Studies on functional models for rapid prototyping.
Nowadays, the industries are in intense pressure to compete in the market, so they need to produce their end- products with shorter lead-time and with lower costs. Rapid prototyping (RPT) or Layered Manufacturing (LM) is a form of additive fabrication process that makes a part by adding material layer-by-layer directly from 3D CAD model directly. This is one of the time compression technologies, which will reduce the product design and development time drastically. The design engineer can easily visualize his newly designed product well in advance using this technology and also can give the recommendations for the form design immediately. It is easy to change the designs at the earliest stage itself, which will reduce the product development time and also cost. Here the 3D solid model can be developed from any one of the solid modeling packages available in the market. Then the model is converted into Stereolithography (STL) files. This is the RP standard representation of CAD data. The STL model is then sliced, by intersecting it with horizontal planes. The sliced data is generally converted to RP machine compatible format and used for fabrication. There are several RP techniques exist, such as Fused Deposition Modeling (FDM), Stereolithography (SLA), Selective Laser Sintering (SLS), 3D Inkjet printing, Selective Inhibition Sintering (SIS) and Stereo-Thermo-Lithography, etc. Based on the application, the researcher or product design engineer will prefer the technique. Complex parts also can be developed easily using these techniques. Fig.1 shows the typical RP process to make parts.
[FIGURE 1 OMITTED]
In RP process planning, slicing is the important step. The 3D model is sliced using .stl files. STL files can only approximate the original model with small facets and triangles. There are several slicing procedures, such as uniform slicing, direct slicing, adaptive slicing, adaptive direct slicing [Pandey et al. ((2003)]. The most common one is to produce cross sectional data from STL files. STL files can only approximate the original model with small facets and triangles. The STL makes the model with gaps, cracks, orphan facets, etc. [Cao et al.(2003)].
During translation of CAD models into .stl format, sometimes there will be a data loss. Hence the user cannot get the original CAD model. Separate .stl viewer and repair software is required to correct those errors. That leads to extra cost to the industries. The repair time and cost also will considerably increase. Slicing of large and complex .stl files takes long time. To avoid the above problems during use of .stl data, several researchers have developed the direct slicing and adaptive direct slicing approaches to use with different CAD/CAM packages.
Jamieson et al(1995) have developed the direct slicing approach for Parasolid files and the sliced data is converted in to CLI, HPGL, SLC formats. Chen et al. (2001) have developed the direct slicing approach based on PowerSHAPE models. In this method, lines, arcs and Bezier curves are used to describe the section contours. Chang et al.(2004) have developed the direct slicing and G code contour for PowerSolution macro commands. Pandey et al.(2003) have developed the software which does slicing of a CAD (surface)model created by sweeping a Bezier curve around an axis. They have considered surface roughness also and the results are compared with the previous researches. Cao et al. (2003) have developed the direct slicing of solid models from AutoCAD. In their work, the sliced layers are saved in ASCII DXF files. Zhou et al. (2004) have developed an algorithm for adaptive direct slicing and used the non-uniform cusp height. The solid model is built in Unigraphics and imported to their stand alone C++ program in STEP format. Gupta et al. (2004) have formulated a procedure for the selection of points approximating slice contours cut in Laminated Object Manufacturing with first order approximation. Starly et al. (2005) have sliced the STEP based Nurbs models with optimal build direction.
From the above discussion, it can be concluded that currently industry is under tremendous pressure to design and manufacture the products in a short time. This lead to the focus of researchers towards developing error free functional models through RP techniques. In this context, this work makes an attempt for AutoCAD based slicing, one of important steps involved in fabricating the error free functional models. This paper is discussing about the direct slicing of the functional model and ADS of some other models developed from AutoCAD2004.The purpose of analyzing the functional model is that the model can be directly used for the real-time applications to implement Rapid Manufacturing. This approach can be used in Fused Deposition Modeling, Stereolithography, Selective Laser Sintering and other RPT processes. The material type is also included in the algorithm. The geometrical and surface analyses are also conducted.
