Dr.-Eng. Ulf Schliephake
Area Sales Manager Scandinavia and Austria, Brechmann-Guss GmbH
Rapid Casting Part 2 Breaking Through Design Barriers with Printed Cores
This article series presents various possibilities along the product life cycle that innovative additive solutions offer to a tradition-conscious industry. In the second part, we focus on freedom of design and bionics.
"The simple reproduction of a conventionally manufactured component is an affront to anyone who works with 3D printing as the advantages lie in the possibilities offered by an almost limitless freedom of design"1. This statement, taken from an overview article on additive manufacturing, highlights a dilemma: For decades, foundries have been trying to teach their customers "casting-oriented design" (for series production) by organizing seminars and customer training courses and distributing technical information sheets2 describing possible tolerances, draft angles, suitable radii and much more. But now, negative drafts (so-called undercuts) or drafts of 0° with printed cores are possible. Therefore, a rethink is necessary3.
The housing shown only in sections in Figure 1 illustrates the possibilities. With conventional production, an (iron) foundry usually assumes a cross-section/depth ratio of 1:1 for safe, stable production. With printed cores up to 1:5 is possible. At the same time, transverse channels less than 1 cm in diameter are formed without mineralization or core breakage. These are design details that will make the hairs of any experienced foundry master stand on end. But, practice shows: it works.
What was previously "uncastable" suddenly becomes castable - if necessary even extremely narrow tolerances that were previously impossible due to the process. Conventionally, complex core packages consisting of various individual cores are often joined together by gluing in a gluing gauge; the chain of tolerances becomes longer with each additional glued core and the internal scrap in the foundry increases to the same extent. The use of printed cores for complex internal geometries represents a real alternative.
In the meantime, indirect additive manufacturing (at least in the automotive sector) has also arrived in series production<4, 5. In the BMW M4, the thermal loads in the cylinder head resulting from the performance increases could no longer be controlled when using conventionally manufactured casting molds and geometries - the necessary geometries could not be produced. Only the complete freedom of shape of printed cores provided the necessary flow and cooling capacities, whereby the designers first had to be made aware of the importance of "freedom of shape", as they were still trapped in the restrictions of conventional production. But the new technical possibility then immediately aroused further desires…
This means that "additive manufacturing begins in the brain"6 and must not be limited to the lifting of conventional, traditional manufacturing restrictions. This is because the freedom of shaping allows for curved, openwork freeform surfaces with cavities behind them - geometries that joining or cutting manufacturing processes usually do not allow.
But not only the design is "turned upside down". Changes in process development procedures also provide significant support. The "Digital Twin" as the starting point of a cycle of topology optimization, load case investigation and foundry process analysis takes into account both the later application and the restrictions of the selected manufacturing process - which are rather small in casting due to the process.
From Lilienthal's flying objects, through velcro, and echo sounder to the lotus effect - many newer technical developments have their origin in phenomena from nature7. "For nature knows no mercy! Those who do not function are at least chased away, overshadowed, repressed or eaten up. It is therefore no wonder that the survivors of evolution are optimized in terms of both shape and material8 - a tree fork, for example, is never radial but usually oval in shape - a geometric form that classical CAD programs did not even know in the past. Topology optimization, i.e. a weight-optimized design adapted to the load case, closes the design optimization loop9, 10. The free-form capability of the additive manufacturing process fuels these optimization strategies (Figure 5 and 6).
All common CAD program packages now offer corresponding work routines, in which the stiffness of a component is usually maximized, or a minimization of the mass is to be achieved with the same stiffness11, 12, Figure 6). At the same time, it is possible to take manufacturing restrictions into account, e.g. the exclusion of individual component areas (such as screw-on geometries) from the topology process or the (production process-related) adherence to minimum wall thicknesses for primary and forming processes, or the specification of demolding directions. In addition to material savings through a load-compatible structural design and the resulting lower use of materials in production, a significantly longer component service life can be achieved at lower costs.
On the basis of defined installation space, load scenarios, boundary conditions and a manageable number of manufacturing restrictions through casting, or only economic constraints when using indirect additive manufacturing, constructive solutions result which can only be realized through casting. If it is possible to close the existing system gaps of most commercial program packages, in example to carry out load analysis, topology optimization and 3D data set for production internally in a very fast process, great potentials become usable. Up to now, the time-consuming data transfer from one computer program to the next and the laborious feedback of the data still represents a real inhibition threshold, also due to the time required and the personal workload. On the other hand, the program systems currently under development13 for direct 3D printing can be easily transferred to indirect manufacturing by replacing the manufacturing restrictions of metallic 3D printing with the material-dependent casting restrictions. In the end, the automated computer-aided optimization is "simply" broken off a few calculation steps earlier with the help of the corresponding restrictions. Topology optimization and 3D printing change our expectations of product design.
