DECOMPOSITION OF FORGING DIES FOR MACHINING PLANNING Laurent TAPIE, Bernardin Kwamivi MAWUSSI, Bernard ANSELMETTI Abstract:   This paper will provide a method to decompose forging dies for machining planning in the case of high speed machining finishing operations. This method lies on a machining feature approach model presented in the following paper .   The two main decomposition phases, called Basic Machining Features Extraction and Process Planning Generation, are presented. These two decomposition phases integrates machining resources models and expert machining knowledge to provide an outstanding process planning.  Key words:   Forging die high speed machining; Feature Modelling; Feature decomposition model.  1.   INTRODUCTION  The   recent   expands   of   High-Speed   Machining (HSM)   technology   and   Computer-Aided   Machining (CAM) software, lead forging die makers to invest in these technologies. Indeed, several works underline the productivity and part quality gains involved by HSM [1]. Yet, the difficulty inherent in machining forging dies   lies   in   developing   an   outstanding   machining process by using HSM. The main expert problem is to make   appropriateness   between   die   topology   and machining resources (CAM software, HSM machine tool, numerical controlled unit (NCU), cutting tools, tool path...) during CAM process planning. Besides, another inherent difficulty is the die model given by forging process simulation and analysis. Indeed, in the past this analysis gave a wire frame model. According to this model machining expert defined both the die design model and the machining planning [2]. By the way   of   recent   development   in   forging   die   process simulation,   the   die   model   is   completely   designed during forging   simulation. Consequently,   machining experts elaborate process planning starting from a 3D surface or solid Computer-Aided Design (CAD) model. In   this   paper   we   are   presenting   an   original approach to elaborate process planning based on the die decomposition during the finishing machining stage. This   decomposition   objective   is   to   provide   an outstanding   process   planning   method   taking   into account   productivity,   part   quality   and   resources constrains. This process planning method is divided in two   stages:   Basic   Machining   Feature   Extraction (BMFE) and Process Planning Generation (PPG).   In the first paper section, a forging die machining features definition model is provided in the case of   three axis finishing HSM. Then, according to this model three levels of decomposition   are   highlighted:   geometrical   and topological decomposition level, machining resources decomposition   level   and   machining   tool   path decomposition   level.   By   the   way   of   these   three decomposition   levels   five   typical   basic   machining features are defined: horizontal plane feature, complex plane feature, flank contoured feature, flank followed feature and blend feature. These basic features are given by the analysis of several forging dies CAD model. First, this paper provides the machining feature model   we   have   develop.   Then,   BMFE   method definition is detailed in section 3. Finally, the PPG stage is provided and applied on a typical forging die.  2.   FEATURE MODELING APPROACH  The feature modelling way is chosen in our method to   associate   CAD   &   CAM   models   and   machining resources.  2.1.   Forging die machining feature model  The core of this model relies on a Unified Model Language (UML) machining feature model (Fig. 1). Four principal classes are defined. The surface class is represented by the CAD die model. Typical surface used to model die part are given in [2]. The topology class   represents   the   relationship   type   between machining feature. Interference topology and proximity topology are the two main topology types. This model is detailed in tab.2. The technology class gives the needed requirements for forging process. In our work this   three   classes   are   considered   as   data   given   to machining expert. The last class, called “machining process”,   represents   the   machining   expert   point   of view. In the case of die machining this class is detailed according to the model represented in Fig. 5. Thus, the machining process represents the aggregation of three classes.   The machining direction class represents the tool axis direction which can be reached according to part shape without any global interference. This class can be represented by a direction map as in [5]. In forging production the part must be extracted in one direction.   Consequently,   an   extraction   direction   is associated to the part. This specificity is integrated in the technology class Fig. 1. So, in the case of forging die machining each machining feature can be machined with   the   same   direction.   Besides,   three axes HSM technology is commonly used in die makers industry. Thus, the model provides a machining direction similar to the spindle axis.