Direct Slicing of FM
AutoCAD is one of the most widely used computer-aided design/drafting (CAD) tool in the market today. One of the main reasons for AutoCAD's popularity is its simplicity and flexibility. With its open architecture, AutoCAD can be customized to fit the needs of almost any CAD user whether that means reconfiguring menus and toolbars, adding specialized commands, or developing add-on programs that can run within AutoCAD. AutoLISP is a programming language designed for extending and customizing AutoCAD functionality. LISP was chosen as the initial AutoCAD Application Programmable Interface (API), because it was uniquely suited for the unstructured design process of AutoCAD projects, which involved repeatedly trying different solutions to design problems. Visual LISP (VLISP) is a software tool designed to expedite AutoLISP program development. The VLISP integrated development environment (IDE) provides features to help ease the tasks of sourcecode creation and modification, program testing, and debugging. In addition, VLISP provides a vehicle for delivering standalone applications written in AutoLISP. In the present work, AutoLISP is used to implement the DS and ADS algorithm.
Direct Slicing (DS) Algorithm
The algorithm for direct slicing of an AutoCAD functional model is summarized in the procedure below.
(1) Design and assemble functional model in AutoCAD
(2) The material can be added to the model, if required.
(3) Check the assembly and make it union.
(4) Define bounding box.
(5) Specify slicing plane.
(6) Specify slicing thickness.
(7) Save the sliced file in the required format
A functional model designed in AutoCAD is shown in Fig.2
[FIGURE 2 OMITTED]
The algorithm has been tested on a part with cylindrical features and functional model. It slices both the models efficiently and uniformly irrespective of the profile as shown Fig. 3. The DS code accurately slices the curved and flat boundaries. A typical layer has been shown in Fig. 4. The layer thicknesses used in slicing are 0.1778mm and 0.254 mm. The number of layers, slicing time and file size details are tabulated and compared with other file format as shown in Table 1.
[FIGURE 3 OMITTED]
In the above table, the size of the .dxf files is larger than the .stl files, due to the addition of material in the models. In .stl files, the material is not included in to the models.
[FIGURE 4 OMITTED]
Adaptive Direct Slicing of FM
In the rapid prototyping process, surface roughness plays a critical role as it affects the part accuracy, post processing costs and functionality of the parts (Pandey et al. 2003). A good surface finish on the parts helps eliminating dimensional inaccuracy and costs due to subsequent post-processing of the part to attain the desired surface finish. The surface finish can be achieved choosing minimum layer thickness; this leads to increase in build time and costs. To overcome these issues an alternate ADS algorithm has been proposed in this work. ADS will take care of build time and surface finish. The proposed ADS will use the maximum layer thickness prescribed by the user for regular boundaries and use the variable layer thickness for curved boundaries based on its curvature so that the staircase effect can be reduced. This will reduce the build-time considerably.
The proposed ADS strategy consists of three stages:
* Detecting the features of the model
* Dividing the model into corresponding blocks
* Optimizing the layer thickness in each block based on the allowable staircase tolerance.
The adaptive slicing of solid model is shown in the Fig.5.
[FIGURE 5 OMITTED]
Algorithm for Proposed Adaptive Direct Slicing
STEP 1: The zero value is assigned to the variables maxthk and minthk as initial layer thickness. The default system variables are changed to get required accuracy of the Slice. The default system variables are osmode, osnap, aperature and lupre.
STEP 2: The input values for the variable like maxthk, minthk are received from user input function like getdist. The values assigned to variables maxthk and minthk are converted into two decimal places by using of RTOS and ATOF function.
STEP 3: The object to be selected is assigned a variable objcen with its bottom center point to zero value. The UCS icon is assigned to the objcen variable.
STEP 4: The UCS is then position the origin point of the screen for slicing purpose. The system variables are modified in order to achieve the accuracy.
STEP 5: The looping function is used to create a slice in new layer with different pattern of color. The cross section of the object is created in different layer for slicing the solid model.
STEP 6: The line is created in between the start point and end point of sliced object. The while loop is used to slice the object into maximum and minimum thickness by using conditional loop.
STEP 7: The various section is created at a new AutoCAD layer at different thickness and the layer is turned off to show the sliced portion of solid object. The sliced object is then saved in export file format such as .dxf.
STEP 8: The .dxf file will act as input file for the FDM machine for manufacturing of functional parts.
Where, maxthk- Maximum thickness, minthk- Minimum thickness osmode. aperature, luprec, RTOS, ATOF-AutoCAD default system variable.
