Research on Framework of Digital Process Planning Platform

Research on Framework of Digital Process Planning PlatformMA Yumin' , FAN Liuqun2 , ZHU Zhihao' , ZHANG Hao'( 1. CIMS Research Center , Tongji University , Shanghai 200092 , China , E-mail : ymma@ mail. tongji. edu. cn ;2. College of Chinese-German ,Tongji University , Shanghai 200092 , China ;3. Shanghai University of Electric Power , Shanghai 200090 ,China )Abstract : Digital factory technology is a research focus in academe and industry , which is an advanced manu facturing tech-nology that is proposed to bridge product derelopment and manufacturing . For applying digital factory technology in machiningdomain ,a concept of digital process planming and its framework are suggested ，its components including machining domainknowledge model , machining knorwledge base , machining resource base and process planning system are studied. The framerwork ofdigital process planming is of vulue for implementing digital factory technology in machining industry.Key words: digital process planning platform ; machining knorwvledge base ; digital factory1 IntroductionDigital factory technology( DFT ) is an advanced manufacturing technology for bridging the gap betweenthe product development and the manufacturing planning. Fraunhofer institute for industry engineering( IAO)1 defined digital factory as the virtual( three- dimensional and digital )illustration of a real factory , equip-ment and product on the basis of an integrated data model. Now , DF T has been applied in automobile , airplaneindustry , and shortens product delivery time greatly , reduces manufacturing risk and cost. According to relativepaper , study of DFT is mainly focused in assembly domain , while in machining domain there is much less studyon it. Digital process planning platform is just the basis for studying and implementing DFT in machining domain.2 Concept of Digital Process PlanningProcess planning is a bridge between product models and manufacturing planning ,it turns the product' s de-sign information into manufacturing information. The society of manufacturing engineer( SME ) defined processplanning as making systemic decision for a product' s machining method in order to make it economically andcompetently , therefore , product planning is closely related to the resources and machining knowledge that an en-terprise owns. Digital process planning is to plan a product' s process on the digital manufacturing platformwhich means determining product process routes through manufacturing features recognition based on digital ma-chining resources and machining method，including process sequences , resources assignment ( machines , toolsand fixtures ) , cutting parameter' s setting , NC codes' generation ,and etc.3 Framework of Digital Process Planning PlatformDigital process planning platform has close relationship with enterprise. It includes a machining knowledgebase ,a machining resources base and a process planning system , where the machining knowledge base is a rulebase which stores corresponding rules for machining , and the resources base stores the available resources for ma-chining which includes machines , tools and fixtures. The above two parts are the base of digital process planningplatform. It' s based on them that the process planning system can selects proper machining method and tools forparts through a predefined reasoning mechanism , and generates valid process planning documents and NC pro-gram codes. The framework of digital process planning platform is shown in Fig. 1. .4 Constitute of Digital Process Planning Platform4.1 Machining Domain Knowledge Model中国煤化工Process plan is dependent on enterprise. Different enterpri.MYHC N M H Glet ,and the same prod-uct in different enterprise will have different process plan. Due to machining domain' s speciality，the knowledgemodel can be divided into five submodels , which are machining features , tools , part materials , machines andmachine motions. Each submodel can be divided further , and can be expressed in tree form by" OR" andAND" logic. A submodel example of machining features is shown in Fig.2. In this example , features can be.一913一InterpretationnoduleMachiningReasoningdomainengineknowledgerulebasemodelProductResourcedatabaseDomain knowledgetule rditormodel rditoreditorAdministrator interfaceFig. 