Going small with 3D printing

The scale of sizes that 3D printing machines can handle is impressive, from huge wind turbine blades to parts that are small, such as two microns. John Kawola, CEO of Boston Micro Fabrication discusses the microscale side of 3D printing.

Microscale 3D printing is the concept of making parts at millimeter scale or even sub millimeter scale. It’s a technology that has been available for five, maybe even 10 years. What’s new is that this technology is being commercialized for industrial use.

Boston Micro Fabrication (BMF) uses a microscale technology that is a variant of stereolithography. It is called projection micro stereolithography, or PµSL. A Boston MIT professor, Nick Fang, developed the idea. Fang had been working on micro fabrication techniques over the years.

A lot of those micro fabrication techniques are not 3D printing. These techniques include embossing, and etching and photolithography. But for 3D printed micro parts, applications include the semiconductor and the MIM space.

Fang’s idea was to try to move into the 3D printing/additive manufacturing space. The concept was to use the PµSL approach, which is similar to other DLP 3D printers in the market. The approach consists of a resin bath, with light to cure the resin. In BMF systems, the light source is focused down to a very fine resolution and the movement is controlled to ensure precision.
Microscale 3D printing fits a market segment of high-value industries that do product development, manufacturing, and production, but with resolution and dimensional accuracy requirements in the millimeter, and perhaps micron area in tolerances, such as electronic connectors.

Other small size production technologies exist, including photon lithography. One criticism of photon lithography is that it takes a long time to make a very small part. For example, a 20-millimeter size part might take a week to build.

The connector market, worldwide, is valued at $100 billion dollars. And connectors are getting smaller all the time. New markets, such as AR/VR for glasses are emerging that suit this technology.

Design efforts continue to try to miniaturize in the optics and photonics market. In the medical device market, things are getting smaller, whether they’re sensors or drug delivery devices, or implantable devices. There’s a trend around industries wanting to get smaller.

Challenges when designing for the microscale market
Design challenges with micro-scale parts are not much different than those for macro parts. If it’s a connector, it needs to have certain features to be a connector, same with chip packaging. It needs to have certain characteristics to hold a chip, or a lens holder, but there are always considerations about how are you going to make it, including how to mold, machine, and/or stamp the part. One of the realities is the manufacturing processes for these smaller parts are more difficult and much more expensive.

For example, if you’re molding something like a computer mouse, that’s not that hard to do. The injection mold for that mouse is maybe tens of thousands of dollars, maybe $25,000. But if you’re making something really small, that injection mold may no longer be $25,000. It might be $250,000 because you need to machine that mold to get all the tolerances that you need, and you need to be able to build in the ports and all the dynamics that go into an injection mold, but at a very small scale.

Uses of micro 3D printing systems
Most users of these micro 3D printing machines are using them for prototyping. Until about two years ago, affordable micro 3D printers were not available for this application. Users could not get the resolution and detail they wanted one for one.

Today, about a third of users are doing some production. Users are testing materials, their design and then looking into production quantities. Some are finding out it’s less expensive to let the 3D printer handle the mold than to machine or stamp them. The mold materials are predominantly resin-based, with some ceramic materials.

What goes into the development of a macro 3D printer
BMF’s chief technology officer, Chunguang Xia, came from the semiconductor industry. One of the challenges was ensuring high tolerance, to create a platform when people are looking to get tolerances in the plus or minus five-micron area. At this level, measurements are done using technologies similar to a CT scanner.

“We needed to have a platform that could image onto the resin and create the part,” says Kawola. “If you’re making a larger part, our platform actually moves to be able to image in certain sections. All of that had to stack up. You’re also dealing with material shrinkage, so you’ve got all these variables that you’re trying to get to a place where you’re plus or minus whatever that tolerance is. I think it’s just a lot of good engineering by our team here to recognize that’s the target.”

For the DLP system, high precision optics were needed. “We’re really focusing down to a certain pixel size,” Kawola continued. “When we talk about, let’s say 10-micron optical resolution, that’s actually the size of the pixel that’s being imaged onto the liquid. Most DLP systems in the market are upside down, so they’re coming out of the liquid. That has a lot of advantages. We are top down. We do that for a couple reasons. One is we need to control layer thickness. The thicknesses of our layers are in the 10-to-20-micron range, so we need to be able to highly control that. Two, all DLP systems, all resin-based systems, when that light comes down and polymerizes the resin, that’s a reaction and there’s heat. Most 3D printing systems, there’s heat being generated and that’s one of the challenges, actually, for a lot of DLP platforms.

“A company in Chicago, Azul, has some new breakthroughs in how to control that heat and try to cool that process as it goes, but it’s a challenge and heat is bad for trying to maintain dimensional accuracy. That’s another reason why we are top down because when you’re bottom up, you only have a thin layer of resin. There’s heat there and that can build up. When the system is top down and you have a vat, it acts as a heat sink, so you don’t really have that buildup of heat that affects most other technologies.”

A materials ecosystem
An ecosystem for materials has been evolving for about five years. BMF is an open platform. “For prototyping, arguably the materials don’t matter that much. They need to be close enough and most people can be pretty happy, but when you get into manufacturing, they need to be really close or exact. I think the collective wisdom of more companies coming into the 3D printing material space, in my opinion, has significantly moved the industry forward over the last five years, at least.

“Now, it’s not just the 3D printing companies with their material science departments within each company. Now you’ve got Henkel, and Covestro, and BASF and Evonik and all these big companies coming in and developing materials to help push that along. We’ve tried to take advantage of that. We’ve tried to shop around and see what’s available and partner with as many people as we can. I think what’s attractive also in that open business model is if resins work on some other DLP printer, and have proven to be successful and attractive to customers, they will probably work on our platform too. We can take things that have been developed. We need to modify them a little bit, primarily because our layers are thinner. We have to tune them essentially, but then we can offer that wider range of materials for the customers.

“The other important thing that we believe in for the open materials model is if big customers are really going to go into production, they’re going to want to have more control over the material supply chain. That’s our opinion. They’re going to want to be involved. I mean, we even see it when we work with some of them. We’re still a small company and when we go to a big chemical company and ask them to do something for us, they’re polite, but they don’t jump, but if we go there with a customer that is already one of their customers on the thermoplastic side, who already buy tons of plastic pellets, they are more willing, I think, to think about working together to develop 3D printing materials for that customer.”

The future of micro 3D printing
There’s been a significant improvement in materials over the last ten years for nearly all 3D printing platforms.

A key trend now is to make sure that all the systems, the materials, and the software are designed with manufacturing in mind so that the systems can be put into a factory and put in series or parallel with other manufacturing technologies, and can work with the quality management system that the company may have.

Users are realizing that for most 3D printing technologies, the very large majority of the cost is not the labor. It’s the amortization of the machine, it’s the materials and the labor’s a relatively small component.

The automation piece is important when it adds real value, but it should be balanced with the cost and flexibility that comes with a person. “A person’s pretty flexible. If he drops something on the floor, he picks it up. If the robot drops something on the floor and it didn’t mean to, it doesn’t know how to pick it up. There’s that balance there. I’ve been in injection molding houses and I’ve been in 3D printing service bureaus where there’s 20 machines running and there’s one or two people. I think the automation will help, but I think there’s still lots of room to go in terms of getting the machines to the point where they may need more automation.”

Kawola thinks that for all of 3D printing, whether it’s resin-based, or filament-based or powder-based, the key is to get those additive manufactured materials, those 3D printing materials to be sufficiently the same as molding materials or machine materials.

“Our challenge and the industry’s challenge is to try to get that part to be close enough to the thermoplastic to be suitable for the application.”