Int J Adv Manuf Technol (1996) 11:186-197 © 1996 Springer-Vertag London Limited
The lntemntim~ Journal of
Rdvanced manuPacturin[! Technolouu
Tool Paths and Cutting Technology in Computer-Aided Process Planning
R. M. Boogert, H. J. J. Kals and F. J. A. M. v a n H o u t e n
Laboratory of Production and Design Engineering, University of Twente, Enschede, The Netherlands
This paper reports on the development of a module to calculate automatically tool paths and cutting conditions for metal cutting operations. Process planning must select correct cutting conditions to minimise disturbances on the shop floor owing to tooling problems. Tool path and cutting condition algorithms to generate reliable NC programs have been designed. The algorithms have been implemented in the framework of a generative computer-aided process planning system, called PART. Geometrical requirements to avoid chipping of cutting teeth are considered in tool-path calculation. The cutting conditions are calculated using metal cutting process models. A method has been developed to calculate cutting forces for milling operations' based on experimental data of cutting ]brces in turning. In the process models, various constraints of the machine tool, cutting tool, and the workpiece are considered.
Keywords: CAPP; Cutting conditions; Milling
Metal cutting is one of the most important processes in part manufacturing. Traditionally, considerable emphasis was placed on how to control the workpiece flow within manufacturing systems. It is only during the last decade that the progress of the production processes without disturbances has received serious attention. Both in small and large batch part manufacturing, the availability of tools and the correctness of the technical machining information, e.g. the NC programs, are critical for a reliable execution of production processes. Computer integration and flexible manufacturing are key items for today's companies to remain competitive. Highly automated machining processes and the continually increasing quality of cutting tools and auxiliary equipment reduce the total machining time of products more and more. The increasing demands on product variety, shorter throughput
Correspondence and offprint requests to: Dr R. M. Boogert, Laboratory for Production and Design Engineering, University of Twente, PO Box 217, 7500 AE Enschede, The Netherlands.
times and smaller batch sizes require a minimisation of disturbances on the shop floor. In small batch part manufacturing, characterised by a high product variety, small lot sizes and universal applicable resources, the reliability of in-time delivery of the products is essential. Owing to smaller batch sizes, shorter throughput times and improved manufacturing technology, machine tools must be supplied with new machining information on a more critical time basis. This can result in a situation in which the preparation time of a manufacturing job exceeds the actual machining time. This problem can be eliminated by the application of highly automated process planning systems incorporating process models , which can quickly generate correct and reliable machining information on a continuous basis. The geometrical and technological correctness of automatically generated NC programs is important to avoid extensive machine test runs. Despite some two decades of research, the implementation of computer-aided process planning systems is still far behind their true potential. Only a few systems have actually been used by industry and even fewer have reached the stage where they cotfld make a significant impact on manufacturing practice [2, 3]. Process planning is closely related to the urgent need for factory integration and integrated manufacturing and can be seen as the critical link between computer-aided design and manufacturing. Ironically, most planning systems themselves are not integrated even though they were motivated by such a need. Current research tends to focus on specific process planning functions. These micro viewpoint approaches produce systems which perform individual process planning tasks in isolation from other planning activities. They result in many small and still isolated islands between the islands of design and manufacturing. The planning of manufacturing activities is a system problem, and therefore capacity, process and operation planning must be addressed in an integrated manner. At the highest level, capacity planning must be tied together with long term strategic factors in order to arrive at decisions which are optimal from engineering, economic and social points of view . Decisions made at the operation planning level must be based on physical models of the manufacturing processes being planned. The process models developed so far must be
. The process models developed are based on available literature on metal cutting and are used to predict and to optimise the cutting conditions. machining operations and the selected cutting tools. is a very deep understanding of the manufacturing processes and systems. The calculation of cutting conditions must be based on metal cutting process models that describe the relations between the various machining parameters (e. owing to the complex and diverse geometrical shapes of the volumes to be machined.g. The high variety of products and features in small batch part manufacturing requires extensive calculations of tool paths. 2. has to be determined in the tool path calculation function of process planning. Consequently. In principal. cutting speed. Tool Path Calculation for Prismatic Parts
An NC program contains both geometrical and technological information.
(. Traditionally much time is lost on the shop floor with tests of the NC programs on the machine tools. A prerequisite for the integration of advanced tool path calculation in process planning is the use of already available planning results. The tool path interface extracts the 2D boundaries of the features. This means that the generation of reliable tool path geometry and cutting data must be improved. To decrease the time span of process planning for part manufacturing. regardless of its level of complexity. During these tests the machine tools are not productive and lower the machine tool utilisation. The product quality requirements. Prerequisite for the development of a generative process planning system. Research into and implementations of 2D tool path algorithms are already available and should be integrated in automatic process planning systems. pocketing and facing (see Fig. so the time spent in testing decreases.~J. (b) Profiling..g. tool path calculation amongst other functions must be supported in an automatic and reliable way.Tool Paths and Cutting Technology
used to support the process planning decisions in a more realistic way.)
. the generation of tool paths is time consuming and error prone.
. the dimensional tolerances and the specified surface roughnesses. so less interactive work is required.
