Search

The productivity and capacity release of hot-strip rolling is severely limited by pendular vibration. Previous studies on mill vibration have investigated the influence of rolling parameters on vertical vibration; however, the influence of dynamic stiffness compensation factors on vertical vibration has not yet been considered owing to the limitations of modelling methods. Herein, we develop a simulation model of mill liquid-machine coupling with dynamic stiffness compensation using AMESim software. The established simulation model is used to investigate the effect of this dynamic stiffness compensation on mill vertical vibration under conditions of downfeed thickness difference excitation frequency, thickness difference excitation amplitude, thickness difference excitation phase, and dynamic stiffness compensation signal generation hysteresis. Evidently, the impact of the mill stiffness compensation on the mill vertical vibration differs with the thickness of the excitation frequency, and the impact should be discussed according to the excitation frequency segment. The amplitudes of the roll system vibration and incoming thickness difference excitation are essentially linear, whereas the excitation phase does not affect the frequency or amplitude of the mill vibration. The rolling machine stiffness compensation signal hysteresis also has a significant impact on the mill vibration amplitude. Our findings will improve the mechanism of the hot rolling process and suppress droop vibrations. [ABSTRACT FROM AUTHOR]

Data-based equipment fault detection and diagnosis is an important research area in the smart factory era, which began with the Fourth Industrial Revolution. Steel manufacturing is a typical processing industry, and efficient equipment operation can improve product quality and cost. Steel production systems require precise control of the equipment, which is a complex process. A gearbox transmits power between shafts and is an essential piece of mechanical equipment. A gearbox malfunction can cause serious problems not only in production, quality, and delivery but in safety. Many researchers are developing methods for monitoring gearbox condition and for diagnosing failures in order to resolve problems. In most data-driven methods, the analysis data set is derived from a distribution of identical data with failure mode labels. Industrial sites, however, often collect data without information on the failure type or failure status due to varying operating conditions and periodic repair. Therefore, the data sets not only include frequent false alarms, but they cannot explain the causes of the alarms. In this paper, a framework called the Reduced Lagrange Method (R-LM) periodically assigns pseudolabels to vibration signals collected without labels and creates an input data set. In order to monitor the status of equipment and to diagnose failures, the input data set is fed into a supervised learning classifier. To verify the proposed method, we build a test rig using motors and gearboxes that are used on production sites in order to artificially simulate defects in the gears and to operate them to collect vibration data. Data features are extracted from the frequency domain and time domain, and pseudolabeling is applied. There were fewer false alarms when applying R-LM, and it was possible to explain which features were responsible for equipment status changes, which improved field applicability. It was possible to detect changes in equipment conditions before a catastrophic failure, thus providing meaningful alarm and warning information, as well as further promising research topics. [ABSTRACT FROM AUTHOR]

The steel industry is typical process manufacturing, and the quality and cost of the products can be improved by efficient operation of equipment. This paper proposes an efficient diagnosis and monitoring method for the gearbox, which is a key piece of mechanical equipment in steel manufacturing. In particular, an equipment maintenance plan for stable operation is essential. Therefore, equipment monitoring and diagnosis to prevent unplanned plant shutdowns are important to operate the equipment efficiently and economically. Most plant data collected on-site have no precise information about equipment malfunctions. Therefore, it is difficult to directly apply supervised learning algorithms to diagnose and monitor the equipment with the operational data collected. The purpose of this paper is to propose a pseudo-label method to enable supervised learning for equipment data without labels. Pseudo-normal (PN) and pseudo-abnormal (PA) vibration datasets are defined and labeled to apply classification analysis algorithms to unlabeled equipment data. To find an anomalous state in the equipment based on vibration data, the initial PN vibration dataset is compared with a PA vibration dataset collected over time, and the equipment is monitored for potential failure. Continuous wavelet transform (CWT) is applied to the vibration signals collected to obtain an image dataset, which is then entered into a convolutional neural network (an image classifier) to determine classification accuracy and detect equipment abnormalities. As a result of Steps 1 to 4, abnormal signals have already been detected in the dataset, and alarms and warnings have already been generated. The classification accuracy was over 0.95 at d = 4 , confirming quantitatively that the status of the equipment had changed significantly. In this way, a catastrophic failure can be avoided by performing a detailed equipment inspection in advance. Lastly, a catastrophic failure occurred in Step 9, and the classification accuracy ranged from 0.95 to 1.0. It was possible to prevent secondary equipment damage, such as motors connected to gearboxes, by identifying catastrophic failures promptly. This case study shows that the proposed procedure gives good results in detecting operation abnormalities of key unit equipment. In the conclusion, further promising topics are discussed. [ABSTRACT FROM AUTHOR]