It is justified from the sample model that minimum thickness of slicing is adapted to inclined and curved surfaces, in order to reduce the staircase effect. The maximum layer thickness is adapted for vertical surfaces of model, to reduce the build-time. Thus minimum layer thickness is applicable to curves and inclined surfaces, to reduce the staircase effect and maximum layer thickness is applicable vertical surfaces, to reduce the build up time. The Table 2 shows the comparison of the DS and ADS results with different file format using the above models. The ADS file size in .dxf format is higher than .stl file size.
The customized menu for the slicing techniques of rapid prototype was developed in AutoCAD environment. The menu was customized in the form of screen menu, pull down menu, cascading menu and tool bar menu, which are containing the LISP coding. The script files act as a bridge between the menu and LISP coding. The DS and ADS LISP code are inter-linked with menus and are saved in the menu file of RPTOOL. MNU. The existing ACAD.MNU was modified and added the LISP program of DS and ADS. The customized screen of modified menu is shown in the Fig.6.
[FIGURE 6 OMITTED]
Fabrication of Functional Model
The functional model is designed using AutoCAD software and sliced within AutoCAD environment using DS procedure. As Insight software used by FDM machine at present is supported with STL file format, so the .stl file is taken to the Insight software. The preprocessing steps are carried out and the required format is taken to the FDM Maxum machine for fabrication. The functional model fabricated here is taken 43 minutes to build. Then the model is subjected to post-processing such as support material removal, etc. The fabricated functional model is shown in Fig. 7.
[FIGURE 7 OMITTED]
The dimensional accuracy of rapid prototype is dependent on many factors, and results can vary from part to part or day to day due to environmental effects. Process parameters such as road width, hatching pattern, air gap, orientation, etc also affects the part accuracy. In this work, these parameters are not considered. Some standard parameters values are assigned for the fabrication. To understand the dimensional aspects of RP functional parts, the cylindrical part with different diameters including [empty set] 2mm, and other shapes such as flat, 90[degrees] L corner, etc. are included in the part. The measurements are taken by using Mitutoyo surface roughness tester.
Fig 8 shows the dimensions of the model fabricated. The nominal value, actual value, deviation and percentage of deviation of each representation were measured. The percentage of deviation ranges between 0.058 and 1.4, the average of percentage deviation is 0.55983. The variations in percentage of deviation are shown in the Fig 9. It is observed that negative deviations in cylindrical features and positive in regular flat boundaries. The deviations may be due to the slicing thickness, build orientation, material properties, flow of the material in the nozzle, road width, hatching distance, etc.
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
The surface roughness of the fabricated model is tested and calculated using Mitutoyo Surface Roughness Tester. The roughness can be minimized by using minimum layer thickness and by optimizing the process parameters. The roughness values are measured in four sides of the functional model on the faces A, B, C and D at four different positions in each face as shown in Fig.10. The surface roughness values are shown in Fig 11.The roughness are not consistent throughout the body.
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
The purpose of this work is to develop a more efficient AutoCAD based direct slicing and adaptive direct slicing methods to improve the RP process efficiency as well as surface quality by avoiding STL format. A slicing strategy was described, where each surface of the CAD model was directly sliced in AutoCAD software itself irrespective of the contour of the geometry. This ADS algorithm gives the user flexibility in defining the surface accuracy specifications for the CAD model by means of variable layer thickness.
The DS and ADS code has been developed using AutoLISP language. Two sample models were directly sliced in AutoCAD environment using the DS and ADS algorithm and saved in .dxf format. The file capacity of ADS file larger than DS files. This format can be opened in RP software and then the data can be sending to the RP machines in machine acceptable form. One of the models has been fabricated in Stratasys FDM MAXUM machine using ABS material for the purpose of analyzing the geometrical changes and accuracy through STL files. The purpose of using the Functional Model is that, the manufacturing industries can directly use the functional models as end products for their applications. The customization is also made for RP pre-processing in AutoCAD environment.
The Dimensional changes and accuracy of the functional model is measured and plotted with suitable relations. The maximum, minimum and average deviation is measured and similarly surface roughness also measured in different sides of the functional model. The purpose of measuring the dimensional changes and surface roughness are to improve the accuracy of the model. These two are improved by optimizing the slicing data and the other process parameters such as road width, deposition path, material properties, flow rate, etc.
 Ron Jamieson and Herbert Hacker, (1995), 'Direct Slicing of CAD Models for Rapid Prototyping', Rapid Prototyping Journal,Vol. 1, No. 2, pp. 4-12.