Modern C A D / C A M systems offer enhanced interactive tool path calculation algorithms for prismatic and rotational parts as well as for sculptured surfaces. 1. (c) Pocketing (spiral-outward). the design of the fixtures). The trajectory which the cutting tool must follow to cut all the material.
.g. the composition of tool sets.-
h E _ _ J "
~C . The geometrical data must be derived automatically from an unambiguously defined model of the workpiece. A major characteristic of 2~-D machining operations is the similarity of the cutting tool trajectory on the various levels of cut. 1). no process planning functions are integrated in CAD/CAM systems and the amount of interactive work required to generate NC programs for simple products is still much too high. " '
Fig.. feedrate). Realise an optimised cutting process with respect to time and cost. Tool path strategies in prismatic parts.g. Visual checking of the correctness of the tool path can be performed using those systems. 3. The improvement of machine tool utilisation and the prevention of rejects are goals of tool management. like volumes to be machined. as well as the manufacturing technology concerning the machining processes and the resources. However. must be considered as well in the process models. e. These tests must be carried out because of the risk of damage which can be caused by programming errors. e. (d) Facing (zigzag). depth of cut. Figure 2 shows the architecture of the tool path module. This means that the tool path calculations must be based on solid model descriptions of the volumes to be cut. Avoid machine idle times owing to extensive test runs. The feature boundaries can be constructed from the feature
2.¢./ / /
. contouring. point-to-point. the requirements with respect to the available tooling and cutting technology data are translated into a cutting data model. but the technical correctness cannot be tested. The geometrical tool path calculation and the establishment of cutting conditions are separated for the following reasons: The depth of cut is a major cutting condition and should be considered together with the establishment of the other cutting conditions. The machining information must be geometrically as well as technologically correct to: 1. Achieve a reliable cutting process without unexpected disturbances such as broken cutting tools or bad surface finish. A 2½-D tool path can be constructed from a 2D tool path and a certain depth of cut by copying the 2D tool path on each level of cut until the total depth is reached. (a)Point-to-point. the NC programs. Process models are developed for the automatic calculation of tool paths and cutting conditions for 2½-D machining operations. there exist only a few types of tool path calculation strategies for prismatic parts. The process planning department is responsible for the preparation of the technical machining information (e.
the data. process models of the machining processes are used to calculate the cutting conditions and it has the following advantages: The amount of data to be stored is much smaller and easier to maintain. etc. Tooth breakage versus the exit angle in milling . much research has been carried o u t to establish an adequate description of the cutting characteristics by mathematical models. and optimisation criteria can be considered during the generation of cutting conditions. A prerequisite for process models is the existence of literature with model parameters or the process models must be simple so that the parameters can be derived from machining tests. because the dependencies between the cutting conditions and the influences of the environment on the cutting conditions. These models should reduce the storage requirements and facilitate the automatic generation of reliable cutting data. which is characterised by successive down.
3. This method of establishing cutting conditions has several drawbacks: The storage space of all the data will grow very fast owing to a combinational explosion. can be calculated. cannot be included in data tables. It becomes more and more difficult to ensure the consistency of. with respect to the incorporation of new workpieces. Most of the known research approaches with respect to cutting force models and production optimisation only perform
~. cutting conditions have been established using data delivered by the tool manufacturer or cutting data handbooks [5-7]. 4. 2. Figure 4 shows an example of the variation of the engaged cutting angle in milling tool paths.5 does not mean that all the tool path elements are cut with the same width of cut. which affect the product tolerances and surface finish quality. the cutting tool. A stepover distance equal to half the cutting tool diameter assures that all material is cut. Cutting forces and tool wear. the user needs to interpolate between the upper and lower values and only cutting experts know the mutual influences of the various cutting conditions. surface roughness and tool life. Boogert et aL
3D part feature(s} 2D tool path boundary interface n'~dule software package 2D tool path Condition module 3D path
width of cut for every tool path element. A far more flexible way. e. The cutting conditions can be optimised within constraints set by the product specifications and the available resources. The handbook approach is not very suitable for the automatic generation of cutting conditions.. These handbooks include tables with recommended cutting speeds. and to maintain. M. the optimisation objectives. The boundaries have to be converted into the data structures required by the tool path software package and an offset tool path is generated. exit angle ~exit
Fig. Using CNC machine tools.
. The actual values are input for the generation of cutting conditions and are considered in the calculation of cutting forces.g. depths of cut. the product tolerances. the machine tool. Geometrical requirements with respect to dangerous exit conditions in milling must be considered during the calculation of tool paths. for a wide variety of workpiece materials and cutting tool materials. The recommended values cover a wide range. Figure 3 shows that an exit angle between 30° and 120° may cause tooth breakage. Metal Cutting Process Models
Fig. parameters or can be determined using solid modelling intersection techniques. In this approach.188
R. The tool path algorithms have to be controlled by a stepover distance between two tool path offsets. This tool path is then converted into the required format of the process planning system. 3. Variation of the engaged cutting angle in milling. is the generative approach to the calculation of cutting conditions. cutting tool and operation types. Particularly for zigzag milling. The variation of the engaged cutting angle requires an analysis of the actual
Fig. Equations are used to predict cutting process characteristics. Architecture of tool path module. an unequal offset distance may be required to avoid tooth breakage as a result of dangerous exit conditions. feed rates.and up-milling. cutting processes can be repeated easily and the improvements in modern shop floor control systems allow data acquisition during processing from which process models can be adapted . To overcome the aforementioned drawbacks. such as cutting forces. tool materials. Even a pre-defined stepover-ratio of 0.