+Type  Feature  +Type  Surface  -is commosed of 1 * +Type  Topology  * +Type: forging process +Material +Slope angle +Minimal blend radius +Surface roughness -Extraction direction: vertical  Technology  * +Type: 3 axes HSM  Machining process -has 1 *  Fig. 1.   UML machining feature model The tool path class represents the CAM model associated   to   the   CAD   model.   This   class   model represents the part finishing phase. Thus, this class is composed of tool trajectory during cutting phase and tool trajectory during approach or clearance motions. The cutting path is decomposed in two sub-classes: the direction path and the sweeping mode. Three common path directions are provided in CAM software for three axes   finishing   machining:   along   curves   in   plane containing   tool   axis,   along   curves   in   plane perpendicular to tool axis, along 3D curves calculated on part shape with geometric criterion (iso parametric, iso curvature...) or part quality criterion (iso scallop height). Then, a direction path is associated with a sweeping mode such as zigzag, one-way, morph in, spiral sweeping. These different data are detailed in [6]. Machining   resources   are   also   integrated   in   our model.   The   machine   tool   skills   have   an   important impact in the feed rate, consequently in part quality and machining   time   [7].   Thus,   two   sub-classes   are associated to the machine tool. The structure class group the machine tool data: axis stroke, maximum feed   rate,   maximum   acceleration,   jerk;   spindle maximum speed, power. The NCU class groups the HSM close loop control parameter. A cinematic model of our machine tool, called “performance viewer” was implemented in Esprit CAM software [7] to model these classes. The cutting tools class model is divided in the tool cutting part and the tool body & attachment. This class is commonly used in CAM software in machining   simulation   to   analyse   local   and   global interference of the tools with the part shape.  Table   2  Topological class model 3.   BASIC MACHINING FEATURE EXTRACTION 3.1.   BMFE Method  The BMFE method is composed of four main steps (Fig. 2): (1) CAM tool paths are generated in several forging   die   CAD   models;   (2)   cinematic   tool paths analysis   is   performed   with   CAM   simulation   and performance viewer model. After steps (1) and (2) machining resources decomposition level is defined by iso-resources areas. In step (3), iso-resources areas tool paths are analysed by the way of expert knowledge. This knowledge is based on some machining tests and simulation performed on our machine tool with quality and   productivity   criterions   [6].   Then,   obtained machining features are validated in step (4) and basic machining features are extracted.  Machining features validation Machining resources affectation & basic tool paths generation 3D CAD die Model CAM Basic tool paths Basic tool paths analysis CAM Basic tool paths simulated  Performance Viewer CAM simulation Resources model Basic tool paths model   Expert knowledge   Machining process association Basic tool paths analysis Machining features Basic machining features  Performance Viewer CAM & machining simulation  Fig. 2.   BMFE method  3.2.   Basic machining features models  Five typical forging die basic machining features are obtained according to BMFE method. The “horizontal plane” (HP) feature is composed of one horizontal plane perpendicular to the tool axis.   To avoid machining with a null effective cutting speed, this feature is machined with end mill tool. A surfacing machining process is associated to this feature. The sweeping mode is chosen during PPG according to topological relationship and productivity criterion. The “complex plane” (CP) feature is composed of 1 to several surfaces. These surfaces can be represented by   generated   cylindrical   surfaces   (direction   and generatrix)   or   composed   cylindrical   surfaces   with “lean on” or “lead into”  Sub-type  Island junction Cavity junction Flank junction  Interference relationship  “Junction type”  Attributes   Opened junction Closed junction  Interference relationship  “Intersection type” “Is in intersection with” “belong to” Proximity relationship  “Property type”   Attributes   Total property Partial property  Proximity relationship  “Top type”   “Top”
