The most important class of machines being developed in Europe and Japan are machines that scan a laser using galvo mirrors. Both vector-outline and raster-only types exist within this class, as shown in Fig. 9.3.
In the very simplest case, this type of machine consists of a laser, a focusing lens, and two axes of galvo mirrors, as illustrated in Fig. 9.4. Examination of some of the critical elements of this system and some of the issues that lead to more complexity in this system provides some insight into the equipment being built today.
Fig. 9.4. The simplest configuration of a laser mirror system.
The performance of the scanning galvo mirrors is a factor that limits the performance of the system, as the galvos mandate a trade-off between speed and accuracy in the imaging of a layer. Since this machine element can be partly responsible for the competitive positioning of RP equipment, a crucial question is whether to make the galvos or to buy them. Both decisions are represented in the market. 3D Systems exemplifies both decisions. For the SLA-250 the company started by buying General Scanning mirrors. When it went to the SLA-500, it decided to build its own mirrors because it couldn't get the desired performance in off-the-shelf mirrors. 3D Systems continued to build custom scanners for the SLA-350. Teijin Seiki builds its own scanners in its own plant (the plant is not dedicated to scanners but has a significant effort in building scanners). EOS builds its own scanners. DTM uses General Scanning scanners. This company has ridden the curve of improvement as General Scanning has improved its products -- at least partly in response to the requirements of RP vendors.
The general trend is that the increased accuracy of materials (especially photopolymers) has driven machine designers to improve the accuracy of the mirror scanners. They have responded either by working with an established vendor such as General Scanning or by building their own scanners. When they build their own scanners, they also build their own control electronics, which are often based on digital signal processing.
Table 9.1 shows the stated accuracies for a number of galvo systems at specified scanning speeds. This data is derived from a combination of published specifications and conversations. It is important to note that it is very easy to confuse accuracy (proximity to target) and precision (repeatability) in conversation and probably in specifications as well. It is also easy to confuse static and dynamic accuracy. Thus, some uncertainty must be acknowledged concerning the numbers presented in Table 9.1; nonetheless, the consistency of the data from different vendors lends credence to the numbers. The Teijin Seiki machine is a raster-only machine, and the higher accuracy specification can be understood by recognizing that such a machine does not have to execute coordinated motions of two axes.
The accuracy of the final part is dependent on both the equipment and the materials. With recent improvements in equipment, the predominant limitation lies with the distortions introduced by the materials systems. This is especially true of processes such as selective laser sintering where there are significant distortions, but it is true of stereolithography as well. Hopefully, future improvements in materials systems will demand that the equipment builders revisit their equipment designs and improve them again, as they have in the past.
Accurate calibration of laser scanner systems is key to delivering the improved systems performances achieved today. If a scanner system has an accuracy of 50 microns over a range of 500 mm, the accuracy is 1 part in 10,000. Such accuracies can only be achieved by calibration of the fully assembled machine. Several methods are in use, all resulting in a calibration table that is captured electronically and used by the machine. 3D Systems puts down a plate with precision-drilled holes. Operators shine a laser through the holes, collect the light, and image it onto a photodetector. The machine is then used to maximize the light transmission, and designers then know the machine settings that correspond to the hole positions. DTM burns holes in a mylar film and then inspects that off-line. EOS exposes photosensitive paper and inspects that off-line.
Consistent Spot Size
Laser/galvo systems pivot the laser beam in order to scan it over the working field. In the most typical system, a lens is placed before the mirrors, as shown in Fig. 9.5. The result is that the beam will be in focus at only one distance along the optical path from this lens. As the beam is scanned over the surface of the working area from the center position, its optical path increases; hence, the laser cannot be in focus at all locations. The effect is more pronounced the closer the mirrors are to the imaging plane, for a given size of the imaging surface, due to the larger angles of deflection of the mirrors.
There are several solutions to this problem. In some systems, the distortion is tolerated. Especially in systems with a large focal spot size, the increase in spot size caused by movement of the beam to the edges is considered to be a tolerable effect. In some equipment, the optical path length is intentionally increased by using a focusing lens with a long focal length. Examples include the SLA 500 and the SLA 350 from 3D Systems, which have a mirror at the top of the machine and which double the optical path back to scanners housed below. The result of this approach is to effectively move the scanners further away from the imaging plane, resulting in lower angles of deflection. Thus, the change in optical path from the center of the field to the edge of the field is a much smaller fraction of the optical path, and the effect on the laser spot size is similarly reduced. This approach has the added benefit of providing a beam that is closer to perpendicular when it is incident on the vat surface, resulting in less distortion of the circular spot.
An alternative solution is to fundamentally modify the optical path by using a flat field lens that is placed after the mirror system. Such a lens system has a different focal length, depending on the position at which the laser hits the lens. The flat field lens is thus able to maintain a constant focal spot size at the image plane. This type of optical element is more costly, however. The JTEC/WTEC panel is aware of only two companies that use this approach: CMET and EOS.
Starting and Stopping
In terms of drawing the scan vector, the optimum situation is one in which the vector drawn by a laser/galvo system is uniform along its length and has the same width near the beginning, middle, and end. If the laser is turned on to full power and the mirrors are simultaneously accelerated, the effect will tend to be like that shown in Fig. 9.6, where the vector is wider near the ends where the laser spot is either accelerating or decelerating and the power density of the laser is higher.
A number of solutions to this problem are currently in use. Teijin Seiki uses an acousto-optic modulator (AOM), which is placed between the laser and the galvo system. With this device, it is possible to do pulse-width modulation of the laser power that goes through the galvos; by varying the duty cycle of the laser power reaching the vat, engineers are able to effectively reduce the laser power during acceleration and deceleration of the beam, resulting in a constant power density at the vat. D-MEC uses a "high-speed light modulator" (probably this is an AOM). In its equipment, DTM does pulse-width modulation of an RF-excited laser.
3D Systems uses an AOM, but only for turning the laser on and off at the beginning and end of the vector, not for pulse-width modulation. As noted earlier, 3D Systems builds its own scanners and control electronics, and these high-speed systems allow the company to minimize the impact of the acceleration and deceleration phases of its vectors.