The manufacturing accuracy is influenced by the deformation of these components. The present state of knowledge on metal cutting does not permit an equation to be drawn which considers all factors influencing the cutting forces. k~1.b.5% for every decrease (increase) in k. Fe increases (decreases) about 1. in a complicated mathematical form which in view of the complex calculation requirement must be regarded as impracticable . l . Clearance angle c~: No influence on the cutting force components could be ascertained for clearance angles between 3° and 12° . These cutting tests include many combinations of workpiece and cutting tool material and are performed with applicable ranges of cutting speeds and feedrates.g.
Fp = kpin.g. Kienzle empirically established the dependence of the cutting force on the cross-sectional area of the cut. Reliable information on the expected cutting forces is significant for the following reasons: The forces occurring in metal cutting processes furnish information with respect to the power and spind~ torque requirements of the machine tool.1 related to the cross-sectional area of cut b × h = 1 mm x 1 mm and the rises in force 1-me. Colding recognised the need for a wider approach in the establishment of cutting conditions . however.3)
Rake angle ~: Fc increases (decreases) about 1. Cutting edge angle K: The influence of the cutting edge angle is contained in the definition of the nominal width of cut b and the nominal thickness of cut h.h 1-me
= k f l . with: b = ap/sin(K). the influence of the cutting speed is taken into account in such a manner that for various cutting speed ranges the specific cutting force values are tabulated. the cutting force equations developed by Kienzle are adopted for the following reasons: The formulae do take into account the major factors influencing the cutting force components.5% for every decrease (increase) in h. The influences of the cutting edge geometry. However. The incorporation of the cutting speed into the force equations results. The following corrections on the calculated cutting force components are recommended by K6nig in case of deviations of cutting edge parameters compared with the cutting edge parameters used during the experiments:
Cutting force: Feed force: Passive force:
Fc = kda.1) (3.0% for every decrease (increase) in ~. Results of Witte  show that cutting data determined in turning operations. Fp increases (decreases) about 10. subject to turning. In listing the specific cutting force values. b . for cutting edge angles less than the experimental cutting edge angle (K = 70°). Most of the work is based on the extensive research of Kienzle  and Kronenberg .1 and kpl. h = f. under reservations. The forces cause deformations of the components involved in the cutting process. K6nig recommends to use the following rule of thumb:
F~ = 2 * &K=7o) Ff(K-actual).1. The cutting force component equations are:
The specific force units kcl.
Cutting Force Model
The cutting forces in metal cutting operations are important criteria for the machineability of a workpiece material. and the cutting tool. the workpiece material. As opposed to the cutting force equation.0% for every decrease (increase) in Fp increases (decreases) about 4. Helix angle k: Fc increases (decreases) about 1. In this research.5% for every decrease (increase) in ~/ Ff increases (decreases) about 5. K6nig mentions that the feed force and passive force equations of Kienzle must be regarded as an approximate solution because of a greater dispersion in the measured feed and passive forces compared with the cutting forces. 10]. the fixturing elements and the machine tool itself. Many cutting tests have been documented . the cutting tool.b.
3. The production rate may not be increased at the cost of reliability of the cutting conditions. 1-mr and 1-mp related to the nominal width of cut are entered into the equations as characteristic data. as many of these factors are not subject to any definite law.0% for every decrease (increase) in h. the cutting conditions.2) (3. may be applied to other machining methods. The specific force units are largely dependent on the workpiece material.sin(K). Local optimisations do not contribute to a global economical optimisation of production. considering the cut of a single tool path element or one machining operation) [9. The cutting edge geometry exercises a substantial influence on the quantity of the cutting force components. In this case. The course of the cutting forces related to the width of cut represents itself as a linear function of the thickness of cut in a system of tog-log coordinates (see ). More than 15 different research approaches are referred to. h 1-mr
(3. the cutting tool material and the progressive tool wear on the force components have been quantified. The rises in force characterise the force behaviour of a combination of workpiece and cutting tool material for different thicknesses of cut. e. This behaviour allows systematic modifications and extensions of the existing cutting data set with new experimental results. Many decades of research have been put into the establishment of empirical cutting force equations.Tool Paths and Cutting Technology
a local optimisation of the cutting conditions (e.hl-mp
All the cutting force experiments have been carried out with tungsten carbide as well as high-speed steel cutting tools. An extensive overview of previous work on fundamental cutting force equations is presented by Lau . a purely mathematical consideration of b and h leads to faulty results.
the resulting cutting forces are about 10% lower as opposed to tungsten carbide. A substantial rise in the cutting force components with increasing cutting time is registered.sin(+) sin(K)
The nominal width of cut b is the same as in turning: b = ap/sin(K) When using a helical cutting tool geometry each cutting edge penetrates step-by-step into the workpiece. (3.
. identical cutting edges.e. etc.3) were established for turning operations only. As mentioned before. The superposition of several cutting edges clearly shows the advantage of the helical cutting tool geometry compared with the straight-edge in respect of the cutting force. The cutting tool is divided into several small cutting edge elements.
