al., Elsevier Science B.V., 1994

AUTHORS

Thomas Cord studied Computer Science at the University of Karlsruhe from 1985 to 1992 with the main interest on Artificial Intelligence and Robotics. In March 1992 he obtained his Diploma Degree with a thesis on modelling of dynamical systems for diagnosis. In April 1992, he joined the FZI, in the department Technical Expert Systems and Robotics as a research associate. He is responsible for the ESPRIT project MARTHA (Mobile Autonomous Robots for Transportation and Handling Appli-cations). From 1993 onwards, he has been heading the robotics group. The aim of this group is the development of mobile and autonomous robots for applications in transportation. His research inter-ests are in computer vision as well as in rapid prototyping.

Walter Reithofer obtained his Diploma Degree in Computer Science in 1993 from the University of Karlsruhe, Germany. In his master thesis he developed the basics of the micro stereolithography technology and proved its feasibility. Prior to his computer science studies, he received a Diploma Degree in Mechanical Engineering in 1987 from the Profession Academy Heidenheim, Germany. Since 1993 he has been a research associate at the Institute for Real-Time Computer Systems and Robotics at the University of Karlsruhe. His research interests are in computer integrated manufac-turing, especially in modelling, simulation and control.•Fast production of prototypes

There is no need to develop special tools, moulds or dies. The present material characteristics (strength, elasticity, etc.) allow to produce functional parts as well, rather than prototypes of reduced strength. Stereolithography proved to be very suited for producing dead molds for investment casting.

The implementation of a complete micro stereolithography device requires a lot of experience on the technical and chemical processes. A lot of work still has to be done:

•Optimization of the exposure unit

•Evaluation of the effects of light intensity and exposure time on the mechanical properties of the produced parts

•Development of an elevator unit which is suited for the manufacturing of micromechanical parts

Summarizing, the proposed micro stereolithography technology is a very promising approach to meet future requirements of highly integrating and packaging mechanical parts in microsystems. ACKNOWLEDGEMENT

The authors would like to thank Prof. Dr.-Ing. U. Rembold from the University of Karlsruhe for his advice and support of the project. The work has been carried out at the Forschungszentrum Informa-tik (FZI), Dept. Technical Expert Systems and Robotics.

REFERENCES

[Jacobs 92]Jacobs, P.F.

Fundamentals of Stereolithography

Society of Manufacturing Engineers, 1992

[Kochan 92]Kochan, D.

Innovative Produktentwicklung und Fertigung mittels Solid Freeform Manufacturing

NCG-Hearing “SFM”, Tagungsband NC-Gesellschaft e.V., Ulm, 1992

[Kochan 94]Kochan, D.

Entwicklungszeiten neuer Produkte werden immer kürzer

VDI Nachrichten Nr. 14, 1994

[Kruth 91]Kruth, J.P.

Material Incress Manufacturing by Rapid Prototyping

Annals of the CIRP, Vol. 40/2/1991, 1991

[Reithofer 93]Reithofer, W.

Konzeption und Realisierung der Schlüsselkomponenten eines Stereolithographiesy-

stems

Diploma Thesis, Universität Karlsruhe, 1993

[Reithofer 94]Reithofer, W.; Cord, T.

Mikro-Stereolithography mit sichtbarem Licht und LCD-Maske

Intelligente Produktionssysteme - Solid Freeform Manufacturing, TU Dresden, 1994 [Zissi 94]Zissi, S.; Corbel, S.; Jézéquel, J.Y.; Ballandras, S.; André J.C.

Microstereophotolithographie: a reality or a dream for tomorrow?to the part design together with its functionality. Several knowledge intensive activities must be per-formed during and after the part design process:

•Verification of feasibility

A part designer develops the part using surface, volumetric or solid modelling techniques.

During this process inconsistencies of the CAD data may occur which have to be recognized and repaired.

•Selection of the best manufacturing order and fixture design

Fixture design is a process which involves the selection and the placement of the most suit-able type, shape and size of supports to constrain the individual layers from movement dur-ing the machining operations. The small size of the objects to be produced makes the manual preprocessing operations more complicated than they are in known stereolithography sys-tems. An effective support design concentrating on minimizing defects and the determina-tion of the best way to orientate the part (selection of the Z axis) may affect the preprocessing operations.

