Roll Gap Model Research Articles

Adaptive feedforward thickness control in hot strip rolling with oil lubrication

The application of oil lubrication in hot strip rolling has many advantages for the process, i.e., reduced roll force, reduced energy consumption, better strip surface quality, and reduced roll wear. On the other hand, oil lubrication may also induce severe disturbances for several control loops in a tandem hot rolling mill, which can reduce the final strip quality or might even jeopardize a stable operation of the plant. For example, the widely used automatic gauge control (AGC) has a limited bandwidth to ensure stability in the whole operating range and thus cannot (dynamically) reject the thickness deviations caused by switching on and off the oil lubrication. For this reason, the effect of lubrication is analyzed in detail and mathematically modeled in this work. Based on this dynamic lubrication model together with a hydrodynamic roll gap model and a mill stand model, two feedforward control approaches for the strip thickness are presented. These controllers counteract the negative effects of varying lubrication. Moreover, the feedforward controllers are extended by a moving horizon estimator to track changes of uncertain parameters of the underlying models based on the measured roll force. The estimated parameters are instantly used by the feedforward controller. Simulation studies using an experimentally validated simulation model of a whole mill stand show that the proposed adaptive feedforward control approach provides a significant improvement of the strip exit thickness accuracy for strips with oil lubrication compared to conventional controllers.

A novel forecast model based on CF-PSO-SVM approach for predicting the roll gap in acceleration and deceleration process

PurposeIn the cold rolling process, friction coefficient, oil film thickness and other factors vary dramatically with the change in the rolling speed, which seriously affects the strip thickness deviation. This paper aims to improve the strip control precision with the forecast roll gap model based on CF-PSO-SVM approach in the rolling process.Design/methodology/approachIn this paper, a novel forecasting model of the roll gap based on support vector machine (SVM) optimized by particle swarm optimization with compression factor (CF-PSO) is proposed. Based on lots of online data, the roll gap models regressed by PSO-SVM, genetic algorithm (GA)-SVM and CF-PSO-SVM are obtained and verified through evaluating the performances with the decision coefficient (R2), mean absolute error and root mean square error. In addition, with the good forecasting performances of CF-PSO-SVM, a roll gap compensation model is studied.FindingsThe results indicate that the proposed CF-PSO-SVM has excellent learning regression ability compared with other optimization algorithms. Meanwhile, a roll gap compensation model based on the rolling speed and plastic coefficient is obtained, which has been proved validated in product.Originality/valueIn this paper, the SVM algorithm is combined with traditional rolling technology to solve the problems in actual production, which has great supporting significance for the improvement of production efficiency.

Slip Model Adaptation Based on Measurements of the Strip Velocity

Abstract In hot strip rolling, the strip tension is an important process variable. It directly depends on the mass flow of the rolled material. In this paper, a simple and reliable measurement system for the mass flow in hot strip rolling is described and its accuracy is analyzed. Based on such a measurement, slip models, which are often only a by-product of roll gap models, can be validated and tuned. An adaptation approach for slip models using an additional disturbance input is proposed and a method for identifying this disturbance is presented. This adaptation approach significantly improves the accuracy of the computed strip velocity compared to state-of-the-art approaches relying on roll gap models like Sims’ model or the hydrodynamic model.

Asymmetric hydrodynamic roll gap model and its experimental validation

In tandem hot strip rolling mills, different friction between the rolls and the strip material on the upper and lower strip surface can occur due to asymmetric surface temperatures or different conditions of oil lubrication. To capture these effects, this paper presents a hydrodynamic roll gap model with asymmetric friction. Based on similarities between the rolled material and viscous fluids, fluid mechanics theory is used to derive this model. Due to the nature of this model, the influence of the rolling speed is inherently taken into account, which allows an accurate prediction of the rolling force and the forward slip. As an analytic solution for the hydrodynamic roll gap model is available, it is well suited for online applications in rolling plants. For validation of the proposed model, an experiment with asymmetric work roll roughness was performed. A specimen of steel strip with copper pins inserted was repeatedly rolled to visualize the material flow inside the roll gap for multiple passes. The resulting deformed copper pins were cut out of the strip and show good agreement with the deformation profiles calculated by the developed model.

