First of all, stop thinking about “rapid prototyping”: the processes used for decades now to convert design information from CAD format into solid materials have come so far in terms of their capabilities and applications that creating prototypes is only a part of what is accomplished. For metalcasters, this is a critical point, because the changes are having a profound effect on how the technology can be used in their organizations, and offers potential for altering what “casting” means.
Prototyping is a process step for many casting methods, and CAD technology has advanced prototyping in many ways, just as it has laid the foundation for process simulation. Typically, CAD models are exported to .STL files, and that data is “printed” in three-dimensions in some polymer material. A solid prototype of a casting, mold, core, or tooling remains invaluable to production planning because it helps to determine the form and function of a final product. Investment casters and sand casters especially rely on prototypes to set the stage for actual production of parts.
| Voxeljet’s VX1000 3D printer has a print area of 1,060600500 mm. |
However, the new term of art (though there is plenty of science in the process, too) is “additive manufacturing,” which applies to manufacturing processes like stereo lithography (SLA) or selective laser sintering (SLS), among others, that build parts by depositing fine layers of various materials and using lasers to fuse the layers only where necessary to achieve the finished shape – as defined by CAD. Another term is “3D printing,” and according to additive manufacturing expert and consultant Terry Wohlers, the two are interchangeable.
“The people who are closest to it, the advanced users, and certainly the industrial users, use the term additive manufacturing (AM),” Wohlers said. “And the more casual users, people from the investment community and the popular press, they tend to use term 3D printing to mean the same thing.”
By any name, it’s a booming business. Wohlers and Associates Inc.’s last annual global report (2011) on AM concluded the sector had revenues for products and services totaling $1.325 billion, and its compound annual growth rate was 24.1% for 2010. Over 23 years, Wohlers found the sector has had a CAGR of 26.2%.
Wohlers’ firm advises startup companies and large manufac turers about their AM choices and opportunities. Last year they completed an additive manufacturing technology roadmap for Australia, and presently they’re working on a similar project for the White House Office of Science and Technology Policy. He acknowledged that, “from the beginning, prototyping was and is the most popular application of additive manufacturing, but in recent years companies have found that they are using the technology for a much wider range of applications, beyond making prototypes.
“For example,” Wohlers continued, “drill guides, jigs and fixtures, and end-use products: there’s a lot of jewelry and even aircraft parts produced that are not prototypes. Yes, they use the same process as prototyping, but they build parts that are final products. The term ‘rapid prototype’ could even be an obstacle to growth, because who wants to use a prototyping device to make a final product?”
Expanding the prototype
The metalcasting world has seen its share of advances, of course, notably the use of bonded sand as a build material. Viridis3D is a fairly new name to the industry. It is an “additive manufacturer” that makes sand molds and cores, but also supplies 3D printing machines, software, scanners, and training for metalcasters seeking to adopt their own 3D printing capability. The past two years have brought numerous examples of metalcasting suppliers expanding their prototyping capability and services thanks to the expanding capabilities that AM makes possible.
Another name that will be new to many metalcasters is Voxeljet Technology, which supplies 3D printing systems as well as 3D products, including molds and cores. It has gained notoriety as one company (ExOne Co.’s ProMetal RCT is another) capable of producing large-dimension sand molds by additive manufacturing.
Nijhuis Water Technology is a Dutch manufacturer of pumps, and its foundry at Winterswijk casts pump housings wheels that weigh up to 1,750 lb. It was already a customer for Voxeljet’s 3D-printed sand molds, and the availability of its VX4000 highperformance printer to produce molds measuring up to 421 meters was particularly helpful for Nijhuis’ large-dimension castings. “In particular, companies such as ours, which manufacture many prototypes and small series, can derive tremendous benefits from the possibilities offered by 3D printing, in terms of quality and time,” stated Nijhuis development engineer Luke Vrielink.
Standard prototype manufacturing for such castings reportedly may take up to four months, and the production of the wooden model may take two to three months, followed by several more days to make the sand mold. 3D printing shortens the cycle significantly. Dimensional specifications are e-mailed to Voxeljet as CAD documents, and there the data is reproduced as molds without tooling. The standard molding technology is eliminated entirely, with results that are dimensionally accurate even for complex geometries. Depending on the size of the piece, printing Nijhuis’ molds will take one to two days.
| The VX4000 high-performance printer can produce sand molds in dimensions up to 421 meters. |
Short production times are important for Nijhuis, which always develops and builds pumps made to order. 3D printing allows it to reduce the number of molds in storage and still produce and deliver a pump quickly. “The fact that we can now have the large sand molds printed at Voxeljet at a reasonable cost means that we are able to ship quickly even if the model is not in stock,” explained Vrielink.
He also recognized a quality advantage. “The 3D print corresponds 1:1 with the computer model, while hand-crafted products can always result in small variances,” he said. “The more complicated the shape of the pump wheel, the higher the savings for printing in this production process. A good example is the screwcentrifugal wheel, a tapered wheel that must correspond exactly with the model. The shape of this wheel must be very precise, so as to avoid imbalances. Nevertheless, we have to polish the castings afterwards to remove any existing imbalances. Occasionally, castings must be polished for more than a full day,” he indicated.
