A new benchmarking part for evaluating the accuracy and repeatability of Additive Manufacturing (AM) processes

Transcription

1 A new benchmarking part for evaluating the accuracy and repeatability of Additive Manufacturing (AM) processes Dr Muhammad Fahad, Dr Neil Hopkinson Abstract Additive Manufacturing (AM) refers to a new class of manufacturing processes that build physical parts in a laye-by-layer (i.e. additive) manner. Even though a variety of AM processes are now commercially available which differ from each other in the way they build a part (i.e. sintering, jetting, potopolymerisation etc), AM processes are still in their early stages of commercialization and thus present a high potential for research and development. Therefore, in order to improve the technical capabilities (i.e. accuracy and repeatability) of any AM process or in order to compare the performance of different AM processes, a benchmark geometry is essential. The benchmark geometry provides a common basis for comparison/fine tuning of different processes and must comprise of certain features/dimensions that will ensure that the operation capabilities are fully evaluated. This paper details such a part which is designed to act as a benchmark part and can be used as a standard geometry for evaluating different AM processes. Keywords Additive Manufacturing, Benchmarking, Accuracy, Repeatability. A I. INTRODUCTION DDITIVE Manufacturing (AM) is the name given to the technologies that produce physical parts from CAD data in an additive manner (i.e. adding successive layers of materials). It is a generic term used to represent techniques that can build components without the need of conventional tooling in the first instance [1]. Although, the cycle time, cost of material and equipment associated with AM processes are high relative to those associated with conventional manufacturing processes (e.g. CNC machining, injection molding), the elimination of tooling cost and time, increased design freedom and reduced cycle are the advantages that AM processes offer [2]. Various AM processes have been developed and are commercially available including Stereolithography (SL),Selective Laser Sintering (SLS), Fused Dr Muhammad Fahad is with the Department of Industrial and Manufacturing Engineering, NED University of Engineering and Technology, Karachi, Pakistan, (phone: ; mfahad@neduet.edu.pk) Dr Neil Hopkinson is with the Wolfson School of Mechanical and Manufacturing, Loughborough University, Loughborough, Leicestershire, United Kingdom. LE11 3TU. ( n.hopkinson@lboro.ac.uk) Deposition Modeling (FDM) and Three Dimensional Printing (3DP). Each AM process builds parts in different materials and by using a different working principle such as photopolymerisation (by SL), sintering (by SLS), extrusion (by FDM) and jetting (by 3DP). Since these processes operate in a tool-less, additive manner, the parts built are enclosed in an envelope and the machine vendors offer machines in different build envelope sizes. Benchmarking refers to the comparison of performance of different similar systems (organisations, machines, processes etc) with each other to establish a standard of performance. A benchmark as defined by the Webster s Dictionary is a standard or point of reference in measuring or judging quality, value, etc. In AM, benchmarking is used not only to compare the strength/weakness of the parts but also to measure and compare accuracy, surface finish, repeatability and resolution of the geometrical features of the parts produced. According to Wong et al, benchmarking helps in identifying the highest standards of excellence for different products/processes so that the subsequent improvements necessary to achieve those standards can be made [3]. Wong et al also classified AM benchmarking in three different types as follows: Geometric Benchmark; used to measure the geometric features of a part (i.e. tolerances, accuracy, repeatability and surface finish). Mechanical Benchmark; used to analyze the mechanical properties (tensile strength, compressive strength, creep, etc). Process Benchmark; used to establish process related parameters (part orientation, support structures, layer thickness, speed, etc). This paper is aimed at defining a new type of geometric benchmarking part for evaluating the accuracy, tolerances and repeatability of parts produced by different AM processes. II. ADDITIVE MANUFACTURING BENCHMARKING STUDIES Various studies have been carried out relating to the design of benchmarking parts for AM processes. Kruth [4] performed one of the earliest benchmarking studies in the field of AM. An inverted U-frame shaped part (Fig. 1)which contained different features like vertical and inclined cylinders, flat and inclined surfaces, squares, pegs, embossed letters and overhangs was used. The part does not have any repeated features and the repeatability was tested by producing many parts. Lart [5] designed a small benchmark part (Fig. 2) 234

