Should metal casting via binder jetting sand 3D printed molds be considered a metal additive process? Sand binder jetting of molds (or cores) for metal cast parts can be highly automated and does not require a furnace sintering post-process. In many cases, it requires minimal or no subtractive finishing and no heat treatment. It can be used to produce massively large parts and it is faster than most other metal 3D printing processes. Lately, more and more foundries and even large end-user companies are embracing this technology to produce large and complex metal parts.
While geometry can be limited by the mold, the process still allows for advanced generative designs while using any material currently used for traditional metal casting. Companies like Enable Manufacturing in the UK (with voxeljet technology) and Ventana in France (with ExOne technology) have demonstrated large, highly complex magnesium parts, proving that 3D printing of sand molds can be an ideal intermediate step between existing manufacturing processes and the future of manufacturing. VoxelMatter Research’s recently released market study on the Traditional Ceramic and Sand Additive Manufacturing Market highlights all the players and business opportunities in a market that could be about to boom.
Why shouldn’t metal casting via sand 3D printed molds be considered a metal additive manufacturing process?
While we often would like to think otherwise, 3D printing—especially metal 3D printing—is not a direct, stand-alone manufacturing process. All 3D printing is part of a workflow that includes multiple pre- and post-processing steps. In metal 3D printing, even the most net-shape processes, such as PBF, require multiple post-processing operations ranging from support and powder removal, to heat treatment and polishing. Other processes such as DED are near-net-shape, meaning they require extensive subtractive finishing, while metal binder jetting (and all bound metal processes) require sintering and, in most cases, debinding and infiltrations.
However, even as it continues to gain more and more widespread adoption, binder jetting of sand molds is still a relatively niche area. Today, it is only implemented by a few dozen innovative companies (foundries and sand 3D printing service providers) around the world.
The companies that hold the leadership in the sand binder jetting hardware market, voxeljet and ExOne, are still struggling to emerge as manufacturing powerhouses. They are mostly ignored by financial analysts who do not yet see them as opportunities in their stock portfolios (and rightly so, since their stock value is currently near an all-time low). Yet, they continue to sell machines that can cost well above one million dollars and can generate value for a lot more. Recently, BMW Group virtually opened the doors of its center in Landshut, Germany to show how it uses both ExOne and voxeljet systems to 3D print sand molds and cores for its metal casting business. Other car companies, including Tesla and General Motors, are looking at and implementing sand binder jetting technology as they shift automotive production towards gigacasting, which is defined as the ability to metal cast entire sections of a car by 3D printing extremely large and complex molds and cores (for die-cast prototyping or end-use parts).
This brings us to another related topic, which is the use of polymer 3D printing to produce sacrificial cores. Printing sacrificial cores for metal casting is possible via stereolithographic 3D printers (in fact this is a major application area for large SLA systems or voxeljet’s unique PMMA binder jetting systems), however, in our analysis, we tend to consider these as polymer 3D printing processes and we will address them only marginally in this article. This remains a relevant topic and a valuable application of 3D printing across multiple industries (automotive, energy, aerospace, maritime), especially considering that it is a process capable of large-batch and serial manufacturing.
In the following sections, we will focus primarily on understanding the value proposition of direct sand binder jetting for the casting of metal parts and what it means for the companies that are making it happen.
What happened to voxeljet?
With its stock priced at just over $1 (from an all-time high above $40), voxeljet has been struggling on the US stock market for most of the past decade. These challenges only partially describe the company’s actual business: although voxeljet has not been recording exponential revenue growth, it has gone from being an innovative tech startup to a global company. The company continues to innovate manufacturing both through its own service centers and with its machines installed in foundries around the world. To better understand voxeljet’s trajectory, we spoke with James O’Neil, the fourth-generation owner of O. K. Foundry Co., who has been closely following the company’s evolution in both the real economy and the financial world.
“We started experimenting and using 3D printed sand molds nearly the same year they became available from ExOne (ProMetal), around 2005,” he tells VoxelMatters. “ExOne opened a service center in Detroit or Houston and I was able to get the ballpark pricing information that a mold would be $0.15 per cubic inch, which enabled me to price projects efficiently for feasibility and even sales. This happened not too long after Hoosier Pattern Industries bought their first S-Max printer. To me, this was a milestone, since I viewed ExOne as an equipment supplier, not a mold supplier, and having third-party service centers, like Hoosier, TEI and Humtown, represented a major evolution of the supplier market. We’ve purchased sand molds from Hoosier ever since, they are a superlative supplier of everything foundry pattern related, including 3D printed sand molds.”