.1 mm VB Ff rises about 25% per 0.. Variation in the nominal thickness of cut in milling. To cope with the complex course of the cutting forces in milling operations. equal tooth spacing.
Determination of Cutting Conditions
In this section. 5). reduces vibration and requires lower forces and spindle power during
4.). The usually applied helix angle in the cutting edge causes a gradual entrance of the cutting tooth over the depth of cut.
When cutting ceramics (for example A1203) are employed as the cutting tool material.1 mm VB Fp rises about 30% per 0. Bouzakis & Methenitis  and Lau . The method considers both the step-by-step penetration of the cutting edges in the case of helical cutting tools and the number of teeth in cut for each angular position of the cutting tool. K6nig refers to much research concerning the influence of tool wear on the cutting force components. equally distributed over the total depth of cut (see Fig. to other machining methods like milling and drilling.2) and (3. Boogert et al. When the cutting edges overlap. 6). The variation in the nominal thickness of cut is considered by the re-calculation of h in each angular position of the cutting tooth. The feed per tooth (]~).190
R. the elementary cutting forces must be added up to obtain the total cutting forces. At each angular position of the cutting tool. cutting tools with helicaledge inserts have been developed to combine the advantages of the traditional helical HSS. Recently. M. a three-dimensional numerical integration method. these equations may be applied. the determination of cutting conditions for milling operations is explained. The elementary cutting forces acting on the cutting edge elements are calculated by addition of the incremental cutting force for the angular positions of the engaged cutting tool. The cutting speed (v¢). A cutting edge element is cutting.
sin(~) = h / f z --~ h = fz-sin(~)
Fig.. This well-known solution improves the stability of the milling process. when the angular position of the element is between the entry and the exit angle of the cut. The separation between the geometrical tool path calculations and the calculation of cutting conditions has already been discussed in Section 2. is applied to calculate the cutting forces.. as proposed by Kline . 6. reaching a maximum value of the cutting forces that is obviously lower than that achieved with a straight cutting edge. The calculation of the nominal thickness of cut h must be extended as follows: h = f~. Any influence of high-speed steel or carbide grades could not be demonstrated. the step-over ratio A e D is supposed to be fixed and the following cutting data must be calculated: The depth of cut (ap). the milling operations.
3. Segmentation of a cutting tooth in cutting edge elements . The teeth of the cutting tool are cutting intermittently and the number of teeth in the cut is not always the same. The cutting speed in combination with the cutting forces determines the required
. the following corrections should be made: Fc rises about 10% per 0.
'4. Compared with turning.. with reservations. The nominal thickness of cut h is determined by the position of a tooth in cut (see Fig. The total forces are checked against the previously calculated minimum and maximum force values.
The cutting forces are dependent on the step-over ratio.2 Cutting Forces in Milling
Initially the cutting force components equations (3. As the tool wear (related to the width of flank wear VB) increases. the calculation of the cutting force components in milling is more complicated because: The thickness of cut continuously changes during the engagement of a tooth. 5. brazed carbide or solid carbide end mills and the advantages of interchangeable inserts . The method assumes a regular cutting geometry of the milling tool (i.1 mm VB
ith e l e m e n t
Fig. the depth of cut and the feed per tooth.1). For each tool path element. the cutting forces never drop to zero and the cutting tool is always under load.
It is capable of generating results for both on-line and off-line process planning for machining centres. The various process planning tasks are divided into six functional modules: Machine tool selection. Again. the capacity of the limited resources is considered.
5. considering the economical aspects of minimum cost or time and sufficient tool life. a violated surface roughness constraint can only be solved by reducing the feed per tooth. Capacity planning. the feature recognition (FR) module extracts the manufacturing features. This means that for every tool path element of a machining operation. the depth of cut is calculated with the minimum allowed values for the feed per tooth and cutting speed. The total tool life used for this machining operation is also known now and should be below the specified maximum amount of tool life. The upper bound values are the initial cutting conditions and the force-related and roughness constraints are checked and reduced when constraints are violated. The system can interpret product models (B-rep) and recognise automatically manufacturing features and selects all the necessary resources. independent of the cutting speed. the maximum flute length or insert
length) or by the cutting tool and workpiece combination (e. From the solid model. Jigs and fixtures selection. the feed rate and the cutting speed may vary over the various toot path elements. The calculation starts with the upper bound values for the cutting conditions resulting from the physical limitations of the cutting tool and the workpiece combination. The cutting conditions are known for all tool path elements and the progressive tool wear on each element can be calculated. limited chip capacity). A violation of a forcerelated constraint can only be solved by changing the depth of cut or the feed per tooth. e. dictated by tool management. feed per tooth and depth of cut) may also be required when the minimum cutting speed (e. NC programs and auxiliary information for the manufacturing of prismatic parts in small batches are automatically generated. On the contrary. cutting tool and workpiece. i. In 2½-D milling. should define the definite sequence. the process models calculate cutting forces. Violation of this constraint requires another iteration to establish feedrates and cutting speeds. Implementation
The tool path and cutting condition algorithms have been implemented in a generative computer-aided process planning system. The cutting speed also has a major influence on tool life. As stated before. Selection of machining methods.e. Further for each tool path element. the attainable machining accuracy and the surface roughness. the depth of cut and the feed per tooth determine the cutting forces.g. PART deals with the selection of set-ups. These modules are executable programs which communicate via a relational database. The minimum value of the depth of cut over the various tool path elements becomes the determined depth of cut for all tool path elements. the step-over ratio.g.Tool Paths and Cutting Technology
machine tool power. requires a reduction of the cutting speed. The sequence in which the modules are executed is controlled by the supervisor module. cutting tools and jigs and fixtures. called PART . The cutting force model and the roughness model consider the specified limitations of the machine tool. These constraints are called "force-related constraints". the depth of cut is maximised first. using the bisection method until none of the constraints is violated any more. machine tools. the depth of cut must have the same value for all the tool path elements belonging to one machining operation. the feed per tooth and the cutting speed are calculated in a next iteration using the already determined depth of cut. From these initial values. The determination of cutting conditions is performed using an iteration method. machining methods and sequences and the calculation of tool paths and cutting conditions. A strategy is required to decide which initial cutting condition should be reduced when a certain constraint is violated. to avoid build-up edges) is reached. As a matter of fact. cutting tool and workpiece. tool deflection and attainable theoretical surface roughness and these results are checked against the limitations of the machine tool. i. Violation of mainly cutting speed dependent constraints. The NC programs and the related documents for the work shop. the cutting speed is determined last. to produce the parts. Therefore. power and tool life. that may be used for this operation.g. The CAD Interface (CI) module accepts the solid product models from external CAD systems and converts them into ACIS models. the largest feasible depth of cut should be determined before determining the feed per tooth. Calculation of tool paths and cutting conditions.