Optimization-based feedforward control of the strip thickness profile in hot strip rolling

A new feedforward control approach for the thickness profile of the strip in a tandem hot rolling mill is developed. In industry, the automatic gauge control (AGC) concept is widely used for thickness control. The AGC has the disadvantage that it does not consider known disturbances from upstream entities. This is why a number of disturbance feedforward control concepts have been proposed in the literature. These feedforward control strategies typically rely on linearized models and only provide symmetric control inputs for the mean thickness to the hydraulic adjustment system. In this work, an optimization-based feedforward controller for the lateral thickness profile is proposed that fully exploits all degrees of freedom available, i.e., the hydraulic cylinder positions and the bending forces at the drive side and at the operator side of the mill stand. Moreover, it is shown that by linearizing the mill stand model while keeping the nonlinearities from the roll gap model leads to a numerically efficient optimization problem, which is a good compromise between accuracy and computational efficiency. The feedforward controllers are further combined with the AGC in the feedback path in a two-degree-of-freedom controller structure to account for model-plant mismatch. Simulation results for a validated mathematical model and first experimental results from an industrial pilot installation show a significant benefit compared to the existing AGC without feedforward control.

Optimal Parameter Identification for a Hydrodynamic Roll Gap Model in Hot Strip Rolling

When using modern model-based control concepts in a hot strip rolling mill, an accurate prediction of the roll force is crucial for the achieved control accuracy in terms of the strip thickness. Hence, the influence of various process parameters, e.g., lubrication, on the roll force is of central interest. Commonly used roll gap models are based on the slab method and are often not capable of describing these influences. The hydrodynamic roll gap model, however, can capture the influence of varying friction conditions between the strip and the work rolls. Varying friction conditions can be caused, for instance, by lubrication. This paper presents a combination of the hydrodynamic roll gap model with tailored material models for a more accurate prediction of the roll force in a tandem rolling mill. To parametrize the model, an optimization-based identification algorithm is proposed for unknown friction coefficients and material parameters. The identification algorithm is used on measurement data from an industrial hot strip rolling mill. A comparison with a commonly used roll gap model shows that the parametrized hydrodynamic roll gap model predicts the roll force with significantly higher accuracy.

Model Predictive Control with Linear Programming for the Strip Infeed in Hot Rolling Mills

A model-based predictive approach is proposed for the strip head motion control during the steel strip infeed in hot rolling finishing mills. The design is based on a nonlinear simplified mathematical model in a form of ordinary differential equations. It is already shown that this model captures the behavior of FEM models for the strip and the roll gap in an excellent manner. Then the simplified time variant linear model is considered and the discrete-time model is derived in a straightforward manner. In addition the property of flatness is exploited to derive an efficient formulation for the optimization problems. We develop an optimal controller based on a linear programming and compare it with standard quadratic programming. It is shown that the proposed linear program is both more efficient than the quadratic program and the performance is equivalent. Therefore, the linear approach is convenient for the real-time application even without high performance computers. The proposed controller is tested in a co-simulation environment using FEM element model for the strip integrated in Hotint together with a highly nonlinear roll gap model. The simulation results show the excellent accordance of the controlled infeed process for both models (ODE and FEM), as well as the high performance of the closed loop.

A New Method Predicting Contact Length and Flattening in Temper Rolling

The current trend in temper rolling is to make a small reduction to steel strip in order to achieve higher strength with good formability and toughness. In addition, very high strength steels can be cold rolled twice with very small reductions. This causes problems in setup values for cold rolling. Rolling models are usually overestimating roll flattening in the case of small reductions.In temper rolling thickness reduction is small (0.5 – 3%) and the elastic deformation of the work roll should be taken into account [3]. However, standard circular arc roll gap models (e.g. Bland Ford Ellis combined with Hitchcock model) fail to predict the roll flattening and thus the rolling force [4]. In this work, finite element method has been used to define a simplified model for work roll flattening and contact length. Model describes the effect of reduction, strength of steel strip and roll radius.

Experimental evaluation of contact stress during cold rolling process with optical fiber Bragg gratings sensors measurements and fast inverse method

There is a strategic importance for the steel rolling industry to get a better understanding of the strip–roll interaction to improve roll-gap models, increase strip quality and decrease roll degradation. This requires roll-gap sensors able to measure this interaction under industrial rolling conditions and in real time in order to propose a feed-back control of process parameters. To reach these goals, this paper proposes a new roll-gap friction sensor based on an inverse method that interprets optical fiber Bragg gratings (FBG) strain measurements under the roll surface (fully embedded), which enables to evaluate contact stresses with very short computation times, compatible with real time interpretation. This elastic inverse method is analytical and relies on plane-strain and isothermal assumptions. The experimental apparatus is detailed, technical issues are clearly exposed as well as calibration procedures. Several pilot cold rolling tests have been performed at various rolling speeds and different strip thicknesses in order to demonstrate the industrial feasibility. Resulting evaluations of contact stresses are then compared with numerical simulations. Reasonable agreement is obtained for normal stress (i.e., pressure) but not for shear stress (only an order of magnitude is obtained).