3 surfaces primitives such as plane and revolved cylinder. The machining process associated to this feature is parallel plane machining. This machining process can be   adapted   during   PPG   according   to   topological relationship and productivity criterion. The “flank contoured” (FC) feature is composed of 1 to several ruled surfaces. These surfaces can be generated or composed of primitives (cone, cylinder and plane). The inclination angle of these surfaces is closed to the slope angle given by the technology class. The machining process associated to this feature is a parallel   plane   contoured   machining   process.   An adaptation of this process can be realized according to topological relationship and productivity criterion. The “flank followed” (FF) feature is composed of 1 to   several   ruled   surfaces.   These   surfaces   can   be generated or composed of primitives (cone, cylinder, plane). The inclination angle of these surfaces is closed to the slope angle given by the technology class. The difference between “flank contoured” is the machining process. This feature is machined along 3D surface generatrix curves. The “blend” (B) feature is composed of blend surface.   This   surface   assured   the   topological relationship between the last four features. This feature model is based on the model provided in [3]. This feature is machined along 3D curves or contoured according to the basic features in topological relation tab.3.  Table   3  Blend machining feature process 4.   PROCESS PLANNING GENERATION (PPG) 4.1.   PPG Method  PPG method is divided in four main steps (Fig. 3): (1) the basic features and their topological relationship are identified based on the BMFE stage results; (2) machining resources and basic machining tool path are associated based on the BMFE stage results; (3) Basic tool   path   are   adapted   according   to   models   and experimental analyses. The basic tool paths are adapted to   provide   an   outstanding   appropriateness   between machining constrains and topological constrains; (4) the process planning is generated. During step (3) some basic features are decomposed or associated, so the corresponding new features are associated with their tool paths and each feature operations are organised.  Machining planning Basic machining features & topological relationship identification 3D CAD die Model Basic machining features Toplogical relationship Machining resources affectation & tool paths generation CAM basic machining features  Machining process model Basic machining features model Topological model Expert knowledge   Basic tool paths adaptation Adapted Machining features Machining process planning  Performance Viewer  Fig. 3.   PPG method  4.2.   Application  Our method is applied on an industrial example. In Fig. 4, the example decomposition in basic machining feature   is   given.   Then,   topological   relationship   is deduced; for instance, FF1 “leans on” CP1 with an opened cavity junction and “leads into” FC2 with an opened   flank   junction.   By   the   way   of   this decomposition each basic tool path is generated.   CP1, CP2, CP3 are swept with up milling (one way) in parallel plane containing tool axis and perpendicular to the generatrix of CPi. FC1, B2 are contoured with up milling in plane parallel to CP3. FF1, FF2, B4, B5 are swept with up milling (one way) along generatix of ruled surfaces defining FF1 or FF2. FC2, FC3, FC4, B6, B7, B8, B9, B10 are contoured with Z-level up milling (one way). HP1, HP2, HP3 are surfaced with a morph out spiral and up milling (one way). B1 is swept along is generatrix curve with up milling. Then, basic tool paths are adapted according to topological relationship and the productivity criterion given by the “performance viewer”. For instance, FC1 “leans   on”   CP3   with   a   closed   cavity   junction. Consequently,   to   minimise   tool   retractions   and approaches,   CP3 is machined with a spiral from the centre to the outside limit of CP3.  CP1 CP2 CP3 HP1 HP2 HP3 FC1 FC2 FC3 FC4 FF1   FF2 B1 B2 B3 B4 B5 B10 B6 B7   B8   B9  CP1 CP2 CP3 HP1 HP2 HP3 FC1 FC2 FC3 FC4 FF1   FF2 B1 B2 B3 B4 B5 B10 B6 B7   B8   B9  Fig. 4.   