CAM System for Layer-Based EDM

CAM System for Layer-Based EDM S. Dimov1, D. T. Pham1, A. Ivanov1, K. Popov2 and V. Rangel3 1 Manufacturing Engineering Centre, Cardiff University, Cardiff, UK 2 Rousse University, Bulgaria 3 Guadalajara University, Mexico

ABSTRACT This paper proposes a new tool-path generating approach for layer-based EDM milling. With this approach, electrode wear is compensated for by varying the layer thickness while the material removal volumes are represented by a sequence of 2-D slices. Tool-path generation is performed taking into account the specific requirements of 3-D micro-EDM milling. The capabilities of the developed CAM system have been verified experimentally. 1

INTRODUCTION

The importance of chipless material removal processes such as EDM and ECM has been growing over the last two decades. This is particularly true in application areas where the machining of advanced materials and production of complex parts incorporating 3D surfaces is required. In addition, these chipless processes have attracted the attention of researchers because of their non-contact nature and their ability to machine parts with negligibly small forces. This is of particular importance in micro-machining applications [1-3], where EDM milling is considered a promising manufacturing route for producing complex micro-tooling cavities. A number of researchers have studied the capabilities of the EDM milling process and have reported some successful applications of this technology [1-4]. However, in spite of the availability of CNC controllers capable of performing complex machining strategies, this process is still not widely used. The EDM process is highly automated and could be controlled very accurately. The information required to perform EDM operations is similar to that required for CNC milling but in-process monitoring is much easier compared to the case of conventional material removal. Computer aided simulation of EDM milling operations is possible and the required NC programs could be generated automatically [1]. The main reason for the limited application of EDM milling in machining complex 3D cavities is the difficulty in generating tool-paths using existing CAD/CAM systems [5], although some attempts have been made to adapt existing systems for this purpose [6, 7]. In this paper, a tool-path generation method specially developed for EDM milling is presented. The method utilises the Uniform Wear Method (UWM) [2] to carry out the segmentation process in the step-depth direction and to generate tool-paths within specified tolerances. The efficiency of the EDM milling process is improved further by minimising the

number of layers required to remove a particular volume of material and by reducing the total machining time. The paper first introduces the proposed method for tool-path generation. Then, the implementation of the CAM system for layer-based EDM in the Pro/Engineer environment is described. Finally, the results of three experiments verifying the performance of the developed CAM system are presented and conclusions made about its capabilities. 2

TOOL-PATH GENERATION

The generation of complex tool-paths for machining free-form surfaces is only possible by using CAM systems. This generation process requires segmentation in step-over and stepdepth directions to be carried out. In the case of EDM milling, the step-depth segments are the slices used for removal of material in layers. The path segmentation procedure involves, first generating a set of cutter-contact (CC) points along the machining path, and then, based on this data, through tool offsetting, computing the corresponding cutter-location (CL) points. The continuous lines connecting these CL points are the motion commands for machining the surfaces. The CC data and the corresponding CL data are defined by the 2D cross-sections generated by “slicing” a selected volume. As in traditional slice-based machining approaches, the first step towards the generation of tool-paths for EDM milling is to specify the layer thickness, depending on the process and/or the surface finish requirements. Then, a sequence of slices with the specified thickness is generated in the step-depth direction. The CL data created in this way defines the movement trajectories of an electrode with a particular shape and size. By following these trajectories, the electrode fully covers the surface area of each slice. This is the geometrical data required to create a NC programme. In addition to this data, the NC programme should contain technological data that is process specific. This data could be stored in an internal database of the CAM system or loaded from an external file. To complete the process, the CL data file and the technological data are post-processed to create a part programme for a particular CNC EDM machine. To summarise, the generation of a tool-path for EDM milling comprises the following steps: 1. Slice generation; 2. Generation of a CL data file; 3. Selection of technological parameters; 4. Post-processing of the geometrical and technological data. 2.1