Development and application of an off-line soft reduction model during continuous casting of high-carbon rectangular billet

An off-line soft reduction model was developed in order to improve the inner quality of high-carbon rectangular billet, which includes a heat transfer model, cooling water control model and roll-gap calculation model. The distribution of the total reduction amount in each pinch roll is calculated by using a segmented slope method in the reduction zone. The model application results showed that the inner quality of the high-carbon 82B steel with a section size of 180 mm × 240 mm had clearly been improved. It is evident that the formation of V-segregation in a rectangular billet subjected to soft reduction is suppressed and the centerline shrinkage cavity is clearly decreased. The mean center carbon segregation index was reduced from 1.15 to 1.08 with soft reduction. Besides, the soft reduction process improved the compactness and homogeneity of the central equiaxed dendrite zone and decreased the secondary dendrite arm spacing.

Analysis of Roll Gap Heat Transfers in Hot Steel Strip Rolling through Roll Temperature Sensors and Heat Transfer Models

This paper presents an analysis of roll bite heat transfers during hot steel strip rolling. Two types of temperature sensors (drilled sensor /slot sensor) implemented near roll surface and heat transfer models are used to identify in the roll bite interfacial heat flux, temperature and Heat Transfer Coefficient HTCroll-bite during pilot rolling tests. It is shown that: - the slot type sensor is much more efficient than the drilled type sensor to capture correctly fast roll temperature changes in the bite during hot rolling but life’s duration of the slot sensor is shorter. - average HTCroll-bite, identified with roll sensors temperature signals is within the range 15-26 kW/m2/K: the higher the strip reduction is, the higher the HTCroll-bite is. - scale thickness at strip surface tends to decrease heat transfers from strip to roll in the roll bite. - HTCroll-bite appears not uniform along the roll-strip contact, in contrast to usual assumptions made in existing models - Heat dissipated by friction at roll-strip interface and its partitioning through roll and strip respectively seems over-estimated in the existing thermal roll gap model [1]. Modeling of interfacial friction heat dissipation should be reviewed and verified. The above results show the interest of roll temperature sensors to determine accurately roll bite heat transfers and evaluate more precisely the corresponding roll thermal fatigue degradation.

Numerical Study of Four Roll Coating System

The characteristics of four roll coating system were numerically investigated compared with experimental data to validate theoretical models used in this research. In the theoretical models, a film splitting model using a power-law type equation, a roll gap model based on elasto-hydrodynamics and a flow model from a rotating cylinder system were applied. The parametric computations to each operation condition revealed the steady and dynamic behaviors of a coating film and liquid films on coating rolls. From the results of the frequency response to the speed disturbances of the coating rolls, it was found that the sensitivity of the lowest coating roll to the disturbance was a half of others, which implies that the requirement for the accuracy of a driving system for the coating roll is not so severe compared with others. The experimental data and the numerical results at steady state agreed in large part. The actual properties of materials used in the coating process will be required to predict the coating thickness precisely.

Roll speed and roll gap modelling – A case study for an experimental rolling mill

In this paper the detailed procedure of model development for a laboratory scale rolling mill (the Hille mill) is given. The main framework of the roll speed and gap model is based on the classic Newton motion equation, with the critical model parameters, such as the inertia of the different rotational parts, the frictions, the stiffness, etc. to be determined by different approaches ranging from analytical to experimental approaches as well as know-how. The developed models have been incorporated into a mill simulation package, and computer simulation results show a good agreement with the actual rolling process. The final goal of the models will be two-fold: to generate a decision support for the upgrading and modernisation of the Hille mill, and to develop advanced model-based rolling control strategies.

A Novel Non-Circular Arc Rollgap Model, Designed from the Control Point of View

The mathematical models for the temper and thin strip rolling case typically incorporate a detailed description of the elastic work roll deformations. These modesls, referred to as non-circular arc rollgap models, are either based on the elastic half-space solution or on Jortner’s solution. In this contribution the approximation of the displacement fields by means of the Ritz method is proposed in order to derive an implicit algebraic rollgap model suitable for the simulation and control purpose.

Modelling for Control of a Steckel Hot Rolling Mill.

In this work the derivation of a nonlinear plant simulator of a Steckel Hot Rolling Mill process is shown. The simulator reflects the thickness crown and tension behaviour of the strip, while the temperature, shape and flatness behaviour of the strip fall outside the scope of this work. Roll gap, stand, tension and hydraulic actuator models are incorporated in order to yield the simulator. The nonlinear simulator is used to identify a linear model for control system design. The linear model was identified at a certain operating point associated with a particular pass of a multiple pass rolling schedule. Step tests were applied to the manipulated variables of the simulator, and using system identification (SID) techniques a linear time invariant (LTI) multivariable input output (MIMO) transfer function model was identified. Lastly an initial control problem is formulated.