•Decomposition of the object into slices

All layer-by-layer processes require slicing of the part model and support structure at dis-tances equal to the layer thickness. The slicing is made easy by the STL format, containing only flat triangular faces. A exposure mask can be generated for each layer. The STL inter-face has already become a standard de facto taken over by nearly all producers of CAD sys-tems.

•Prediction and compensation of object deformation

Potential defects such as shrinkage, warpage, airtraps and distortion are difficult to anticipate during the part design process. Other problems may occur due to the small size of the objects and the low viscosity of the resin. A knowledge based system could be used to predict and compensate such effects. The new exposure mechanism allows us to adapt the exposure time of each pixel individually in order to darken bright segments earlier as segments which are not well illuminated.

•Control of the whole manufacturing process

The results of the slicing process are transmitted to the controller of the elevator and the ex-posure unit of the system where the part is produced layer by layer.

SUMMARY AND CONCLUSIONS

A new technique for the fabrication of micromechanical objects called micro stereolithography was proposed.An exposure pattern which is displayed on a LCD is mapped in reduced scale to the sur-face of the resin.This reduces the mechanical effort very drastically since there are no more moved elements apart from the elevator. Several experiments have been performed in order to demonstrate the capability of the micro stereolithography process to manufacture microparts. The limit of the micro stereolithography process lies presumably at structures with measurements considerably smaller than 0.01 mm.

The results obtained have shown that the principle of stereolithography is usable in packaging and interconnection technologies (PIT) in microsystem technology. Therefore, the advantages of solid freeform manufacturing are usable in order to reduce time and cost of microsystem development:•Freedom in geometrical shape

There are almost no limitations to the complexity of the parts. Parts with intricate internal shape, tiny details, workpieces with undercuts and hollow spaces can be produced directly from three-dimensional CAD data without tools.dark regions are the solidified resin. The layer thickness is about 0.08 mm. The soil and the small damages on the surface of the sample result from the removing of the remaining unsolidified resin after the exposure. Furthermore, fine horizontal lines are recognizable on the sample. These lines result from the pixel structure of the LCD. Theoretically each illuminated pixel results in a solidified tile which is separated from its neighbours. But fortunately the experiments have shown that neigh-bouring tiles are linked together since the illumination borders of each pixel are not exact enough to preserve the gap.

The experiments proved the suitability of the proposed process to create microparts with small struc-tural dimensions in principle. One has to take into account that the positive results could be achieved with very primitive hardware. Therefore a large field for optimization remains e.g., the use of an auto focus system. Three-dimensional structures consisting multiple layers were not created because that would require an elevator like it is used in well-known stereolithography systems. The limit of the micro stereolithography process lies presumably at structures with measurements considerably smaller than 0.01 mm.

Figure 3:Microscopic photograph of a solidified structure

CONTROL OF THE MICRO STEREOLITHOGRAPHY PROCESS

One of the major reasons for the success of solid freeform technologies in microsystem technologies is expected to be its integration in CIM environments. Automating the whole process from design to manufacture offers a lot advantages. The complete micro stereolithography system can be splitted into the exposure unit, a motor driven system which moves the object along the Z axis and the com-puter system which receives the CAD data and controls the other components. The system should be able to accept data of different commercial CAD-Systems (e.g., VDA-FS, IGES, STEP).

Prior to the manufacturing process, an effective process plan and fixture design are needed according

Figure 1: The functioning of the micro stereolithography process

light transmission

dark transmission

wavelength [nm]

An indispensable condition for the success of further experiments was the optimal adjustment of all components to each other. A resin was formulated which is sensitive in the range of visible green light. Because of the broadbanded sensitivity of the resin to light the use of a filter is required. On the other hand the transition to visible light effected some advantages, e.g., the possibility to use a cus-tomary photographic objective for the mapping of the LCD.

