Vol.5 No.2 2012

27/66

Research paper : Toward the integrated optimization of steel plate production process (K. Nishioka et al.)−101−Synthesiology - English edition Vol.5 No.2 (2012) in obtaining a satisfactory result. The simulator assumes as given the parameters of daily manufacturing lot sizes, product mix, processing capacity of each finishing process, equipment utilization rate, etc., and has as its objective the fine-tuning of the priority order of processing, which we consider was the fundamental cause of this failure to obtain a satisfactory result.We learned again through the study of the existing cases of the lean production system and also from understanding the limitations of the logistics simulator that it is most important to realize leveled production, together with an appropriate investment in bottleneck processes. Therefore, to ensure the enlargement of manufacturing lot sizes and the leveling of loads on finishing processes simultaneously, we worked on the development of a production lead time model capable of describing comprehensively the relationship between the manufacturing lead time variation and the success rate of delivery time, as well as the relationship between the product mix and manufacturing lot sizes and between the manufacturing lead time and the amount of stock.(3)Required time/ in-process stock modelThe above undertaking identified the relationship between the manufacturing lead time of each product type whose estimation used to be difficult and the corresponding time required to complete each process. However, even at this stage, it was not fully understood how the required completion time of each process is determined and which control factors played a part therein. In contrast, for the analysis of waiting time for cases where there is a variation in the frequency and intervals of events and in the intervals of processing, the queuing system theory was applicable.[12] The development of a model capable of describing the required completion time and in-process stock volume of each process became possible by the development of a required time/ in-process stock model based on the queuing system theory.5.3 Development of an efficiency modelWe have already explained that the plate production is classified largely into the upstream rolling processes and the downstream finishing processes. The rolling processes are comprised of heating, rolling, cooling, etc. and each one of these processes is arranged continuously, in direct connection and in tandem. In contrast, the finishing processes are comprised of heat treatment, ultra sonic testing (UST), coating, gas cutting, cold leveling (CL), oil-press leveling (OL), surface grinding, etc., and each process is arranged independently and in juxtaposition.Like the plate rolling processes where many types of steel mix and flow, changing greatly the efficiency of each process and where the processes are arranged continuously, in direct connection and in tandem, the efficiency of each process changes in response to the process conditions of each material, and the bottleneck processes change sequentially. In association with this, the friction of preceding and subsequent processes occurs frequently, changing greatly the efficiency of the plant as a whole. In contrast, if the processes are arranged independently, it is possible to stock in a sufficient amount between processes as a buffer, and the friction of inter-process processing rarely occurs, resulting almost uniquely in the efficiency of each process by the process conditions of each material.In order to overcome the issues of heavier load in rolling as a consequence of having expanded the application of TMCP/ accelerated cooling technology and of diversified products, the most important issue is to reduce the inter-process friction of processes in rolling caused by the diversification of process conditions associated with the diversification of steel types and by the smaller lot sizes to be processed. We therefore attempted to develop an efficiency model that enables quantitative evaluation of the lowering of efficiency attributable to the friction between individual processes of rolling, by determining the efficiency of each process according to the process conditions matching the material specifications and to the processing lot sizes, and at the same time by identifying bottleneck processes.Each one of the rolling processes has various factors that work against better efficiency. For example, the heating process has the temperature at the start of heating, heating conditions, etc.; the roughing process has the temperature to extract a slab after heating, rolling sizes, etc.; the finishing rolling process has the TMCP temperature, weight of the slab to be rolled, etc. We developed a rolling efficiency model Component ratio %Manufacturing lead time, in daysAchieving delivery due dateManufacturing lead timeApprx. log-normaldistribution Past recordsDelivery due date<1<3<5<7<9<11<13<15<17<19<21<23<25<27<29<31<33<35<37<39<41<43<45<47<49024681012141618 exp (2 µ + 2){exp( 2)-1} exp µ + 2 20222exp21-(ln( )- )> 0xxx < 0x ( )= xfσσ ( )= xE ( )= xVσσσFig. 4 Distribution of manufacturing lead time

※このページを正しく表示するにはFlashPlayer9以上が必要です