Guest Column | February 19, 2015

3D Printing For Medical Device Manufacturing: In-House Vs Outsourcing


By Andres Bernal, Founder, BIONIKO Consulting

Over the past few years, 3D printing has been a hot topic in mainstream media. President Obama even discussed 3D printing during his 2013 State of the Union address, soon after the creation of the National Additive Manufacturing Innovation Institute (NAMII) in Ohio. This hype inevitably resulted in increased awareness from consumers and common investors, both groups waiting for the next industrial revolution to unfold before their eyes, fueling a bubble that peaked early last year. However, consumer patience is short, and the hype faded as 3D printed objects did not instantly become commonplace consumer products, and consumers did not instantly become designers. Likewise, the hype has faded for the investor and mainstream media. Nevertheless, the revolution is still underway.  

For those in the manufacturing sector, news of 3D printing technologies was not new. 3D printers had been around for decades and were primarily used in research labs. As was the case with mainframe computers, it made little sense for the majority of companies to invest in such expensive equipment to meet their needs. However, entrepreneurs realized that the prototyping needs of many companies could keep one, then two, and then 20 3D printing machines busy, creating a profitable business model in the process. Thus, the 3D printing service bureau was born — and then it multiplied.

Now, a large selection of companies specializes in providing 3D printing services on demand. The major bureaus have an impressive array of technologies and finishing processes to choose from, and there are specialty bureaus focused on specific technologies or finishing processes. Bureaus have multiplied into the hundreds and practically commoditized 3D printing services. As a user, you can upload your CAD files to an online wizard, choose from a variety of materials and finishes, and receive quotes almost instantly. You enter your credit card details, and after a couple of days you receive your parts in the mail, ready to go. It couldn't get any easier.

While bureaus multiplied to satisfy the prototyping needs of every industry sector and geographical location, 3D printing was undergoing a gradual transition from the rapid prototyping domain to the additive manufacturing domain. New machines allowed for more durable materials, larger build envelopes, faster build times, and increased detail. At the same time, designers and engineers began to understand the competitive advantages to be gained by capitalizing on the complex forms enabled by 3D printing, the ability to mass customize, the flexibility of tool-less production, and the advantages of built-in assembly, to name a few.

Medical device companies are now looking to expand their use of 3D printing beyond prototyping and incorporate it into their manufacturing processes. However, does this necessarily translate into bringing a 3D printer onto the manufacturing floor?

There are many good arguments to convince management to invest in a 3D printer. Design engineers can benefit from fast and inexpensive iteration and feedback from a 3D printed object. A functional prototype can be valuable for performance testing. Meeting with marketing, management, or even clients can benefit from aesthetically accurate prototypes. And why not directly manufacture some components, jigs, or tools? All are good reasons, but a diligent buyer will quickly discover that there is no single 3D printer capable of making all these different types of parts efficiently.

There is a dizzying array of 3D printers on the market today, but most are based on three main technology families: powders, extrusion, and photopolymers. Within each of these groups are many flavors, with different combinations of build envelope, resolution, footprint, cost, material selections, etc. Hence, any given design has a particular 3D printer that is best suited to make the part for its intended application. Machines that build inexpensive prototypes may not be suitable for aesthetic prototypes. The machine that produces aesthetic prototypes may not be able to print a functional, load-bearing part. The machine that prints functional parts does not print fine detail. The machine that prints fine detail prints only small parts.

Plastic or metal, detail or strength, cost or performance — these are all tradeoffs that need to be considered when choosing a 3D printing platform. Then again, why purchase one type of 3D printer if the bureaus have them all? Why not build your parts with the printer (and with the bureau) that best suits the current need?

To be able to replicate the capacity and versatility of a 3D printing bureau is certainly cost prohibitive for all but the largest manufacturers. Even if they focus on a single printing technology, a bureau may still be the most cost-effective choice. In addition to the initial capital expense, production-grade 3D printers (we’re not talking about MakerBots here) require floor-space, maintenance contracts, skilled operators, material handling, and finishing stations — and they depreciate quite rapidly. 

Let’s assume for a moment that your engineering department has developed an innovative design for additive manufacturing, and has identified the 3D printing system that best fits the application (by first trying it out with a few bureaus). You will use the same machine to produce similar types of printed parts, and you forecast needing a steady volume of these parts to keep the machine busy. You do the math and show the board it makes financial sense to buy the same 3D printer the bureau used to make your pilot parts. The board members unanimously agree, and with a snap of the fingers it gets done.

The machine finally arrives and is installed. Your team gets training, you get material, you print your first batch of parts, and … they don’t look like the parts you received from the bureau. Your parts have support structures that are difficult to remove, the surface finish is poor, the mechanical properties are anisotropic, the transparent part is opaque, and fine details are lost during finishing. You are left wondering if it may need post-curing, heat treatment, or some unknown secondary operation. What do you tell the board now?

Bureaus have a lot of experience. 3D printing is not as plug-and-play as the media or some salespeople will have us believe. Aside from machine settings, build orientations, and other hardware/software optimizations that may affect print quality, parts rarely (if ever) come out of these machines ready to use. They must undergo varying degrees of post-processing. This entails all the methods required to finish the part from the time you take the build out of the machine  — from support removal to polishing, plating, sintering, etching, painting, machining, and in some cases using the 3D printed parts as masters to produce derived parts in casting or molding processes. It is in the post-process methods where bureaus have a big edge. Working on so many different projects over time gives them tremendous experience and process control, not only in 3D printer settings but in the finishing processes. This is especially important when handling complex and detailed geometries, which are usually the ones that justify 3D printing as a manufacturing process in the first place.

