Subjects Covered

An holistic knowledge of rolling theory, mill operating practices and product metallurgy is desirable for engineers, operations and technical personnel charged with the responsibility of optimising mill performance and product quality. The final material properties are a function of pre-rolling processing, the rolling process and post-rolling processing. The development of material properties through the various manufacturing stages, including rolling, are considered. The main sheet products are considered: steel, aluminium and copper/ brass.

A review of actuators and mill responses to various process and actuator disturbances leads to the design of thickness, tension and temperature controls for tandem hot finish mills. Both mills with interstand loopers and tensiometer rolls are considered.

This introductory lecture introduces a range of important physical concepts whose understanding is critical to an appreciation of the rolling process. The notation, units and symbols employed in the Hatch IAS course and published information are also described so that all delegates have a common “language” in which to converse. The various phenomena occurring in the rollgap deformation zone between the work rolls are explained in detail and include: yield stress behaviour of plastic deformation of common materials, physical properties, frictional effects, surface interactions between the strip and the work roll, definition of forward slip. Finally a summary of the characteristic differences between the different types rolling is presented, covering hot, cold, tinplate, temper and foil mills.

The equations describing the two dimensional deformation of the metal in the rollgap and their use in predicting force, torque and slip are presented. This lecture focuses on the circular arc models, providing a brief derivation the model developed by Orowan in 1943. This model employs the iterative solution of the roll deformation as described by the Hitchcock (1935). The contributions of the elastic compression and recovery regions and inhomogeneous deformation in rollgap modelling are included. The lecture also presents a simplified graphical model which highlights the effect of changes in tension, yield stress, friction and strip thickness on the roll force and slip. Typical pressure distributions for difference rolling scenarios and expected model accuracy conclude the presentation.

Rollgap sensitivity analysis allows the use of a linear model to represent a non-linear rolling process. A rollgap sensitivity is a single number representing the change in one rolling condition with respect to another. A computer model can generate a sensitivity coefficient, which will be applicable for the simulated rolling condition. The sensitivity can be used to estimate for example the change in roll force cause by a change in entry thickness or the change in entry thickness required to achieve a specified roll force change. Alternatively a sensitivity coefficient could be measured on a rolling mill and used for future reference on a particular product.

This lecture extends on the material covered in the rolling theory lecture, examining those circumstances where the circular arc model fails to adequately predict the strip deformation. Alternative models are examined and the non-circular arc model proposed by Fleck et al. is presented. A comparison of the circular arc and non-circular arc models highlights their differences. A number of particular rolling scenarios including skin pass rolling, asymmetric rolling foil and temper rolling are covered. The impact of lubrication and friction on foil rolling and surface roughness in temper rolling are studied. The lecture concludes with a brief coverage of roll touching or kissing.

The trends towards tighter tolerances for width, thickness and finish temperature of hot rolled strip, and towards thinner finish thicknesses, have driven the development of more accurate rollgap models for online use. These new models are replacing older, less sophisticated models which were generally tabular in form, schedule specific and heavily dependent on adaption for effective performance. The influence of temperature, recrystallisation, precipitation and phase transformations complicate the modelling effort. The key issues to be addressed are considered and developments in the modelling of hot yield stress, friction, slip and strip temperature are reviewed. An indication of the expected performance that can be achieved in the strip quality parameters with these new models is provided.

The thermal flows within the mill impact upon shape and profile via the thermal expansion of the roll, on the strip properties via the strip temperature change, the lubricant via the heat generated during rolling and thermal fatigue of the rolls. Mill thermal analysis examines the rollgap heat flows which allows the derivation of strip temperature models for the rollgap and between the stands. Particular attention is given to the interstand behaviour of hot steel and aluminium and runout table strip cooling. This often requires the two dimensional analysis of the temperature distribution in hot rolling mills.