As a stockholder, Mr. O’Neil has been tracking everything that happened with voxeljet since the company went public. “I think voxeljet made significant CAPEX investments into R&D, infrastructure and business development every year,” he says, “[the company] has executed flawlessly on reaching the objectives and milestones of the decade-long R&D investments. They now have substantial sales and manufacturing presence in three global regions, IP and products, and operational capabilities.” James also believes that “the IPO in 2013 was brilliantly timed to capitalize at the peak of the 3D printing investment bubble. The $10 million raise in 2021 was not nearly as rich, but still very well timed to coincide with a stock market rally in 3D printing stocks. Another $4.4 million raise in 2022 at $3.44 per share was timed to cure a problem with working capital instead of exploiting a rallying appetite in markets. It was not a good deal for shareholders except as a way to stave off a cash crisis,” he argues.
This brings us to the latest capital raise with Anzu Partners in December 2022, which only netted $.8 million for voxeljet in exchange for 10% voting rights. O’Neil thinks that this capital raise was an agreement with Anzu Partners to find a buyer for Voxeljet or more private equity. Recently voxeljet initiated a formal review process to “evaluate strategic alternatives”, which includes merging or selling the company but also possibly acquiring another company to strengthen its position on the global market.
Voxeljet’s “core” potential
When discussing the voxeljet’s potential, O’Neil refers to the company’s VX1000 printer and its capabilities for investment castings using PMMA patterns. He first saw it in operation at Aristo Cast in 2016 and was convinced that binder jet printing was ready for volume production and that investment casting would be the first application to use it. That time may now have come as voxeljet revealed it is working with Sharrow Marine to use its technology for serial production of a marine propeller to be distributed by Yamaha distribution deal.
“Even though it’s a long time coming, I think this is a huge milestone,” O’Neil says. “What’s important to recognize is that for every different horsepower engine Yamaha offers, and other marine motor makers, and for different applications, there will need to be a different impeller. The part number proliferation here is extreme and that’s important for additive adoption. With 3D printed PMMA patterns for investment casting Yamaha can launch a new propeller design that ostensibly will replace all but the cheapest, lowest-performance propellers, with a new impeller without needing to build new hard metal tooling for casting wax patterns for investment. It will save them millions.”
This is not the only application that shows the potential of voxeljet’s technology. Several others are related to sand binder jetting for indirect production of metal parts (still mostly cores, for now). The most important one for serial production has been developed in collaboration with Loramendi and BMW, leveraging the highly automated VJET-X platform. voxeljet and Loramendi implemented the first ICP production line at BMW Group’s plant in Landshut, Germany which supplies all of its vehicle and engine plants worldwide, including for nearly all BMW, MINI and Rolls-Royce vehicles, as well as for its motorcycle brand, BMW Motorrad.
The solution integrates voxeljet’s high-speed VX1300-X (VJET-X) 3D printers into a fully automated pre- and post-processing workflow, which also includes industrial microwaves for curing the 3D printed cores. Printing rates were increased by a factor of ten with the latest generation of VJET-X 3D printers, and the toolless design of the sand cores allows for variant changes at unprecedented speed without time-consuming tool changes and production downtime. The unused material is 100% recycled and reused in the production process.
Together, voxeljet and Loramendi are revolutionizing the industrialization of core printing. The production of inorganic 3D printed cores has enabled BMW Group to advance the design of its engine components. For example, the cylinder head for BMW’s B48 engine has been significantly improved by using 3D printing to produce water jacket-outlet combi cores. Additionally, 3D printing allows BMW to produce sand cores in one piece, reducing the complex design of engine components while optimizing the engine’s efficiency and fuel consumption. The inorganic 3D production line also significantly reduces the foundry’s emissions, as only water steam is produced during the casting process.
The sand binder jetting business
“We have been unable to really create any kind of profit-generating business model around 3D printed sand molds, but we’ve used 3D printed sand molds for a very large range of applications, from very small castings to the largest molds that can be printed,” O’Neil tells VoxelMatters, showing a high-end audio product as an example of where this process and technology has been pushed to its very limits.
“For us to be profitable with the process, we must have some volume. We do a lot of lower volume work for historic restoration castings, and they can be profitable in a large project, but stand-alone small-volume castings for prototypes, repair parts or legacy obsolete industrial castings are not profitable at any price point. We’ve made prototype castings for suppliers to large OEMs like CAT, Deere and Ford, but this is limited to prototypes and there’s not enough volume to generate profit. There are no pure-play prototype foundries that I know of, it is always some kind of loss leader business or the occasional pet project or project for marketing.
On the other hand, larger companies like Hoosier, TEI and Humtown are likely conducting profitable businesses around sand binder jetting. Hoosier has grown from one S-Max printer to five. These companies are financially conservative and have volume customers from companies like CAT, Deere, Ford, etc. In some cases, they are operating the machines directly for their customers. Hoosier, in particular, was a hard tooling pattern shop that got into patternless printing of molds and cores, while Humtown was a high-volume core-making shop that extended their core making into 3D printed cores and molds. TEI is similar, although maybe more aerospace-focused.