. minimum production costs or minimum production time. Selection of cutting tools.e. on which most of the process planning functions are based. Both the production costs and the production time decrease with the increasing depth of cut and feed per tooth but the depth of cut has a greater influence than the feed per tooth . Consequently. like tool lists. The feed per tooth and the cutting speed are temporarily set to the minimum values to avoid violation of force-related constraints owing to these two cutting conditions. Figure 7 presents the PART reference scheme. the cutting conditions cannot be determined simultaneously. The optimisation objective.g. the force-related constraints are checked for violations due to higher cutting forces as a result of tool wear. During these selections. A reduction of the force-related cutting conditions (i. The maximum value for the depth of cut is physically limited by the cutting tool (e.e. When a constraint is violated. The roughness is dependent on the feed per tooth as long as the cutting speed is high enough to avoid build-up edges. As a result of the various metal cutting process models. According to the roughness model. A certain sequence to determine the conditions is required. The influence of tool wear on the cutting forces is accounted for in a third iteration. the cutting conditions are reduced.
modeller independent manner of extracting geometrical boundary information and the possibility of generating a boundary for features which only exist temporarily as a result of a premachining operation and do not have a direct link with the solid model geometry. operation sheets and set-up sketches. Intersect the feature model faces with an imaginary plane perpendicular to the depth orientation of that feature. This is called the slice technique. Advantages of the feature parameter method are the solid
. 19-22]. the boundary contains gaps. The second method of tool path boundary extraction is based on intersection techniques available in solid model systems. More detailed descriptions of the modules can be found in [1. the 2D tool path is copied several times until the total depth is reached. unless an open boundary is required and gaps are closed by extension of boundary elements or addition of extra elements.192
Generation of Milling Tool Paths
Prismatic parts are characterised by 2½-Dmachining operations. The methods described above suppose that the side tolerances are symmetric and the boundary geometry always corresponds
User I UserInterface
Fig. As a result of intersecting features in the workpiece model. More complex features have to be treated with the intersection method. The construction of a tool path boundary using feature parameters is restricted to feature types which can be defined unambiguously with a fixed set of parameters. Boogert et al. the tool path boundaries are extracted by the following methods: Calculate the boundary using the feature parameters. PART reference scheme. For the generation of tool paths and cutting conditions this means: the trajectory of the tool path is calculated in a 2D plane and depending on the depth of cut and total depth to be machined. 7. M. Depending on the complexity of manufacturing features. The faces which bound a manufacturing feature are intersected with an imaginary plane temporarily constructed somewhere between the feature position and total depth. are created by the NC output module. The PART system uses ACIS for the representation of workpieces. Open boundaries can be extended to assure that the cutting tool starts at a position outside the workpiece geometry and results in a tangential entry. Additional problems in the extraction of tool path boundaries are the non-symmetric tolerances on side faces of features. the situation can occur that at the level of the faces/plane intersection. and perpendicular to the depth orientation. Boundaries are assumed to be closed.
.>. A tool path element is split up when the deviation of the width of cut exceeds a predefined limit. depth of cut. use in principle one pre-defined fixed stepover distance for the boundary offset calculation.Tool Paths and Cutting Technology
with the nominal dimensions of features.>'\i /\ \ I \I
/'\1 /\ \ I \I
". 5. During the calculation of cutting conditions half the tolerance bandwidth is used as a maximum value for the cutting tool deflection. The cutting edge material type and the geometry of the cutting edge (e. The model constants which have been established in turning experiments. The cutting forces on the elementary cutting edge elements are compensated for with the factors defined in Section 3.
. 8. helix angle) are the characteristics in the search for the correct constants.
