Tetsu-to-Hagane
Online ISSN : 1883-2954
Print ISSN : 0021-1575
ISSN-L : 0021-1575
Volume 90, Issue 11
Displaying 1-18 of 18 articles from this issue
  • Shigeru ANDO
    2004 Volume 90 Issue 11 Pages 860-869
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    This article reviews the developments of the measurement technologies in the Japanese steel industry based on the papers published in the biannual conferences of The Iron and Steel Institute of Japan. We classify the roles of measurement technology into the one for automatic control, the one for inspection and quality improvement, and the one for scientific and technological developments. The needs for the measurement technologies shows a clear shift from the first one to the second one in this decade. Towards the next decade, we describe the increasing demand for the quality assurance and specialization, the micro parameterization and the multifactorial characterization of the quality, and the cooperation of multiple sensors in the ubiquitous network environments.
    Download PDF (3071K)
  • Akira KAZAMA, Hiroyuki SUGIURA, Takahiko OSHIGE, Masaichi INOMATA, Mit ...
    2004 Volume 90 Issue 11 Pages 870-876
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    It is necessary to detect certain types of "shallow defects", which can be harmful in the press processes of steel sheets customers. Since it is difficult to detect them, the thresholds of automatic inspection systems should be lowered, which may cause enormous excessive detection of harmless patterns, such as "shallow" oil patterns.
    In this paper, an automatic inspection technique to discriminate harmful defects from harmless patterns clearly in steel strips is studied.
    By investigating the microscopic surface structure of defects and harmless patterns, we have found that micro flat elements of metal are dominant in the defect surface, while the harmless patterns have coatings of dielectric materials in the surface layer. Both surfaces show similar mirror reflection characteristic due to their surface structures. This fact accounts for the difficulty of the discrimination between defects and harmless patterns.
    Considering above facts, we have adopted ellipsometry to obtain polarized images of steel sheet samples for quantitative analysis. By utilizing linearly polarized light with an azimuth angle of 45 degrees as the incident light, the following results are obtained.
    (1) Alternation to elliptical polarization of reflected light occurs to more degree in the case of the defect than in that of the harmless pattern.
    (2) Reflected light from the harmless pattern keeps linearity of the polarized incident light.
    (3) Defects and harmless patterns can be discriminated by ellipsometric parameters Δ and Ψ.
    Based on these results, we have designed a practical instrument for in-line automatic inspection with three polarization cameras for three different azimuth angles. This inspection system is successfully under operation in West Japan Works of JFE Steel Corporation.
    Download PDF (10234K)
  • Yasuhiro WASA, Akira OKAMOTO
    2004 Volume 90 Issue 11 Pages 877-882
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    An automatic ultrasonic inspection system for crank throws of built-up type crankshaft has been developed. Several kinds of UT-probe for each part of throw are used to detect faults in important area. Probe scanning technology with proper contact force and noise eliminate technology are newly developed, so that the stable inspections for all types of throws are realized.
    Download PDF (4070K)
  • Hajime TAKADA, Takuya YAMASAKI, Yasuo TOMURA, Hideaki UNZAKI, Toshihir ...
    2004 Volume 90 Issue 11 Pages 883-889
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    A new ultrasonic immersion testing method for the detection of internal flaws in a running steel strip has been developed. It consists of a transmitting probe array and a receiving probe array arranged face-to-face on opposite sides of the strip in water, and two flaw echoes given as follows are received by the receiving probe array.
    (1) Flaw echo reflected first at a internal flaw and next at the surface wall of the strip.
    (2) Flaw echo reflected first at the back wall of the strip and next at a internal flaw.
    A linear area in the strip can be tested within a cycle of ultrasonic pulse repetition by integrating transmission of line-focused ultrasonic beam (25 MHz in frequency) using all elements in the transmitting probe array at once with parallel processing of signals received by the receiving probe array.
    In succession, an ultrasonic detecting system has been installed in No. 6 Pickling Line at JFE Steel Chiba District. Steel strip being tested is immersed in water by using 6 additional deflector rolls. The transmitting probe array and the receiving probe array cover whole width of the strip so that testing of whole volume of the strip is realized. It is confirmed that an inclusion with 5×10-5mm3 in minimum volume can be detected with signal-to-noise ratio in the range of 9-10 dB. The system is being used for the evaluation of commercial products without trouble, and very useful for internal quality control and assurance in the production line.
