As a direct consequence of technological developments, the competitiveness and optimization of machining operations have been studied over the past years in order to understand the representativeness of each variable (labor, cost of acquisition and maintenance of machines, raw material, tool costs and others) involved in this important transformation process. In general, many studies concluded that the direct costs with the tools are responsible for something around 5% of the total cost involved in this operation . However, the inappropriate usage of tools can significantly negatively affect the other variables, causing a high consumption of machine hours and labor, generating high maintenance costs, and causing irreparable damage to the part, that could even generate scraps. As previously mentioned, during the 80's and 90's, several studies were performed about establishing automatic criteria for the end of tool life. However, anyone could indicate an ideal method for tool change in real production, without the participation of the machine operator. To implement this automatic methodology, it is necessary to apply scientific and technological resources based on the operational excellence approach and also Industry I4.0 (sensors, software’s, auxiliary electronic equipment, communication interfaces for data acquisition / processing and others). This procedure provides solid technical basis for an automatic and accurate intervention in respect to the end of tool life criteria, confirming benefits for HSE (Health, Safety and Environment = sustainability aspects), quality, deliveries and costs. Furthermore, it is also very important to mention that this work, in spite of having an automatic procedure to define tool life and replacement, did not use a complex solution, such as neural network or any kind of mathematical model.
The aim of this work is to develop a monitoring system able to automatically define the end of tool life and to perform the automatic tool change in a rough milling operation, completely independent of the cutting parameters inside a production environment and without the participation of the machine operator. In order to develop and implement this very important technological step the experiments were carried out in a five-axis milling operation of blades for hydraulic turbines.
In general, blades for hydraulic turbines are produced in cast stainless steel material and during the milling process of such components (i.e. large size: superficial area around 25m²) more than three (3) tons of chips can be generated in a few tens of machining hours .
As contextual information, the energy generated by hydraulic turbines represent around 60% of the Brazilian energy matrix and talking about the whole world, the hydraulic turbines are responsible for something around 30% of the total world energy matrix .
The amount of publications referring to topics connected with the automatic tool change and definition for end of tool life under heavy roughing five-axis milling process in production environment is very scarce, which hinders the bibliographical survey and the technological development of five-axis milling in heavy roughing operations with focus in high removal volume (tons of chips). This fact once more demonstrate the relevance and important contribution of the present work which also confirms the originality proposal as a strong development for the scientific and industrial communities that can replicate this approach for further developments in other machining operations, getting as main benefit to minimize the underutilization of tools, which generates unnecessary machine downtime and consequently, several other losses.
Figure 1 shows the main themes investigated during the literature review in order to obtain updated knowledge based on publications in manufacturing technology journals, university research and benchmarks with industry developments.
According to the respective literature review, the key boundary aspects have been observed. Those are presented in the subchapters below.
1.1 The cutting tool
The most important diameter to be considered when selecting cutters with round inserts (toroidal) for mold milling process, aerospace industry and also, in milling process of hydraulic turbine blades (organic shapes) is the outer diameter of the cutters, as it is used as a fundamental parameter for generation of milling path and strategies inside the CAD-CAM-CAV software .
The axial depth of cut (ap) is another very important parameter in the milling process and can be defined as the difference between the machined and un-machined surface in the axial direction. The maximum cutting depth is mainly limited by the insert size and machine power. The milled width also known as radial depth of cut (ae) is an extremely relevant parameter for the productivity aspect. By increasing the number of tool teeth (z), the feed rate (vf) can be increased, while the cutting speed (vc) and feed per tooth (fz) remain unchanged [5 - 7].
Other critical factors for the cutting tools in roughing operations are the temperature, power consumption and vibration. In this case, the selection of the inserts have a fundamental contribution in the machining performance and characteristics of high hardness, good wear resistance at high temperatures, combined with toughness are extremely important .
As the experiments have focus in roughing operation, the tool toughness is an indispensable property and thus, the insert grade ISO-M35 are commonly used, especially in the rough milling of hydraulic turbine blades. This insert has an aluminum oxide coating with optimized microstructure, applied using the CVD process, combining layers of Titanium Carbonitride and Titanium Nitride (TiCN+Al2O3+TiN), which provide for the tool conditions to work with high cutting speeds . Moreover, due to the fact that the machined material is ductile, the tool geometry must be positive.