This paper presents a new method for the construction and training of interval type-3 fuzzy logic systems whose inputs are uncertain type-2 non-singleton numbers (IT3 NSFLS-2). The proposed methodology is divided in two processes: 1) The novel construction of the structure of the IT3 NSFLS-2 systems based on: a) The level-alpha-0 of the interval type-2 fuzzy logic system (IT2-alpha-0 FLS), and on b) The secondary membership function using Gaussian modeling to construct each rule of the alpha-k fuzzy rule base (FRB), the firing intervals of the antecedent and the centroids of the consequent, and 2) The training methodology based on gradient descent algorithm to train the antecedent and consequent parameters of the alpha-0 FRB. The primary membership functions (MF) of the antecedents of the IT3 NSFLS-2 system are modeled as Gaussians with uncertain means and with common standard deviation. The proposal was applied and tested with the prediction of a transfer bar’s surface temperature in an industrial hot strip mill facility located in Monterrey México. The modeling results show that the proposal supports the stability required by this critical process and shows the best performance when compared with similar methods.

Data-based equipment fault detection and diagnosis is an important research area in the smart factory era, which began with the Fourth Industrial Revolution. Steel manufacturing is a typical processing industry, and efficient equipment operation can improve product quality and cost. Steel production systems require precise control of the equipment, which is a complex process. A gearbox transmits power between shafts and is an essential piece of mechanical equipment. A gearbox malfunction can cause serious problems not only in production, quality, and delivery but in safety. Many researchers are developing methods for monitoring gearbox condition and for diagnosing failures in order to resolve problems. In most data-driven methods, the analysis data set is derived from a distribution of identical data with failure mode labels. Industrial sites, however, often collect data without information on the failure type or failure status due to varying operating conditions and periodic repair. Therefore, the data sets not only include frequent false alarms, but they cannot explain the causes of the alarms. In this paper, a framework called the Reduced Lagrange Method (R-LM) periodically assigns pseudolabels to vibration signals collected without labels and creates an input data set. In order to monitor the status of equipment and to diagnose failures, the input data set is fed into a supervised learning classifier. To verify the proposed method, we build a test rig using motors and gearboxes that are used on production sites in order to artificially simulate defects in the gears and to operate them to collect vibration data. Data features are extracted from the frequency domain and time domain, and pseudolabeling is applied. There were fewer false alarms when applying R-LM, and it was possible to explain which features were responsible for equipment status changes, which improved field applicability. It was possible to detect changes in equipment conditions before a catastrophic failure, thus providing meaningful alarm and warning information, as well as further promising research topics.

The steel industry is typical process manufacturing, and the quality and cost of the products can be improved by efficient operation of equipment. This paper proposes an efficient diagnosis and monitoring method for the gearbox, which is a key piece of mechanical equipment in steel manufacturing. In particular, an equipment maintenance plan for stable operation is essential. Therefore, equipment monitoring and diagnosis to prevent unplanned plant shutdowns are important to operate the equipment efficiently and economically. Most plant data collected on-site have no precise information about equipment malfunctions. Therefore, it is difficult to directly apply supervised learning algorithms to diagnose and monitor the equipment with the operational data collected. The purpose of this paper is to propose a pseudo-label method to enable supervised learning for equipment data without labels. Pseudo-normal (PN) and pseudo-abnormal (PA) vibration datasets are defined and labeled to apply classification analysis algorithms to unlabeled equipment data. To find an anomalous state in the equipment based on vibration data, the initial PN vibration dataset is compared with a PA vibration dataset collected over time, and the equipment is monitored for potential failure. Continuous wavelet transform (CWT) is applied to the vibration signals collected to obtain an image dataset, which is then entered into a convolutional neural network (an image classifier) to determine classification accuracy and detect equipment abnormalities. As a result of Steps 1 to 4, abnormal signals have already been detected in the dataset, and alarms and warnings have already been generated. The classification accuracy was over 0.95 at d=4, confirming quantitatively that the status of the equipment had changed significantly. In this way, a catastrophic failure can be avoided by performing a detailed equipment inspection in advance. Lastly, a catastrophic failure occurred in Step 9, and the classification accuracy ranged from 0.95 to 1.0. It was possible to prevent secondary equipment damage, such as motors connected to gearboxes, by identifying catastrophic failures promptly. This case study shows that the proposed procedure gives good results in detecting operation abnormalities of key unit equipment. In the conclusion, further promising topics are discussed.