When voxeljet refers to parts for aerospace in their earnings reports, they are likely referring to TEI and SpaceX. TEI is also the company using voxeljet’s VX4000 to produce cores for the series production of large-format, weight-saving structural components for the Cadillac Celestiq. They just added another system in the first quarter and now own three VX4000 printers. Using 3D printing, the novel underbody structure consists of only six large precision sand-cast aluminum parts. Each of the six castings reduces the number of parts by 30 to 40 components compared to a typical sand construction.
Then there is the GE Research collaboration on the development of the gigantic VX9000 system to manufacture massive sand-casting molds. The new manufacturing technology will produce metallic near-net shape (NNS) components for the wind and hydro energy sectors, reducing production time and costs. voxeljet will develop and build a 3D sand printer with breakthrough size for the additive manufacturing of sand molds for casting parts ranging from 10 tons to over 60 tons. The project aims to produce large-scale 3D printed sand molds to cast components for the nacelle of the GE Haliade-X Offshore Turbine. The nacelle, where mechanical components are housed, can weigh more than 60 metric tons. The goal is to reduce the time it takes to produce this pattern and mold, from around ten weeks to two weeks.
This novel manufacturing technology has the potential to reduce overall hydropower costs by 20% and lead times by four months. The project will also include the production optimization of a 16-ton rotor hub using the ACC as well as the development of a robotic welding process for the assembly of a >10-ton Francis runner. To help ensure the successful implementation of ACC, an advanced manufacturing curriculum is being created for local workforce development to train and engage workers on the specifics of this 3D printing manufacturing technology.
ExOne’s thousands of metal parts
In a recent exclusive interview with VoxelMatters, CEO Ric Fulop said that “Desktop Metal believes it has more printed parts in cars than any other company” With this statement, Fulop was referring primarily to metal cast parts produced using ExOne’s sand binder jetting technology.
These include metal cast parts produced by BMW using 3D printed sand cores and molds. The company revealed it is using four ExOne Exerial systems (with two more on the way) to cast the Series 3 engine cores, streamlining highly automated core and mold production. When combined with a microwave, a desanding station and a fully automated conveyor system, the complete Exerial system runs high-speed 24/7 production and delivers high-accuracy parts.
Using terminology loosely borrowed from Tesla’s Giga Factories, in our interview and the relative analyst presentation, Fulop spoke specifically about advancements in giga-casting by Desktop Metal/ExOne clients around the world. Today cars are still built with a process called Body in White (BIW), where sheet metal is stamped and eventually welded by robots in an assembly line. Automotive OEMs such as Tesla, Mercedes and BMW are working with Desktop Metal/ExOne and its customers to offer giga-casting, where portions of the car are made as a complex casting replacing thousands of welds and hundreds of stamped parts.
For example, Grainger & Worrall in the UK uses ExOne’s S-Max to 3D print giga-casting molds used for homologation and design before locking expensive and time-consuming tooling for die casting, enabling dramatic savings and fast development timelines. Higher complexity geometries are also possible by combining die casting with binder jet 3D printed sand cores.
ExOne’s S-Max and Exerial machines thus continue to represent the lion’s share of revenue scale with 3D printed sand cores and molds as indirect 3D printing of castings. These are also the processes used for the production of the Mercury Maring and Caterpillar engines.
The biggest (value) in sand binder jetting
According to VoxelMatters research data, Humtown is currently the largest provider of sand 3D printing services using binder jetting technology. This started when Mark Lamoncha, president of an Ohio business making patterns, molds, and cores for the shrinking foundry industry, made a risky decision that he hoped would re-energize the business founded in 1959 by his father.
On paper, the math didn’t make perfect financial sense, but the industry was changing and presenting new challenges. The number of U.S. foundries was dropping, from about 5,000 to 900. In 2016, Lamoncha partnered with Youngstown State University and America Makes to install a sand 3D printer—the S-Max from ExOne—in his facility. The new machine directly printed sand molds and cores from digital design files, which his foundry customers would use to pour metal castings.
Relatively speaking, the S-Max was much more expensive than the equipment that Humtown typically purchased to create sand cores. But Lamoncha saw the new technology as simplifying the whole end-to-end process in a way that would ultimately save money. The new machine promised to reduce labor, scrap, repair and inventory, and allow him to deliver products to customers much faster, including parts that were once impossible to make, ultimately enabling him to win new business.
