels obtained when sized lump ore was used as iron oxide feed material. The plant trials established the suitability of composite pellets for sponge ironmaking with enhanced kiln productivity.

A mathematical model simulating flow fields was developed and used in a study of the influences of scaffold and coal injection on gas and liquid distributions in the blast furnace. It was found that the peripheral gas flow in the upper part of the furnace would become more pronounced with scaffold formation in the walls of the bosh. It was also revealed that total pressure drop and the central gas flow at the furnace top would increase at high coal injection rates. The simulation mode is useful for acquiring a deeper understanding of the complex phenomena in the blast furnace and for determination of appropriate operational actions under different production conditions.

The effects of flow control devices and buoyancy forces on the melt flow in a large tundish have been mathematically simulated using a k–ϵ turbulence model. Flow control devices included arrangements consisting of a pair of weirs and a pair of dams, a turbulence inhibitor and a pair of dams, and only a turbulence inhibitor. Buoyancy forces were simulated using step inputs of temperature and inputs of varying ladle stream temperature into the tundish. It was found that with inputs of hot steel, flow control devices improve performance by driving the melt upwards through the action of buoyancy forces. A bare tundish is less sensitive to buoyancy forces and shows greater thermal mixing than any other arrangement. Inputs of temperature steps promote higher temperature gradients of liquid steel inside the vessel than inputs of varying ladle stream temperatures. A turbulence inhibitor delays the thermal disturbance compared with a bare tundish or a tundish with a weir–dam arrangement. When using a turbulence inhibitor higher volume fractions of melt obey a plug like flow. The dimensionless quantity Gr/Re ², where Gr and Re are the Grash of and Reynolds number respectively, quantifies buoyancy forces: high values indicate that buoyancy forces have more effect than inertial forces on fluid flow. When Gr/Re ² <5, buoyancy forces have no noticeable influence on fluid flow.

Unacceptable shape defects in light gauge strip products may be introduced as a result of differential reduction being applied across the width of the material at rolling mills. Consumer specifications and quality procedures demand that the resulting defects are removed by subsequent processing operations. Such an operation is tension levelling, which attempts to remove the resulting defects such as edge waves by ensuring that all longitudinal material 'fibres' possess the same length. Shape removal is achieved by subjecting the material to a series of alternate bends under superimposed tension. Consequently, imbalanced residual stress fields may be present in the as levelled strip, resulting in bowed material in a further downstream slitting operation. Much work has been published on the mechanics of the process, and many advances have been made, especially in the design of levelling systems to improve shape removal and reduce internal stress levels. However, residual stresses and specifically the imbalance of these stresses post-levelling is an issue that has failed to be addressed in its entirety. To investigate the generation and effects of residual stresses owing to tension levelling, a finite element (FE) model has been developed with validation carried out on an experimental laboratory leveller. The analysis is performed in real time using the commercial FE code Abaqus. The present paper gives details of investigations carried out to examine the effectiveness of material work hardening models and the associated implications for residual stress fields; validation of the geometric modelling and boundary conditions are also discussed. Results for the analyses reported here show that a characteristic (longitudinal) residual stress field is produced across the width of the material as a fundamental consequence of the levelling process. The current model produces predictions comparable to experimental results from the laboratory leveller.

High noise levels and edge overcoating problems have been encountered in the gas wiping of hot dip galvanised layers on steel strip in a continuous line. These have been examined and analysed. Noise levels were decreased by up to 10 dBA by using edge baffles to reduce direct collision of wiping air jets. Average overcoating of zinc in the zone within 50 mm of the strip edge could be decreased by 50% or more during air wiping by using edge baffles with a baffle to strip distance of 20 mm or less. The extent of edge overcoating has been related to the edge drop of the gauge of the substrate. If the combined profile of the edge overcoating and the edge drop of the substrate exhibited a noticeable high spot of about 10 μm in the edge zone, edge buildup formed along the high spot, mostly in the edge area 10–30 mm from the edge. Edge baffles have effectively reduced this type of edge buildup.

The suitability of temperature as a parameter to determine strain failure mode in press forming is examined by thermographic evaluation of the forming limit diagram (FLD). The work reported in the present paper forms part of a research programme to develop condition monitoring technology for press forming, the highlights of which are discussed. Experimental methodologies are described. The inability to use temperature to distinguish between the left and right hand sides of the FLD is reported. A strong correlation between major strain and fracture temperature is also observed. From this, it is concluded that it may be possible to perform condition monitoring of press forming, by a technique termed 'thermal line analysis'.

A new drawing die designed to reduce the residual stresses in bar after drawing is presented. The die design consists of two consecutive straight taper portions, and the outlet portion has a very small die angle. The marked beneficial effect of adding the outlet portion was predicted by finite element modelling (FEM), and subsequently verified by measuring the residual stresses in drawn bars. The optimum range of the outlet die angle was determined using FEM. The new design requires a slightly longer die length than the conventional one, but it is short enough to be adopted in current production lines.