Additive manufacturing (AM), also called 3D printing, is likely to transform the production of physical goods, including valves and actuators, in the same way the Internet drastically changed the information business model. It is now possible through AM to manufacture goods at locations where they are needed rather than where labor costs are lowest. That kind of model could change the concept of “just-in-time delivery” to “just-in-time manufacture where it is needed.”
The challenge for industry is to find the most economically viable fit of processes and products while maintaining quality, safety and design standards essential to the end use.
THE HISTORY OF THE TECHNOLOGY
Generally, AM technologies employ the process of joining materials to make objects from a three-dimensional model, usually adding layer upon layer to create the product, as opposed to the traditional subtractive manufacturing methodologies.
While the development of inexpensive consumer technologies has brought 3D printing to the consciousness of the general public only in the last decade, the technologies have been around since the early 1980s.
Initially, these technologies were called rapid prototyping, named because the process was originally conceived as a faster and less expensive method for creating prototypes of industrial products. The first patent was issued to inventor Charles Hull in 1986 for a stereolithography apparatus. Hull went on to co-found 3D Systems Corporation, which introduced its first commercial system in 1987. That company is still very active today.
The next patent was awarded to Carl Deckard in 1989 for the process of selective laser sintering (SLS). In that same year, Scott Crump, a co-founder of Stratasys Inc., filed a patent for fused deposition modeling (FDM). Crump’s patent was issued in 1992; and the FDM process is still used by many of today’s entry-level machines.
Not all of the work in 3D technologies was being done in the U.S., however. Hans Langer formed EOS GmbH in Germany in 1989, a company that sold its first “Stereos” system in 1990. It focused on the laser sintering process, including direct metal laser sintering.
The term “3D printing” is generally associated today with hobbyists and consumer-oriented models that use fused deposition modeling, a special application of plastic extrusion. The term “additive manufacturing” comes into play when people are referring to industrial processes. The two processes are generally the same. Whether polymers or metals are the ingredients, the technologies share the common practice of sequential-layer material addition—a joining throughout a 3D work envelope under automated control.
The AM process can either result in a finished product so that it qualifies as direct production, or it can result in a mold or product that requires heat, finishing or assembly, which means it’s considered indirect production. Within each of these general production types, several technologies are used.
Following are the main production technologies and the applications in which they can be applied:
Powder bed fusion can be used with metal or plastic. This is a layer-additive process using a laser to produce functional parts, complex geometries and end-use production parts.
Material extrusion is when a filament, generally thermoplastic, is fed into a heated liquefier where it is turned into a semi-liquid. The material is then extruded by a nozzle mounted to a mechanical stage, which is moved back and forth on computer-controlled tool paths to build parts. These parts, which can vary in size from microns to meters, are very sturdy and functional and have a high strength-to-weight ratio. In industrial applications, the parts are generally used for prototyping form and fit.
Computer numerical code (CNC) is cast machining of polymer or metal parts to refine shapes or surface finishes or to achieve tighter tolerances. CNC machining is incorporated into a hybrid additive manufacturing process that combines laser processes with a machining process. The chips produced by the machining process are often a size that allows them to be automatically recycled into the laser process. Chips that are larger than allowable for laser sintering are swept aside as the next layer is deposited.
Cast urethane is a process in which liquid silicone rubber is poured around a master pattern. The resulting mold is pulled from the pattern and cured, then subsequently used to cast urethane parts.
Binder jetting involves a machine that distributes a layer of powder onto a build platform. A liquid bonding agent is applied through inkjet print heads to bond together the particles. The build platform is then lowered and the next layer of powder is laid on top. By repeating the process of laying out powder and bonding, the parts are built up in a powder bed. Materials used include plastic, metal, ceramic and sand, which form both parts and molds.
In sheet lamination, sheets of paper, plastic or metal are fused together with glue, heat welding or fasteners (bolts or rivets) to make parts.
Vat photopolymerization is a process in which a pre-deposited photopolymer in a vat is selectively cured by an ultraviolet laser beam. This creates cross linking of adjoining polymer chains. This process is commonly used in medicine and dental appliances such as mouth guards.
Directed energy deposition (Figure 1) is where focused thermal energy is used to fuse materials by melting as the materials are deposited. Wire and powder materials are used, which are fused using lasers or electron beams. This process is effective for adding features to already manufactured parts and for repairs.
BENEFITS AND DRAWBACKS
The main advantage of AM over more conventional manufacturing processes for valves is that the need to produce a number of parts using different methods is eliminated. With AM, the entire valve can be produced by the same method so no assembly is necessary. This is possible because AM processes can produce accurate, intricate internal passages and complex geometries.
In a recent presentation, Sheku Kamara, director of the rapid prototyping consortium, the Milwaukee School of Engineering, referred to a fabrication exercise conducted at that university. In 2002, a linear motion valve was designed, then fabricated using AM. The valve’s performance was evaluated and compared to that of a traditional gate valve. The test showed that making moving parts, integral O-rings and threaded connections for the valve were feasible, and the entire valve was prototyped in a single build using a vat photopolymerization process. This illustrated that it is possible to directly build functional prototypes of valves using this method.
