【作者】 樊少华；

【导师】 申桂香；

【作者基本信息】 吉林大学 ， 工业工程， 2011， 博士

【摘要】 数控装备是制造业所必须的重要设备,更是先进制造技术的核心载体,汽车、航空航天、机械工程、船舶等行业对数控机床的需求日益增加。电子信息、生物工程等高新技术产业的兴起和发展对数控机床提出了新的需求,需要更多精密、高效和专用性机床。高档数控机床和基础制造装备是实现工业制造和工业现代化的重要保障,世界各国都非常重视高档制造装备行业,都在努力通过采用现代化技术积极发展制造业装备。中国是一个制造业大国,全面发展国内的数控装备制造业不仅有利于满足我国制造业对数控装备的需求,同时还可以提升我国的综合竞争力,具有深远的战略意义。结合国家项目,本文以12台某系列重型立式数控车床为研究对象,搜集整理可靠性数据并进行故障模式及影响分析,求出每个子系统的风险值;对整机和各个子系统进行数据拟合,拟合出整机和各个子系统的可靠度函数;根据所选择的影响度指标分别求解各个子系统的影响度,进行对比分析;依据第二、三、四章的结论,综合考虑影响整机可靠性的各个影响因素,对某系列数控车床进行可靠性分配。对数控车床子系统进行了故障模式分析,确定故障模式排序及子系统的风险排序。首先采集了12台某系列数控车床的故障数据,对故障数据进行了故障模式分析,将整机分为10个子系统分别为:液压系统(D)、伺服单元(F)、数控刀架(M)、电气系统(V)、主传动系统(S)、工作台(T)、横梁(B)、润滑系统(L)、CNC系统(NC)、其它(R),对每个子系统的故障模式进行分析。针对每个子系统的故障模式构建求解模糊风险优先数的模型,在求解风险优先数的过程中,对三个风险因子O、S、D赋予不同的权重,各位专家的评价结果也都采用模糊数表示,专家分配不同的权重,权重也用模糊数表示。这种方法能够有效克服传统RPN模型的缺陷,在很大程度上克服了评价的主观性并且提高了评价的准确性。对FRPN值进行去模糊化,根据去模糊化的值对各个子系统进行排序。对现场采集的故障数据进行初步的可靠性数据处理,根据第二章中对子系统的划分,分别求出整机和子系统的可靠性函数,采用比较分布函数曲线的拟合“相关指数法”对分布模型进行优选,确定整机和子系统均服从威布尔分布,并使用图检验法和D检验法检验模型的拟合优度。求出整机和子系统的可靠度函数便于第四章中的可靠性影响度分析,分别计算出整机和子系统的MTBF。求解子系统的可靠性影响度。重点考察了可靠性影响度的三个影响指标——故障次数影响度指标、故障停时影响度指标和动态影响度指标。整理计算这三个指标所需要的故障数据,计算出10个子系统在这个三个指标下的值,画出10个子系统的动态影响度曲线图,计算当T=1000时的影响度值。将这10个子系统的故障次数影响度、故障停时影响度和T=1000时的影响度值作对比分析,找出该型号数控机床的关键子系统是:液压系统(D)、伺服系统(F)、数控刀架(M)、电气系统(V)、主传动系统(S)、工作台(T)。对该系列数控车床进行可靠性分配。探讨了可靠性分配应该考虑的影响因素,最后确定用于指导该系列数控车床可靠性分配的因素有7个:子系统影响度、子系统风险值、复杂度、平均维修时间、工艺技术水平、环境条件和费用;针对这些影响因素对一些数据进行二次加工,综合考虑这些影响因素并构建一个可靠性综合分配模型;对于一些找不到相关可靠性数据支持的影响因素采用专家打分法,将数据代入可靠性分配模型,求出在下一批数控车床可靠性设计中所应分配的可靠性指标值。对子系统分配前后的可靠度进行对比,可以发现横梁、润滑系统、液压系统和分配后的目标值相差最大,需要进行重点的设计改进;CNC系统、电气系统和伺服系统和现有可靠性水平相差也很大,但是由于CNC系统的多数故障都是操作人员的误操作,电气系统的故障大都是由于元器件损坏,需要对操作人员进行培训,提高操作工人的业务素质,而电气元件主要是外购获得,所以企业应加强外购件的采购质量管理,从质量优选手册中优选厂家,如果不在手册中,应加强质量管理,实现质量控制。每个子系统的可靠性水平并不是越高越好,每提升一个等级所需要花费的费用则是呈几何级数增加的,每个子系统的可靠性都存在一个最佳值,超过这个最佳值便不会得到最优的经济效益。考虑到目前该系列数控车床的发展水平,并且目前还处于设计的初期,可靠性分配只能进行到子系统和部件级,随着我国数控车床的发展以及相关可靠性数据的积累,可以在此基础上将各个子系统的可靠性分配到各个零件。更多还原