2 comprising fine and medium sized features to compare the ability to produce fine details. The study comprised circular features (holes and cylinders), flat and angular surfaces in different directions but large scale accuracy and warping. Juster and Childs [6,7] used a square shaped part (Fig. 3) to compare the accuracy and repeatability of parts produced by different AM processes. Although repeatable features were produced in this part but the part did not contained any plane of symmetry to evaluate the repeatability of features. Iuliano, Ippolito and de Filippi [8] also used a square shaped part (Fig. 4) to evaluate the accuracy and tolerances of various AM processes. The part was not found suitable for evaluating the accuracy of non-flat surfaces. Therefore, a shell of uniform thickness, comprising of a cylinder merged with a sphere (Fig 5) was used to evaluate non-flat surfaces. Shellabear [9] designed a simple part that contained surfaces at different angles along the vertical direction. The part could only be used for linear measurements as no circular/curved features were included in the design. Mahesh, Wong, Fuh and Loh [10] designed a square based part containing various features such as square bosses, cones, cylinders and free-form surfaces (Fig. 6). Although, the part geometry comprised of different features, the design did not show a plane of symmetry as the features were not distributed equally onto the square base. Hopkinson and Sercombe [11] used a square shaped part with steps of uniform thickness in X, Y and Z directions (Fig. 7). The part could only be used for linear accuracy and repeatability was evaluated by building more than one parts. Fig. 1. Part used by Kruth [4] Fig. 4. Part used by Iuliano, Ippolito and de Filippi [8] Fig. 2. Part used by Lart [5] Fig. 5. Part used by Iuliano, Ippolito and de Filippi (For non-flat surfaces) [8] Fig. 3. Part used by Juster and Childs [6,7] Fig. 6. Part used by Shellabear [9] 235

3 Fig. 7. Part used by Mahesh, Wong, Funh and Loh [10] III. DESIGN OF NEW BENCHMARKING PART On the basis of literature review of various benchmarking studies related to AM processes, it was noted that all the designed parts contained various features to evaluate/compare accuracy of features but none of these parts contained features in a truly repetitive manner. Due to this reason, more than one parts are required to build to evaluate the repeatability of a process. Therefore, it can be said that true repeatability (i.e. symmetry of features within the part) is an important requirement in order to compare the repeatability within a single build. Also, the literature review revealed the fact that some features are necessary to evaluate the accuracy/tolerances of parts produced by AM processes. These features, along with their intended purpose [19] are summarized in the table below: FEATURE Flat Base Cube Cylindrica l Hole Sphere Solid Hollow TABLE I GEOMETRIC FEATURES AND THEIR INTENDED PURPOSE PURPOSE Flatness and straightness Squareness, parallelism, linear accuracy and repeatability Roundness, cylindricity, accuracy and repeatability of radius (internal) Sphereness, relative accuracy and repeatability of a continuously changing sloping surface Roundness, cylindricity, accuracy and repeatability of radius (external) Roundness, cylindricity and coaxiality of cylinders Fig. 8. Part used by Hopkinson and Sercombe [11] A new benchmarking part was designed keeping in view all the above features as well as making sure that the features are repeated symmetrically in one single part (Fig. 8). The part was designed on a flat base of size 270 mm x 50 mm. This base size was selected to make sure that the part can be built on most of the commercially available AM machines. To ensure the true repeatability of all the features, the base was divided into three section, each being 90 mm long. Each of these 90 mm long section contained same features at exactly same location/distances from the centre of respective section. Table II summarizes the sizes of different features of the designed benchmarking part. TABLE II GEOMETRIC FEATURES INCLUDED IN THE NEW BENCHMARKING PART AND THEIR SIZES FEATURE NUMBER AND DIMENSIONS (mm) Base 1 (270x50x5) Cube 12 (15 x 15 x 15) Cylindrical Hole Sphere Solid Hollow Cone Angled Surfaces 3 (X direction), 3 (Y direction), 3 (Z direction), All holes with diameter 10 mm and depth 15 mm 3 (15 mm diameter) 3 (10 mm diameter, 12 mm height) 3 (Outer diameter 10 mm, inner diameter 7 mm, height 16 mm) 3 (9.65 mm base diameter, 18 mm height) 3 (30 degrees), 3 ( 60 degrees) Cone Angled Surfaces Concity, sloping profile and taper Angularity, accuracy and repeatability of angled surfaces 236

