The optimization of surface quality in rapid prototyping processes.
Ancau, Mircea
1. INTRODUCTION
In the optimization field of RP technologies, the researcher's
interest concentrates on few major directions such as: STL-based slicing
algorithms, process parameters, modeling and simulation, part
orientation, packing problem. Depending on the source data, there are
two methods to slice the geometric model of a part into layers, i.e. the
STL-based slicing and direct slicing based on different CAD systems with
different data formats (Haipeng & Tianrui, 2007; Chang, 2004; Cao
& Myamoto, 2003). Compared with STL-based slicing, direct slicing
avoids some approximation which exists in (STL) format file. Some
researchers proposes a direct slicing method that can provide more exact
laser beam paths by slicing a constructive solid geometry (CSG)
representation of a part. A severe disadvantage of direct slicing is the
capability among various CAD systems, in other words, it can only be
used for a specific set of CAD software and machine, and is not
applicable to any other CAD combinations (Haipeng & Tianrui, 2007).
As a matter of fact, STL-based slicing is still the commonly used method
in processing the problem of layered manufacturing. The advantage of
slicing a STL file is that the problem is reduced to finding
plane-intersections. The relationship between the main process factors
and surface roughness distribution is analyzed theoretically by modeling
the stair stepping effect, which results from stacking layers. Based on
this investigation an optimum part orientation for fabrication is chosen
(Ahn et al, 2007). While surface of stereolithography parts become rough
due to the stair-stepping effect and burrs from the support structure,
most parts need some finishing work for further applications. Because
post-processing operations require additional time and cost, the
reduction of post-processing time and cost by fabrication-direction
optimization become an important concern (Kim & Lee, 2005). Among
the criterions taken into account by (Hur et al, 2001), it can be
mentioned: minimum part height and the minimum value of the part
dimensions ratio x/y (i.e. the part length considered into the direction
of powder disposal on the platform must not overstep the part width).
Referring only to selective laser sintering on Sinterstation 2000, (Hur
et al, 2001) state that the first section of sinterized material must be
less than a specific value denoted minimum specified area, in order to
avoid to high values of part deformation. The problem of part
orientation in the machine chamber is solved via an empiric algorithm so
that the total volume occupied by multiple parts is minim and mean time
the parts to be grouped as much as possible in the near centre of the
work chamber. The concepts of orientation optimization are implemented
and solved with a genetic algorithm. Concerning the process parameters
some researchers frequently used a statistical analysis of the rapid
prototyping processes, in order to find out the combination of
parameters leading to the best accuracy of the manufactured parts
(Campanelli et al, 2007). In the study of (Zhang et al, 2002), a
simulated annealing algorithm was applied to find the optimal batch
configuration layout for the minimum cost of production for solid ground
curing processes. They take into account three kinds of objectives:
fitting models into the specified container, avoiding any overlap
between models and achieving high packing density, in other words,
achieving the minimum overall height. A detailed study of some important
build parameters which affect the quality and accuracy of the final
stereolithography parts, such as the layer thickness, resultant
overcure, hatch space, blade gap, and part location can be found in
(Zhou et al, 2000). Their investigation suggests the best setting of
these control factors for different 20 individual features.
The main objective of this research is to find the optimum 3D model
orientation on the RP working platform so that the surface quality will
result as good as possible. We will follow two possible approaches. The
first one will establish the orientation in an automate manner, while
the second approach will involve human-computer interaction.
2. THE OPTIMUM PART ORIENTATION
2.1 The automated part orientation
The surface of the 3D model in stl format is approximated with
small triangle facets. For each facet there are known the (x,y,z)
coordinates of each vertex, and the projections on the axes of the
normal unit vector to the triangle surface, pointing outside the model.
The position of the normal vector remains unchanged relative to the
triangle position. As a consequence, if we rotate the 3D model with an
angle, this will involve the rotation of all normal vectors with the
same angle.
[FIGURE 1 OMITTED]
To test our concepts for optimum part orientation, we consider as
example the upper case of a mobile phone as you can see in Figure 1.
The drawing position was taken as reference for further
orientation, according to the surface quality criterion. In order to
establish optimal part orientation on the working platform of the RP
system, we will give an incremental rotation to the 3D model around Ox
and Oy axis. For each incremental position we will calculate the number
of triangular facets whose slope is between some initial established
limits. Only those values which correspond to a minimum number of facets
which respects initial conditions will be taken.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
Fig.3. shows the facets which do not satisfy the criterion of slope
value, if the part orientation is made automatically.