In short, bureaus can make your 3D printed parts cheaper and better than you can. However, if you see additive manufacturing as a potential game-changer in your particular industry, you should seriously consider bringing a 3D printer in-house soon. (Wait … what?) If your long-term strategy involves exploiting the advantages promised by additive manufacturing to make end-use parts or parts that work within your manufacturing process, you may want to absorb the learning curve cost, if at all possible.  After you have identified the technology that best suits your needs (by first engaging some friendly bureaus), becoming an expert in the technology of choice will eventually give you an edge over the bureaus and your competitors in your particular application. The same factors that make the bureaus so much better than you today will become the factors that make you better than your competitor tomorrow.

It's important to note that, for the most part, bureaus are 3D printing generalists. They are experts at the 3D printing process itself, but not in any particular 3D printing application. In other words, although bureaus may be specialized in some type of 3D printing — medical, small, cheap, transparent, metal, casting, etc — they are not experts in specific applications, like heart valves. It is their clients that come up with the designs to satisfy needs within a specific market. Companies (or expert individuals) have the know-how to identify a need and translate it into a product concept and then a design. But can you develop an optimum design if you don't understand/control the manufacturing process?

There is a lengthy body of knowledge regarding best practices and DFM (design for manufacturing) rules to follow when designing for conventional manufacturing methods such as molding, stamping, machining, etc. Hence, it is commonplace to design a relatively optimized part for these processes without needing to have the capability in-house. Take, for example, injection molding. There are several software optimization tools based on geometry analysis, finite elements, and mold flow simulations that can help you optimize your part for manufacturing. However, a good designer is one that really understands the DFM rules for a given process and therefore optimizes the design with the best balance of manufacturability (cost) and performance.

Today, the body of knowledge in 3D printing is very superficial and generalized in comparison to conventional methods. Specific geometrical constraints or DFM rules are difficult to define, since ideally a 3D printed part has no geometrical constraints. Known constraints are machine specific and deal more with build envelope, resolution, wall thickness, speed, and material used, but there are no geometry or complexity boundaries, at least on paper. When you send your pilot design over to a 3D printing bureau, there will be little, if any, manufacturability feedback. The bureau will make a "best effort" to manufacture according to your .STL file. If the part works as intended with only a few iterations, great. Chances are that most bureaus can make it, and you can shop around for the best pricing.

However, if the part is truly innovative, it will likely require a great deal of optimization. Optimization relies on feedback mechanisms, and at the end of the day, the bureau doesn't know what the purpose / use-modes of the part are — and you probably don't know how they make it. There is a feedback block, where you cannot help them manufacture it more efficiently by modifying the design, and their lessons learned in manufacturing are not incorporated into your design. Of course, you can always work with contract manufacturers to bridge that gap, but this entails greater dependence on a particular vendor, not to mention intellectual property (IP) considerations.

3D printing should be chosen as a manufacturing method when conventional manufacturing methods are not feasible or cost effective, most commonly with complex and innovative geometries. These geometries could pose design and manufacturing challenges that may transcend the capability of a bureau. If a design cannot be build to spec or efficiently by a bureau, it is tempting to abandon the idea. However, for some applications, this could be viewed as an opportunity to bring 3D printing in-house and try it on your own. When you control the whole process from design to manufacturing to post-process, the feedback loop is fast and powerful. Optimization comes in many forms: cost, throughput, performance, quality, post-processing methods, etc. You can experiment and iterate many design factors and options in parallel, and quickly understand how design decisions and features affect all these factors. Hopefully, you will discover methods and best practices unique to your part and application, which may constitute a more important form of IP than the geometry itself.

In industries like medical, there are additional reasons beyond technical ones to bring 3D printing in-house. This is a regulated industry in which quality control is of utmost importance, and federal agencies require validated manufacturing processes and controls. Again, working with a contract manufacturer may be feasible, but in the long run it is desirable to develop and retain the know-how in-house. Validating 3D printing processes for manufacturing is one of the main hurdles in the path to commercialization for printed medical devices, but then again, it may present an opportunity for those willing to bear the risk and undertake the challenge.

In a very generalized conclusion, if your purpose for a 3D printer is prototyping a varied assortment of part types to fill different needs (aesthetic, functional, tool, jig, one-off end use) and your end-use parts can be made with a conventional subtractive or molding manufacturing method, you do not need to have control over the 3D printing process. You need the best performing/looking parts fast. Basic (desktop) 3D printing capabilities and a list of trusted bureaus is the way to go.

If, on the other hand, you have identified an opportunity to use 3D printing to solve a manufacturing challenge, first use bureaus to explore printing technologies and materials. If the results are promising, bring the technology in-house for further development and integration of design, printing, and post-processing. The biggest bureaus should not have a problem with helping you get started with a machine of your own these days, since some of the major ones are owned by the companies that make and sell the printers.

It may take a couple more years, but as companies bring 3D printing capabilities in-house, they will be able to take full advantage of the design possibilities it affords, make it part of their continuous improvement process, and inevitably create the innovative 3D-printed products that patients and end-users have been waiting for. Those who wait for additive manufacturing to “make sense” will find themselves trying to catch up to those who took a risk when it didn’t.

About The Author
Andres Bernal is a biomedical engineer and founder of BIONIKO Consulting LLC. He provides consulting services in design and additive manufacturing applications in the medical field. Bernal has been actively designing for additive manufacturing technologies for almost 10 years. Aside from providing consulting services to industry and academia in the use of additive manufacturing technologies, he conducts original research in applications of multimaterial 3D printing for medical modeling.

Image credit: BIONIKO Consulting LLC