Optimization of Cutting Parameters
in Machining of Compacted
Graphite Iron (CGI)
Mulugeta Berhane Haile
Optimization of cutting parameters on
Compacted Graphite Iron (CGI)
Machining
Master of Science Thesis
Mulugeta Berhane Haile
May 26, 2011
School of Industrial Engineering and Management
Royal Institute of Technology
Table of Contents
1 Introduction ... 8
1.1 Project background ... 9
1.2 Objective of the project ... 9
1.3 Problem approach ... 9
1.4 Limitations of the project ... 9
2 Theory ... 10
2.1 Milling process ... 10
2.2 Cutting parameters ... 11
2.3 Machining Performance Measures ... 12
2.3.1 Cutting force ... 12
2.3.2 Surface roughness ... 12
2.3.3 Material removal rate ... 13
2.3.4 Tool life ... 14
2.3.5 Mechanisms of Tool Wear ... 14
2.3.6 Tool Wear Types ... 15
2.3.7 Taylor’s Tool Life ... 16
2.3.8 Economic tool life (T
e) ... 17
2.4 Compacted graphite Iron ... 17
2.4.1 Cast Iron... 17
2.4.2 Material properties of CGI ... 18
2.4.3 Foundry Technology of CGI ... 19
2.4.4 CGI Machining Process ... 20
2.4.5 CGI for automotive of today ... 21
2.5 Optimization of cutting parameters ... 22
3 Experimental Procedures ... 23
3.1 Experimental setup ... 23
3.2 Design of Experiment (DoE) ... 24
3.3 Milling tools and inserts ... 26
3.4.1 Tool life criteria and Measurement of Flank Wear ... 27
3.5 OPTIMA Sweden workpiece ... 28
4 Modeling and Experimental Analysis ... 29
4.1 Development of Mathematical model ... 29
4.1.1 Tool life model ... 30
4.1.2 Cutting force model ... 31
4.1.3 Validation of model ... 33
4.2 Experimental Analysis ... 34
4.2.1 Tool life experiment analysis ... 34
4.2.2 Predictability of tool life ... 35
4.2.3 The influence of cutting parameters on tool life ... 36
4.2.4 Cutting force experiment analysis ... 39
5 Optimization of Machining Parameters ... 40
5.1 Optimization Technique ... 40
6 Conclusions ... 42
7 Future work and recommendation ... 43
Recommendations to increase tool life ... 43
Bibliography ... 45
8 Appendix ... 46
8.1 Matlab Scripts for cutting forces ... 46
8.2 CNC programming for 1 run ... 47
8.3 Tool life curves for all experiments ... 48
8.4 Tool wear progress measured for all experiments ... 57
8.5 Tool wear images for three inserts at the end of tool life ... 61
8.6 Prediction vs observed at each combination of cutting parameters ... 63
Abstract
Compacted Graphite Iron offers mechanical properties in between of gray cast iron and ductile iron. Thus, the material is seen as a hopeful alternative for engine cylinder blocks and heads satisfying environmental and performance objectives. Nevertheless, CGI is more difficult to machine than conventional gray cast iron due to the presence of MnS and thin wall section of probing high strength. This problem of CGI then led to the initiation of large Optima Sweden project to study machinability and optimization of related to CGI.
The thesis is centered on a study of tool life, cutting force and MRR with regards to machining parameters mainly feed rate and cutting speed for CGI milling. Tool life is measured; flank-wear is observed and compared between several combinations of cutting parameters.
Similarly cutting forces were measured using LMS software for full factorial design experiments. Extensive machining experiments were carried out. Machining tests was done based on Design of Experiment (DoE) for high cutting data and lower data are performed separately. Comparison is made for tool life and cutting forces. After data collection, analysis of tool life and force has been followed. Once the data is analyzed and checked its consistency. An approximate model is developed using MODDE software. Further, multi objective optimization of tool life and Material removal rate (MRR) using cutting parameters mainly feed rate and cutting speed are investigated. Working on optimal parameters will allow for CGI is to be competitive in manufacturing with gray CI, aluminum alloy, magnesium.
Keywords: Compacted Graphite Iron (CGI), face milling, tool life, MRR, optimization, cutting force
Acknowledgments
First of all, I would like to thank my main supervisor, Anders Berglund, PhD student, for his constant guidance and help throughout the project. Without his consistently support and encouragement, the most frozen winter might have been a major obstacle to complete my work.
I would like to take this opportunity to express my sincere gratitude to my supervisor Prof. Mihai Nicolescu who has spent his valuable time answering my questions.
I would like to extend my gratitude to Jan Stamer, workshop supervisor for his consistent support during experiments.
My final appreciation goes to all my colleagues who have made my stay so comfortable and friendly environments while doing in and outside of my thesis.
1 Introduction
Machining is one of the most important and widely used manufacturing processes in engineering industries. Today’s machining processes are caught between the growing need for quality, high process safety, minimal machining costs, and short machining times.
Tool life is one of the most important economic considerations in the metal cutting industry. Tool life not only effects other machining performance measures, but in today’s automated industry it is very essential to predict tool life to insure timely tool changes for uninterrupted machining and to avoid loss of production due to tool breakage.
Compacted Graphite Iron (CGI) is an interesting material that could be use in diesel engine blocks. One reason why engine manufacturers are interested in the material is because of its mechanical properties (higher hardness, tensile strength, etc). CGI has strength of at least 75% greater than gray iron and a roughly twice higher in fatigue strength; hence, less material is required for automobile engine. In addition, it enables thinner wall sections without losing its mechanical strength. The main disadvantage with this material is its poor machinability. Compared with machining gray iron, tool life for milling operations in CGI are half (1).
This main focus of this project is to examine which tool life can be expected in milling of CGI workpiece.
The workpiece is a special designed and it has similar features with actual cylinder block. The wear behavior of the inserts when milling CGI has been investigated. Once a predictable tool life has been achieved, then a model and tool life analysis is followed by. A predictable and stable process is assumed if spontaneous tool breakage is not happened. The most influencing and controlling parameters for machining of CGI are cutting speed and feed rate. The study of these parameters was carried to investigate their influences on machining of CGI on cutting forces, and tool life.
High productivity is the goal of any company so as to achieve high rate of profit. To get higher production rate, it is then necessary to shorten the machining time or in other words increase MRR (Material Removal Rate). That means that you will need to use higher insert densities and higher feed rates to achieve the desired MRR’s for CGI. The correlation between tool life with MRR is also thoroughly studied during the period of this project.
In order to achieve high production rate and higher tool life, optimal cutting parameters have to be chosen.
