Intrepid Automation: How Investment Casting Benefits From High Speed DLP

3D printing offers a way to achieve mass customization — production in which every part is custom and different. Meanwhile, investment casting is a potentially economical method for straightforward mass production. Can the two be brought together, so casting can be efficiently mass customized?

Intrepid Automation’s answer is yes — so long as investment casting patterns are 3D printed fast enough to keep pace with casting production. The startup in San Diego, California, has developed digital light processing (DLP) systems aimed at production speed and scale. After five years of development, the company introduced its first commercial production DLP systems in 2022, and investment casting is one of the major opportunities it has seen.

Justin Swartz is director of commercial solutions for the company. He describes how mass customization could work in the case of medical implants made via casting. A replacement knee, for example, could have its dimensions fine-tuned to measurements of the patient’s anatomy. The implants themselves would be mass produced in a traditional investment casting operation, but the patterns at the heart of this casting would change piece by piece, assuring the customization.

Tooling is the obstacle to casting with this kind of design variety today. The wax patterns for investment casting (also called lost wax casting) are injection molded. A mold makes the wax pattern that makes the one-use ceramic mold, and the wax pattern melts out of this ceramic mold so it can make the cast metal part. The fixed mold tooling at the start of this chain fixes the design. Using DLP to make patterns instead offers the benefit of removing this mold from the process, and therefore removing the need for fixed part geometry.

Indeed, the freedom to proceed without tooling in this way is potentially valuable even where the design does not have to change for every single part, Swartz says. Intrepid has seen this in discussions with sporting goods producers: “Golf club makers begin developing new club designs two and a half years ahead of the release date,” he says. That is, they have the new club models for three different years’ products in development at a time. Given this long lead, “a new design that takes away a popular feature is a risk with every product launch,” he says, so club designers are cautious. Relying on 3D printing in place of tooling addresses this risk: The design of a cast product could be changed well after the product’s introduction just by changing the design of the pattern. Club design concepts can be adjusted based on customer response.  

For investment casting users, and for foundries performing this work, this mode of patternmaking opens new possibilities. 3D printing patterns is not new — 3D printing them at speed, repeatability and economy suited to production potentially is. Swartz says Intrepid has been talking to leading investment casting providers to explore the promise of augmenting casting with design flexibility in this way. This mode of patternmaking also potentially brings a material change: Intrepid’s system makes patterns from thermoplastic. Thus, the 3D printed replacements for molded wax patterns are not made of wax, and one potential consequence —more on this below — is a reduction in the amount of material used to make the patterns.

Projectors and Printers in Tandem

Intrepid has developed two different DLP systems: its “Range” machine, which prints onto a build surface that descends into the resin, and its “Epic” system, which encloses multiple parallel DLP machines that suspend the 3D printed part from above. The Range model allows for larger parts and batches, as well as printing with elastomers (which would otherwise be vulnerable to the pull of the suspended mode of printing). The Epic system allows for fully automated production of smaller parts.

On the Range machine, speed comes from six light projectors working in tandem, combined with software coordinating their positioning relative to one another and material formulations developed by Intrepid that are tuned for fast build rates in this machine. A pattern for a cast automotive wheel could be produced at the rate of two per build in about four hours to build both parts, Swartz says. The machine also features automatic door opening for robot access, and a robot on a track serving several of these machines is one likely scenario for applying this system in production.

The Epic system operates as a production cell, employing multiple DLP 3D printers in an enclosed unit. The robot serving them is built in. So is the postprocessing — a robot moving on a track at the top of the system is able to load and unload build cartridges from various DLP stations, and also load and unload them from integrated spin units that remove extra resin from the printed parts.

Intrepid plans to offer a development kit for production users of either system to enable close connection between manufacturing and design, Swartz says. The development kit is a non-production machine using the same system as the production units to allow design engineers to trial and fine-tune 3D printed parts. Once the design is finalized in this way, the same 3D printing process proven on the development kit can be sent with no change for use on the production machines — perhaps located in another facility.

Investment casting is far from the only application. Polymer end-use parts can be produced this way as well, including mass-customized products. For example, dental aligners (tailor-made to every patient’s teeth) represent another potential market. An Epic system with eight DLP units is capable of producing aligners at the rate of about 180 per hour, Swartz says — all of them different, but with the production nevertheless requiring no operator intervention.

Hollow Patterns

For the investment casting application, Intrepid is focusing its material development efforts today not just on continuing to advance the material’s suitability to production speed, but also on tailoring it so that it can serve investment casting foundries with as little disruption as possible.

That the 3D printed patterns offer different possibilities comes across tactilely, with the very heft of the parts. Compared to wax patterns, the new patterns are light. 3D printing plus the stronger polymer used to produce them mean these patterns can be made mostly hollow, full of air.

This is not the case with wax patterns. Injection molding produces solid forms. These patterns melt in an autoclave, and one of the cost-saving measures investment casting foundries pursue is capturing and reusing the melted wax.

By contrast, Intrepid’s investment casting material, a thermoplastic rather than wax, is designed to be burned away in a furnace. The material is not recaptured. Swartz says, “Our approach to material savings is to not use the material in the first place.” The hollow 3D printed patterns use far less material than the solid forms they replace.

And yet, investment casters have autoclaves. They use wax. The material change for a pattern is an adjustment the foundry may or may not be willing to make. So, one active area of engineering development for Intrepid is the pursuit of a material that performs more like wax and is suitable to an autoclave. This challenge is still ongoing, Swartz says, but the company believes this new method of investment casting — a toolingless approach that brings design freedom — ultimately can serve even more users if the adoption path can be made as easy as possible.