2.2 The manual part orientation
Some surfaces need a better roughness due to its functional task,
while others need the same roughness values for aesthetic purpose. In
every case the engineer have the mission to decide in this problem.
Despite engineers experience, when the part has a complicated geometry,
is difficult to appreciate which is the most appropriate orientation so
that some surfaces will result with some prescribed roughness. We need
an instrument to identify those triangular facets which have a specific
slope, for a given orientation of the 3D model. That specific slope will
produce a roughness above the input limits. This is why we made a
program which shows the triangular facets with a specific slope in a
different contrast color than the other facets, so that these facets may
be easy identified. Both faces of the 3D model have triangular facets
with a slope which involve a roughness higher than that the admissible
roughness. If there is an automate process which count the number of
these facets, then will be taken into account all the facets which do
not satisfy the criterion of slope value, even if the facets situated on
the back side of the model do not matter if we are interested only in
the surface quality on the front surface.
It can be simply seen that at an automated orientation of
135[degrees] rotation around Ox and also 45o around Oy, that even if the
facet number which will result with a higher staircase effect is
minimum, their placement is asymmetric. This situation will involve
asymmetric surfaces which will require post-processing, which may lead
to geometric errors. This is why, a manual orientation of the part on
the RP system working platform is almost always essential.
3. CONCLUSION
There exists a connection between the surface roughness and the
staircase effect in RP processes. The staircase effect greatly depends
on the layer thickness. As a consequence, the decrease of the layer
thickness will decrease the staircase effect and will improve the
surface roughness. The decreasing of the layer thickness will lead
implicitly at a considerably reduction of RP process productivity. The
surface quality in RP processes depends on the 3D model orientation on
the working platform.
This paper presented two different alternatives to find out the
optimal orientation of the 3D model on the working platform of the RP
system. The first variant is an entirely computerized one. The method is
based on the minimum number of the triangular facets whose slope
accomplishes an initial condition. A disadvantage of the automated
method is that it does not take into account that some surfaces are very
important concerning the functional or aesthetic role, while other
surfaces are not. The second method requires the manual part orientation
by the designer. This is why a computer program which shows in a
contrast color those triangular facets whose slope angle fulfill an
initial condition, was made. By rotation of the model around its
horizontal axis, those regions of the surfaces which will result with a
higher staircase effect, can be easy visualized. Future research will
take into consideration the productivity of RP processes as a very
important optimization criterion.
4. REFERENCES
Ahn, D., Kim, H., Lee, S. Fabrication direction optimization to
minimize post-machining in layered manufacturing. International Journal
of Machine-Tools & Manufacture, Vol. 47, 2007, p.593-606
Campanelli, S.L. et al. Statistical analysis of the
stereolithographic process to improve the accuracy. Computer-Aided
Design, Vo.39, 2007, p.80-86
Cao, W., Myiamoto, Y. Direct slicing from AutoCAD solid models for
rapid prototyping, Int. J. Adv. Manuf. Technol. Vol.21, 2003, p.739-7422
Chang, C.C. Direct slicing and G-code contour for rapid prototyping
machine of UV resin spray using PowerSOLUTION macro commands, Int. J.
Adv.Manuf. Technol. Vol.23, 2004, p. 358-365
Haipeng, P., Tianrui, Z. Generation and optimization of slice
profile data in rapid prototyping and manufacturing. Journal of
Materials Processing Technology, Vol. 187-188, 2007, p.623-626
Hur, S.M., Choi, K.H., Lee, S.H., Chang, P.K. Determination of
fabricating orientation and packing in SLS process. Journal of Materials
Processing Technology, Vol.112, 2001, p.236-243
Kim, H.C., Lee, S.H. Reduction of post-processing for
stereolithography systems by fabrication-direction optimization.
Computer-Aided Design, Vo.37, 2005, p.711-725
Zhang, X., Zhou, B., Zeng, Y., Gu, P. Model layout optimization for
solid ground curing rapid prototyping processes. Robotics and Computer
Integrated Manufacturing, Vol.18, 2002, p.41-51
Zhou, J.G., Herscovici, D., Chen, C.C. Parametric process
optimization to improve the accuracy of rapid prototyped
stereolithography parts. Int. J. Machine Tools & Manufacture,
Vol.40, 2000, p.363-379