". and feedrate. The extracted boundary always corresponds with the nominal feature geometry.5 .75. the cutting forces and consequently the machining accuracy and material removal rate.25. Therefore. The automatic calculation of 2D milling tool paths given a set of boundaries has been studied by many researchers  and many implementations are available in CAD/CAM systems. perpendicular
.g. For nonsymmetric tolerances a shift of the boundary geometry with the tolerance offset is required to assure that the correct tolerance bandwidth is used for cutting tool deflection. The given depth of cut is divided into several small cutting edge elements. except for zigzag toot paths which can have two pre-defined values for the stepover distance. The shift in angular position of a cutting edge element depends on the height of the cutting edge element as seen from the cutting tool tip and the helix angle of the cutting tool. The elementary cutting force components are split up into two orthogonal components (i. The tool path algorithms.. The engaged angle dominates.e.. D • ~ 1
0.1 \ 1 \ /\ /\ \ \ I \ I \ \ \1 \l \
width of cut 0.. D-
Fig. The actual geometry being cut is intersected with the tool volume at several discrete points on the toot path element and the length of the highlighted intersection (see Fig. In spite of a fixed stepover distance. as described above.3 Generation of Cutting Conditions
A tool path element which is input for the cutting force model has a constant width of cut and initiated values for the cutting speed. 8) is converted into a width of cut. together with the depth of cut and the feedrate. The cutting forces on the elementary cutting edge elements are calculated for several angular positions between the entry and the exit angle of the engaged tool. equally distributed over the total depth of cut. Nonsymmetric tolerances can be considered in the construction of tool path boundaries from feature parameters.D
0. The stepby-step penetration of the cutting edges in the case of helical cutting tools is considered by the correction of the angular position of the cutting edge elements. It is therefore important that the actual engaged angle on every tool path element is known. For cutting tool deflection. . Non-symmetrical tolerances cannot be handled using the slice technique. are retrieved from the database. rake angle. half the tolerance bandwidth is used and the tolerance offset can only be considered in the final NC program using a cutting tool diameter compensation. The actual geometry being cut results from a solid model intersection of
/ f t f / I I I L___L / I I I
the tool swept volume and the actual geometry of the intermediate workpiece model. "-. The actual engaged angle on each tool path element is calculated using the actual geometry which has been cut by the tool moving along the tool path element.. The axial rake angle of milling tools is compared with the rake angle of the turning cutting tool which has been used during the cutting force experiments. Determination of the actual width of cut./ /x I
.-"-. no new tool path algorithms are developed for spiral outward pocketing and profile contouring but existing tool path software for these strategies has been integrated into the PART system. the engaged cutting angle over the various tool path elements is not constant. The position of the boundary elements is shifted so the tolerance bandwidths become symmetric.
200.1) at lower cutting speeds. The depth of cut is maximised first and therefore the cutting speed and the feed per tooth are temporarily set to the minimum values to avoid violation of force-related constraints owing to these cutting conditions. This ensures that when the feedrate and the cutting speed are determined and reductions are required to obey force-related constraints. Violation of the required machine tool power is solved primarily by a reduction of the cutting speed.
and parallel to the feed direction). The selected cutting conditions represent the v~. Because of systematical influences which are not included in the model or which were not considered during the establishment of the model constants. the cutting forces are calculated and the forcerelated constraints are checked. the piezo-electrical elements generate an electric charge. The cutting forces were measured during a straight line cut.~
Kist~er meter Dynamo
Fig. Schematic representation of the measurement configuration.e. Boogert et al. The cutting forces were measured with a Kistler dynamometer containing piezo-electrical elements. When a force is applied on the dynamometer. e. ap.
.e. The experiments have been carried out on an MAHO MC 400 milling machine. Because of the linearity of the dynamometer . The actual process data can then be compared with the predicted force values. From these calculations results the maximum permitted feed rate for further calculations. Figure 10(a) shows the results of the experiments with a low cutting speed (i. M. The sampling frequency was 10 kHz and the signals were analysed for 5s. and several combinations of feedrates and depths of cuts. 200 m/min) and various combinations of depth of cut and feedrate. adaptation of the constants may be required. 600 and 1000 m/min) . A e D and information with respect to the type of cutting tool material and geometry. K~H and Kn. workpiece material group. an analyser was used to determine the cutting forces. From literature. Figure 10 shows the forces calculated with the cutting force model versus the measured cutting forces. the depth of cut never violates these constraints. fz.e. plece
Spectral. minimum production costs or minimum production time). extra tool changes and new cutting tool or inserts) are considered. The maximum feedrate. having a fixed width and depth of cut. Now the cutting force model overestimates the cutting forces by about 15%. for every reduction of v~. Cutting forces and roughnesses are calculated with the maximised depth of cut and minimised feed and cutting speed. The feedrate and the cutting speed for a given tool path element are initiated at the maximum values defined for the combination of cutter class. The evaluation of cutting forces can be performed on three hierarchical levels .g::~
X Y Z
worv. a certain offset in calculated forces compared with the measured forces can be seen. is determined now and consequently the cutting speed is calculated in accordance with the objectives (i. The predicted forces are about 10% too low.g.e. In the experiments. the specific cutting forces were given for three different cutting speeds (i. The surface roughness values are calculated with the given maximum feedrate and are checked against the required roughness values. machining process characteristics. The behaviour of the cutting force model corresponds with the measured cutting forces. When the cutting speed cannot be adequately reduced. The workpiece material was low alloyed aluminium. the cutting forces must be decreased and this situation is handled as if the force-related constraints were violated. The used cutting tool was a face mill 80 mm with insert type TFAN 2203 PFR and material type HW-(K10). The selected cutting conditions are input for the cutting process.