In most research on the hot strip mill production scheduling problem (HSMPSP) arising in the steel industry, it is accepted that a schedule with lower penalty caused by jumps of width, hardness, and gauge will result in lower roller wear, so it is regarded as a better schedule. However, based on the analysis of production processes, it is realised that rolling each coil also cause roller wear. In order to assessing the roller wear associated with production scheduling more precisely, it is necessary to consider it as another factor besides those jumps, especially when complicated constraints are involved. In this paper, an improved method is proposed to quantify the expected wear of the rollers done by those jumps and rolling processes. Then the HSMPSP whose objective is to maximise the total length of all scheduled coils is formulated as a team orienteering problem with time windows and additional production constraints. A heuristic method combining an improved Ant Colony Extended algorithm with local search procedures dedicated to HSMPSP is developed. Finally, computational results on instances generated based on production data from an integrated steel mill in China indicate that the proposed algorithm is a promising solution specific to HSMPSP. [ABSTRACT FROM AUTHOR]

An online sequential extreme learning machine (OSELM) algorithm with self-learning capability based on improved sparrow search algorithm (ISSA) is proposed to solve the problem of feedback lag, strong coupling, and large thickness deviation in thickness prediction in hot continuous rolling process. Firstly, an online thickness prediction model of OSELM is established based on the data collected from hot rolling site. Next, in order to improve the model accuracy and stability, an ISSA method is proposed to optimize the weights and the biases of OSELM to reduce the error between the model and actual production process. On this basis, a self-learning method is applied to further improve the prediction accuracy. Finally, the validity of the ISSA-OSELM with self-learning was verified by simulation. The prediction results can be used to improve the control accuracy of the automatic gauge control (AGC) system, and provide a firm basis for improving the quality of rolled strip products. [ABSTRACT FROM AUTHOR]

• An interpretable roller's health indicator construction method is proposed. • The reason for multi-stage roller degradation process formation is analyzed. • An RUL prediction architecture is built for multi-stage roller degradation. • A multi-scale LSTM-MHA is proposed to extract different temporal information. Accurate prediction of a roller's remaining useful life (RUL) is significant for a hot strip mill to avoid major safety accidents and substantial economic losses. Since rollers' degradation processes represent multi-stage characteristics, traditional single-stage models cannot acquire accurate RUL prediction results. Therefore, this paper proposes an adaptive staged RUL prediction method based on multi-scale long short-term memory network with multi-head self-attention (LSTM-MHA). The roller's production data and operation data are fused to construct interpretable health indicators (HI), which can represent the roller's remaining rotatable angle. Then the degradation process is divided into multiple stages and analyzed. The proposed multi-scale LSTM-MHA can adaptively update the model weights based on changes in the roller degradation stages by the multi-scale memory structure. The MHA embedding mechanism filters important temporal information for LSTM units. The trained model can predict the roller's RUL by inputting the current roller's HI. The proposed method is verified on an industrial hot strip mill roller dataset from a well-known steel company. The validation results show that the prediction accuracy of the proposed method is higher than 98.98%. Compared with existing deep learn-based methods, the proposed multi-scale LSTM-MHA method has significant advantages in roller's RUL prediction. [ABSTRACT FROM AUTHOR]