The objectives are to maximize both tool life and MRR. Full factorial experiments are performed during machining of CGI. The experiment results are then analyzed using MODDE software. Optimization
1.1 Project background
The work of this project is included to the OPTIMA project. OPTIMA project started in 2006; it was motivated by the Swedish truck industry’s investigations into the possibility of using CGI in the future production of truck engines. Thus, the main goal of the project has been the development about the interaction between the machining process, the material and the casting process of CGI. Three Swedish universities; The Royal Institute of Technology (KTH), Chalmers University of Technology and Jönköping University, one research institute; Swerea SWECAST, two truck manufacturers; Scania and Volvo Powertrain, one cutting tool supplier Sandvik Coromant and two CGI process suppliers; SinterCast and NovaCast, have involved in the project.
1.2 Objective of the project
The main focus of this thesis is to study the tool life of the inserts for machining CGI. Furthermore an attempt to study the effect of the cutting parameters on machining performance such as cutting force and Material removal rate (MRR) has also been pointed out during the thesis project. The aim of the project can be summarized as:
• Study the tool wear mechanism and tool life by varying the cutting parameters
• Study of the effect of combination of cutting data on cutting forces and MRR and Tool life
• To recommended the possible optimal cutting data for CGI machining
1.3 Problem approach
Though various variables can affect machining process for CGI but some of them are assumed to be constant. For instance, depth of cut (3mm) is kept constant as well as the material property. Machining experiment of CGI workpiece has been done until the tool wear reaches the end of its life. The flank wear land (VB) is often used as criterion because of its influence on workpiece surface roughness and accuracy. The wear of the inserts were examined under optical microscope after certain machining cycles.
When the amount of wear reaches the permissible tool wear (VB=0.3mm), the tool is said to be worn out.
The main criteria are either 2 inserts or average of 3 inserts reach the tool wear limit and hence machining will be stopped.
1.4 Limitations of the project
As any other projects, this project has limitations. The first limitation is that machining tests were carried out with one type of insert and hence comparison with other recommended insert is not possible in this project. Hence it will have an effect on the optimal parameters solution of tool life. The second limitation is that the project was solely performed on one milling machine and similar work holding device. The third limitation is that CGI was the only workpiece utilized during experiment. Comparison with gray iron has not been attempted. Since gray iron is still competitive material used for some type car engine blocks.
2 Theory
The aim of the literature study is to discuss about the basic theories about metal cutting operations and to summarize what has been written about tool life and optimization of cutting parameters of CGI. This chapter begins with general theoretical background of machining process. Later, material properties and machining specific to CGI is explained.
2.1 Milling process
In general, there are two possibility of milling process. These are down-milling and up-milling.
Up-milling: Also called conventional milling, the feed direction of the workpiece is opposite to that of the cutter rotation. Each tooth of the cutter starts the cut with zero depth of cut, which gradually increases and reaches the maximum value as the tooth leaves the cut. The cutter has to be forced into the material, creating a rubbing or burnishing effect with excessive friction and high temperatures. The resultant cutting forces are directed upward and tend to lift the workpiece upward from the table, and therefore, more secured clamping of workpiece is required.
Figure 2-1 Up-milling (left) and Down-milling (right)
Down-milling: Also called climb milling, the feed direction of the workpiece is same as that of the cutter rotation. The maximum thickness of the chip at the start of the cut decreases to zero thickness at the end of the cut. The resultant cutting force in the down-milling are directed downward into the machine and tend to drag the workpiece into the cutter. If the workpiece is not securely held, it will be drawn into the cutter so fast that the cutter teeth are unable to make the cut and may break. The advantage of the down milling is that the cutting force tends to hold the work against the machine table, permitting lower clamping forces. This type of milling produces better surface finish and dimensional accuracy.
2.2 Cutting parameters
The cutting parameters that influence machining behavior of any materials are feed rate, cutting speed and depth of cut.
Cutting speed
This is the speed at which each tooth cuts through the material as the tool spins. This is measured in metres per minute. Typical values for cutting speed used for this project ranged from 120m/min to 250m/min for CGI 400.
∗ ∗ Eqn 2-1
Feed rate
Is the feed of the tool against the workpiece in distance per time-unit. It is also called the table-feed and machine feed. Feed rate (f) in mm/min is calculated by
∗ ∗
Eqn 2-2Where is feed rate in mm/tooth, z is number of inserts.
Feed per tooth ( ) is the linear distance moved by the tool while one particular tooth is engaged in cut. It can be also defined as the distance of table feed covered between the engagements of two consecutive cutting edges.
Depth of cut (ap)
This is how deep the tool is under the surface of the material being cut. This will be the height of the chip produced. Typically, the depth of cut will be less than or equal to the diameter of the cutting tool
Width of cut (ae)
This is the radial depth of cut, measured along the diameter of the milling cutter. It is the distance the milling cutter cover across the workpiece in face milling. In face milling, a good value is approximately 75% of the diameter of the mill.
2.3 Machining Performance Measures
The performance of machining process can be evaluated with respect to cutting force, tool life, MRR and surface roughness etc. lowering cutting force and increasing tool life is the required for any manufacturing process. Furthermore, comprise of between higher MRR, higher tool life and better surface quality meets the need of some processes.
2.3.1 Cutting force
The understanding of cutting forces has led to well balanced cutting edges in regards to positive cutting action and cutting edge strength. Cutting forces can be estimated theoretically or be measured using dynamometer sensors.
By studying the cutting force, machining process can be monitored and control. By measuring the cutting force and it has direct effect on tool wear. That means indirectly tool wear is controlled or monitored by measuring the forces. Progress increment of cutting forces indicates the tool wear is taking place rapidly.
2.3.2 Surface roughness
There are two types of surface roughness specified for the surface of interest, the arithmetical mean roughness (Ra) and the mean roughness (Rz). In this case Rz is the limiting requirement.
Ra is specified as:
. | | Eqn 2-3
where r is the test length and y= , as it can be seen figure 2-2. Rz is defined with the mean line between peaks and valleys (see the relation below).
… …
Where is the peak value and is the valley value as shown in the figure below.
2.3.3 Material removal rate
The MRR is a volume of metal removed per time unit. It is the depth of cut times the width of cut times the feed rate of the tool. The unit for MRR is cubic mm per minute. The general equation for material removal rate, represented by Q is:
∗ ∗
Eqn 2-4Where Q is the removal rate in mm /min , is the depth of cut in (mm), is the width of engagement in (mm) and is the feed rate of the table in (mm/min)
Using Feed rate equation from Eqn 2-5 and the spindle rotation (n) is defined as
∗ ∗ Eqn 2-5
Hence
Eqn 2-4
c
an be rewritten as∗ ∗ ∗ ∗
∗ ∗ Eqn 2-6Figure 2-2 Surface roughness description, m is the mean line
2.3.4 Tool life
Tool life is defined as the productive time available during which the edge will machine components to be acceptable within the limiting parameters. The tool life of the cutting edge is decided by the ability of the edge to maintain values within the operational criteria. Reliability and predictability is also important factor for deciding at which point the inserts needs to be changed. It is because most machining of today is behind a closed door or even fully automated.