【Abstract】 NC machine tools are very important equipment for manufacturing, they are also the core of advance manufacturing technology with the growing of air space, mechanical engineering, vessels etc. The electronic information, biology, the rise of the high-tech industries give out more needs to the development of NC machine tool, now more accurate ,efficient machine tools are needed very much. High-tech NC machine tools are the basis for fulfill industrial manufacturing and trades, many countries are trying to make high-tech machine tools. China is a manufacturing countries, the development of its manufacturing equipment is not only meet the need of its own industry development, but also can improve the overall competitive, with profound strategic significance.This paper studied 12 sets a series of heavy vertical NC lathes, gathered reliability data and analysis these data,some failure model analyze had been done,and we can get the risk data of every sub-systems; According to the data fitting of whole machine and sub-system, we can get the reliability function of them; at the same time the three reliability importance index will be calculated, compare there of them, some useful conclusions will be obtained. According to the above conclusions and take the whole machine’s reliability influence factors into consideration, and give out a reliability allocation decision of this type of machine tool.This paper make a failure mode analysis on sub-systems and point out the order of sub-system. First of all, gathered the failure data of one type of machine tool, and the failure mode analysis have been made, the whole machine was divided into 10 sub-systems, they are electrical system, tool carrier, spindle system, CNC system, spindle system, hydraulic system, lubricate system, collecting and transferring system, external interface and other else. To target each subsystem failure patterns of risk preference for a solution as a model of risk preference, in the solution of the process, the risk factors O S D are given different weights, and the weights are expressed as fuzzy numbers. This method can effectively overcome the defects of traditional RPN method, and overcome the subjective of evaluations made by experts. De-fuzzy the fuzzy number and we will get a real value, and sort the sub-system according to these real values.We preliminary processed the on-site data failure data, according to the classification of subsystems in the second chapter, and then calculated reliability functions of whole system and subsystems respectively. Next, we chose the best distrubution model through comparing with the fit correlation index of distribution functions, and comfirmed that both whole system and subsystems would obey the weibull distribution. Finally, we used a chart Examination methods and D Examination methods to test the fit of model. Calculating the eliability functions of whole system and subsystems would be benefit for reliability importance analysis in the fourth chapter, so in Chapter 4 we could figure out the MTBF (Mean Time between Failures) of whole system and subsystems.In order to figure out the reliability importance of subsystems, we focus on three importance indicators of reliability importance, including failure number importance index, failure downtime importance index and dynamic importance index. Based on the failure data needed by above three indicators, we computed values of this three indicators for 10 subsystems, drawn up the curve graphs of dynamic importance of 10 subsystems, and then calculate the value of importance as T=1000. Comparing the failure number importance and failure downtime importance of 10 subsystems with the values of importance as T=1000, we found out the key subsystems of this type NC machine tools, including hydraulic system(D), servo system(F), CNC system (NC), electrical system (V), the transmission (S), worktable (T).In order to allocate the reliable allocation for this type NC machine tool, we first studied the impact factors consided by reliable allocation, and found out the significant impact factors of reliable allocation are importance of subsystems, risk value of subsystems, c omplexity of subsystems, mean of maintenance time, level of technology, environmental conditions and cost. Then based on these impact factors, we taken on secondary processing for given data, and constructed a reliable allocation model. Third, we use expert evaluation method to deal with the impact factors which could not be supported by relative reliable date, and put date into reliable allocation model to figure out the values of reliable index which should be allocated in reliable design of next NC machine tool. Fourth, comparing with the reliabilities before allocation and after allocation, we found that the mamium difference between the actual value and target value would be existed in beam, lubrication system and the hydraulic system, so we should focus on the impovement in design for these three systems. We also found that, in CNC system, electrical system and server system, the differences between actual value and target value would be large too. Because most of the failure of CNC system are fault operation of operator, the cause of most electrical system failure are damage of components, so the operator should get more training and should improve the Standardization Management of the purchase components according to quality management system, the unqualified supplier should be correct once the components which are undesirable were found. Manufacturers should make a further improvement on confirming the law of quality acceptance clauses and the acceptance method, and at the same time reinforce the work concerning on debugging and detection.The reliability of subsystem is not the higher the better, and when the reliability is improved a little high, it also need a large amount costs of geometric progression, and every subsystem have its own optimum value, over this value will not get the optimum economic returns. Considering it is the first level of design of this type of NC lathe, the reliability can only implement to subsystem, with the development of technology and the gathering of failure data, we can divide the reliability to components.更多还原