An assessment of geometrical features for a new rapid prototyping benchmark part.

Popescu, Diana ; Ghinea, Mihalache ; Chircor, Lidia 等

1. INTRODUCTION

Due to the existence of a relative large variety of test parts for evaluating RP processes, the literature is offering information about the prototypes different geometrical features and their manufacturing resolution. Each of the test parts presented in the literature contains a certain number of features (holes, slots, threads, chamfers, slope walls, overlap features, very thin walls, freeform features, etc.) of different sizes and placed at different locations, which can be used by a designer (both in technical and medical field) in defining the geometrical form of a product. But there are also geometric features that have not been studied yet and which should be identified and evaluated, thus offering useful information for an objective evaluation of RP processes capabilities, in terms of position and form accuracy.

In order to complete the data set gather from literature, the first step is to consider the independent benchmark parts which were used to test more than one RP process. Further a broad list of geometrical features is consider, each with the corresponding evaluating purpose. Based on this information, a benchmark part containing new geometrical features was designed.

2. RP GEOMETRICAL BENCHMARK PARTS

An extensive study of the RP benchmarking and comparative studies helped distinguish the following important issues (Ispas et al. 2002), (Gatto & Iuliano, 1998), (Childs & Juster, 1994), (Makesh et al., 2004):

* There are more than 20 benchmark parts available in RP industry, designed by both RP machines producers and researchers in the field. Benchmarks developed by RP systems' producers tend to have a design which underlines the advantages of a particular process and machine. Most of the time these are user-dependent, system-dependent or process-dependent benchmarks;

* There is a lack of standardization regarding both the design of benchmark part and measurement/test methods, which makes difficult the evaluation of the results of the comparative studies;

* There are three types of benchmark parts: geometric (used for evaluating dimensional and geometric accuracy, parallelism, symmetry, concentricity, flatness, repeatability, etc.), process (for assessing aspects like parameter settings, building orientation, build styles, etc.) and mechanical (used for evaluating strengths, curling, shrinkage, swelling, etc).

* In the literature, comparative analyses have been made considering mostly the most common RP processes (SLA, 3DP, FDM, LOM, SLS). The purposes of benchmarking these RP processes are to determine feature size limitations, build orientation issues or surface and dimensional accuracy capabilities, and to make comparison between different processes. Also benchmark studies are very useful for determining the optimal RP process for a certain application;

* The benchmark studies performed so far involves building of only one sample for each process and material, when there are other factors to be considered: machines parameters settings or building styles;

* Geometric benchmark parts are designed to include many different geometric features (spheres, cones, parallelepipeds, cylinders, cubs, slots, slope and thin walls, draft angles, freeform surfaces, etc.). These features are built in different positions and number in order to check characteristics such as: repeatability, relative position, parallelism, symmetry, etc.

In this context, in order to avoid the use of system, user or process-dependent benchmarks, rules for building an "ideal" benchmark part for RP were developed (Gatto & Iuliano 1998):

* Its dimensions must be proportionally distributed in the modeling envelope of the RP machine, in order to accept the performances in the central and extreme zones;

* Small, medium and large dimensions uniformly distributed; building time not too big;

* Easy to measure using a coordinate measuring machine;

* Not be specific to a certain RP process.

For the present study, based on the above mentioned observations and developed criteria for designing a benchmark part, five independent test parts were considered as relevant because they gather the biggest number and variety of geometrical features.

Benchmark parts presented in the analyzed studies were used for assessing different aspects regarding RP processes:

* Linear accuracy and feature repeatability for: SLA, FDM, SLS, EOS, SGC and LOM (Childs & Juster, 1994). The benchmark has a plane base on which are placed different geometric features: parallelepipeds, cylindrical holes, free form features, slope and thin walls. Also, the test part has small, medium and large size features;

* In (Ippolito et al., 1995) on the based of the 3D Systems benchmark parts, the study was also focused on analyzing the surface accuracy for different RP systems;

* Accuracy of curved surfaces and surfaces roughness. The benchmark part studied in (Gatto & Iuliano 1998) consists in a cylinder combined with a sphere through sloping and planar surfaces;

* Accuracy and surface quality. The test part contains only planar surfaces, oriented at different angles to the vertical building direction: 0[degrees], 10[degrees], 40[degrees], 80[degrees] and 90[degrees]. 44 prototypes were built using the following RP processes: SLA, SLS, LOM, EOS, 3DP and FDM. An important particularity of this test part is that different materials were used with the same process, i.e. for SLA--resins: SL5185, SL5195, for SLS--resins: Somos6120, Somos7120, Somos8100, for FDM: ABS P400, ABS401, wax ICS400, for EOS: PA2200, PA3200GF, PS2500, etc. (Shellabear, 1999).

* In the test part presented in (Kruth et al. 2005), the researchers introduced the slope walls for evaluating the staircase effect, sharp edges to check the heat accumulation at the angle tips, small dimensions holes and thin walls for determining processes accuracy and resolution on x, y, z axes.