Optimizing tool life is important in machining since considerable time is lost whenever a tool is replaced and reset. The process of wear and failures of cutting tools increases the surface roughness, and the accuracy of workpiece deteriorates.
2.3.5 Mechanisms of Tool Wear
Tool wear leads to tool failure. According to many authors, the failure of cutting tool occurs as premature tool failure (i.e., tool breakage) and progressive tool wear. Figure 2-4 shows some types of failures and wear on cutting tools. Generally, wear of cutting tools depends on tool material and geometry, workpiece materials, cutting parameters (cutting speed, feed rate and depth of cut), cutting fluids and machine-tool characteristics.
Following are the commonly known wear mechanisms which may occur simultaneously during machining operations
(1) Adhesion: This simple wear mechanism is based on the concept of formation of welded joints between the two metals.
(2) Abrasion: This type of tool-wear is caused by the relative motion between the underside of the chip and the face, and the newly cut workpiece surface and the flank.
(3) Diffusion: Diffusion wear occurs when atoms in a metallic crystal lattice move from a region of high atomic concentration to a low concentration region.
(4) Fatigue: Fatigue wear occurs when surfaces are repeatedly subjected to loading and unloading and may gradually fail by fatigue leading to detachment of portion of the surface.
(5) Erosion: Erosion wear may occur due to contact with a fluid in relative motion. In the case of flood cooling, chemical reactions may take place at the tool tip, which is at high temperature, resulting in formation of oxides. These compounds are loosely bound to the tool and can easily get abraded along with the fluid.
2.3.6 Tool Wear Types
Normally, tool wear is a gradual process. There are two basics zones of wear in cutting tools: flank wear and crater wear. Flank and crater wear are the most important measured forms of tool wear. Flank wear is most commonly used for wear monitoring.
According to the standard ISO 3685:1993 for wear measurements, the major cutting edge is considered to be divided in to four regions, as shown in Figure 2-3:
Region C is the curved part of the cutting edge at the tool corner;
Region B is the remaining straight part of the cutting edge in zone C;
Region A is the quarter of the worn cutting edge length b farthest away from the tool corner;
Region N extends beyond the area of mutual contact between the tool workpiece for approximately 1–2 mm along the major cutting edge. The wear is of notch type
Figure 2-3 Types of tool wear according to standard ISO 3685:1993
The width of the flank wear land, VB, is measured within zone B in the cutting edge plane Ps (Figure 2-3) perpendicular to the major cutting edge. The width of the flank wear land is measured from the position of the original major cutting edge.
Figure 2-4 Types of wear on cutting tools (Adapted from sandvik)
2.3.7 Taylor’s Tool Life
Tool wear is always used as a lifetime criterion because it is easy to determine quantitatively.
Taylor presented the following equation (2):
Eqn 2-6
Where Vc is the cutting speed (m/min), T is the tool life (min) taken to develop a certain flank wear (VB), n is an exponent that depends on the cutting parameters and C is a constant. Note that C is equal to the cutting speed at T = 1 min. Therefore, each combination of tool material and workpiece and each cutting parameter has it is own n and C values, to be determined experimentally.
According to the original Taylor tool life formula, the cutting speed is the only parameter that affects tool life. This is because this formula was obtained using high-carbon and high-speed steels as tool materials.
With the further development of carbides and other tool materials, it was found that the cutting feed and the depth of cut are also significant. As a result, the Taylor’s tool life formula was modified to accommodate these changes as (2):
Eqn 2-7
where d is the depth of cut (mm) and f is the feed (mm/rev). The exponents a and b are to be determined experimentally for each combination of the cutting conditions.
Although cutting speed is the most important cutting parameter in the tool life equation, the cutting feed and the depth of cut can also be the significant factors. The tool life depends on the tool (material and geometry); the cutting parameters (cutting speed, feed, and depth of cut); the brand and conditions of the cutting fluid used; the work material (chemical composition, hardness, strength, toughness, homogeneity, and inclusions); the machining operation (turning, drilling, and milling), the machine tool (for example, stiffness, runout and maintenance) and other machining parameters. As a result, it is nearly impossible to develop a universal tool life criterion.
2.3.8 Economic tool life (Te)
The economic tool life can be calculated by the adjoining formula (3) as follows
Eqn 2-8
∗ Eqn 2-9
Where TCE is economic tool life in terms of machining cost and TCQ is tool life at maximum production rate.
The other variables are α, CT, Cm, tc are slope of the curve tool life against cutting speed, tool cost per cutting edge, overhead cost per minute and tool changing time respectively.
The number of components per hour can be calculated by using adjoining formula as
Eqn 2-10
Where is pieces per hour, T is tool life and tp is total time per piece (including machining and down time)
2.4 Compacted graphite Iron
Compacted Graphite Iron has been increasingly accepted as automotive manufacturing material; showing its efficiency for several components, as brake discs and brake drums, exhaust manifolds, engine heads and diesel engine blocks, traditionally manufactured from gray cast iron. More recently compacted graphite iron has been used for diesel engine blocks. It has proven to be useful in the manufacture of V topology diesel engines where the loading on the block is very high between the cylinder banks, and for heavy goods vehicles, which use diesel engines with high combustion pressures.
2.4.1 Cast Iron
Cast iron (CI) contains 2-4% of Carbon as main alloying element. Mechanical properties of CI are affected by cooling rate. A high cooling rate and a low carbon equivalent favors the formation of white cast iron whereas a low cooling rate or a high carbon equivalent promotes gray cast iron. Varying of sectional thickness in castings affect the cooling rate, this affects the state of carbon. Thick section will solidify into gray iron while thin ones will chill into white iron. Basic structural constituents of different types of cast iron are Ferritic, Pearlitic or mixture. Cast iron with ferritic matrix are easy to machine because of low strength
however Pearlite has a stronger, harder and less ductile structure than ferrite, hence it is difficult to machine (2).
2.4.2 Material properties of CGI
The graphite particles in CGI are clustered as worm-shaped or vermicular. The shape of these particles determines physical and mechanical properties. Among the properties, good thermal conductivity and vibration damping are important for engine block application.
While ductile iron became a manufacturing staple, CGI never was seriously utilized despite possessing some very interesting properties. While not quite as strong as ductile iron, CGI is 75 percent stronger and up to 75 percent stiffer than gray iron (4).
The thermal and damping characteristics of CGI are mid-way between ductile and gray iron. It is five times more fatigue resistant than aluminum at elevated temperatures, and twice as resistant to metal fatigue as gray iron (4).
Gray iron is used extensively in cylinder blocks because it's easy to machine, good castability has good vibration damping, good thermal properties and low production cost. However, due to gray iron's structure it's not as strong as CGI.