    Download PDF (1615K)
  • Yasuaki NAGATA, Yusuke KONNO, Toshiharu KINO, Jun HIROWATARI
    2004 Volume 90 Issue 11 Pages 890-896
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    In order to improve the defect detectability of Lamb wave inspection system conventionally used for on-line detection of internal defects of steel strip, application of non-linear signal processing, SSP (split spectrum processing), to Lamb waves and the optimization of the SSP parameters were investigated. Sample strips with natural defects were used to evaluate the performance of SSP in this research. As a result, the improvement of the sensitivity of Lamb wave inspection was confirmed by optimum selection of SSP parameters (filter number 10, filter tap number 500, frequency range 2.1-2.4 MHz, algorithm : combination of minimization algorithm and polarity thresholding algorithm, S/N : about from 2 to 5 times improvement). Furthermore, real-time processing of SSP was realized. Consequently, the newly developed Lamb wave can perform SSP at a 500 Hz repetition rate that is sufficiently high repetition rate for the on-line Lamb wave inspection system.
    Download PDF (3585K)
  • Mitsuo HASHIMOTO, Koichi ASAI
    2004 Volume 90 Issue 11 Pages 897-901
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    High quality products are demanded in steel production. Therefore quality measurement technology corresponding to it becomes more and more important. Eddy current testing is generally used for surface inspection. Because this technique enables high speed inspection without contacting objects. However, eddy current testing using magnetization is used because magnetic noises are large in steel samples. On the other hand, the technique is governed by an electromagnetic phenomenon, but it is difficult to understand this phenomenon intuitively. In recent years the phenomenon elucidation of the eddy current testing by numerical analysis was became possible by combining with development of numerical analysis technology and exploiting of markedly improved computers. Therefore establishment of the numerical analysis for this testing is expected to provide a useful method of separating signals of defects from magnetic noises.
    Download PDF (4572K)
  • Yoshito ISEI, Tatsurou HONDA, Kazuyoshi KIMURA, Yukihiko YAKITA, Yasuh ...
    2004 Volume 90 Issue 11 Pages 902-908
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    To measure strip velocity accurately between the stands of the hot strip finishing mill is effective in keeping the stable production and improving the strip dimension control accuracy. Two laser doppler velocimeters has been installed at F4 stand exit and F5 stand exit for evaluating their durability and measurement accuracy. These velocimeters have achieved accurate and stable measurement in bad environment with water splash and high temperature at interstand of hot strip finishing mill by implementing the effective countermeasures to environment and improving the digital signal processing. F5 stand mass flow gage is calculated by using the F4 stand exit gage and the strip velocities measured by these velocimeters. The F4 stand exit gage is measured by existing interstand X-ray thickness gage. The evaluation results indicate that the mass flow thickness has been measured within ±0.46% accuracy and will be able to apply to the mass flow AGC (Automatic Gage Control) at the hot strip finishing mill.
    Download PDF (1666K)
  • Akira KITAMURA
    2004 Volume 90 Issue 11 Pages 909-916
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    This paper describes the application of system control technology and the future view in a steel manufacturing process. In the steel manufacturing process of Japan, new system control technologies have greatly contributed to the quality of a product, or improvement in the yield. Then, in steel making, a rolling process, etc., it looks down at the application example of control theory or intellectual system technology. Specifically, the optimal flux control system for converter, looper control of hot-rolling by adapted type robust control of and an is introduced. Mill setup modification control of hot-rolling and mill balance control of tandem cold mill are explained as know-how fusion type control. Moreover, expectable technologies will be introduced from now on for achievement of the left-behind subjects. Specifically, the view is described about the application in consideration of restricted conditions of the nonlinear optimal control, mixed logical Dynamical system (MLD), and reinforcement learning. As an application example, the learning result of the function by reinforcement learning is shown. Finally, rolling is made into an example and the paradigm of the next-generation intelligent control system of a steel process is introduced.
    Download PDF (1681K)
  • Masahiro ITO, Shinroku MATSUZAKI, Harutoshi OGAI, Naoki ODATE, Kenko U ...
    2004 Volume 90 Issue 11 Pages 917-924
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    With the advance of computer hardware performance and database system technology, it has been practicable to store and to search for a large quantity of data. So that Local modeling method, such as Just-In-Time modeling, has become very attractive in recent years. We have developed "Large scale database-based Online Modeling" as the practical method based on the Just-In-Time modeling concept on blast furnace operation, which has very complicated physical phenomena and strong non-linear specific characteristics. The validity of the developed modeling method has been confirmed by the study with blast furnace operation data, then the past similar operation data have been searched and the prospective operation data have been estimated very quickly and precisely.