Separate layers have been produced with the outlined hardware configuration. The main goal was to achieve a solidification of the light regions within a bearable period (below one minute) whereas the dark regions should stay liquid. The solidification of the dark regions after a certain period of time is caused by the effect that the LCD is not able to block the light completely but allows always a low light flow (dark transmission). One result of the performed experiments is shown in figure 3. The microscopic photograph displays a grid which has been created with 22 seconds exposure time. The1)The higher resolution can be achieved by stronger focusing the laser beam and more accurate

positioning of the scanning mirrors. This possibility which has led to first encouraging re-sults [Zissi 94] involves only little risk because it is based on the relatively good controlled point-by-point exposure. But one particular disadvantage is the further increased mechanical effort. To limit this effort in [Zissi 94] is proposed to move the part with the whole resin vat under the fixed laser beam. However, for us it seems to be very difficult to ensure the re-quired plane surface of the resin because the vat is moved during the exposure.

2)A higher resolution can easily be reached by mapping an exposure mask in reduced scale to

the surface of the resin. Under the mechanical viewpoint this solution requires only a more precise part positioning system.

The micro stereolithography process described in this paper is based on the second suggestion. A further advantage is the possibility to use a Liquid Crystal Display (LCD) as exposure mask. This reduces the mechanical effort very drastically since there are no more moved elements apart from the elevator[Reithofer 94].

THE MICRO STEREOLITHOGRAPHY PROCESS

The functioning of the micro stereolithography process is depicted in Fig. 1. The radiation is gener-ated by a rod-shaped lamp (e.g., a halogen lamp). A Liquid Crystal Display (LCD) is used as expo-sure mask. To achieve a homogenous illumination of the LCD consisting of parallel rays of light, a optical system has been developed which is based on a Fresnel lens. Since the contrast of the LCD depends heavily on the wavelength of the radiation a special coloured filter is needed to narrow the radiation spectrum. The LCD is controlled by a micro computer and displays the cross section pat-tern of the layer which should be exposed. An objective maps this pattern to the surface of the resin. Variants are possible e.g., a zoom lens allows to adjust the mapping scale to the desired resolution. The exposure of larger cross sections can be performed partly in a stepper process (Fig.1). DESCRIPTION OF THE EXPERIMENTS

We performed several experiments in order to demonstrate that the micro stereolithography process is suitable to manufacture micromechanical parts. First we experimented with a commercial resin. Its maximal sensitivity to light is in the range of about 355 nm. We used ultraviolet fluorescent tubes as light source. A monochrome LCD with a resolution of 0 x 480 pixel (pixel size about 0.3 x 0.3 mm) served as exposure mask. The LCD is normally used for overhead projection. It is connected to the video port of a personal computer.

It turned out that the used LCD does not allow any transmission in the ultraviolet range (wavelength shorter than 390 nm). Fig. 2 displays the exact analysis of the transmission spectrum. Clearly recog-nizable is that no transmission with wavelengths shorter than 420 nm is possible. It is also obvious that the maximal contrast which is the ratio between light transmission and dark transmission is reached at a wavelength of about 530 nm (visible green light). Furthermore, we performed transmis-sion measurements with other LCDs e.g., active matrix (TFT) displays. However, in any other case the maximal reachable transmission was lower than 5%. Compared to the transmission of the mono-chrome display this is only a quarter. Therefore the power of the light source would have to be multi-plied by four to illuminate the resin with the same luminous intensity. Consequently we kept the monochrome LCD and chose 530 nm as the new characteristic wavelength.

kind of functions they perform. Microsystems are combining microtechniques (e.g., micro-electron-ics, micro-optics, and micro-mechanics) functionally. There is a wide field of applications such as micro-actors like valves and pumps which can be transplanted in organisms. Piezoceramic tubes can be used to miniaturize mobile robots.

The use of packaging and interconnection technologies (PIT) in microsystem technology is charac-terized by the need to advance the miniaturization, complexity and reliability of microsystems. Highly complex microtechnologies are utilized with system technologies to create ideal electronic and mechanical connections, to control the thermal behaviour of the entire system or to establish suitable electronic and optical signal fluxes. Over the last few years a number of concepts and ideas for microsystem technology have been developed. However, the transformation of all these activities into marketable products still leaves a lot to be desired.