The basic difference between CGI and gray iron is the shape of the interconnected graphite matrix. In a CGI matrix, the graphite is more compacted and rounded in what is referred to as "vermicular" or "worm- like" in shape.
The new super metal CGI is reengineering the car and truck industry; with 75% higher tensile strength and 45% greater stiffness than gray cast iron, CGI provides stronger and lighter engine blocks (5).
Table 2-1 CGI properties
Property Gray Iron CGI Ductile Iron
Tensile strength (MPa) 250 450 750
Elastic Modulus (GPa) 105 150 160
Elongation (%) 0 1.5 2.5
Tension Compression 0.20-0.30 0.25-0.35 0.65-0.75
Thermal Conductivity (W/m-K)
48 38 32
Figure 2-5 Gray iron, compacted graphite iron, ductile iron differentiated by their particles shape (6)
Figure 2-6 CGI microstructure containing 10% nodularity and 3D microsctructure (7) 2.4.3 Foundry Technology of CGI
Like Gray iron, casting process is the primarily manufacturing method for CGI to make near net shape production. It is cost effective and most widely used for engine blocks due to the fact that CGI has good castablity properties. The Swedish company SinterCast® licenses a foundry technology to accurately produce CGI (4).
The choice of process used for production of compacted graphite iron influences machinability in an analogous way. Treatment of base iron to produce CGI can basically be made in three ways:
1. Using a Two-Step method: where the iron is first under-treated with magnesium. A sample is taken from the iron and analyzed in various ways. Based on the analysis, the metal is then (after about 3-5 minutes) treated again with magnesium and an inoculants in order to reach the final specification. This method allows a long time for the magnesium to act as an element that reduces MnS that decreases tool life, as MnS particles are known to act as a lubricant for the cutting tools. The double treatment with magnesium that promotes carbides that further reduces machinability
2. Using a One-Step method: where the chemical and metallurgical status of the base iron is measured while still in the melting or holding furnace. Based on this information the ideal addition of magnesium can be calculated and used for the treatment. This method is much faster than the two-step method, which is beneficial for machinability as there is less time for reduction of MnS that is present in the base iron.
3. Graphite Flow method: where the magnesium alloy is added in a reaction chamber in each mould and the base iron is conditioned to have a constant metallurgical quality before pouring. The treatment in the mould offers two advantages regarding machinability compared to other batch treatment methods. First the kinetic effect of the treatment, which takes place 0.5 to 1 second before the iron is filled into the casting cavity, creates a high nucleation effect. The occurrence of micro-carbides is therefore non- existing. Wall thickness down to 2 mm can be cast without any carbide. The extremely short time between treatment and pouring preserves most of the MnS that is present in the base iron and acts as a lubricant.
The disadvantage of foundry technology is that the dimensional accuracy and tolerance of the casted product is poor relative to other manufacturing processes. Therefore, In most cases machining operations are followed by so as to get better surface finish and accuracy.
2.4.4 CGI Machining Process
Machining can be defined as the process of removing material from a workpiece in the form of chips. The term metal cutting is used when the material is metallic. However, the term machinability is used quite generally to describe the capability of a material to be subjected to cutting processes. The machinability of a material is judged in the context of the machining processes involved the tool material, and the cutting parameters. The four evaluative criteria for machining process are tool life, cutting force, surface finish of the part, and chip geometry.
The machinability of materials is influenced by their microstructure and their mechanical properties. The most significant factors affecting microstructure and hence mechanical properties are: carbon content, alloying elements, and heat treatment.
For dry milling at a speed of 150 m/min, a feed of 0.2 mm/tooth and an axial depth of cut of 3 mm, the number of passes for equivalent wear is 21, 40 and 70 for nodular iron, CGI and gray iron, respectively (8).
Machining of CGI probes a difficulty in the car manufacturers. One reason is that of due to the absence of MnS compound in CGI unlike gray iron. Another important reason is that CGI is tougher and higher strength than gray cast iron. Hence, the production rate would be lower for CGI because of the above two reasons, the speed and feed rate will essentially be cut in half, compared with the speeds and feeds normally used with gray iron.
2.4.5 CGI for automotive of today
In 1949 a now well-known material called ductile iron was patented. At the same time, a lesser-known material called Compacted Graphite Iron (CGI) was also patented, though it was just considered a curiosity at the time (4).
An assembled automotive engine can be made 9 percent lighter with CGI. The engine block weight alone can be reduced by 22 percent (4). This corresponds to a 15 percent reduction in length and a 5 percent reduction in height and width. More examples of weight reduction are; 2.5-liter V6 Calibra of the DTM race series, Opel AG cut engine block weight 20.4% and weight-to-power ratio 70.7% (8). A prototype of its 1.6-liter Family 1 block was 29.4% lighter in weight. Audi’s 3.3-liter V8 TDI (turbo direct injection) diesel and BMW’s 3.9-liter V8d are both new designs in CGI (8).
DAF Trucks in Eindhoven in the Netherlands changed a 12.6L, 480hp cast iron engine into a 530hp engine merely by changing the material to CGI. That equated to a 10 percent increase in power and a significant decrease in weight. Now all of its production is being switched to CGI, creating a tremendous market advantage
Figure 2-7 Engine block (SinterCast®-CGI)
Nearly all NASCAR teams are running CGI engine blocks, or blocks with CGI liners. These liners are usually plated with a hard-surface coating. It has been reported that some NASCAR teams are able to run a whole season without having to re-bore the blocks (4).
A recent 500cc Suzuki Grand Prix motorcycle engine had a crankcase fabricated from CGI. Nothing is put on these racing machines that would pose any kind of a weight penalty, and this is an extreme example showing the real potential of CGI applications (4).
Other than Car applications of CGI were the high-speed rail (175 mph+) trains in Europe (4). Initially they had cast iron disc brakes that were simply not up to the task. They suffered severe heat-checking and
cracks, which were potentially catastrophic. CGI cured all that, and has since been found to be the perfect solution for other real world applications
2.5 Optimization of cutting parameters
Due to the enormous complexity of many machining processes and the high number of influencing parameters, conventional approaches to modelling and optimization are no longer sufficient. As the machining processes get increasingly complex, the process models are possibly discontinuous, non- differentiable, or non-explicit with the design variables. As conventional gradient-based nonlinear optimization techniques have difficulty in solving those optimization problems, one must resort to advanced optimization techniques such as evolutionary algorithms, meta-heuristics, neural networks, etc.