4 Fig. 9. Design of New Benchmarking Part for AM processes IV. FABRICATION OF BENCHMARKING PART The designed benchmarking part was fabricated using Selective Laser Sintering (SLS), a well established AM process. The part was built using DuraForm (Nylon) powder to concept evaluation purpose and is shown in Fig. 9. By evaluating the physical prototype fabricated using SLS process, it can be said that the part design is suited for fabrication on different AM processes such as Fused Deposition Modelling, Stereolithography, 3D printing, etc. Fig. 10. New benchmarking part built using SLS V. CONCLUSIONS A benchmarking part is an important aspect of evaluating the performance characteristics of various AM processes. Although, many authors/researchers have proposed various designs of benchmarking parts for AM processes, none of these parts comprehensively included all the features necessary to establish the desired accuracy/repeatability related parameters. In this study, a new type of benchmarking part has been designed and fabricated (using an AM process). This new benchmarking part not only includes all the necessary features in a very compact manner, but it also allows the measurement of repeatability of features by incorporating the features in a symmetrically repeatable manner. This part will not only help in establishing the performance parameters for any AM process, but it will also help in comparing various new and/or well established AM processes in terms of their accuracy and repeatability. 237

5 REFERENCES [1] S. Upcraft, R Fletcher, Rapid prototyping technologies Assembly Automation. Vol. 23, Number 4, 2003, pp [2] N. Hopkinson, P. Dickens, Rapid prototyping for direct manufacture Rapid Prototyping Journal, Vol. 7, Number 4, 2007, pp [3] Y. S. Wong, Y. H. Fuh, H. T. Loh, M. Mahesh, Rapid prototyping and manufacturing benchmarking, in Sofwtare Solutions for RP, 1 st Edition, Editor: I. Gibson, 2002, [4] J. P. Kruth, Material incress manufacturing by rapid prototyping techniques CIRP Annals, Vol. 40, Number 2, 1991, pp [5] G. Lart, Comparison of rapid prototyping systems, Proceedings of First European Conference on Rapid Prototyping, University of Nottingham, July 1992, pp [6] N. P. Juster, T. H. C. Childs, Linear and geometric accuracies from layer manufacturing CIRP annals, Vol. 43, Number 1, 1994, pp [7] N. P. Juster, T. H. C. Childs, A comparison of rapid prototyping processes Proceedings of Third European Conference on Rapid Prototyping and Manufacturing, University of Nottingham, July 1994, pp [8] NR. Ippolito, L. Iuliano, A. de Filippi, A new user part for performance evaluation of rapid prototyping systems Proceedings of Third European Conference on Rapid Prototyping and Manufacturing, University of Nottingham, July 1994, pp [9] M. Shellabear, Benchmarking study of accuracy and surface quality in RP models RAPTEC, Task 4.2, Report 2, 1999 [10] M. Mahesh, Y. S. Wong, Y. H. Fuh, H. T. Loh Benchmarking for comparative evaluation of RP systems and processes Rapid Prototyping Journal, Vol. 10, Number 2, 2004, pp [11] T. B. Sercombe and N. Hopkinson, Process shrinkage and accuracy during indirect laser sintering of Aluminum Advanced Engineering Materials, Vol. 8, Number 4, 2006, pp