1 Framework of digital process planning platformdefined as two types' Hole" and' Slot" , where the' Hole"type can be described by section shape( Section" ) and bot-9 Features)R~tom shape(' Bottom" ). The section shape(' Section" ) can8 Hole0 Slotbe divided into" Round” and" Rectangular" further , andBottom”can be divided into' Through" type and' Blind"Section 只AND只Botom .type.' That means that the hole has four types which are6 ORO6OR 8round through hole , round blind hole , rectangular throughRoundRectangular ThroughBlindhole and rectangular blind hole.Fig.2 Example of features submodelEach node in the submodel has parameters describes itsattributes which could be either a variable or a constant. Each Parameter of a node could be inherited by the chil-dren nodes , which is also called as parameter inheritance. For example , Parameters of node” Features" havebeen defined as" Position" " LengthC )rientation" and" DepthC )rientation" ( 3D vector ), parameters of nodeRound”are defined as”Diamete" , while parameter of node' rectangular" are' Length" " Width" and' Cor-nerRadius" . The two nodes of" Round" and" Rectangular" also have parameters" Position" ”LengthOrien-tation" and' DepthC )rientation”due to the inheritance ,and it' s no need to define them again.In the same way , other submodels , such as Material" " Tools" " Machines" and"' Motions" , can be de-fined.4.2 Machining Knowledge Base4.2.1 Creation of machining ruleCreation of machining rule is based on the knowledge model described above. In machining rule editor , ma-chining domain knowledge model is loaded and' features" submodel is copied as”井 Feature" submodel whichrepresents the less worked feature( LWF ). The former feature' ' submodel represents the more worked feature( MWF ). Machining rule is used to describe the machining method , which represents that a MWF in a kind ofmaterial can be produced from a LWF using a tool and a motion on a machine. So it should connect correspond-ing nodes together using a curve , as shown in Fig.3. At the same time , some restriction should be applied tomachining rules，such as machining precision , tool' s size，tool' s hardness and etc.After flling all of the machining rules and compiling them ,a machining domain knowledge base is availablefor the process planning system.4.2.2 Reasoning mechanismThe reasoning mechanism used in the digital process planping .olatform. . which means how to select a ma-chining method for a feature，is an object-driven reasoning m中国煤化工onverse reasoning ) anddeep first and backtracking , an example is shown in Fig. 4 2]MHCNMHGsentsafeaturethathasto be machined. The reasoning process will start from the upper corner. Firstly ，the candidate methods that canget the feature A can be found and sorted by priority. The method 1 has the highest priority and method 4 haslowest priority. The method 1 is applied firstly ,and if it is approved which means the restriction is meet , andthere' s a suitable tools in the magazine , and then generatesa LWF , feature B. The system will check whether_ 918 MethodQFeatures Q ToolsQ Materials Q Machines Q Motions 0 #FeaturestOHSS 184-AX[Mi6又HoleBlankMillTapdrill AlBWDrill6δNon-Thread TheadFig.3 Definition of machining rulethe feature B can be found in blank model or not. It'Method 1Method 10Method 30()s not found ,so it has to be machined further. ForFeatherA Method2 Feather BFeatherC Mehod 31feature B , the system continues the same reasoningprocess as feature A. the method 10 is applied , andMethod 3Method 11Method 40the restriction is also meet , and generates feature C.Method 4Method 12 Feather DMethod 41( ) BlankFeature C cannot be found in blank model too ,so ithas to be machined. The candidate method for theFig.4 Object. driven reasc oning mechanismfeature C is method 30 and method 31. The method30 is applied , but is not approved , and then method31 is applied , but is also not approved. This means there' s no more methods for the feature C and have to go onestep back ( backtracking ). The method 10 is approved to be invalid , and method 11 is applied , the process con-tinued. Finally , method 41 generates a Blank feature which is assumed to be present in the blank model. So afeasible process route has been found ,that is method 1- +method 11- -method 41. When applying in the prac-tice , the process route is in reverse order.An example of reasoning process of a step hole is used to ilustrate a reasoning engine' s running process ,asshownin Fig.5. If a step hole needs to be machined , the process route is"' blank- > center hole- > Hole- > Stephole" .P13(lank)AP. 10AF 2F_9Fig.5 Reasoning process of machining a step hole4.3 Resource BaseThe machining resource base is a basis of process planning in an enterprise. It comprises machines , tools ,fixtures and relative cutting technology.A Machine consists of machine body , control system , tool changer and pallet , so machine can be definedthrough the above four parts. Due to the system is oriented to box type part , there' s only one kind of machine ,that is milling machine( machining center ) , and the operation that can be performed is corresponding operationsof