According to state of the art of literature reviews, there are various modeling and optimization techniques with applications to the machining processes. The summary of them are ( (9), (10) ):
Response surface methodology
Taguchi’s robust design method
Fuzzy set theory based modelling
Artificial neural networks (ANNs)
Quadratic or linear programming
Genetic algorithm (GA)
ANOVA and regression approach
3 Experimental Procedures
The objective of the experiment is to study the trends on the wear and tool life of carbide inserts, in an intermittent cutting operation of CGI. The determination of tool life is based on the amount of flank wear on the tool insert during machining. The experiment work procedures consist of preparing of Design of Experiment (DoE) and workpiece, writing CNC programming (see appendix 8.2), measuring flank wear on microscope, and machining.
3.1 Experimental setup
The face milling tests were conducted on MAZAK CNC vertical machining centre using Sandvik Carbde milling inserts (K20W) on CGI 400 block material with dimensions of 400 x200 x 100 mm. The cutting speed ranges from 120-250 m/min. The feed rate was selected to be 0.15- 0.3 mm/tooth. While a constant depth of the cut was chosen as 3 mm for entire experiment. It is important to note that the values for these cutting conditions were selected with respect to the tool insert manufacturer’s (Sandvik) and Volvo and Scania recommendations. Therefore, the variable parameters are going to be feed per tooth (fz) and cutting speed (Vc). The workpiece is clamped on a magnetic table of milling machine as shown in the Figure 3-1. The tool path orientation is shown Figure 3-2.
Figure 3-1 Overview of MAZAK vertical milling machine
Figure 3-2 Tool path direction workpiece setup
Tool direction
3.2 Design of Experiment (DoE)
Designed experiments were carried out for a selected combination of cutting speeds and feeds as shown in Table 3-1. The most preferred method of experimentation is full factorial experiments where experiments are carried out for all combinations of variables. In cases where many variables are involved, full factorial experimentation is not feasible because of the time constraints and the cost involved in performing these experiments. The most efficient way of carrying out experiments is by using the Design of Experiments (DoE) method. With the full factorial designs, in each complete trial or replication of the experiment all possible combinations of levels of the factors are investigated. Generation of DoE is created according to the procedures from Umetrics MODDE 9 and State-ease of Design-expert 8. A full factorial of two factors and three levels are utilized for the experiments. A three level provides an insight if it is linear or not. An extensive machining experiment with full factorial (i.e. 32 =9) and each run with two replicas hence the total number experiments are 18. A replica of two for each experiment is done to study for repeatability and fitting to a model. According to DoE, Table 3-2 is filled out with machining order and experiment results of tool life.
Table 3-1 Design of experiments
0,1 0,15 0,2 0,25 0,3 0,35
120 140 160 180 200 220 240 260
Feed rate (mm/tooth)
Cutting speed (m/min)
New DoE
high cutting data DoE
Table 3-2 Experimental results
Std Block Run Order
Cutting feed [mm/tooth]
Cutting speed [m/min]
MRR [cubic mm/min]
Spindle speed [rpm]
Table feed [mm/min]
Tool life [min]
1 Exp 1 1 0,15 120 41,25 305,58 137,51 102,78
11 Exp 1 2 0,25 160 91,67 407,44 305,58 55,92
3 Exp 1 3 0,2 120 55,00 305,58 183,35 107,95
7 Exp 1 4 0,15 160 55,00 407,44 183,35 114,99
13 Exp 1 5 0,15 200 68,75 509,30 229,18 71,70
17 Exp 1 6 0,25 200 114,59 509,30 381,97 48,27
9 Exp 1 7 0,2 160 73,34 407,44 244,46 113,50
15 Exp 1 8 0,2 200 91,67 509,30 305,58 62,47
5 Exp 1 9 0,25 120 68,75 305,58 229,18 128,17
6 Exp 2 10 0,25 120 68,75 305,58 229,18 141,38
18 Exp 2 11 0,25 200 114,59 509,30 381,97 48,43
14 Exp 2 12 0,15 200 68,75 509,30 229,18 57,46
10 Exp 2 13 0,2 160 73,34 407,44 244,46 104,41
8 Exp 2 14 0,15 160 55,00 407,44 183,35 93,79
12 Exp 2 15 0,25 160 91,67 407,44 305,58 66,50
16 Exp 2 16 0,2 200 91,67 509,30 305,58 63,59
2 Exp 2 17 0,15 120 41,25 305,58 137,51 116,91
4 Exp 2 18 0,2 120 55,00 305,58 183,35 102,51
3.3 Milling tools and inserts
Proper selection of tools and inserts is critical for achieving maximum productivity during machining of CGI. The choice of tool material and cutting geometry are crucial. Nevertheless, no matter how right selection of the tool is, if the machining conditions are not up to standard, especially the cutting data and general stability of the machine, hence optimum productivity might not achieved. Vibrations and lack of rigidity in tool holder and clamping can also enhance the tool wear and hence affects on reaching a higher productivity.
According to Sandvik recommendations, the possible inserts that can be used for cast iron are R365- 1505ZNE-KM K20D, K20W and GC1020 (coated with PVD). These are well-proven inserts, extra close pitch face mills can now be operated with new K20D and K20W inserts featuring new KX geometries specially optimized for CGI. However, in our case, Sandvik insert type K20W is employed for the all experiments as it is shown in Figure 3-3.
Figure 3-3 Sandvik insert (R365-1505ZNE-KM K20W) with 8 cutting edges (Sandvik)
The face-milling tool is equipped with diameter of 125mm cutter. This cutter is picked out based on the dimension of the workpiece and number of inserts. The face mill can hold 12 inserts but only reduced numbers of inserts are used to minimize waste of workpiece, time and effort. Three inserts are chosen because a stable process and more predictable tool life can be attained than only 1 insert.
3.4 Measurement methods
Machining tests were carried out to determine the flank wear progression of the cutting inserts. A suitable magnifying glass or microscope is worthwhile equipment. Flank wear and chipping are measured under the Light-optical microscope. The flank wear is measured from the original cutting edge. The measuring setup consists of PC, tool holder and optical stereo microscope as shown in the Figure 3-4.
Figure 3-4 Tool wear measuring microscope
The cutting force measuring setup contains LMS software, PC and dynamometers. A Kistler quantz 3- component dynamometer (Type 9265B) was used to measure the cutting forces.
3.4.1 Tool life criteria and Measurement of Flank Wear
Tool wear is determined by inspection and measuring the wear as it develops in relation to the cutting time which elapses before a certain degree of wear is reached.
The criteria used in this project for measuring tool life are as follows:
The maximum permissible flank wear (VB) is 0.3mm for useful tool life.
The first criterion is when the average of three inserts have reached the criteria.
The second criterion is when the flank wear for two inserts reaches 0.3 mm.
Occurrence of catastrophic failure of the cutting edge.
To calculate the correct tool life end we have to interpolate the line between the two measuring occasions.
This is done by linear interpolation method as far as the two points are available. Hence the tool life is estimated by:
. ∗
Eqn 3-1 Where x1=time for measuring point one, x2=time for measuring point two, y1=tool wear at measuring point one, y2=tool wear at measuring point two.