In 2020, Humtown had three S-Max 3D printers, and that number has grown since. With its 3D printing technology, Humtown supplies sand molds and cores for metal casting to foundries all over the United States, as well as the rest of the Americas, Asia and even occasionally to Europe. As of 2020, Humtown was already the largest producer of sand cores in the United States, with over 125 tons of sand 3D printed each month into molds and cores for foundries to pour metal parts. That’s equal to about a rail car each month.
Looking at cores alone, for example, a core shooting machine tool retails for about $100,000 and it costs an estimated 2-3 cents per cubic inch for a part. These machines essentially blow sand mixed with wet binder into a mold, and they can quickly churn out sand cores for metal casting. At face value, these machines may seem much more affordable than a 3D printer that retails for over $1 million, which would seem to put the cost per cubic inch at a noncompetitive value of 12-14 cents per cubic inch for the 3D printer.
But when taking into account the broad range of savings for complex molds and cores, the technology is highly competitive and also opens up new business opportunities. When producing a core with a core shooter, you start with a mold, which costs extra time and money to create. These molds are typically CNC machined out of wood, plastic or metals, depending on the required volumes. Whenever you create a mold using a subtractive method, you are paying for material cost for the mold, machining time and expenses, as well as the material waste that comes with machining because all of the waste must be disposed of or recycled.
If the core is complex, the desired core shape is usually broken into several parts that will later be assembled. This requires extra design work and more mold parts for each separate piece, and also the cost to assemble separate core pieces. This requires labor, as well as jigs and fixtures that must be manufactured to help ensure the proper placement of the assembled pieces. These additional parts come with the same costs as a mold: materials, machining time and waste. What’s more, they’re built for a single purpose and aren’t needed after the job is complete.
Then there are the quality considerations. If the part is truly complex, the failure, scrap and repair rate of the part coming out of the core shooter can be high. Depending on who you ask, the waste can range anywhere from 5-20%, and even more for very complex designs. That is simply because it can be challenging to extract a complex core from the mold. Additionally, molds for traditional core shooters deteriorate very quickly because they are being blown full of wet sand, which is akin to sandblasting. That means that sandblasting often deteriorates as volumes increase.
With 3D printing, every one of these costs and concerns is reduced, if not eliminated. There are no patterns, molds, jigs or fixtures. Core shapes don’t need to be broken into pieces for manufacturability. There’s also no labor to assemble pieces or repair a core part that might not have been extracted perfectly from the mold. Foundries partner with Humtown because they produce parts very rapidly and can create designs that cannot be achieved with conventional core-making processes.
It’s just beginning
Shortly, voxeljet and ExOne will no longer be the only companies offering sand binder jetting technology for metal casting—and that’s not necessarily a bad thing. Many other players are appearing in the market. Some of these are in China, where both ExOne and voxeljet also have a strong presence. VoxelMatters Research’s report lists as many as six Chinese companies offering sand binder jetting technology and either selling it in the domestic market or using it to provide services. A few other small companies also already exist in Europe and the US.
A machine from one of these companies, Kocel, was installed by the Austrian company Voestalpine at its Gießerei plant in Traisen. Voestalpine has been collaborating with Kocel since 2014, when the group expanded its special steel activities in the Chinese market, targeting the Chinese premium market with its €140 million investment in Yinchuan (Ningxia province) and working together with Kocel to produce tool steels and forged materials for the automotive, consumer goods and mechanical engineering industries.
The new additive manufacturing technology for the production of castings using silica sand was installed in Austria to reduce production times via a more environmentally friendly process. The 3D printer model is the Foundry Sand 3D Printer-AJS 2500A/2600A with EU CE and ISO 9001 certification. It features a resin binder recovery system for low resin consumption and enables the use of new sand, reused sand and reclaimed sand.
The two available models have build volumes of 2500 × 1500 × 1000 mm and 2600 × 2000 × 1000 mm respectively, measure six meters in length (and almost six meters in width) and weigh 30 and 35 tons respectively. The process is primarily used to produce castings for the energy industry and the automotive and railway sectors. One order, for example, was for rotor discs for water turbines.
The next phase of segment evolution is coming, as really large groups begin to invest in developing this technology. One of these is Laempe Mössner Sinto GmbH. The company demonstrated its complete portfolio of solutions for core-making technology at GIFA 2023, the world’s largest foundry trade fair. With a 3D printer developed in-house, Laempe has strengthened its claim as a global full-range supplier and technology leader in core manufacturing. The first six 3D printers have already been delivered and are now in use at one of the world’s largest car manufacturers.
In the end, it’s interesting to think about the existing opportunities and future potential of 3D printing sand cores within the framework of metal additive manufacturing. Although still niche in terms of applications and adoption, many industries are starting to see how the indirect metal AM process can transform core production and metal casting and the market is starting to evolve accordingly.