The development of the roll thermal expansion or camber is studied. Experimental measurements on roll temperature are used to study the internal heat flows which occur in the roll. The solution of the roll temperature field and subsequent thermal expansion are described for the case of continuous steady state rolling. Simple algorithms to predict the thermal camber behaviour are described. The lecture moves on to show how the development of thermal camber over time for both continuous and batch rolling can be predicted. The response of thermal camber to a single spray, a situation which occurs in shape spray control systems is highlighted. The lecture finishes with practical issues in the measurement of parameters required for the prediction of thermal camber.

This lecture gives an introduction to definitions and causes of thickness profile, strip shape and flatness of strip. A parametric representation of measured flatness and thickness profile is described. The origins of thickness profile and shape in both the hot mill and cold mill regimes are discussed. Apart from the typical rollgap interaction affecting these defects, other causes and effects due to friction, lubrication, sprays, product qualities, mill alignment and temperature are important. Causes of flatness defects that may not necessarily be manifest in a shapemeter display but affect final product quality can be critical and difficult to diagnose. Shape parameter sensitivities to roll force, roll bending and thermal camber are important for calculating controller gains for shape control systems.

While modelling of rolling is well covered in the literature there is little published about the strategies behind the calculation of pass thicknesses and drafts to satisfy mill equipment constraints and strip flatness and profile requirements, or the calculation of pass speeds and spray patterns to achieve the required strip surface quality and last pass exit temperature. Several drafting strategies are compared, and the normalised drafting strategy is considered in some detail, as it is a robust, versatile and practical method for online use. A limited discussion of width drafting strategy is provided. Some results from an online scheduling algorithm based on a normalised drafting strategy are presented.

The objective of the mill setup strategy is to provide a consistent and reliable prediction of threading references which, in conjunction with the experience and skill of the roller, will achieve the most efficient rolling operation. Constraints for the scheduling algorithms are defined. There are two major components – the strategies for scheduling tensions, thickness reductions and last stand roll force and the calculation of actuator references based on models and physics. In particular, a tension strategy based on mill exit thickness, a reduction strategy based on a normalised reduction technique and last stand roll force based on specific force is described. A calculation of control actuator references for threading of tandem cold mills to minimise off-thickness material and threading delays can significantly effect the efficiency of operation. Schedule dependent controller gains are calculated to improve thickness and tension control.

While modelling of rolling is well covered in the literature there is little published about the strategies behind the calculation of stand thicknesses and reductions, interstand tensions or the calculation of mill exit speed to satisfy strip temperature requirements. Different thickness reduction algorithms are compared and an algorithm for interstand tensions is provided. The influence of temperature on the setup model calculations for a hot finishing mill is considered. A number of algorithms for calculating the runout table spray quantity for steel rolling are presented.

The operation and performance of a rolling mill is influenced by many factors, a key one being the design, speed, accuracy and robustness of the actuators. An introduction to AC and DC drive motors, hydraulic and electric screw rollgap position actuators, roll bending actuators and electric, hydraulic and pneumatic interstand loopers in hot finishing mills are presented. The merits of each type of actuator are discussed, with typical response times and design limitations.

The modelling of strip flatness and profile is described in terms of its component models. These include the three dimensional deformation in the rollgap, the rollgap entry and exit tension stress analysis, the roll stack deflection and the interstand creep. One important aspect of shape and profile development is the tension stress feedback mechanism which occurs in the rollgap. This feedback loop is a predominant mechanism in the development and control of shape and profile. After development of the model a number of model applications are studied including the control of profile and shape in hot and cold rolling tandem mills and the impact of spread and thermal camber on shape and profile.

A conceptual design of width control systems, typically employed at the roughing and finishing stands of steel hot strip mills, is described. The lecture focuses on the many physical phenomena present in rolling which modify the final strip width. These include: spread in flat rolling, edging pass efficiencies, slab end effects in reversing mills, width changes due to interstand creep and bending stresses and thermal expansion behaviour. Typical practices and the control of slab end effects by dynamic edger position control is addressed.