3.5 OPTIMA Sweden workpiece
A replica of the actual engine block is imitated and produced in small size ratio as shown in the figure 3-5 below. The workpiece sizes are dimension with 400*200*100mm. The two holes refer to the cylinder holes on a real cylinder block as well. Due to this geometry, the machining is more intermittent, it induces vibrations, and creates entry-exit problem of the cutting inserts. Therefore, tool wear can be observed with a better accuracy and conclusion can be obtained for the machining of real cylinder blocks. Table 3-3 indicates the mechanical property of CGI 400 which has been used during experiments.
Table 3-3 Mechanical properties of the CGI workpiece
Material composition Properties
Nodularity [%] 5 – 10
Pearlite [%] 90 – 95
Rm [MPa], Ultimate tensile Strength 438
Rp0,2% [MPa], Yield strength 335
Elongation to Fracture [%] 1,7
Figure 3-5 Dimension of a 2D and 3D drawing of the CGI workpiece
4 Modeling and Experimental Analysis
In order to meet high production rate and stable process, machining parameters have to be chosen in the best possible way. For such optimization, it is necessary to represent the machining process in a model.
Accuracy and possibility of determining global optimum solutions depend on the type of modelling technique used to express the objective functions and constraints in terms of the decision variables.
Accurate and reliable models of a machining process can compensate for the inability to completely understand and adequately describe the process mechanism. Thus, formulation of an optimization model is the most important task in the optimization process. It involves identifying the decision variables to be optimized, expressing the objective functions and constraints as function of decision variables, setting up of feasible ranges for decision variables, and finally expressing the optimization problem as a mathematical model in a standard format, which can be solved by an appropriate optimization algorithm.
Predictive modeling of machining processes is the first and the most important step for process control and optimization. A predictive model is an accurate relationship between the independent input variables and dependent output performance measures. There are two well known approaches to obtain this relationship: the experimental approach and analytical approach. In this project, we focus only on experimental approach modeling. This method is considered as short-term and practical method, and it is the most suited approach for industrial applications. Large amount of experimentation is required to establish the empirical relationships between various influencing independent operation variables and the technological performance measures. In most cases, tools such as designed experiments and curve fitting are used to obtain the empirical relationship from experimental data. Therefore empirical modeling approach has been used to develop predictive models for tool life and cutting forces in terms of machining parameters such as cutting speed and feed.
4.1 Development of Mathematical model
Regression model are mostly adopted for fitting cutting data for milling operation. The general model in terms of the cutting parameters can be expressed as second order polynomial function as
∗ Eqn 4‐1
Where Y refers to the true response of experimental results for either tool life or cutting force. The variables are x0 = 1 (dummy variable), x1, x2 represents of speed, feed, and depth of cut respectively, while β0, β1, β2, β3 β4 and β5 are the coefficient of parameters to be estimated. The experimental result model may not exactly the same as Eqn 4-1 because the cross and square relations may be eliminated if they are insignificant. These are explained in the sequel sections.
4.1.1 Tool life model
The output of the experimental result is fitted with second-degree polynomial equation. The summary of the fitting for tool life is shown in the figure 4-1 below. R2 is a measure of fit, i.e. how well the model fits the data. Since R2 is large in number (0.764), hence the model linearity is good enough. Besides that Q2 tells you how well the model predicts new data. A useful model should have a large Q2. This is also proven true. The Reproducibility bar shows greater than 0.85, the pure error is low. This means that repeatability of tool life experiment is excellent.
Figure 4-1 summary of fitting to experiment output data
The Umetrics MODDE software generates the coefficients of the model based on Analysis of Variance (ANOVA) and Regression Analysis (RA). The effects of the cutting parameters on the experimental model are shown in Figure 4-2. The bar for cutting speed is inside the box, meaning that the most important factor having an effect on the tool life is cutting speed. The mixed relationship has slightly influences the tool life but the two square relations of speed and feed rate indicates less significant. Hence, these coefficients are eliminated and the reduced model will be
. . . . ∗ Eqn 4-2
Where TL is tool life response, fz is feed rate in mm/tooth and Vc is cutting speed in m/min.
-0.2 0.0 0.2 0.4 0.6 0.8 1.0
Tool life
Investigation: modde analysis (MLR) Summary of Fit
N=18 Cond. no.=4.243 DF=12
R2 Q2 Model Validity Reproducibility
MODDE 9 - 2011-03-29 18:45:49 (UTC+1)
Figure 4-2 Parameter importance for tool life
4.1.2 Cutting force model
In CGI machining, the cutting force is expected to be higher than gray iron due to the fact that the round edges of the compacted graphite does not initiate cracks as easy as the sharp edges of the flake graphite in gray iron (11). Besides that the tougher strength of the CGI material should also result in larger forces.
A three component piezoelectric quartz crystal Kistler dynamometer was used to measure the forces. The three components of forces, the cutting force in X direction; force Fx , cutting force in Y direction; force Fy and the axial component in z direction; force Fz were measured. The resultant force (Fr) will be:
Fr= Eqn 4-3
-40 -30 -20 -10 0 10
fz vc fz*fz vc*vc fz*vc
min
Scaled & Centered Coefficients for Tool life
N=18 R2=0.774 RSD=16.72 DF=12 Q2=0.506 Conf. lev.=0.95
Cutting force model and analysis is performed based on the DoE for higher cutting data region shown in table 3-1. Literary, it is possible to reach similar conclusion for new DoE of lower cutting data. The influence square factor as shown in Figure 4-3 has negligible effect hence it is eliminated from the model.
The average resultant force model is estimated in terms of the cutting parameters. The force model which has fitted well with the data will be:
605.12 17.02 ∗ fz 58.77 ∗ Vc 9.23 ∗ fz ∗ Vc Eqn 4-4
Figure 4-3 parameters Effect on cutting force
Figure 4-4 influence of feed and cutting speed on cutting force in 3D dimension From the above Figure 4-4, we can see that as the cutting speed increases the average cutting force
0.2 0.25
0.3
150
200
250 500
550 600 650 700 750
Average of resultant force
A: Cutting speed B: Feed rate
Table 4-1 Comparsion of predicted vs actual average cutting force values
Run Order
Actual Value
Predicted Value
Run Order
Actual Value
Predicted Value 1 559.57 554.15 10 602.69 605.52 2 548.73 554.15 11 597.08 590.51 3 536.65 546.95 12 583.94 590.51 4 557.26 546.95 13 706.75 685.05 5 536.80 537.95 14 663.35 685.05 6 539.11 537.95 15 658.12 653.93 7 631.42 627.23 16 649.73 653.93 8 623.05 627.23 17 651.29 644.78 9 608.36 605.52 18 638.26 644.78
From Table 4-1, we can see that the difference between prediction and actual average cutting force is very small. Therefore, it tells us that the experiment result gives very good repeatability.