6. However. and ranges of depth and width of cut. For every tool path element. cutting edge material type. and other relevant data. The minimum depth of cut for all tool path elements becomes the final depth of cut. 9. All experiments were carried out with a single tooth cutting tool and because of the rotation frequency of the cutting tool. different cutting speeds. Figure 11 shows a feedback loop with the cutting force diagnostic module to evaluate structural deviations between measured and predicted forces and to determine adapted model constants. A similar overestimate was found for a cutting speed of 1000 m/rain.194
R. workpiece material. Because of higher specific cutting forces (i. Such a geometrical cut can be reproduced easily and the width and depth of cut can be varied in many ways. The orthogonal cutting forces on each angular position are added up to obtain the total cutting forces. A1MgSil. A real-time control loop is necessary to invoke immediate
6. The force measurements should be invoked for particular blocks in the NC program. cutting forces were measured for a fixed width of cut. the cutting forces can be derived directly from the amount of charge.e. Violation of a force-relation constraint is solved by setting the cutting speed to the minimum value and reducing the feedrate using the bisection method. The tool life equation is used to calculate the theoretical tool life for the given cutting conditions and the costs resulting from the tool wear (i. Therefore experiments were carried out to determine differences between the calculated forces and the measured forces in milling.1 Adaptation of Cutting Force Model Constants by Feedback
The cutting force model uses experimental turning constants which are available from cutting force literature . Figure 9 schematically illustrates the measurement configuration. not violating the force constraints. Figure 10(b) shows the results of similar experiments with a higher cutting speed (600 m/min). the force-related constraints are checked again.
Evaluation of the Cutting Force Model
The correctness of the cutting force model is very important for a reliable prediction of cutting conditions.
/ .~ % .1.. This is represented by the feedback loop between the cutting process and force measurement... .35
500N 4501 4001 3501 3001 2501 2001
600 m/rain AeD = 0. In milling.: % =2.. r y /
Adapted model constants
( e. . 11.~ . 11) can monitor the optimisation control loop and may change the structure of the"applied models or choose a different model when the optimisation control loop tends to become unstable.
experimental turning constants
Cutting Predicted Cutting
] r °
t rx c.15
t 0. .. C a l c u l a t e d
Selected cutting conditions
cutting forces versus the mean measured
0. / .~//
.0mm . The cutting force equation is:
= kcl. Kcl.15
* O. Commercially available tool or process monitoring equipment can be applied to perform the real-time control loop.g.30 fz
F i g .125 ..~r
AeO = 0.X = model • = process .25
/ //. . In this section.05
0. The second control loop is the optimisation control loop consisting of the cutting
force diagnostics module which adapts the cutting force model constants in case of structural deviations between predicted and measured force values.. / "
b.. action in order to avoid damage when a force threshold is exceeded... Kf11. ./ I~'~. the maximum thickness of cut at angle qbhmcan be approximated by the mean thickness of cut hm along the active width of cut
.0 mm :a~ =3. The cutting force model uses a numerical method to calculate the maximum cutting forces Fc and Ff.25
0. Feedback loop with cutting force diagnostics module. I. 1-mc.. The third control loop (not depicted in Fig..~ 0.Tool Paths and Cutting Technology
200 rdmin : ap = 2. / / i I~s t i f t t 4< 1/~
0.. 10. hl-m~
The maximum cutting forces occur at the maximum value for h. only the optimisation control loop is considered. Constant
v¢ a n d
0.0 mm :ap=4. . 1 roff)
.0 mm :ap=3. . ~ . the nominal thickness of cut is not maximal at a certain angular position of a cutting tooth...0mm :ap=4..
Machining Science and Application. Metal Cutting Principles (Oxford series on advanced manufacturing. J. \\. A much more reliable prediction is attainable through the use of the adapted model constants.
Fig. (i. The experiments show a larger decrease in cutting forces than the model does. 5. The cutting conditions are calculated based on the machining capabilities of the resources. A. M.
process model and a method has been proposed to adapt constants used in the process method. 1982.0 mm \~NN. Colding. J. Cutting force measurements are carried out to verify the reliability of the
1. "Determination of forces and power for cutting tools and machine tools" (in German: Bestimmung von Kraften und Leistungen an spanenden Werkzeugen und Werkzeugmaschinen). Oxford. pp. Dammer. 2. pp. pp. the influence of the cutting speed is not included sufficiently in the cutting force model. Annals of the C1RP. 12. A. 600 and 1000 m/ rain. B. Cincinnati. I. Cutting data for manufacturing from the INFOS database). "Computer aided process planning: the present and the future". Shaw. The model constants Kcla and Kn. Ohio. 27(1). 10.