 Xue Yan and P Gu, (1996), 'A Review of Rapid Prototyping Technologies and Systems', Computer Aided Design,Vol. 28, No. 4, pp.307-318.
 W. Cao and Y. Miyamoto, (2003), 'Direct Slicing from AutoCAD Solid Models for Rapid Prototyping', Int. J. Adv Manuf Technology, Vol. 21, pp. 739-742.
 X. Chen, C. Wang, X. Ye, Y. Xiao and S. Huang, (2001), 'Direct Slicing from PowerSHAPE Models for Rapid Prototyping', Int. J. Adv Manuf Technology, Vol. 17, pp. 543-547.
 Prashant Kulkarni and Debasish Dutta, (1996), 'An Accurate Slicing Procedure for Layered Manufacturing', Computer Aided Design, Vol. 28, No. 9, pp. 683-697.
 B. Starly, A. lau, W. Sun,W. Lau, T. Bradbury, (2004), 'Direct Slicing of STEP based NURBS Models for Layered Manufacturing', Computer - Aided Design,Vol. 37, pp. 387-397.
 P.M.Pandey, N.Venkata Reddy, Sanjay G.Dhande, (2003), Slicing procedures in layered manufacturing: a review, Rapid Prototyping Journal, Vol.9, No5, pp.274-288
 P.M.Pandey,N.Venkata Reddy, S.G.Dhande, (2003), Real time adaptive slicing for fused deposition modeling, Int.J.Machine Tools & Manuf, Vol.43, pp.61-71
 M.Y. Zhou, J.T. Xi and J.Q. Yan, (2004) 'Adaptive Direct Slicing with Nonuniform Cusp Heights for Rapid Prototyping', Int. J. Adv Manuf Technology, Vol. 23, No. 1-2, pp. 20-27.
 Z. Zhao, L. Laperriere, Adaptive direct slicing of the solid model for rapid prototyping, http://ecoleing.uqtruquebec.ca /geniedoc / gmm / productique / ADSlice.pdf.
 Paul alexander,seth Allen and Debasasish Dutta,(1998) 'Part orientation and build cost determination in layered manufacturing', Computer Aided Design, Vol. 30, No. 5, pp. 343-356.
 C.C.Chang, H.W.Chiang, S.H.Sun, (2004),Direct Slicing and G-code contour for Rapid prototyping machine of UV resin spray by PowerSolution macro commands., Int. J. Adv Manuf Technology,Vol. 23, No. 5-6, pp.358-365
 Autodesk, Inc., AutoCAD 2004 Manual
M.Sendilkumar (1), P.S.S.Prasad (2) and K.A.Jagadeesh (3)
(1) Research Scholar, Rapid Prototyping Laboratory, (2, 3) Asst.Prof.,Dept. of Mechnical Engineering, PSG College of Technology, Coimbatore-641004
(2) Corresponding Author Email:firstname.lastname@example.org
Table 1: Comparison of DS with QuickSlice Time in File size Layer No. of sec. S.NO Model Thickness Layers DS QS .dxf .stl 1. Part with 0.1778 587 4.6 2.42 5.24 1.14 cylindrical MB MB features 0.254 347 3.52 2.04 3.12 825 MB KB 2. Functional 0.1778 57 5.04 1.19 1.67 119 model MB KB 0.254 40 3.84 0.92 1.16 88 MB KB Table 2: Comparison of DS and ADS Layer No. of File size S.NO Model Method Thicknes Layers .dxf .stl 1. Model 1 DS 0.254 1082 10.24 MB 4.14 MB ADS 0.1778-0.254 1426 12.85 MB 6.21MB 2. Model 2 DS 0.254 1246 7.642 MB 3.86 MB ADS 0.1778-0.254 1428 8.36 MB 4.28 MB
|Printer friendly Cite/link Email Feedback|
|Author:||Sendilkumar, M.; Prasad, P.S.S.; Jagadeesh, K.A.|
|Publication:||International Journal of Applied Engineering Research|
|Date:||Dec 1, 2008|
|Previous Article:||Mechanical properties and durability of rice husk ash concrete.|
|Next Article:||Programmable speed controller of AC servomotor using FPGA.|
|Rapid prototyping slices into new uses.|
|RP improvements suggested.|
|High-impact ABS arrives for FDM rapid prototypes.|
|Stereolithography resin resembles white ABS.|