4.1.3 Validation of model
The accuracy of the models for each combination of cutting condition was tested by first comparing the prediction with the actual measured values of the response variables for the experimental conditions, which are tool life and cutting forces. Tests were repeated twice for all to check the validation of the experiments. Small error of repeatability is registered as indicated on Figure 4-5. The relative error ranges from smallest relative error 5% to highest relative error 20%.
Figure 4-5 Repeatablity of the experiments 102,8
55,9
108,0 115,0
71,7 48,3
113,7
62,5
128,2 116,9
66,5
102,5 93,8
57,5 48,4
102,4 63,5
141,4
0,0 20,0 40,0 60,0 80,0 100,0 120,0 140,0 160,0
Tool life [min]
Set of Cutting parameters
Repeatablity of tool life experiment
exp1 exp2 error
4.2 Experimental Analysis
4.2.1 Tool life experiment analysisAll cutting tools wear during machining process. The tool wear continues until it comes to the end of its tool life. The life of a cutting edge is counted in minutes. The tool life is one of the most important values for determining the productivity level of machining process. Measuring the amount of tool wear generated, and then making analysis is best approach for optimizing tool life.
The level of wear for cutting edge of an insert used for finishing or roughing can vary. For finishing operation, the tool is considered to be worn out when it can no longer generate a certain surface texture.
Just a little wear along a very small part of the insert nose means the edge of the insert needs changing.
In rough operation, wear develops along a much longer part of the edge. Significantly, more wear can be tolerated as far as fine surface finish or high accuracy is not needed.
Run order
Cutting time [min]
Cutting parameters
Tool wear [mm]
MRR Tool life
[min]
Vc fz Insert 1
Insert 2
Insert 3
Average
2 60.54 160 0.25 0.27 0.67 0.38 0.44 91.67325 56
Figure 4-6 Experimental Tool life curve of ( feed 0.25mm/tooth,speed 160m/min) 0
0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8
0 20 40 60 80 100 120
Maximum flank wear [mm]
Machining time [min]
Test 2
Insert 1 Insert 2 Insert 3 Average
4.2.2 Predictability of tool life
The graph between predicted and observed data is displayed in the Figure 4-7 and Figure 4-8 for a sample of feed 0.2mm/tooth and speed 160m/min. But for all combinations of parameters are available in appendix 8.6. The prediction plot shows that the points are floating around the linear line with R2 of 0.764.
Figure 4-7 Prediction plot with 95% confidence interval
Figure 4-8 Observed vs predicted data for tool life
Table 4-2: lower and upper boundaries of tool life for each combination of parameters values
Feed rate Cutting speed Tool life Lower Upper
0.15 120 113.284 93.299 133.27
0.25 160 83.0772 70.4374 95.7171
0.2 120 117.82 105.18 130.46
0.15 160 94.5739 81.9341 107.214
0.15 200 75.8635 55.8782 95.8488
0.25 200 43.7993 23.814 63.7847
0.2 160 88.8256 80.8315 96.8197 0.2 200 59.8314 47.1916 72.4712
0.25 120 122.355 102.37 142.34
50 100
0,15 0,16 0,17 0,18 0,19 0,20 0,21 0,22 0,23 0,24 0,25
Tool life
50 100
120 130 140 150 160 170 180 190 200
60 70 80 90
0,15 0,16 0,17 0,18 0,19 0,20 0,21 0,22 0,23 0,24 0,25
MRR
Feed rate
60 70 80 90
120 130 140 150 160 170 180 190 200
Cutting speed
Prediction Plot
Feed rate = 0,2 Cutting speed = 160
40 50 60 70 80 90 100 110 120 130 140
40 50 60 70 80 90 100 110 120 130 140
Observed
Predicted Tool life with Experiment Number labels
N=18 R2=0.764 RSD=15.81 DF=14 Q2=0.632
1
2 3 5 4
6
7 8
9 10
11 12
13 14
15 16 17
18 y=1*x+8.046e-006 R2=0.764
4.2.3 The influence of cutting parameters on tool life
A full factorial experiment has been performed during machining experiment in order to study the influence of cutting data variation on tool life and MRR. From the experiment results, one can easily notice tool life is highly dependent on machining parameters. The higher the cutting data is used the shorten tool life results but the MRR is higher vice versa. Therefore, a compromise between MRR and tool life is needed to get optimal parameters in terms of productivity of machining CGI.
The outcome of the experiments indicates that the most influencing factor for tool life is the cutting speed.
This conforms to the previous work which says that cutting speed is the main factor which influences the tool life for carbide grade cutting tools (12). Reducing cutting speed while increasing the feed rate provides a better machining solution for CGI. Because higher feed rate compensates the negative effect of lower cutting speed on MRR.
Machining
time 30.23min Machining
time 48.37min Machining
time 60.46min Machiningtime 66.50min
Figure 4-9 Progress of tool wear for setting of feed (0.25mm/tooth) and speed (160m/min)
Figure 4-9 indicates that the tool wear progress for a typical cutting condition of feed (0.25mm/tooth) and speed (160m/min). All the experiments tool wear results can be found on Appendix 8.4. We can easily notice that the tool wears uniformly according to the wear pattern in each run. However, at certain part of the image reaches the flank wear criteria.
Less tool wear is observed for high speed and high feed when the same amount of material removal is considered for comparison as shown in the Figure 4-10. Figure 4-11 depicts the tool wear images when
Cutting speed in m/min
run16 fz0.15 Vc 120 run18 fz0.2 Vc 120 run16 fz0.25 Vc 120
run18 fz0.15 Vc 160 run18 fz0.2 Vc 160 run18 fz0.25 Vc 160
run 19 fz0.15 Vc 200 Run16 fz0.2 Vc 200 Run 20 fz0.25 Vc 200
Feed rate in mm/tooth
Figure 4-10 Tool wear measured at approximately equal amount volume material removed Cutting speed 120m/min Cutting speed 160m/min Cutting speed 200m/min
Feed 0.2
fz0.2vc120 test 1 insert 1
fz0.2vc160 test 2 insert 1 fz0.2vc200 test 2 insert 3
Feed 0.15
fz.15vc120 test 1 insert 1
fz.15vc160 test 2 insert 3 fz.15vc200 test 2 insert 1
Feed 0.25
fz.25vc120 test 2 insert 3 fz.25vc160 test 2 insert 1 fz.25vc200 test 2 insert 3
Figure 4-11 Measured flank wear tool wear at the end of tool life
fz0.2vc120 test 1 insert 1
fz0.2vc160 test 2 insert 1 fz0.2vc200 test 2 insert 3
fz.15vc120 test 1 insert 1
fz.15vc160 test 2 insert 3 fz.15vc200 test 2 insert 1
fz.25vc120 test 2 insert 3 fz.25vc160 test 2 insert 1 fz.25vc200 test 2 insert 3
Figure 4-12 Measured rake face tool wear at the end of tool life
For further investigation of tool wear behavior, rake face wear measurement is taken at the end of the tool life as shown in figure 4-12. Rake face wear of 0.14mm is the maximum tool wear registered and it indicates that also the rake tool wear is uniformly distributed. By Comparing of the above two figures, it shows that flank wear is more dominant than rake face wear during CGI machining.