* The benchmark part in (Ispas et al. 2002) has a geometrical form obtained by intersecting a cylinder and a cone. The dimensions are grouped in small, medium and large, a comparative analysis of dimensional and form accuracy in vertical and horizontal planes being performed.

3. DESIGNING A NEW GEOMETRICAL BENCHMARK PART

Analyzing the information collected in table 1, the following geometrical features were not considered so far to the best of our knowledge:

* Inclined cylindrical surfaces for studying the relative position of inclined axis

* Hexagonal hole with inclined axis for evaluating the need for support structures and form tolerances such us angularity

* Torus for assessing the ability to build this type of feature and other similar features like helicoidally channel with circular profile (for ball screws)

* Pyramid for measuring the flatness and straightness of sloping planes

* Shoulder hole for assessing for the same feature coaxiality, concentricity and perpendicularity

* Completely enclosed spherical holes Also, the slope plane and the edge fillet are considered for

comparatively assessing the staircase effect innate to all layered manufacturing processes. All the geometrical features can be used for measuring the position accuracy on x, y, z directions.

[FIGURE 1 OMITTED]

4. CONCLUSION AND FURTHER WORK

The research presented in this paper focused on designing a new benchmark part for assessing the form and position tolerances of different geometrical features not analyzed in the RP literature. This benchmark part can be used for all processes, its features not being particular to a certain technology or RP process. By using this type of benchmark parts and this approach, geometrical features which can be encountered in functional parts can be assessed.

The test part proposed contains features as completely enclosed spherical voids, torus, pyramid or shoulder holes, which can be used for assessing on one hand the ability of RP processes to build these features, and on the other hand evaluating the form and position tolerances.

Further work will consider building the benchmark part using 3DP, FDM and SLA processes and comparatively analyzing the results. Moreover, an analysis of the causes determining the defects could be significant.

5. REFERENCES

Childs, T.H.C. & Juster, N.P., (1994), Linear and Geometric Accuracies from Layer Manufacturing, Annals of CIRP, vol.43/1/1994, pp<.163-166

Gatto, A. & Iuliano, L., (1998), Prototipazione rapida. La tecnologia per la competizione globale (Rapid Prototyping. Technology for global competition.), Tecniche Nuove, Milano ISBN 88-481-0294-8

Ippolito, R., et al. (1995), Benchmarking of Rapid Prototyping Techniques in Terms of Dimensional Accuracy and Surface Finish, Annals of CIRP vol.44/1/1995, pp.157-160

Ispas, C., Popescu, D., Rigal, J.F. (2002), Research on the dimensional and form accuracy of a FDM1650 system, IDMME 2002, CD

Jurens, K.K., (1999), Standard for the rapid prototyping industry, Rapid Prototyping Journal, vol.5, no.4, pp.169-178, ISSN 1355-2546

Kruth, J.P. et al., (2005), Benchmarking of different SLS/SLM processes as Rapid Manufacturing Techniques, Int. Conf. Polymers&Moulds Innovations, Gent Belgium, April 20-23, pp.1/7-7/7

Makesh, M., et al., (2004), Benchmarking for comparative evaluation of RP systems and processes, Rapid Prototyping Journal, vol.10, no.2, pp.123-135, ISSN 1355-2546

Popescu, D. (2007), Design for Rapid Prototyping: Implementation of Design Rules Regarding the Form and Dimensional Accuracy of RP Prototypes, Annals of DAAAM for 2007 & Proceedings of The 18th International DAAAM SYMPOSIUM, pp.591-592, ISSN 1726-9679, ISBN 3-901509-57-7, Zadar, Croatia

Shellabear, M., (1999), Benchmark study of accuracy and surface quality in RP&M models, RAPTEC, Task 4.2, Report 2

Tab. 1. List of geometrical features and the form and position
tolerance objectives envisioned from their built

Geometrical feature Form and position tolerances--
 objective

Plane in parallelepipeds and Flatness,parallelism, linear
cubes accuracy, straightness,
 repeatability, perpendicity,

Cylindrical surfaces in solid Concentricity, roundness,
or cylindrical holes with coaxiality
horizontal and vertical axes

Thin walls Capability to build this type of
 features

Spherical surfaces in sphere Roundness, symmetry
or hemispheres

Overhang features such us in Capability to build this type of
horizontal holes etc. features

Sloping surfaces at different Linearaccuracy, relative
angles position, angularity

Completely enclosed features Capability to build this type of
such as spherical or cubic features
voids

Circular holes Cylindricity, roundness,
 repeatability, symmetry,
 coaxiality

Hollowparts: cylinders, Straightness, ability to build thin
squares walls, concentricity

Cones Sloping profile

Mechanical features: Capability to build this type of
chamfer, blending or fillet features