1.2 Stainless steels machinability
The machinability of martensitic stainless steels are influenced by the concentration of carbon content, nickel content and metallurgical structure . As in the most materials, increasing hardness typically reduces tool life and machinability. Increasing the carbon content increases the proportion of abrasive chromium carbides in the matrix and, consequently, reduces tool life . The metallurgical factor that has the highest influence on machinability is the proportion of free ferrite in the matrix. Machinability generally increases with free ferrite content . Tool life is related to annealed hardness, which increases with increasing nitrogen content. Increasing the carbon content also increases the work hardening rate that decreases machinability . Abrasive carbon / nitrogen compounds can be formed in the material matrix and, consequently, reduce tool life; these can be controlled by adding titanium or niobium. Hardness increases and machinability decreases with increasing nickel content. In general, stainless steels can contain alloying elements and additives such as sulfur, selenium, tellurium, lead, bismuth and phosphorus intended to improve machinability .
1.3 Tool wear and damage
During the milling process, the tool cutting edge is subjected to numerous efforts and friction, which generates wear. This continuous wear on the tool faces is one of the factors that determine the end of tool life . In conventional machining processes, there are several factors, called malfunctions, which determine the necessity of cutting tool replacement. They can be defined as:
a) Breakdown – happens suddenly. Types of damage are breakage, chipping, plastic deformation or built up edge (BUE). Damages are more common in interrupted cutting processes like milling, in which tools are subjected to extreme demands, mainly with regard to thermal and mechanical shocks .
b) Wear – continuous and microscopic removal of particles from the tool under the action of the chip and/or the workpiece. If the wear is too high, it may compromise the quality of the machining process and lead to edge breakage. It can occur in both machining conditions, interrupted and continuous cutting .
A cutting tool, even if it has enough toughness to withstand the cyclical load and temperature variations that could cause it to malfunction, will not be free from a progressive loss of material, which leads to the necessity of tool replacement .
1.4 Considerations for the end of tool life
The precise definition of the end of tool life is very difficult to be made inside the production environment, since it is usually done by the machine operator. The result, in the majority of the cases, is a big waste of tools, due to the difficulty that the machine operator encounters in establishing the end of tool life. Therefore, he (or she) does it very conservatively, in order to not damage the part or to not cause any other damage to the production. The solution for this problem lies in monitoring the tool wear using different indirect methods like vibration sensors, temperatures, acoustic emission, electrical parameters and machining forces, which will provide the necessary subsidies and, based on a scientific approach, indicate the ideal moment for tool change .
1.5 Methods for monitoring the end of tool life
The end of tool life can be evaluated through direct and indirect methods. In direct methods, geometry (the element to be controlled is measured) and tool wear are measured using optical system. The indirect method uses the acquisition of measured values of process variables (such as cutting forces, temperatures, electrical parameters, vibration and others) which can be correlated with tool wear .
The dimensional laser tool monitoring system is a direct monitoring method that allows the check of wear and/or small damage to the tools during the machining process, which are vital for the safe operation of the system machine-workpiece-tool . The wear measurement is carried out through brief tool travel movements passing through the laser. This is an intrusive monitoring process that generate stops in the machining operation and, consequently, adds unproductive times to the process .
In favor of indirect monitoring systems is the fact that it is not necessary to interrupt the process for the control to be carried out. This simultaneity also allows the monitoring of the tool's status in real time, enabling its breakage to be avoided . The most common indirect monitoring systems observed in the literature review are:
Monitoring of electrical parameters: Measuring the electrical parameters of the machine's motor is measuring the cutting efforts. This is because the machine motor, when generating the mechanical power necessary to carry out the machining operation, consumes electrical current in an amount directly proportional to the power and cutting force generated . As the tool wears, the cutting force increases and this increase can be correlated with the tool wear value. The same effect holds for the feed rate drive motor (X-axis), that it is possible to observe the effect of the increased friction force due to the gradual increase in tool wear on the power consumption of t7he machine motor [15, 17].