Depending on various properties such as hardness, roughness, friction coefficient, wear resistance, and oxidation behavior, different grades of steel can be used in hot strip mills. In this paper, two grades were investigated: one grade of semi-high-speed steel and one grade of high-speed steel. They differ in their chemical composition, principally the carbon and the chromium content, and the presence or not of other alloying elements such as vanadium, niobium, and tungsten. Their oxidation behavior was studied at 576 °C and 600 °C, temperatures reached in the F1 finishing stand and the roughing stand, respectively, by high-temperature oxidation tests in a wet atmosphere. The sliding wear resistance was followed by pin-on-disc experiments. The evolution of the microstructure was observed by optical and scanning electron microscopies. The oxides formed on the surface of the samples were analyzed by XRD and EDS. The thickness of the oxide layers and the mass gain were measured from oxidation tests. The results showed that for both grades, the surface roughness Ra after oxidation tests is high enough to prevent slipping. The low friction coefficient will increase the wear resistance of the roll. The hardness of the steel grades decreases but remains high enough to ensure wear resistance. Consequently, those grades can be used in the roughing stand or the F1 finishing stand of hot rolling mills. [ABSTRACT FROM AUTHOR]

In hot strip mills, prediction of the necking width of the hot strip is a fundamental step in the hot strip mill process. However, owing to the large number and complexity of the variables involved, this prediction remains a challenging problem. In this article, we propose a deep neural model with an attentional residual network that combines an attentional network to calculate feature importance and a residual network to estimate the necking value. When a hot strip mill dataset from a South Korean steelmaking company was evaluated, the proposed model showed a higher performance than several machine-learning methods. Furthermore, the importance of the features selected by the attentional network outperformed those by other feature selection methods. Our approach is useful for necking predictions and can be applied to determine feature importance. [ABSTRACT FROM AUTHOR]

Hot-rolled strips are cooled on the run-out table to achieve the customer-required mechanical properties. Cooling reduces the oxidation, which can assist the coiling operation. The ability to obtain a range of mechanical properties from a single steel grade reduces the use of alloying elements (elements added to a metal or an alloy which are incorporated in the metallic structure and change in the properties of the basic alloy) and the size of the slab inventories. So in this paper, the high-temperature steel strip cooling pattern with the temperature of 870 °C and reaching the temperature of 630 °C using water and air is studied. Moreover, the numerical simulation of the high-temperature strip cooling and its cooling method is carried out. The results are compared with the similar sampler ones. The changes of three key parameters such as change in strip speed and its effect on cooling rate, change in pattern and arrangement of water showers and its effects and change in the cooling water temperature are investigated. It was found that the uniformity of the heat distribution inside the strip was increased, but the temperature of the strip was lower than the required temperature. The middle and up surface temperatures of the strip were performed at a speed of 3.48 (m s−1), which showed a very good match with the actual tested sample. [ABSTRACT FROM AUTHOR]

Copyright of Mining & Metallurgy (10057854) is the property of Beijing Research Institute of Mining & Metallurgy Technology Group and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

High-speed steel type cast iron rolls (HSS-rolls) used in the hot rolling process of steel are composite rolls manufactured by the CPC method or the centrifugal casting method (CCM). Comparing the two, the centrifugal cast HSS-roll is advantageous in terms of production cost, and since the shaft material is ductile cast iron, the thermal crown during rolling is small, and the plate threading property is excellent. For this reason, Centrifugal cast HSS-rolls are often used as standard type HSS-rolls, and some of them are also used in the later stands of the hot strip mills. On the other hand, forged steel is used for the shaft material of CPC-HSS-rolls, and the strength of the shaft material is superior to that of ductile cast iron, so there is very little risk of fracture accident (thermal breakage) from the center. The thermal breakage of the roll is defect that occurs when the tensile residual stress of the shaft and the thermal stress generated during rolling exceed the material strength of the shaft. Therefore, in order to prevent thermal breakage, it is considered extremely effective to suppress the tensile residual stress of the shaft portion. Therefore, in this study, an attempt was made to significantly reduce the tensile residual stress of the shaft by improving the heat treatment conditions. Furthermore, the influence of heat treatment conditions on the compressive residual stress of the outer layer was also investigated. As a result, it was confirmed that by applying the surface rapid heating method to heating conditions before quenching, the tensile residual stress inside the roll can be reduced by about 30% compared to the conventional uniform heating method. Effective results were obtained by deciding the manufacturing conditions for significantly improving safety against thermal breakage in centrifugal cast HSS-rolls. [ABSTRACT FROM AUTHOR]