A tool path and cutting technology module for process planning has been designed and implemented to generate reliable toot paths and cutting conditions to reduce the time spent for testing the NC programs on the shop floor. "An adaptive control module for ROUND". The predictions for low cutting speeds are too low but they become too high for higher speeds (>400 m/min). University of Twente. Process analysis and cutting condition generation for face milling). The constants for intermediate cutting speeds are interpolated. 1978. 1991. wt-Z. 827-832.. "Evolution and future perspectives of CAPP". Hoff and H. 1987. 12. M. Figure 12 also shows the predicted cutting forces using adapted model constants. J. 1984. Lu. Generation of optimised cutting data with technology oriented databases). ind.196
R. The helix angle of the cutting tool should be considered during the numerical establishment of hm.. 1966. 58-59. 1 metal-cutting). Machining Data Handbook. "Bereitstellung optimierter Zerspandaten durch technologisch orientierte Datenbanken" (i. E1Maraghy. into the maximum cutting force Fc and feed force Ft. using +hm. A metal cutting process model is used to calculate cutting forces in milling. 1952. Tiemersma. ~ 1O0
~ 200 300
. Ph D thesis. L. "Zerspanwerte ffir die Fertigung aus der INFOSDatenbank". 11. C. "Ein Beitrag zur Prozessanalyse and Schnittwertvorgabe beim Messerkopf-stirnfr/isen". Kronenberg.~. Annals of the CIRP. Clarendon Press. 299-305. J. Pennsylvania State. C. NL. (i. 4. L. M. Verein Deutsche Ingenieure 94. 7. Metcut Research Associates Inc. 3. Y. 591-601. Ham and S. Annals of the CIRP. Prediction of cutting forces using adapted model constants. 1993.. Pergamon Press.~ \A ~ AeD = O. W. F. 739-751. Fertig. A similar procedure must be followed for the feed force. van Houtcn and J. 453-458. Oxford. J.
. For various combinations of b and hm a log-log diagram can be constructed. Proceedings 19th CIRP International Seminar on Manufacturing Systems.
(b). pp. VDI-Z 123(20). Z. 1979. 8. M. 9. K6nig.e. Boogert et al. 69(1).M. The measured maximum Fx and Fy values must be converted.m e can be determined graphically.
7. Kienzle. 42(2).e. 1981. 37(2). Dammer. O. 1988.3 mm a O = 3.
400 ~. 1966. 6. 125 x = model & = process • = model with adapted constants
150 ~ . The constants Kcl. pp.1 and 1 . F. van Houten. D. PhD thesis.1 are available for a cutting speed of 200. A. Wesch. University of Aachen (RWT). M. pp. "Relative effects of shop variables on manufacturing cost and performance". As can be concluded from Figs 10 and 12. ~z = 0. H. USA. A strategy is developed to calculate 2k-D geometrical tool paths. "PART a computer aided process planning system".e.
1987. Verlag Stahleisen m. pp. NL. Lau. 1994. Specific cutting forces in turning and drilling). September 1993. D. 23. University of Aachen. Switzerland. face and corner mills". A. "Tool management in computer aided process planning". Proceedings CIRP/VD1 Seminar: Developments in Cutting Technology. 20. "New cutting tool geometries for shank-. Springer Verlag.b. T. "Optimization of milling conditions". L. 19. Satran. W. Bouzakis and G. University of Twente. "A software architecture for CAPP systems" PhD thesis.
f f~ F
F¢ Fe r~
mm mm N
N N i
h hm k~1. Witte. 1990. width of cut/cutting tool diameter) cutting tool diameter feed feed per tooth resultant cutting force cutting force feed force passive force nominal thickness of cut mean thickness of cut specific cutting force unit specific feed force unit specific passive force unit rise in cutting force related to the nominal width of cut rise in feed force related to the nominal width of cut rise in passive force related to the nominal width of cut cutting speed flank wear clearance angle helix angle rake angle cutting edge angle angular position of a cutting tooth angular position of a cutting tooth for the mean thickness of cut
. Jonkers. University of Twente. International Journal of Machine Tool Design and Research. F. Boogert.. 1985. R. NL. "Spezifische Zerspankr~ifte beim Drehen und Bohren" (i.H. A. On the Computational Geometry of Pocket Machining.1
kn. M. K6nig. Lenderink. Winterthur. Essel and L.e. Operating instructions Three Component Dynamometer type Z 11206 SN 108482.~
kpH 1-me 1-mr t -mp vc VB k 2t
K (~hm o~
mm mm N/ram2 N/ram z N/mm 2
o o o
depth of cut step-over ratio (i. UK. R. D. R. Witte. Boerma "The design of fixtures for prismatic parts". "Determination of the values of the technolgical parameters which are used to describe the time course of cutting force components in milling". 1992. 34(1). J. K. DeVor and R. PhD thesis. NL. 1991. D. PhD thesis. 1982. Wertheim and A. M. pp. Kline.Tool Paths and Cutting Technology
13. Methenitis. 15. "The integration of process and production planning in small batch part manufacturing". NL. 22. "Specific Cutting Force Data for Metal-Cutting". Dfisseldorf. 22. University of Manchester. Lindberg. 21. K. Dfisseldorf. Kistler. 17. 1978. 18. 16. 7-22. 14. R. W. University of Twente.e. M. 1980. Annals of the CIRP. C. 1994. Held. PhD thesis. 24. E. J. PhD thesis. 85-101. "The prediction of cutting forces in end milling with application to cornering cuts". 1982. PhD thesis. L. University of Twente.