Sand casting is an ancient method, it’s been used since the Shang Dynasty (c. 1600 to 1046 BC) and is used in industrial manufacturing to this day. During the years and as industrial production developed it has adapted to the needs of manufacturing yet it’s interesting to see how the principle of it remains the same. In recent years with the 4th industrial revolution, the combination of the ancient technique with today’s additive manufacturing solutions brings new possibilities and takes this industrial process even further in creating a digital solution for the traditional foundry. How do these two technologies come together? What can and is achieved by the combination?
Sand Casting – The Basics
Sand casting is a production process widely used in creating large metal components in various industries. In the Automotive industry, sand casting is often used in the manufacturing of engine blocks, engine manifolds, cylinder heads, and transmission cases. It’s used specifically for large parts that can’t be mass-produced using other manufacturing technologies but is also used for smaller intricate parts such as gears, valves, pulleys, and crankshafts. Typically, a replica of the final part is covered with sand, usually in a two-part mold. The sand is tightly packed so that when the initial part is removed, the cavity remaining in the sand becomes the mold. Inner parts are also created using sand prior to forming the mold. When the mold is finalized (sometimes using heating) the cores and inserts are placed in position and the parts of the molds are clamped together. The molten alloy is then poured in and cooled to take the shape of the cavity. The final step is breaking the mold and retrieving the final part. The fact that the sand mold is destroyed during part removal and requires building again to create another part makes it suitable for low to medium production batches, compared to injection molding, for example. Yet the same quality makes it suitable for the production of complex parts with internal areas.
What happens when you add AM to the mix? While the difference in the manufacturing method of an AM produced part versus an injection molded part is substantial, that’s not the case when it comes to sand casting with and without AM. At its core – creating sand molds by using 3D printing over traditional methods doesn’t differ in results but it does differ in costs, time and details. When creating a sand mold the materials used are various types of sands combined with binding materials. If that sounds familiar it’s because it is similar in a way to AM binder jetting methods. Another similarity regards a common deterrent when it comes to integrating AM in industrial production – quantity. For mass production, as in the production of hundreds of thousands of parts, AM isn’t ideal, it is mostly suitable for low to medium-size production runs. But when it comes to sand casting which is meant for low to medium quantities AM doesn’t create a limitation. On top of that, because creating the mold is just one step of the process, the process as a whole doesn’t change drastically, removing another obstacle – the fear of disturbing the operation.
So what are the differences, the reasons to integrate AM into an already functioning production line? Let’s start with shape – 3D printing the sand mold enables the creation of more complex geometries, intricate details and undercuts are not an issue when 3D printing the mold. Next, are the automation and digitization of industrial manufacturing. Sand molds have been manufactured pretty much by hand, with AM the physical piece that creates the cavity is replaced with a digital file, creating a simplified and accurate process. Even only additively manufacturing the cores and inserts while still using a conventional sand mold can simplify the process. This allows for integrating complex cores with undercuts, reducing the number of parts as well as the time spent on the assembly and construction process. Using a digital file reduces costs, it is said to save up to 75% in sand casting costs, it reduces the need and cost of tooling as well, and subsequently the production time. Time is always an important variable in the production process, while the conventional sand casting mold can take weeks or months to produce printing the same mold will take a day or two, according to the size and complexity of the part. PumpWorks Castings, for example, utilized ExOne S-max for the production of sand molds, reducing lead time from 17 weeks to 8 weeks. All these make additively manufacturing sand molds an economically worthy alternative (sand casting mold by Voxeljet below).
Not a Hypothetical
The use of AM for the production of sand casting molds actually isn’t new, a number of foundries have been applying the technology in their process for quite some time now. Some use AM for the entire mold and some just for the production of cores. One example is Cox Powertrain, a propulsion system manufacturer and Grainger & Worrall (GW). Together they have been developing engine castings using additive manufacturing. GW’s use of digital manufacturing enables the company to speed up time to market of new products, specifically in this case – the goal is producing 70 engine sets per week. “The innovative casting technologies employed by our company makes us the ideal supplier for cox’s ambitious program, which requires its first engine casting deliveries to be made by year-end. our agile, flexible approach is the perfect solution for manufacturers and suppliers working to short timescales who require reduced lead-times without jeopardizing quality control” said Matthew Grainger, director of GW (below sand casts from GW, photo via GW). Another collaboration was presented at the GIFA 2019, the International Foundry Trade Fair, where Loramendi, Voxeljet and ASK Chemicals exhibited an additive manufacturing integrated solution for foundry serial-production. AMRC Castings is yet another example, the company currently owns 2 AM machines that are capable of producing one-piece sand molds and cores. Using these machines in their already established process reduces the costs of equipment and assembly required for each cast. “…depending on print density, the typical turnaround time is between 12-24 hours for the S-Max, and 24 and 50 hours for the S15. Such performance can be transformational for clients, who typically invest tens-of-thousands of pounds in dedicated, inflexible tooling that takes longer to manufacture than it takes us to deliver a finished casting. In addition, we can easily incorporate minor design revisions without any further costs being incurred.” says AMRC Casting Group’s Anthony Kenney.
The Big Picture of Manufacturing
3D printed sand casting molds is an excellent example of the possibilities of hybrid manufacturing processes. As part of one process, It doesn’t need to be one or the other – traditional vs new isn’t a relevant dichotomy, it’s all about finding the right combinations to achieve best results in minimum costs and time. Of course, where there is a need, there is a machine. In a recent post, we listed a number of new AM machines, two of those designed for industrial use and are specifically geared for sand casting. The ExOne and the Voxeljet both offer industrial sand 3D printers, which as you can see from the examples above are already in use (part printed by Exone above and mold printed by Voxeljet below). While sand casting and additive manufacturing are seemingly opposed technologies, they complete one another. The integration of the two can be described as disruptive in terms of the possible effects but it does so without actually disrupting the production process. The casting itself hasn’t changed, it’s just about finding a better way to produce the mold. The integration of the two is more than suitable it is a necessity for the traditional foundry moving forward.
For those of you attending Formnext this week, come by our stand (B81A in Hall 12) or follow LEO Lane on Facebook, Twitter, and LinkedIn for live updates from the exhibition and events and stay tuned for our report from Formnext in a couple of weeks.