The fact that each mold is custom-designed means that those imbalances can be resolved during the design stage, and the finished casting is much more precise, requires much less finishing work, and wastes less metal than if it had been cast in a standard mold made by hand. The fine quality of the sand used in 3D printing is a factor, too; it leaves a smoother surface on the final casting than the standard-grade molding sand.
Nijhuis’ particular needs — speed, volume, and design detail — make AM effective for sand molds. For another foundry it may make sense to produce cores this way, but the production processes are already so advanced that some metalcasters find it effective to produce finished products this way. Wohlers noted that jewelry and dental implants are two categories that have seen a noticeable impact by additive processes. His colleague Tim Caffrey explained that product volume is a determining factor in the viability of additive manufacturing for finished metal products. For small volumes of higher-value products, tool-less production has “real cost advantages on the capital investment side,” according to Caffrey. For some aerospace and medical/surgical products, AM may already be considered a disruptive technology to traditional metalcasting.
The potential is great for more of this type of disruption. Caffrey pointed first to the fact that the range of metal materials that are viable in AM processes includes tool steels, stainless steels, some grades of aluminum (though not high-performance aerospace aluminum alloys, he allowed), titanium, and titanium alloys. Reactive metals can be processed by establishing an inert atmosphere within the “build box” of the 3D printer.
| This medical implant was produced by selective laser melting, and exhibits an internal lattice structure that only AM can accomplish. |
AM systems can produce fully dense metal parts, and Caffrey pointed out that in some cases these products exhibit new microstructures because the cooling pattern is different than it would be for a cast shape. “You have these very thin layers of material being melted and essentially welded together again,” he explained, noting that the performance of these products generally “is as good as cast materials, and in many cases as good as wrought materials.”
Andrzej Grzesiak, an engineer who heads Germany’s Fraunhofer Additive Manufacturing Alliance, said the great potential of AM is that “any conceivable shape that can be created in a 3D CAD program can actually be produced. There are no restrictions in terms of manufacturing transparent or hollow structures. Nor are there any problems with complex geometries and freeform designs.”
Grzesiak said successful applications of AM parts in aircraft construction are the result of the ability to work with lightweight materials like titanium as well as the “limitless freedom in terms of shape and design” that allow optimized products to be created. Even complicated functional details like snap-fit connections, form-locking elements, and geometries like leaf springs or helical springs, can be accomplished within the design of a single component, meaning fewer parts have to be mounted or connected with tools. He expects similar progress in medical/surgical components, mechanical structures, and tool-and-die products.
Caffrey listed three attributes of AM that are compelling for the task of designing better components. First, “topology optimization” involves letting a mathematical algorithm determine an optimal geometry for any particular loading and anchoring condition, leading to a part that satisfies those conditions at the minimal weight. These “very organic” looking structures may never result from casting or even machining, Caffrey contended, but can be produced by AM.
Another attribute is “functionally graded materials,” meaning that AM can optimize alloys and materials for optimal placement within a component design. The result is not just a “homogenous blend” of an alloy, like most castings. “You could, potentially, when building a tool, give it a very hard surface but also a very conductive metal on the interior, so that it cools more quickly,” Caffrey said.
| A part of a sand mold for a pump wheel printed in sand from a CAD file by Voxeljet Technology. |
A third design attribute with AM is “lattice structure,” which Wohlers offered as complementary to topology optimization: within the shell of the AM part, only the most essential volume of metal material needs to be established. The finished part is not really solid, but has a mesh or lattice of material located only where necessary for structural integrity. “Let mathematics decide where to place the material,” Wohlers said. This cuts down manufacturing time, reduces product weight, and conserves material.
In some ways, the problems that AM solves have been addressed by metalcasting: sand cores create internal cavities; investment casting can achieve high-volumes of intricate shapes; heat-treating or hard facing can enhance mechanical properties of metals. But, advanced manufacturing represents is a way to produce components that function as a casting would, except that it is not really cast. And, as the processes become more functional, flexible, and affordable, they will challenge castings for market share because they promise faster production at lower cost, with greater customization.
How far off is this future? And what steps may be taken between now and then? Terry Wohlers said that today’s AM successes are defined by the volumes of finished products, the size of the product and the order volume (which influence manufacturing time), and the cost of the materials. Large-volume foundry customers like automakers are not likely to switch to the new technology soon.
But, Tim Caffrey advised casting design engineers to learn how AM can help their efforts, and how it cannot. “For small parts and low-volume parts, or if they have highly complex parts, it’s worth taking a look at it, for example to bypass corebox tooling, and going straight to these processes to produce cores.”
Caffrey also observed that a lot of the iterations of castings that foundries do to determine gating and risers, for example, can be simplified with AM so that they don’t have to build molds or tools for test runs. “And when their production tool is finally ready to be made they have all these questions answered already, so it can still be used as a prototyping tool, and it can be used for preproduction parts.