Returning to the classification scheme, machines that use lasers constitute a broad class of systems that cut across categories of the matrix presented here. (The laser/galvo systems make up a subgroup of these systems.) This class is shown in Fig. 9.7.

Fig. 9.7. Systems with lasers.

Spot Size

Table 9.2 shows the spot size achieved in various laser-based RP machines, including both mirror and gantry types of machines. As can be seen, most systems have spot sizes of 150-200 or 250 microns. There are a few exceptions, notably the DTM machine that has a significantly larger spot size (400 µm), and two machines with significantly smaller spot sizes: a 3D Systems Beta machine (80 µm) and the Nagoya photomolding machine (5 µm).

Table 9.2
Spot Size

* older, smaller machine has variable spot size
** variable spot size, 130-1500 microns

An important source of perspective on the data of Table 9.2 is to recognize that there is a fundamental lower limit to the spot size that can be achieved from a collimated light source. This is given by the expression

Minimum Spot Size = (wavelength * focal length)/diamter

where lambda is the wavelength of the light, f is the focal length of the focusing lens, and d is the diameter of the collimated beam that enters the lens. Thus, the smallest focal spot will be achieved with a large-diameter beam entering a lens with a short focal length.

As noted above, in a laser/galvo system, there is an advantage to moving the mirrors (and therefore focusing lens) far from the imaging plane in order to achieve better control over the size and shape of the focal spot. However, as can be seen from the equation above, a longer focal length acts to increase the minimum spot size that can be achieved. In order to compensate for this effect, a beam expander may be placed in the optical path before the focusing lens. This solution is employed in many laser/galvo systems, as shown in Fig. 9.8.

Fig. 9.8. A laser/mirror system with a beam expander in the optical path.

This strategy, too, has its limits. As the beam diameter is increased, the mirrors in the galvo system must be increased in size and mass, potentially resulting in a loss of performance. However, most laser/galvo systems do use a beam expander in order to bring the focal spot down into the typical range of 150-200 microns.

The larger, 400-micron spot size of the DTM machine can be understood by noting that no beam expander is used in this machine. The laser sintering machine at the Bavarian Laser Center uses a concentric dual beam. The larger beam preheats and anneals the powder, while the inner beam performs the sintering.

A standout on the smaller end of the scale is the small, 80 µm spot size of the 3D Systems Beta machine, which is a retrofitted SL250 machine in beta testing. In this machine, several modes of the HeCd laser are filtered out, resulting in a beam that can be focused to a smaller spot size. In the process, most of the beta energy is lost.

Achieving a small and well-controlled spot size is less difficult in a gantry machine. For example, in the LOM (laminated object manufacturing) machine, the focusing lens is carried on the gantry and is much closer to the imaging plane (approximately 10-15 cm distance) than it is in a galvo mirror machine. For this reason, no beam expander is needed before the focusing lens. Further, there is no change in angle of incidence of the beam, and hence no distortion of the shape of the beam. The photomolding machine at Nagoya is intended for MEMS (microelectromechanical) application and so uses a very small spot indeed. The gantry construction of this machine helps in achieving this small spot size.

A possible strategy with a laser machine is to change the size of the spot, depending on the geometry being imaged. For example, a small spot size might be used to scan the perimeter of a part in a vector motion with fine detail, followed by use of a large spot to achieve a raster fill of the geometry. An older DMEC machine and a CMET machine have such capability.

Recoating in Laser Photolithography

Many machines in the laser class are photolithography machines (both vector and gantry), and recoating is an important aspect of the design and performance of these machines. The technique used can influence (1) the range of resin properties that are allowed in a machine, (2) the flatness of the resin surface produced, and (3) the speed of recoating. Recoating has been an important source of differentiation among competing laser photolithography products. There are essentially four approaches used in systems that build parts in a vat of polymer: deep dip, inverted "U," viscous retention, and positive displacement (Fig. 9.9).

The "deep dip" approach is perhaps the most widely used recoating method. As shown in Fig. 9.9a, the part is lowered into the vat by a distance greater than the intended layer thickness to promote the flow of a viscous resin over the surface of the part. The part is then raised to the height that it should be when the next layer is imaged, and a mound of photopolymer is created above the free surface of the resin in the vat. A blade then moves across and wipes the excess resin off to the side, leaving the desired layer thickness. This method is a substantial improvement over the previous method where gravity was relied on to level the surface (this was slow); however, it can be sensitive to part geometries such as that shown in Fig. 9.9a where there is a "trapped volume" of resin. As the blade wipes across, the resin in this volume may be shifted, with the result that the layer thickness after equilibration is not quite as desired. This method is used by 3D Systems and DMEC (DMEC's blade flips up at the end of a pass).

Fig. 9.9. Recoating methods in laser photolithography.

Fig. 9.9b shows the "inverted 'U'" approach, where a coating device is filled with resin above the free surface of the vat, and this resin is then dispensed as the coater traverses the vat. Filling of the inverted "U" can be by capillarity, electrostatics, or vacuum. The trailing edge of the recoater acts as a doctor blade. This method is relatively immune to the problem of trapped volumes and can be faster than the deep dip, since no stage is strongly dependent on gravity-induced flow. CMET uses capillary filling of the recoater (although there may have been a communication problem on this issue). Fockele and Schwarze uses the electrostatic method, although it apparently does not work equally well with all polymers. 3D Systems uses the vacuum method.

Fig. 9.9c shows the method of "viscous retention." A brush or mesh (depending on the resin) is supported between two doctor blades. While the layer is being imaged, this device is submerged in the vat. When it is time to recoat, it comes up and over the surface of the vat, and the material drains out by gravity as a rate determined by the viscosity of the resin. The final surface is created by the trailing doctor blade. Teijin Seiki practices this method.

Fig. 9.9d shows a positive displacement pump used to bring material up for the new layer. This method is used by EOS.

[Previous Section][Top of Report][Send Your Comments][WTEC Welcome Page][Next Section]

Published: March 1997; WTEC Hyper-Librarian