Stereolithography for microparts?
Recent movement in the medical, aerospace and consumer-products industries toward smaller parts that incorporate microscale features and exacting tolerances would seem to eliminate stereolithography (SL) as a prototyping option for development engineers. Not so!
Granted, the process—developed more than 20 years ago—historically has served as a time- and money-saving tool for producing macroscale prototypes to relatively relaxed tolerances. And, it’s true that the technology’s early years were marked by poor material properties and inadequate accuracy. Those issues are now behind us.
SL just needs a few measured nudges to produce detailed, accurate microparts. And, it’s worth noting, SL technology has not yet reached its technological limits.
Stereolithography is an additive manufacturing process that builds up parts one layer at a time. The thinnest layers today are typically 0.002". CAD-generated part design data are electronically sliced into 2-D cross sections, which are then imaged onto the surface of a liquid resin by a UV laser. Wherever the laser contacts the resin it initiates polymerization, which solidifies the resin (see Figure 1 below). As each layer solidifies, it builds upon preceding layers, adding height to the part. The SL machine then adds fresh resin for the next layer and repeats the process until the full height of the part is reached.

Figure 1: Schematic of the stereolithography process.
The limitations of SL technology stem from the two main processes taking place in the machine: the laser’s movement in the X/Y plane of the part, and application of fresh resin to form thin layers that stack up in the Z-direction (see Figure 2 below). Currently, the Viper si2 from 3D Systems Inc., Rock Hill, S.C., is the SL machine capable of producing the finest details. In high-resolution mode, the machine can focus its UV laser down to a diameter of 0.004" (100µm). It can solidify features such as walls or ribs as thin as 0.004" with outside radii as small as 0.002" and can reliably produce negative features (holes and slots) as small as ~0.007".
Figure 2: Cross section showing thicker first layer in standard SL resins.
In the Z-axis, the machine is mechanically reliable down to a layer thickness of ~0.001". The most significant limitation in the Z-axis, however, is the material itself. SL resins, to date, have been formulated with an eye toward speed—to allow the machine to build as quickly as possible.
This emphasis on speed, however, is counter to requirements for fine features in the Z-direction. It results in a deep cure depth in the resin—in other words, thick layers. A conventional resin formulation will have a minimum first layer thickness of 0.012" to 0.016", even as subsequent layers are only 0.001" thick (Figure 2 above).
In order to take current state-of-the-art SL machines to the next level—i.e., to meet microscale part-feature production requirements — modification of two key systems are needed: the focusing optics of the laser and formulation of the resin. The laser focus can be made finer to allow drawing of smaller features with tighter tolerances, and the resin can be formulated to produce a thinner initial layer.
Lasers for industrial use generally output a beam having a diameter from 0.040" to 0.050". This is far too large for fine drawing. To tighten the focus, a beam expander is first used to make the spot large, then the beam is passed through a powerful lens that angles it down to the minimum diameter at the drawing plane (Figure 3 below).
Figure 3: Beam expander employed to focus beam to smaller spot.
The size of the spot at its waist (the point where beam radius is smallest) depends on several variables, most notably the expansion ratio. To get a smaller spot, users must start with a larger expansion. That sounds simple enough, but it might require a new beam expander, larger and more-powerful output lens, larger scanning mirrors or all three. However, once the needed modifications are done, there is almost no limit to how small the waist can be. Thus, the modified SL machine can make smaller and smaller features in the X/Y plane.
But getting down to a small laser spot only solves the problem in the drawing plane of the process. It doesn’t help the situation in the Z (or layering) direction. The issue here is the photospeed of the resin used—that is, how fast it cures when exposed to the UV laser.
Stereolithography resins behave much like photographic paper. There is a certain exposure they must receive in terms of laser energy for the liquid monomer to polymerize and solidify. This exposure is called “critical energy,” or EC. Associated with EC is another parameter, “depth of penetration,” or DP. This is the resin depth that will turn solid once the surface receives the proper critical energy. Most commercially available SL resins have a DP that allows rapid building of parts—in other words, they cure to a relatively thick depth.
With microparts, however, a deep cure is not desirable. It leads to the phenomenon shown in Figure 2 above. So, making parts with small features inthe Z-direction requires a custom-blended SL material, one that minimizes DP so the resin cures upon exposure but does not solidify the material too deeply. A resin with DP in the 50µm range is ideal for producing microparts.
With these modifications, an SL machine can build parts with extremely fine details. Notice the detail in the opening photo of the chess set image below. The parts on the pencil eraser include a microgear and a tetrahedral scaffold with 75µm connections between nodes.
All images courtesy FineLine Prototyping
Do these parts represent the ultimate limits of the SL process? Certainly not. It is only a matter of time before further improvements in beam focus and resin formulation eclipse these results. µ
About the author: Rob Connelly is president of FineLine Prototyping Inc., a Raleigh, N.C., company offering rapid-prototyping services including SL, selective laser sintering and 3-D printing. Telephone: (919) 781-7702. E-mail: rob@finelineprototyping.com. Web: www.finelineprototyping.com.
- 830 reads





