Accuracy

For example, consider the number of variables involved in exposing a photopolymer with a laser as in stereolithography. While one can precisely determine where the center of the laser strikes the surface of the photopolymer, the situation is a lot more imprecise than when using a cutting tool. The energy from the laser causes a geometrically and chemically complex chemical reaction within the fluid. This is due to the fact that the laser has a non-uniform energy profile and the fluid itself has optical properties which affect the reaction. The photopolymer changes phase and becomes a solid and in the process may shrink, possibly causing internal stresses in the part. The way the polymer cures will be affected by the previous layer and possibly by the time of exposure to the remaining liquid photopolymer. And so on. It's not a wonder that rapid prototyping is less accurate compared to CNC, it's more of a wonder that the accuracy is as good as it is. This is true to a lesser or greater extent with all rapid prototyping methods.

Those RP technologies that incorporate some form of subtractive technology as part of the process, will consequently have greater absolute accuracy because the subtractive technology can be used to correct the compounded errors made by the additive process. For this reason, the inkjet-based method employed by Solidscape has an edge. It can produce parts that approach or equal CNC accuracy. The tradeoffs are speed and materials versatility relative to other RP methods.

Most of the other methods are in the ballpark of "a few mils" Taking stereolithography as a starting point, as shown in the table, one can get at least a qualitative idea. A good rule of thumb is that stereolithography will produce accuracy results of about + - 0.004 inches over a dimension of about six inches. LOM and powder-based methods will generally be somewhat less accurate than that. The inkjet methods mentioned previously, somewhat better. Since the final numbers depend on the geometry of the part, the particular dimension measured, the material used and other factors it's difficult to make a more definitive statement.

A comparison of rapid prototyping technologies, part of a study for NASA's Marshall Space Flight Center [1], was recently published by Ken Cooper, Glenn Williams and Pat Salvail. They used several RP technologies to create patterns for casting an approximately 6 inch diameter diverter valve component. Table 1 shows a comparison of the average of 11 measurements ranging from 0.3 to 5.5 inches, build times and costs.