Mathematical models for rolling force calculation during hot rolling are crucial for both automatic mill operation and prediction of steel behavior and mill capacity, especially for development of new steel grades and improvement of the existing ones. In literature, the models are based usually on the calculation of steel mean flow stress (MFS), firstly, followed by a force model based on MFS. This approach was applied to a large volume of industrial data of coils produced at Usiminas’ plant in Ipatinga, giving unsatisfactory results. Then, the empirical model developed by Schultz was applied to calculate directly the force, with further fine tuning the model with a linear regression taking into account the steel chemical composition. This approach led to a force prediction capability better than the traditional models based on the binomial MFS-force presented in literature, markedly for plain CMn and IF steels. In such cases, more than 90% of force predictions were inside the ±10% error margin, considering the front stands of the hot strip mill. [ABSTRACT FROM AUTHOR]

Scheduling of a hot strip mill is an important decision problem in the steel manufacturing industry. Previous studies on the hot strip mill scheduling problem have mostly neglected the random factors in production. However, random variations in processing times are inevitable due to unpredictable delays and disturbances. In this article, we adopt a robust optimization approach to deal with the uncertainty in processing times. The advantage is that no assumption has to be made regarding the distribution of random data, and the obtained schedule will remain strictly feasible when the variations have not exceeded a predefined uncertainty set. First, a mixed-integer linear programming model is presented to formulate the robust scheduling problem. Then, a hybrid metaheuristic algorithm, which combines ant colony system (ACS) and enhanced local search, is proposed to provide an efficient solution to the problem. Finally, extensive computational experiments involving both randomly generated and real-world instances have been conducted to verify the effectiveness of the proposed algorithm. It is shown that the algorithm achieves optimality for small instances and outperforms two state-of-the-art metaheuristics when used to solve large instances. [ABSTRACT FROM AUTHOR]

Existing method of run-out table drive control on the hot strip mill ensures the required strip tension by setting a difference between speeds of the delivered strip and roller bodies. Such difference results in a higher wear rate of roller bodies and their more frequent replacement, entailing high expenses. Thus, a currently important objective is to increase efficiency of run-out table drives on hot strip mills. The paper describes a method to calculate required torque of an electric drive of run-out table rollers in a variety of strip delivery modes, using results of experiments carried out on mill 2000 at public company "Magnitogorsk Iron and Steel Works" (MMK). We adjusted mandatory process requirements for a run-out table roller drive with regard to a decreased wear rate of roller bodies and the improved strip surface quality. The authors developed a method to control a drive of run-out table rollers, ensuring compliance with new process requirements. Efficiency of the offered technical solutions was under an experimental evaluation. [ABSTRACT FROM AUTHOR]

To ensure the formation of a sound shell-core bond interface free of defects between the shell and the core in work rolls used in the finishing stands of hot strip mills, a complete fusion of this interface must be achieved, avoiding excessive mixing of the two components and the formation of hard, fragile microstructures. The shell is made of white cast iron, alloyed with Ni and Cr, and the core is manufactured of grey cast iron spheroidal graphite in a pearlitic matrix. It is thus advisable to inoculate the shell with 0.6 kg/T SiCaMn, as this promotes discontinuity in the carbide network and leads to an increase in the impact toughness of the bond interface. Furthermore, inoculation of the shell with FeSi-La should be avoided, as this inoculant leads to an increase in graphite counts, promoting it with a lamellar morphology at the edge of the bond and hence reducing the impact toughness in this interface. Addition of Mg to the shell has been found to produce an increase in hardness in the regions adjacent to the bond interface.

The mechanism of the position-mode side guide in correcting slab centerline profile and camber in the roughing process of a hot strip mill (HSM) was analyzed using finite element simulation. The finite element model was established based on the actual size of the roughing mill and on the actual actuating time sequence of the roughing mill in China Steel Corporation (CSC), Kaohsiung. This work could be the first to give an insight into the mechanism of side guides in correcting the slab camber. Time sequence analysis was explored to visualize the progress of centerline profile variation and the interaction between the slab and the related roughing mill components at different moments. The history of reaction forces exerted on the slab was analyzed to explain the interaction between roughing mill components and the slab. The effect of the separation distance of side guide and the effect of the slab wedge on the centerline profile was investigated. A schematic model illustrating the reactions and the resulting moments exerted on the slab was created. By examining the force history, the cross-sectional strain/stress distribution, and the roll force across the horizontal roller, the correcting mechanism of the side guide could be elucidated. The simulation results provide further knowledge in selection and dimension design of side guide to improve the effectiveness of side guide in correcting the slab profile.