Based on the ANOVA and RA graph shown in the Figure 4-13, the followings are observed from the experiment results
– Variables which affect Tool life, and are significant for mathematical model are: cutting speed Vc, feed rate fz , the interactions of cutting speed Vc and fz. The squares are removed due to their insignificance.
– Cutting speed Vc has the most significant influence on tool life, hence feed rate can be increased according to the surface roughness requirements
– R2 is 0.761 and Q2 is 0.629 with confidence interval of 95%,
Figure 4-13 ANOVA for tool life
Higher tool life is registered towards the right bottom of the contour plot shown in Figure 4-14. The bar on the right shows different colors related to the amount of tool life measured. The red color shows higher tool life where as blue indicates lower tool life region.
0 10 20 30 40 50 60
SD-regression RSD RSD * sqrt(F(crit))
ANOVA for Tool life
N=18 R2=0,761 RSD=15,93 DF=14 Q2=0,629 Conf. lev.=0,95
Figure 4-14 Contour plot of tool life experiment result
4.2.4 Cutting force experiment analysis
If the influences of the cutting parameters on the cutting force are to be determined, the cutting force data should be measured accurately or further analysis. Force measurement is performed during machining using 4 rotating Kistler dynamometer and the force data is collected in the computer mounted with LMS Test.Lab software. The force cutting force versus time shown in Figure 4-14 is plotted using Matlab script (see appendix 8.1).
Figure 4-14 Resultant cutting force graph
0 10 20 30 40 50 60
0 500 1000 1500 2000 2500 3000
3500 Resultant force
data 1
5 Optimization of Machining Parameters
After developing models to predict the machining performance measures the next logical step is optimization of these machining performance measures with respect to cutting conditions. Umetrics MODDE software was used as optimization tool based on the objectives (MRR, tool life) and cutting condition constraints.
5.1 Optimization Technique
Based on the model
developed in the previous chapter for tool life of Eqn 4-2 which is obtained from the experiment results and the mathematical relation for MRR, the optimization problem can be defined as follows.
Objective function
Maximize . . . . ∗ Eqn 5‐1
And
Maximize . ∗ Eqn 5‐2
Subjected under constraints
0.15 fz 0.25
120 Vc 200
All Vc and fz is greater than zero.
Figures 5-1 shows the dependency of tool life on the two factors. The optimal region is close to high feed rate and low cutting speed.
For equal weight factor of (1,1) for both Tool life and MRR, the optimal values are:
Feed rate Cutting speed Tool life MRR 0.2499 148.684 96.192 85.1682 0.2499 148.091 94.776 84.8244
If MRR value minimum is 100 then optimal solution is Feed rate Cutting speed Tool life MRR
0.25 169.314 73.9377 96.9932
By varying the weight factors for MRR and Tool life, the range of objective function optimal cutting parameters for maximum value of tool life and MRR are:
Vc=148-170 m/min, fz=0.25 mm/tooth.
For best selection of which optimal parameters to use depends on the assigned weight factor to the objective function either to MRR or tool Life (TL). If one is interested in higher tool life and slightly lower in MRR, and then more weight factor should be given to TL so that the optimal parameters values will change accordingly.
6 Conclusions
The “OPTIMA SWEDEN” workpiece design, that has similar geometry with the real engine block, is expected to give similar tool wear behavior and tool life compared to actual engine block. Hence, the results, which are obtained from the experiments using this workpiece, can directly be used in machining of real cylinder blocks.
An extensive machining tests based on full factorial design of experiments was performed inorder to investigate the effect of cutting parameters on tool wear, tool life and cutting forces. From the experiments, one can easily deduce that cutting data are the most influential on CGI tool life. For higher cutting data (speed 200-300 m/min, feed 0,2-0.3mm/tooth), sudden tool break was repeatedly observed but later the DoE region was modified as it is shown in Table 3-1 Design of experiments. In this case, a stable and predictable tool life is found out from the experiment results.
Among the two factors, cutting speed is the most influencing parameters for tool life compare with feed rate. A close look at cutting speed is highly needed in order to avoid sudden tool break. Unpredictable tool life should be minimized because it costs a huge amount of money changing setup frequently. In order to get higher MRR, one has to increase feed rate though it has negative effect on surface roughness.
The optimal cutting data suited for CGI are obtained but further steps are needed before successfully introducing the cutting parameters into automotive industry. It is therefore evident that additional investigations regarding machining of the CGI material family is required before a final optimization of machining can be done. A study of cutting force has been attempted but still more research work is needed to come up with good result and conclusion.
7 Future work and recommendation
Recommendations to increase tool life
1. Constant engagement: Keep cutter constantly engaged with workpiece during machining because in and out of cutter reduces tool life.
Figure 7-1 Constant engagement of workpiece(adapted from Sandvik)
2. Roll on technique: the tool enters around external corners and rolls around the corners. This way the tool life can be enhanced. It is also possible to use 45 degree entering angle for optimizing tool life and cutting process. This is because of the advantageous balance of cutting forces and the chip-thinning effect of the long cutting edge that the inclination gives. This means good machining stability as well as the capability for high feed rates resulting in shorter machining times. There are two solutions to remedy this common problem allowing optimised feed rate for when the cutter is fully engaged
1) Program straight into cut but with the feed reduced to 50% until the cutter is fully engaged.
2) Roll into cut in a clockwise motion (anti clockwise will not solve the thick chip thickness problem).
It can be seen that by rolling into cut the chip thickness on exit is always zero allowing higher feed and longer tool life.
Figure 7-2 Roll on technique at the entry of the workpiece
2. Program CNC codes around interruptions and holes by changing feed and speed.
3. Cutting tool positioning: Three passes on the workpiece gives better result for higher cutting data than two passes. An experiment test was carried out for cutting parameters of speed 250m/min and feed rate 0.3mm/tooth. The finding was that the tool breaks suddenly with in the first 5 runs for two pass setup (see figure). However, the tool wear indicates gradual for three passes. The possible reason is due to the holes and thin section of the workpiece geometry.
Where entry and exit took place at the same time that induces high vibration during machining.
Figure 7-3 Two passes machining on CGI
Figure 7-4 Three passes machining on CGI