    Download PDF (1737K)
  • Jun-ichi IMURA, Akira KOJIMA, Shiro MASUDA, Kazuro TSUDA, Kazuya ASANO
    2004 Volume 90 Issue 11 Pages 925-931
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    This paper proposes a new looper control design method of a hot strip mill, where the strip tension and looper trajectories between the noncontact mode and the contact mode are simultaneously optimized. First, based on the model linearized around some equilibrium point in each mode, a discrete-time piecewise affine model, which is one of standard hybrid system models, is derived to describe the discontinuous dynamics of this system. Next, a model predictive control method for this model is proposed. The proposed method provides an exact solution to this control problem. However, it will not be easy to calculate an exact solution on-line via the usual personal computers. So an approximate solution method is also presented, where a solution can be obtained on line. Furthermore, the proposed method can be used to generate a feedforward control input, where optimality of trajectories between the non-contact mode and the contact mode is considered. This will be practically useful because it is executed via off-line computation. Finally, numerical simulations show the effectiveness of the proposed method, compared with the constant-torque based control.
    Download PDF (1294K)
  • Shuya IMAJO, Masami KONISHI, Jun IMAI, Tatsushi NISHI
    2004 Volume 90 Issue 11 Pages 933-940
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    In recent years, many control systems have been developed and applied to hot rolling in order to improve the quality of the finished products. Due to the plant characteristics and the control system performance variations over time, human experts intervene and modulate control gains to maintain and to improve the control system performance. Through a breakthrough, we can expect to attain the fully automated modulation of control gains without human intervention. A neural network model representing gain modulating actions by human was developed for a looper height controller in hot strip mills. The developed neural network model is a recurrent type neural network (RNN) which calculates the appropriate PID gains of the looper height controller based on the modification data of human operations as training data. Further, a learning algorithm for the RNN model was developed to accelerate convergence of the gain modification process and to stabilize the looper movement. The neural gain tuning model was applied to the inter-stands looper height controller in hot strip mills. The usefulness of the developed model was checked through numerical experiments. From the experimental results, it was verified that the tuning actions by humans can be realized by the model. Through its learning mechanism, the model could also cope with disturbances such as changes in roll gap. This may lead to the stabilization o f threading operations of hot strip mills.
    Download PDF (1272K)
  • Yoshiro WASHIKITA, Yasunori KADOYA, Kazuyoshi KIMURA
    2004 Volume 90 Issue 11 Pages 941-946
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    This paper describes a system identification method on the gage control system of steel rolling. System identification is a useful method for modeling a controlled object to design and tune a high performance controller. However, in online identification during commercial rolling, it is usually difficult to identify exact open loop properties of the system, because the power of the observed signals is relatively small in low frequency, especially in cold rolling. Therefore, we developed an identification algorithm based on auto-regressive moving average model with exogenous input (ARMAX) model to satisfy the specified steady state gain constraint, which is often obtained from the physical model or constants. Furthermore, to remove influence of the causality between input and output signals yielded from feedback gage control, we firstly identified tension response properties using the above new algorithm, and secondly calculated gage control properties by way of interactive model between tension and gage. The effectiveness of the proposed method was demonstrated on the feedforward gage control system of a reversing cold mill.
    Download PDF (1252K)
  • Yasunori KADOYA, Toshitsugu OKADA, Ryuji HAMADA, Yoshiro WASHIKITA, Ka ...
    2004 Volume 90 Issue 11 Pages 947-952
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    A new strip width control technique has been proposed for a cold tandem mill.
    First, the phenomena of the strip width change during the cold rolling were investigated. The results are as follows. The strip width changes on materials and sizes. Moreover, when rolling conditions like the rolling speed are changed, the strip width deviations are caused for the same grade strips. These mechanisms can be explained based on the relationship between the transverse metal flow and the change of the strip profile at the edge.
    Second, the strip width control technique to compensate for the strip width deviation caused by rolling speed change was examined. The relationship between the rolling speed and the target strip crown for the strip width control has been derived since the strip width change is related with the integration of the change of strip profile. The strip width control is adjusted the work roll bender in response to rolling speed, so as to become the strip crown to the target crown. The effectiveness of this control technique was confirmed through experiments.