It is important to realize the transformation behavior of wüstite, because it greatly affects the final structure of the oxide layer and the surface quality of steel products. In the present study, the transformation behaviors of the wüstite layer are examined under nearly-oxygen-free conditions in order to simulate the cooling processes after the hot rolled strip is coiled. As the single phase wüstite is prepared at 950 °C, the 460 °C transformed oxide layer is composed of a mixture of iron and magnetite formed through eutectoid reactions. For the 750 °C-fabricated wüstite, only magnetite is observed after transformation, without iron precipitates, and residual wüstite. It is speculated that the unusual transformation behavior of the low-temperature-made wüstite results from the pseudo-structural intermediate phase transformation between wüstite and magnetite. This pseudo-structure is a pre-transformed wüstite and contains various concentrations of ferrous ion, which is determined by the fabrication conditions. During the hot strip mill process, wüstite is produced continuously from finish mill to laminar flow sections and culminated at 570 °C. Consequently, the final eutectoid transformation below 570 °C is dominantly controlled by the surface temperature ranging from 750 to 950 °C for low carbon steel. [ABSTRACT FROM AUTHOR]

Controlling the looper height and strip tension is important in hot strip mills because these variables affect both the strip quality and strip threading. Many researchers have proposed and applied a variety of control schemes for this problem, but the increasingly strict market demand for strip quality requires further improvements. This work describes a dynamic matrix predictive control (DMC) strategy that realizes the optimal control of a hydraulic looper multivariable system. Simulation experiments for a traditional controller and the proposed DMC controller were conducted using MATLAB/Simulink software. The simulation results show that both controllers acquire good control effects with model matching. However, when the model is mismatched, the traditional controller produces an overshoot of 32.4% and a rising time of up to 2120.2 ms, which is unacceptable in a hydraulic looper system. The DMC controller restricts the overshoot to less than 0.08%, and the rising time is less than 48.6 ms in all cases. [ABSTRACT FROM AUTHOR]

The purpose of the paper is to achieve a constrained estimation of process state variables using the anisochronic state observer tuned by the dominant root locus technique. The anisochronic state observer is based on the state-space time delay model of the process. Moreover the process model is identified not only as delayed but also as non-linear. This model is developed to describe a material flow process. The root locus technique combined with the magnitude optimum method is utilized to investigate the estimation process. Resulting dominant roots location serves as a measure of estimation process performance. The higher the dominant (natural) frequency in the leftmost position of the complex plane the more enhanced performance with good robustness is achieved. Also the model based observer control methodology for material flow processes is provided by means of the separation principle. For demonstration purposes, the computer-based anisochronic state observer is applied to the strip temperatures estimation in the hot strip finishing mill composed of seven stands. This application was the original motivation to the presented research. [ABSTRACT FROM AUTHOR]

Hot rolling is an essential industrial process in the production of sheet steel, a widely used product in manufacturing and construction. The hot strip mill is a complex unit including many processes and control loops and its modeling is a challenging task owing to the variety of phenomena that occur within the mill. Model validation is also challenging owing to a scarcity of measurements. In this study, a computer simulative emulation that can analyze the dynamic behaviors of the tandem hot rolling process with original non-linear rolling models has been developed and validated against a linear model and measured data. Moreover, the influence coefficients of rolling process parameters were calculated by analyzing the rolling data. [ABSTRACT FROM AUTHOR]

An evaluation is made about the precision of hot strip rolling mill loads at the F1 stand calculated according to the theoretical models of Orowan, Sims, Alexander-Ford, Orowan-Pascoe, Ekelund and Tselikov in comparison to real values got for carbon-manganese steels. In the deterministic approach, without any fit of the calculated values to real data, Orowan, Sims and Alexander-Models show best levels of precision, as expected from the information got in the literature. However, in the semi-empirical approach, after a linear fit between calculated values and real data, Tselikov and Ekelund models show better adequacy to the industrial data, a fact that can be attributed to more significant errors occurring in the sub-models of temperature, tribology and hot strength than in the rolling load models. For its turn, neural network models show the best levels of precision which make very attractive the adoption of this approach.