However, one fact should be noted at this point: only the designer of the assembly or machine knows (hopefully) all load and application cases of his component. Every component supplier (especially in the non-automotive sector) is groping in the dark there; that means that the design responsibility must either be rewarded accordingly (a common field of activity for engineering and design offices) or simply remain directly with the buyer. The supplier knows his manufacturing restrictions (such as minimum wall thicknesses depending on the surfaces, demolding chamfers and/or tolerances) - the load case, especially of abusive applications, is hardly known. Only with open communication and rapid data exchange between load analysis, topology optimization (both at the customer's end) and simulation of the manufacturing process (at the supplier's end) can the potential of bionics be exploited.
And this also requires changes in procurement processes. It is not conducive if the customer's design department quietly designs a casting with bevels, wall thicknesses and material in detail (up to the meaningless dimensioning of non-loaded radii with R2), passes the internally approved data set to the purchasing department, which in turn distributes it worldwide as a request and the supplier is then determined on the basis of the lowest price. The supplier of the future can no longer be the cheapest actor, who produces and delivers any imaginative design without his own constructive input.
No competent foundryman needs a fully designed component for cost estimation. In iron sand casting, for example, it is sufficient to know what batch sizes are planned, how many parts can be produced in a mold (this gives you the fixed costs per part), how many cores and/or cooling molds are required and the approximate weight of the part - because the costs of the hot metal often make up barely 25 % of the component costs. By specifying the additional/reduced costs for weight differences (and other possible cost-influencing factors), a supplier can already be determined at the sketch stage, who then, on the basis of a corresponding commitment, contributes his services (and his special manufacturing know-how) to the component development in the sense of simultaneous engineering. The procurement organization will also have to change under the pressure of bionic approaches and find new forms of cooperation.
1Käfer, S.: Die neue Freiheit der Konstrukteure, MM MaschinenMarkt, https://www.maschinenmarkt.vogel.de/index.cfm?pid=5112&pk=61346, 05.06.2018
2Brechmann-Guss: Bearbeitungszugaben und Toleranzen nach DIN EN ISO 8062, Rev. 2, September 2018
3VDI, Fachbereich Produktionstechnik und Fertigungsverfahren: Handlungsfelder Additive Fertigungsverfahren, 2016, p. 31 ff.
4Richter, D.: GF Casting Solutions Leipzig GmbH opens Innovative 3D Core Printing Center, spotlightmetal, jun19, 2018
5Ségaud, Jean-Marc, Vortrag: Der neue Zylinderkopf BMW M4 - Serieneinsatz der Additiven Fertigung von Kernen, 10. VDI-Tagung Gießtechnik im Motorenbau, Magdeburg 29.-20.01.2019, siehe auch VDI-Berichte 2339, p. 129
6Kaliudis, A.: Form folgt Funktion, ADDITIVE FERTIGUNG, 2, Mai 2018, p. 34 - 38
7Bittermann, P.: Ikarus ist sanft gelandet, PERSPEKTIVEN, Hüthig-Verlag 2015, p. 174-178
8Mattheck, C.: Universalformen der Natur, Sonderdruck aus labor&more, Jan. 2012
9VDI 3405 Additive Fertigungsverfahren, 2013-2015, insbesondere Blatt 3: Konstruktionsempfehlungen
10VDI 6224. Bionische Optimierung; Anwendung biologischer Wachstumsgesetze zur strukturmechanischen Optimierung technischer Bauteile
11Padmanbhan, H.: „Highlights SOLIDWORKS Simulation: die neue Topologiestudie“, http://blogs solidworkscom/solidworksdeutschland/author/solidworksde, 20.12.2017
12Kluge, S.: Leichtbau in der Landtechnik - dank Topologieoptimierung und neuen Fertigungsverfahren, Firmenschrift, HyperWorks - an ALTAIR division, 2016
13Reiher, T.; Vogelsang, S.; Koch, R.: Computer integration for geometry generation for product optimization with Additive Manufacturing, Austin, Texas, August 2017