3D printers may have come on in leaps and bounds in recent years, but most are one trick ponies in that their computer-controlled syringes extrude only one material at a time to build up an object. It's a process that's slow, imprecise, and often requires items to be printed in separate pieces and then assembled. MIT's Computer Science and Artificial Intelligence Lab's (CSAIL) MutliFab printer takes 3D printing technology a step further by combining 3D optical scanning with the ability to print using 10 different materials on the same job.

Multi-material 3D printing already exists, but it's a very expensive process with machines costing in the neighborhood of US$250,000. Also, these machines only use three materials at a time. Part of the reason is that different materials require different temperatures and pressures for extrusion. These limitations usually means items still need to be printed in separate parts and then assembled. In addition, current technology is imprecise and many projects require a certain amount of tweaking and a number of iterations to get right.

According to the CSAIL team, MultiFab uses off-the-shelf components, which keeps the machine's cost down to $7,000. It uses 10 different materials because, instead of melting plastic filaments and squirting them out, MultiFab mixes microscopic droplets of photopolymers and sprays them through inkjet printheads.

This principle also allows MultiFab to gain a much higher resolution of 40 microns compared to typical resolutions of around 100 microns. CSAIL says that it's the first 3D printer to use 3D-scanning techniques from machine vision and is also self-calibrating and self-correcting. During the printing process, MultiFab scans the 3D geometry of components and develops a printing strategy. Using a feedback loop to 3D scan and detect errors for each layer, the system then generates "correction masks" using dozens of gigabytes of visual data.

The upshot of this is that the MultiFab is able to print items in many different materials at the same time instead of in different parts for assembly. One example of its capabilities is that the MultiFab can print a case directly around a smartphone. It can also directly embed circuits and sensors into printed objects, allowing for more complex objects.

The team sees MultiFab as a rapid prototyping tool for designers and manufacturers, as well as on-point manufacturing at the consumer level. It's already been used to produce items such as phone cases and LED lenses, but the team hopes to use it in more complex products, such as microlens arrays, metamaterials, printable fabrics, consumer electronics, and medical imaging systems. One particular long-term goal is to embed actuators and motors for printed robots.

"Picture someone who sells electric wine-openers, but doesn’t have $7,000 to buy a printer like this," says Javier Ramos, a research engineer at CSAIL. "In the future they could walk into a FedEx with a design and print out batches of their finished product at a reasonable price. For me, a practical use like that would be the ultimate dream."

The team's results (PDF) were presented at the SIGGRAPH computer-graphics conference and Ramos demonstrates MultiFab in the video below.

View gallery - 2 images