Basic machining features identification For CP2 basic machining process, the “performance viewer”   reveals   areas   where   the   feed   rate   is   not reached. CP3 interference topological relationship with FF1 and FF2 induces these feed rate loss. So, feed rate loss areas detected are machined with FF1 or FF2. This adaptation is validated by the “performance viewer”, machine tool cinematic behaviour is better. Then, the machining process planning is totally defined (Fig. 6).  CP   FC   FF SP   No existence   Contoured   3D curve  CP   Contoured or 3D curve   3D curve  FC   Contoured or 3D curve
+Type: 3 axes HSM  Machining process  +Type: 3 axes  Tool path  +Type: 3 axes HSM  Machining resources  +Type: 3 axes -Spindle direction  Machining direction  1 1..*   1   1..* +Type:unspecified  Cutting path  +Type: 3 axes  Approach/Clearance path  1 1 1   1..* +Type:unspecified  Path direction  +Type:unspecified  Sweeping mode  1 1..* 1 1 +Type: 3 axes  Machine tool  1 1 +Type: HSM  Cutting tools  1 1..* 1 1 +Type: 3 axes  Structure  1 1 +Type: HSM  NCU  1   1 +Type: unspecified  Body & attachment  1   1 1 1 +Type: unspecified  Cutting part  Fig. 5.   Machining process model  Tool path adaptation Machining feature adaptation  Tool path adaptation Machining feature adaptation  Fig. 6.   Machining process planning.  5.   CONCLUSION  In   this   paper,   a   method   of   decomposition   for machining planning is provided. This method lies on a machining feature model including geometry, topology, technology of forging die and high speed machining resources. This method is divided in two stages: Basic Machining   Features   Extraction   and   Process   Planning Generation. Our   future   works   will   be   focused   on   defining topological rules by the way of the importance of the linkage between basic machining feature topology and machining process. REFERENCES [1]   P. Fallböhmer, C.A. Rodriguez, T. Özel, T. Altan, (2000),  “High-speed machining of cast iron and alloy steels for die and mold manufacturing” , Journal of Materials Processing, vol. 98, pp. 104-115. [2]   K.   Mawussi,   A.   Bernard,   (1996),   “Complex   dies representation   and   definition   with   complex   feature-based model” ,   Proc.   of   the   15 th   International   Computers   in Engineering Conference (ASME), September 17-21, Boston (USA). [3]   K.   Mawussi,   (1995) , “Modèle   de   Représentation   et   de Définition d'Outillages de Forme Complexe. Application à la génération   Automatique   de   Processus   D'Usinage”,   PhD dissertation, LURPA - ENS de Cachan.  [4]   Bernard, (1989) , “Usinage tridimensionnel d'outillages de topologie complexe: analyse des contraintes de production et contribution à l'optimisation du processus d'usinage”,   PhD dissertation, LURPA - ENS de Cachan.  [5]   Omar   Zirmi,   Henri   Paris,   (2005),   “Modélisation   de l’usinabilité   des   pieces   aéronautiques   dans   un contexte   de conception   de   gamme   d’usinage” ,   9ème   colloque   AIP PRIMECA La Plagne, 5-8 Avril 2005. [6]   P-Y. Pechard, L. Tapie, B. K. Mawussi, Y. Quinsat, (2006),  “Complex feature machining: a methodology for machining strategy   choice” ,   (2006),   15th   International   Conference   on Manufacturing Systems – ICMaS, 26-27 October, Bucarest, Romania. [7]   L.   Tapie,   B.K.   Mawussi,   B.   Anselmetti,   (2006),  “Machining strategy choice: performance viewer” , IDMME 2006, 17-19 May, Grenoble, France. [8]   L.   Tapie,   (2006),   “Die   machining   feature   definition”, Internal report LURPA Ens-Cachan. Author(s):  Laurent   TAPIE,   PhD   student,   Laboratoire Universitiare   de   Recherche   en   Production Automatisée   ENS-Cachan,   61,   Avenue   du   Président Wilson,   94235   Cachan   Cedex   France. tapie@lurpa.ens-cachan.fr Dr.   Bernardin   Kwamivi   MAWUSSI,   Assistant Professor,   LURPA   ENS-Cachan,   61,   Avenue   du Président   Wilson,   94235   Cachan   Cedex,   France. mawussi@lurpa.ens-cachan.fr. Pr.   Bernard   ANSELMETTI,   Professor,   LURPA   ENS- Cachan,   61,   Avenue   du   Président   Wilson,   94235 Cachan   Cedex   France.   anselmetti@lurpa.ens- cachan.fr.