One of the reasons for this delay is the lack of suitable packaging technologies. Furthermore, manu-facturing technologies in microsystem technologies largely depend on PIT, a fact that is also reflected in the cost relations. Today, packaging and interconnection, testing and trimming account for 70 percent of system costs at the moment. There are some techniques for the production of microparts (e.g., LIGA-Process) which are very expendable and expensive. Conventional production methods are not able to meet the stringent requirements of accuracy. This fact clearly illustrates the great strategic relevance of advanced PIT concepts. Thus, research and development in PIT is increasingly concentrating on methods and concepts that will facilitate the flexibility of manufactur-ing processes, thus aiming at cost optimization and a better system performance at the same time. SOLID FREEFORM MANUFACTURING

Several new manufacturing technologies which are building up the workpiece by adding layers of solidified material appeared in recent years. These technologies facilitate the realization of complex geometries, e.g., workpieces with undercuts or hollow spaces, directly from three-dimensional CAD data without any tools. The best-known generic terms for these new technologies are Rapid Proto-typing or Solid Freefom Manufacturing (SFM). These terms emphasize two fundamental aspects:•The technologies are mainly used to create a prototype quickly.

•It is possible to generate workpieces with nearly arbitrary geometrical shape.

A classification of various rapid prototyping technologies and a description of the basic principles can be found in e.g., [Kruth 91] or [Reithofer 93]. The stereolithographic technologies form a subset of the rapid prototyping technologies. They build up the model by solidifying a liquid resin which is sensitive to light. The solidification is initiated by the exposure with light of a certain wavelength. Commercial stereolithography systems can be split up into two groups:

1)Systems exposing the surface of the resin point-by-point with a bundled laser beam (e.g.,

SLA systems, STEREOS), and

2)systems exposing the resin layer-by-layer through an exposure mask (e.g., Cubital).

The prototypes are used as design models for the illustration of blueprints, as function models to prove the functionality or as a first model for subsequent casting processes. In future SFM will not only be used for the creation of models but also for the production of final products [Kochan 94]. Limitations to the use of stereolithographically created parts in micro systems result from the proper-ties of the resin and the structural resolution. In general, today’s commercial systems reach resolu-tions in the range of 0.1 mm. That is not enough. Therefore it is absolutely necessary to increase the maximal reachable resolution to open up this new field of application. There are two possible solu-tions corresponding to the exposure variants mentioned above:MICRO STEREOLITHOGRAPHY

USING A LIQUID CRYSTAL DISPLAY AS EXPOSURE MASK

ISATA Paper 95RP055

Thomas Cord

FZI – Forschungszentrum Informatik

Dept. Technical Expert Systems and Robotics

Haid- und Neustr. 10-14, D-76131 Karlsruhe, Germany

Walter Reithofer

University of Karlsruhe, Faculty for Informatics

Institute for Real-Time Computer Systems and Robotics

Kaiserstr. 12, D-76128 Karlsruhe, Germany

ABSTRACT

The automotive industry is of key importance in all European countries, in the USA, and in Japan. In order to fulfill the technical and economic requirements of future cars, a consequent utilization of technological developments is necessary. Of paramount importance are miniaturization technologies for sensors, microelectronics and their integration into microsystems. At the moment, a micro fabri-cation technique for three dimensional structures with high resolution is demanded. Stereolithogra-phy allows to manufacture 3D objects made of polymers directly from CAD data. The basic principle is the solidification of a liquid resin which is sensitive to light. The technology tends to be limited to manufacture small objects (e.g., micromechanical parts). Nevertheless, it seems important to broaden this principle to manufacture elements directly usable in microtechnology.

The micro stereolithography process described in this paper is mapping an exposure mask in reduced scale to the surface of the resin. A further advantage is the possibility to use a Liquid Crystal Display (LCD) as exposure mask. This reduces the mechanical effort very drastically since there are no more moved elements apart from the elevator. The radiation is generated by a rod-shaped lamp (e.g., a halogen lamp). The LCD is controlled by a micro computer and displays the cross section pattern of the layer which should be exposed. An objective maps this pattern to the surface of the resin. Several experiments have been performed in order to demonstrate that the micro stereolithography process is suitable to manufacture micromechanical parts.

INTRODUCTION

A micro-engineered device is defined as a miniaturized operational assembly, whose dimensions range from submicrometers up to millimeters, integrating electrical, mechanical, optical, chemical or other functionality (e.g., sensing, micro-mechanical actuating, packaging) for specific industrial applications and new products. Microsystems are gathering and processing information from their environment and are performing actions independently. They are often highly diverse in number and下载本文