    Download PDF (898K)
  • Akira MURAKAMI, Makishi NAKAYAMA, Mitsuo OKAMOTO, Yoji ABIKO, Kenichi ...
    2004 Volume 90 Issue 11 Pages 953-957
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    Pass schedules for a tandem cold mill affect the productivity and quality of rolled strips. This paper describes optimization of pass schedules, which are optimized via Sequential Quadratic Programming method. Performance functions and constraint conditions are chosen to achieve desired rolling conditions such as rolling forces, motor electric currents and reductions in thickness. Consequently, strips with small gage tolerance can be produced at higher productivity.
    The newly optimized pass schedules are applied to a 5-stand tandem cold mill. The results showed 3% decrease in off-gage length and 0.4% increase in productivity.
    Download PDF (902K)
  • Kazuya ASANO, Hiroyuki TAKAHASHI, Takeshi MIYATA, Yoshitake KOHIRO
    2004 Volume 90 Issue 11 Pages 958-963
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    In reverse cold rolling, tension control between the rolling stand and reels plays an important role in assuring stable rolling of strips. While in tandem mills tension control is performed by manipulating the roll velocity, the reel motor torque is manipulated to control tension in reversing mills, which characterizes the tension dynamics as a second-order system. As a result, when the dumping ratio of the second-order system is small, the tension dynamics becomes oscillatory and hard to control. To this end, a new control scheme is proposed to stabilize tension between the rolling stand and reels in reversing mills. In the proposed control scheme, a feedback loop of the derivative of the tension to the reel motor torque is newly introduced to the existing feedforward control. Since this feedback is equivalent to a state feedback of the strip velocity to the reel motor torque and increases the dumping ratio, it is effective to alleviate the oscillatory nature of the tension dynamics and to stabilize the tension. Simulation and experimental results are shown to demonstrate the efficient performance of the control scheme.
    Download PDF (995K)
  • Masami KONISHI
    2004 Volume 90 Issue 11 Pages 964-969
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    Production management technologies such as production planning and scheduling are reviewed together with planning and scheduling for logistics of products in steel industry. First, research topics in last decade on fundamental theories and related technologies concerning optimization such as scheduling, meta-heuristic methods for optimization, branch and bound method, Lagrange relaxation, multi agents and inference system by AI are described. Next, applications of optimization technologies to production planning and scheduling in steel industry are introduced. Reports on production managements for steel industry such as iron and steel making, rolling, surface finishing and inventory management together with transportation of steel products are reviewed. These technologies have been applied to production management of steel industry in various ways from single individual process to combined plural production processes. Finally, future challenge items are proposed. These are intelligence for production system, large scale optimization for production management cooperating plural processes, supply chain management system for steel production and accumulation and reuse of human knowledge for advanced steel production.
    Download PDF (1623K)
  • Toshio OKAWA, Shigeto HOJO
    2004 Volume 90 Issue 11 Pages 970-976
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    In most of production planning problems, it is required to manage multiple objective functions. An autonomous decentralized planning algorithm has been developed to apply to multi-objective lot-making problems. The methodology is to improve multi-objective function values through iterative negotiation among the planning modules which represent individual objective functions. The idea of planning 'share' is introduced as the measure to adjust the priority of multiple objective functions. The proposed algorithm has been applied to the plate design problem in steel manufacturing process and its effectiveness has been shown.
    Download PDF (3538K)
  • Youichi YOSHINAGA, Toshio OKAWA, Minoru TANABE, Yoshiaki NISHINA, Masa ...
    2004 Volume 90 Issue 11 Pages 977-982
    Published: November 01, 2004
    Released on J-STAGE: May 29, 2009
    JOURNAL FREE ACCESS
    It is required to establish the "Recycling-based society" in the 21st century. Development of effecive inverse logistics networks and recycling system in consideration of cost and environmental burden is a key issue. This paper describes planning system of inverse logistics networks. Application of the planning system makes it possible to evaluate optimized logistics network and also capabilty of recycling facilities by using actual operational data.
    It is difficult to get the optimal solution for the large-scale and complex problem of inverse logistics supply chain, because it involves specific constraints and the recycling requirements often vary. The proposed logistics planning system adopts effective discrete optimization.
    High performance inverse logistics network (flow and storage recycling facilities network) can be designed by using the planning system which is combined with the Geographical Information System (GIS). The planning system has been applied to planning and evaluating inverse-logistics networks for recycling municipal waste plastics.
    Download PDF (5957K)
feedback
Top