Also, by building prototypes this way, it is possible to check flow (whether liquid, gas or oil) through the valve by using translucent polymers, which allow the prototype creator to see what’s going on inside the valve (Figure 2.)
Indirect AM production of valves is also being discussed and used. In fact, Nathan VanBecelaere, manufacturing engineer for the American Foundry Group, says AM manufacturing is the first real innovation for the sand casting process to occur in 200 years (Figures 3-6). While it can be costly in comparison to conventional methods, many steps and much time are taken out of the process.
VanBecelaere points out that there are many advantages to using AM to create molds for casting valves. “There are no restrictions on gating systems, the potential for more feeding capabilities and locations is expanded, and it is more accurate when changing material types,” he says. Also, “while the machines are a set size, the size of the product is not limited by machine capacity. The molds can be made in sections and pieced together to become as large as necessary,” he says.
However, he also points to limitations. “Insulation cannot be printed, although the foundry can insert insulating sleeves by modeling in a counter-bore in the mold to receive the sleeves,” he says. Also, excess sand has to be removed from a mold created with AM, “and, though we are currently undergoing case studies to blend additives with the silica sand, nothing has proven successful yet,” he adds.
While the molding process is definitely shorter with AM, prices can be prohibitive. Dean Markle, project manager–foundry specialist, Emerson Process Management, who was involved in foundries for more than 30 years, offered situations in which it could make economic sense, however.
For example, the models are expensive because they are one-off. “However, if you are replacing a valve that hasn’t been made for 30 years, and you are only going to need this once, it’s much more efficient than conventional manufacture,” he points out. Instead of hoping the 2D model exists somewhere, “We can make a 3D model in a few days, rig and gate it,” then have the valve made shortly thereafter, Markle said.
AM also makes sense for specialty items that are in locations great distances apart, he says. “You can have the computer file sent to a partnering foundry anywhere in the world that uses AM and have that product built in that location,” he says. In other words, “Instead of sending a pattern from foundry to foundry, you have the engineering done in a 3D model and you don’t have the transportation fees to get a model to where you need the valve,” he points out.
According to David Leigh, senior vice president of Engineering for Stratasys Direct Manufacturing, AM can offer many other benefits, including substantial weight savings for many products. It also can allow consolidation of parts or cost savings for low quantities, one-offs and spares. When parts undergo several revisions or customization, AM can mean substantial savings of time and money, allowing the product to get to market faster.
Leigh also said these products can be easier to use and install. “There is no concern about overhangs, undercuts, gating or venting and products can be designed around functionality, rather than tooling constraints,” he says. Because there are no constraints from tooling, “creative contours and internal features can be designed into what can ultimately be a one-part product rather than several parts that have to be assembled,” he added.
James Sears, senior mechancial engineer, additive manufacturing laboratory, GE Global Research, pointed out that AM could also reduce energy use by half, and, since no milling away of excess material occurs, the process can reduce material costs for the right products by up to 90% compared to traditional manufacturing.
While AM offers many advantages, it also has limitations to its use.
As Leigh pointed out, custom AM machines are expensive, machine producers are small and producers cannot support multiple development efforts. Raw materials are expensive, and the process offers limited recyclability for used material. Also, as-built tolerance is less for AM products than products that are tooled. Processing after production is needed for dimensional tolerance and cosmetics, and there can be anisotropy (the property of being directionally dependent as opposed to isotropy, for which a product has identical properties in all directions).
In the case of “just-in-time manufacture” at locations not owned by a patent holder, there also is the issue of protecting intellectual property. This is because, as with other technologies, advances in AM have come about far faster than legal and societal structures can be created and implemented to protect interested parties.
Sears pointed out that getting materials qualification for created products is expensive, and can take five or more years. He said understanding the interaction between the laser and the material being used is critical. Considerations include the location of maximum power absorption inside the powder bed, the impact of particle size and distribution on melt kinetics, and the percentage of laser power absorbed. With AM, distortion also can occur in the thermal process, which can result in cracks or undesirable shapes and fits.
All of these challenges must be weighed against conventional manufacturing, where every part starts with a pre-formed billet that gets reformed and machined. With traditional manufacturing, material properties are known and cannot be changed, whereas with additive manufacturing, the material properties are created as the part is built. That reality can be both positive and negative. In AM, for example, because the properties of the material can be created as the part is built, those properties can be adapted to a specific location.
As with cell phones and many other newer technologies of today, AM technology will continue to improve, and the costs will come down. The process has definitely gone past the point of fascination with the technology, but it’s still suitable only for the research labs to test. Some applications for the technology make sense and are economically and commercially justifiable. It also is probable that if an article like this one is written a year from now, a few valve and actuator companies will have moved this technology out of the lab and onto their production floors.
Editor’s Note: Additive manufacturing technologies were discussed by David Leigh, James Sears, Sheku Kamara and Nathan Van Becelaere in presentations at the 2015 VMA Technical Seminar & Exhibits. Dean Markle works closely with VanBecelaere to develop the indirect AM model for sand casting valves.