Slice generation

The process of generating parallel cross-sections of a 3D object is called slicing. The resulting 2D slices with a particular thickness are considered an approximation of the object. Since this is an approximation, the user can specify the maximum deviation of these slices from the original contour of the object that is acceptable from an application point of view. If the thickness is the same for all layers, the process is called uniform slicing. In this case, the layer thickness is determined by the slice introducing the largest deviation from the original contour. Hence, to limit the error, a smaller layer thickness should be applied, resulting in a longer processing time. An alternative approach is to vary the layer thickness during slicing in order to keep the deviations from the original contour within acceptable limits. This approach is called adaptive slicing. For example, if both approaches are applied to slice a hemisphere (diameter of 10 mm) and the specified maximum deviation is 0.781 mm, then 81 and 20 slices are generated respectively [6]. In the proposed CAM system, the adaptive slicing approach is adopted.

2.2

Generation of the CL data file

Most current CAD/CAM systems have built-in tools for generating uniform thickness slices, but cannot perform slicing with varying layer thickness. Adaptive slicing is, nevertheless, a generally better and faster way of performing 3D volume removal, which can also satisfy particular process requirements. In the case of EDM milling, the adaptive slicing approach should comply not only with the requirement to keep the deviation of the slices from the original contour within specified limits but also to satisfy the rules of UWM. In particular, this approach should allow the electrode wear in the step-depth direction to be compensated for by varying the slice thickness. The implementation of the adaptive slicing approach is not discussed further in this paper. Regarding segmentation in the step-over direction, there are four approaches for path generation [3]. In the proposed CAM system, the “to-and-fro” approach is applied to form each successive sweep of the electrode. With this approach, the in-process error due to tool wear is compensated for by varying the path starting point after each increment in the Z direction and also by changing the direction of the electrode movement (the machining angle in the step-over plane) every other slice. The layer thickness, or increment applied in the step-depth direction (ΔZ), is calculated using equation 1 [2, 3]: ⎛ vSwi ⎞ ΔZ i = Lwi ⎜1 + ⎟ Se ⎠ ⎝

(1)

where: Lwi is the thickness of the ith layer calculated using the adaptive slicing approach (mm); ν is the volumetric relative wear of the electrode; Swi is the area of the ith layer in the step-over direction (mm2); and Se is the cross sectional area of the electrode (mm2). This equation allows electrode wear to be taken into account when the thickness of each layer is computed during the slicing process. Figure 1 presents a summary of the proposed slicing procedure. Since Se and ν are known before starting the slicing process, the increment in the Z direction (ΔZ) for each layer is computed depending on Sw and Lw. In most cases Swi will be significantly larger than Se and, according to equation 1, ΔZ will be always larger than 2Lw. This shows once more the importance of tool wear compensation for each slice. 2.3

Selection of technological parameters

Technological parameters depend on the EDM machine selected to carry out the operation. In the proposed CAM system this data is loaded from a database stored in a separate file. To retrieve the volumetric relative wear (ν) from the database, the user specifies the particular combination of materials for the electrode and the workpiece, and the surface finish to be achieved after the machining. In addition, to avoid significant wear on the side of the electrode, which affects the machining accuracy (for example, when machining a vertical wall), the maximum increment in the step-depth direction should be restricted to ΔZmax. The value of ΔZmax is specified by the user depending on ν, Se and the surface area of the largest slice for the volume to be removed.

2.4

Post-processing of the geometrical and technological data

After verifying the electrode path based on the generated CL points, a postprocessor is used to convert this data into a part-programme for a particular machine. The postprocessor reflects the technological capabilities of a specific EDM machine. A generic version of a postprocessor could be easily created within a commercially available CAD/CAM system and then tuned for particular machines. START Input: Zmax, Zmin,

δ max, ΔZmax,Se,ν, Vr Z=Zmax

Calculate Sw at Z depth Calculate Lw at Z depth based on δ max (adaptive slicing)

⎛ ν .Sw⎞ ΔZ = Lw.⎜1 + ⎟ Se ⎠ ⎝ ΔZ > ΔZmax

Yes

ΔZ = ΔZmax

No Z = Z - ΔZ No

Yes Z