The principle concepts and design issues of the various thickness controllers employed on single stand mills are presented, through the development of process and control models. The strengths and weaknesses of each controller are compared. Model based time delay compensation for faster tuning of conventional feedback controllers with feedback path time delay is explained. Peripheral issues such as mill stretch, speed effects and tension interactions are discussed also. This lecture is an excellent introduction for the more challenging issue of designing thickness and tension controls for tandem mills. The thickness controllers described are also used on tandem cold mills and hot mills.

Many of the concepts and issues in the design of thickness and tension controls for tandem mills are not intuitive. Unlike in single stand mills, there is scope for loop coupling and interaction. The design of an integrated, non-interacting tension and thickness control scheme is presented for both cold mills and hot mills. The controller design is developed from a consideration of the uncontrolled tandem mill response to a range of disturbance types, and the response to the use of different rollgap and motor actuators for control action. Other issues examined are the use of schedule dependent controller gains and the potential interaction of multiple thickness controllers.

The online measurement of strip shape which began in the early 1960’s has seen extensive development in the area of shape control as shape has become one on the important strip quality parameters. The design of shape control systems commences with an examination of the major shape disturbances and the different types of actuators available. Roll bending, roll crossing, roll side shifting, hydraulically modified rolls, segmented rolls and cold and hot sprays are all used for shape control. The sensitivity of strip shape to these actuators and the common forms of disturbance form the fundamental basis for the design of the shape control system. Other important aspects covered include shape measurement and signal processing. An examination of the performance of different levels of automation concludes this lecture.

The lecture starts with a discussion of the physical phenomena which affect the thickness profile. The mathematical representation of typical thickness profiles is discussed and common methods of measurement described. The design of automation systems for profile control is reviewed and the characteristics of currently available measurement and actuator systems described. The mathematical formulation of a vector based profile setup model algorithm is developed which has many important advantages over the systems installed at most mills. Typical performance of online thickness profile models and control systems is given as well as a discussion of practical constraints and future directions for further improvements.

There are a number of model coefficients that vary over time as mill conditions change or are not accurately known for every product. If these are not addressed adequately, the model errors lead to operating difficulties. The concept of adaption, or model learning is discussed using simple algorithms for achieving enhanced on-line model performance. A number of more recent techniques, such as Kalman Filtering and Neural Networks are introduced. Many practical aspects of model adaption schemes for reversing and tandem mills and methods of tuning adaption filters for optimum performance are presented. Ensuring that the number of adaption parameters is minimised and the parameters are not interdependent is of prime importance.

Mill vibration has become an increasingly important operational issue in both cold and hot rolling mills as mill size and speed have increased. Torsional, third octave and fifth octave chatter, common vibration modes which occur in cold and hot mills, are studied. For each of the three modes of vibration the underlying mechanisms involved are discussed. For third octave chatter a stability criteria for the critical rolling speed is presented. This allows the effect of the major parameters on the critical speed to be determined. Methods of suppressing third octave chatter are examined including the design, development and testing of the vibration inhibiting piston. The current understanding of the mechanism and methods for suppression of fifth octave chatter are then presented.

Lubrication and cooling remains a highly specialised field in the rolling industry. This lecture provides an overview of the major principles of lubrication and cooling, their impact upon the automation of rolling mills and some of the practical aspects of this subject. It commences examining factors which impact upon the selection of the lubricant and describes the different modes of lubrication. Differences between water based emulsions, dispersions, stabilised dispersions and petroleum based solutions are highlighted. It also examines the design of roll coolant application and solution systems.

This lecture discusses the issues involved in automating single stand rolling mills, highlighting the areas of similarity and difference with tandem cold mills. The strategy for selecting suitable tensions and reductions is described. A discussion is included on the importance of selecting the correct winding tensions to avoid coil collapse and inter-wrap sticking during batch annealing in downstream processing. The components of a state-of-the-art dynamic thickness and flatness control are summarised. A discussion on the relative merits of different types of measurement transducers is included. Finally, notes on the design of performance reporting and fault diagnosis systems are provided.