There have been accounts of repeated failure of crane hooks at the coil yard of a Hot Strip Mill which pose a serious threat to safety in the area. More than 4 hooks failed in less than 5 years. The crane hook (rated for 36000 kg) failed from the threaded shank while lifting a load of 18143 kg. The metal in the hook was revealed by chemical analysis to be killed IS: 4367 20C15 steel. The hook rod failed from a step where there was a cross sectional change and the locations were associated with machining and chatter marks. Such cross-sectional changes are the potential sites of stress concentrations leading to crack initiations. Fracture surfaces of broken pieces of hook reveal initiation of beach marks from both sides with granular rough surface at the middle of fracture zone. Beach marks initiated from both sides indicate origin of reverse bending fatigue. Distinct granular rough zone at the middle is due to final brittle fracture. Microstructure of the polished sample revealed numerous inclusions which indicate that the steel was not clean .Such a huge number of inclusions are not desirable as they can act as stress concentration sites and lead to fatigue crack initiation. Etched microstructure of failed hook reveals coarse cast structure having inhomogeneous microstructure with a mixture of ferrite and pearlite (which meant lower fatigue strength). This inhomogeneous coarser cast structure is outcome of lower reduction ratio during rolling followed by improper heat treatment process. The fracture was concluded to have occurred due to stress concentration from the step region due to inferior material (Inclusion and Improper Heat treatment process). Preventive maintenance and condition monitoring procedures should be applied to identify and minimize the risk(s) Establishment of an NDT procedure for regular basis inspection of the structural members (welded joints and hookshaped steel rods during incoming inspection and in-service).

En etkili üretim proseslerinden biri olan haddelemeyi gerçekleştirmek için kullanılacak malzemelerin seçimi ve bu malzemelerin merdanenin özelliklerine etkisi önemli bir parametredir. Aynı zamanda uygun malzeme ile çeşitli üretim metotlarının kullanılması da bu özellikler üzerinde önemli bir etkiye sahiptir. Hangi uygulamalarda ne çeşit bileşimde merdane malzemesine ihtiyaç bulunduğunu anlamak için, haddeleme koşulları ile merdane malzemesi bileşimlerinin ve özelliklerinin bilinmesi gerekmektedir. Bu çalışmada klasik malzemelerden, son zamanalarda geliştirilen malzemelere kadar olan gelişmenin değerlendirilmesi sunulmuş ve bi-metal merdane üretim teknolojileri irdelenmiştir. Ayrıca deneysel olarak dökümü gerçekleştirilen perlitik ve martenzitik merdane mikroyapıları incelenmiştir. Elde edilen sonuçlar ile mikroyapı oluşumlarının merdaneden beklenen özellikler üzerine etkileri analiz edilmiştir.

En etkili üretim proseslerinden biri olan haddelemeyi gerçekleştirmek için kullanılacak malzemelerin seçimi ve bu malzemelerin merdanenin özelliklerine etkisi önemli bir parametredir. Aynı zamanda uygun malzeme ile çeşitli üretim metotlarının kullanılması da bu özellikler üzerinde önemli bir etkiye sahiptir. Hangi uygulamalarda ne çeşit bileşimde merdane malzemesine ihtiyaç bulunduğunu anlamak için, haddeleme koşulları ile merdane malzemesi bileşimlerinin ve özelliklerinin bilinmesi gerekmektedir. Bu çalışmada klasik malzemelerden, son zamanalarda geliştirilen malzemelere kadar olan gelişmenin değerlendirilmesi sunulmuş ve bi-metal merdane üretim teknolojileri irdelenmiştir. Ayrıca deneysel olarak dökümü gerçekleştirilen perlitik ve martenzitik merdane mikroyapıları incelenmiştir. Elde edilen sonuçlar ile mikroyapı oluşumlarının merdaneden beklenen özellikler üzerine etkileri analiz edilmiştir.