face and hole such as milling , boring , drilling , reaming and tapping. For machine body , some parameter suchas axis configuration( X ,Y Z A B ,and etc. ) , axis movement range , machine end type , and other parametersshould be defined.Tools are defined in the form of tool assembly , which consist of edge ,cutter body , adapter and holder. Thecollection of tools is called toolset. Tools are assigned to machine in the form of toolset ,so it can be regard asmagazine.A Fixture consists of base component，positioning component，clamping component and intermediate , andalso a fixture can be defined through the above four parts. Since the fixture participates in machining processand collision between cutting and fixture should be avoided w中国煤化Iated，the fixture modelshould be in three- dimension format.Relative cutting technology includes milling , boring , drilMHCNMHG。technology. In prac-tice ,definition of cutting technology is dependent on cutting type , part material and cutter edge' s material ,socorresponding feed rate and spindle speed and requirement of cooling fluid can be given.Relationship of resources is shown in Fig. 6..一915一区Tool Bierarchy白口区Tool setsMachine toolsTool assonbliesTool changer'sPallotCutting toolsHachine conF lgurat lonsViseHachine holdersPouver diogranBase plateAdoptersHachine endClanping toolsPosit ioning toolssystonsRnsetCancel J elpDRIL12.00CancelFig. 6 Relationship of resources4.4 Process Planning SystemProcess planning system is a system that faces user directly in digital process planning platform. It' s basedon the machining knowledge base and resource base , and defines machine , fixture and clamping position appliedin the part' s machining through human -machine interaction. Fig.7 is a workflow of process planning for multi-setup part. The process is described as following :CAD interfacePooduct baseF eature recognitionMachine (production line) definitionResource baseMethode and tools selectionNC sisimulationTool path generationKnowledge baseProduction bine simulationLine balancingReport outputFig. 7 Worklow of process planing( 1 )Process planning of parts is based on machining feature recognition , therefore ,a part model and a blankmodel should be imported in system firstly and the machining features are then recognized automatically by com-paring the two models.( 2 ) According to the primary process planning , the Planner should assign machine , magazine and fixture toeach setup of part , adjust the position relationship among machivtrs and,rort and define the machin-ing direction( normal vector of machining face ). The system中国煤化工:hosen diretion can bereached by the machine kinematics. For example ,a hole or a SMHCNMHGpartcan'tbemachined.by horizontal milling machine. If the direction cannot be reached ，a warning is given and the followed task can't be continued.( 3 ) When completing the definition , the system will reason a corresponding method and tool for each ma-chining feature based on machining knowledge base and resource base. Then , the planner can adjust it if re-- .918quired , such as method adjustment or tool adjustment.( 4)Tool path is then generated by system according to the selection of methods and tools , and the cllisionduring machining can be detected through simulation. Thus , the basic step of process planning has been fin-(5 ) Due to the assignment of machining feature to machine is based on the position relationship betweenmachine and part , the load of each machine can' t be balanced , therefore , a line balancing is needed. The linebalancing is reallocation of features' machining while meeting the technical requirements. Finally , a process planreport is output , including the operation sequence , resource lists ,APT file , NC code , machining time and etc.5 Conclusions .Digital factory technology is a research focus in academe and industry. It provides strong support for enter-prises to enhance their competence. For complicated parts , such as automotive engine cylinder body , cylinderhead ,and etc ,a framework of digital process planning platform is provided ,its constitutes including machiningdomain knowledge model , machining knowledge base , resource base and process planning system are analyzed.The framework is the basis of some intelligent CAPP system , and be helpful for process planning of complicatedparts and production planning.AcknowledgementThis work was supported by Sino German project , grant No.2002DFG00027.References[1] D. Spath ,T. Potinecke. Virtual Product Development- -Digital Factory Based Methodology for SME[ A] 37th CIRP Interna-tional Seminar on Manufacturing Systems[C] Budapest , 200453-60.[2] Tecnomatix Company. The Complete eM- Machining Methods Tutorial [ EB/OL] 2000.中国煤化工MHCNMHG一.917一