Temperature monitoring: Every machining operation generates heat during its execution. This fact is mainly due to two factors, which are the friction between tool-workpiece-chip and the internal shear to which the workpiece material is subjected during the formation of chips. The machining process generates a significant amount of heat. The resulting temperatures around the cutting edges have a direct influence on the rate and manner in which tool wear occurs, as well as the friction between the tool and the chip, and also, the tool and the newly formed surface. Radiation measurement methods consist of non-contact thermographic techniques to measure temperature from the thermal energy emitted by the cutting zone. The methods that use these techniques are the most used for measuring the cutting temperature. The main advantages of this technique are the quick response to temperature variation, non-physical contact to carry out the measurement, in addition to allowing the measurement of difficult-to-access objects, for example, in cases where there is some movement that prevents fixing the sensor [14, 16].
Vibration monitoring: Vibration is defined as the movement of a dynamic system around its static position. For a mechanical system to vibrate it is necessary, and sufficient, that it has the capacity to store kinetic energy (which has inertia), elastic potential energy (which is flexible) and a time-dependent external excitation. Consequently, in practice it is very difficult to avoid vibration in machining processes. It usually occurs because of the dynamic effects of manufacturing tolerances, clearances, contacts, friction between machine parts, part material heterogeneities, and due to the effect of unbalanced dynamic forces of rotating and alternative machine components. Again, as the tool wears, the cutting forces increase and, consequently, either the tool or workpiece may increase. In addition to the aforementioned factors, the cutting tool wear increases, vibration amplitude increases in all frequency bands, demonstrating a good correlation between wear and vibration which vary according to the case under study. This fact also justifies the use of the vibration signal in this research [14, 17].
1.6 Case of study - Heavy roughing five-axis milling process in production environment
As already mentioned, to develop and implement the monitoring and automatic change of the worn tool in the shop floor inside the industry, a bottleneck five-axis milling operation in blades for hydraulic turbines was used. This process represent well the end of tool life under heavy roughing five-axis milling process in production environment, because during the milling process of blades for hydraulic turbines (superficial area around 25m²) more than three (3) tons of chips can be generated in a few tens of machining hours .
The publications referring to the subject of milling of hydraulic turbine blades are very scarce, which hinders the bibliographical survey and the technological development of five-axis milling of these important components. This condition demonstrates the important contribution of the present work and confirms its originality proposal.
The milling process in hydraulic turbine blades has developed mainly during the last two decades. There are several variants in this process, which aim to give the blade the required geometric shape through the milling operation in machining centers with computer numerical control (CNC) equipped with three, four or five-axis, using in most cases, a cast blade as raw material. Other fabrication concepts using formed sheets are also under development and analysis. However, the traditional and most used manufacturing concept around the world has been the use of cast solution as a raw material for the production of hydraulic turbines blades . Five-axis milling techniques have been widely used in the power generation segments and in the aerospace industry. In part, because of the complexity of the geometries found in turbines and critical aircraft components. Pieces produced in five-axis milling reduce the number of manufacturing steps, avoiding additional assembly’s steps in order to obtain the final product. In the aerospace industry, components that require five-axis milling often have less weight, which is a very important to obtain a good performance. In the heavy mechanics industry where hydraulic turbines for hydroelectric power plants are manufactured, turbine blades are generally large, providing a high amount of chip removal. Thus, five-axis milling reduces manufacturing time and, consequently, the final cost of the product .
Further studies concludes that despite the technological advantages of the five-axis milling, the use of this technique is still limited, due to the complexity and difficulties in generating collision-free CNC programs. The most of CAD-CAM-CAV systems have limited functionality for generating milling paths in five-axis. The CNC programmer is often forced to intervene in the tool orientation and manually adjust the respective milling path. In most cases, the CNC programmer uses as base the local region of the surface to be milled with the heights concavity. More advanced CAD-CAM-CAV systems have more customized algorithms for interpolation that optimized the tool movement and orientation .
Another very important aspect studied by researchers that must be observed when carrying out five-axis milling is the clearance angle between the cutter and the surface that is commonly called by CAD-CAM-CAV software’s as “lead angle” (details in the section 2). The “lead angle” refers to the angle of inclination from the tool axis in relation to the direction of movement and it is intended to ensure that the back of the cutter does not collide with the back side of the workpiece, which can have a high elevation / curvature (surfaces with different geometric and/or organic shapes) than the front side .