Additive manufacturing (AM) technologies continue to mature, and the size of metallic parts that can be produced is yearly increasing.
To produce large parts, so-called Direct Energy Deposition (DED) manufacturing processes are very attractive. Those methods involve the use of a welding arc or laser mounted on a robot arm.
One advantage of AM is that parts can be produced, which would otherwise be difficult or even impossible to manufacture using traditional techniques. This allows us to carry out (in practice) often-complex lightweight designs that are generated via structural optimization. However, while it is sometimes suggested that with the AM “complexity comes for free”, the reality is usually different as process and machine related constraints can influence what can and cannot be carried out in practice.
As part of the INTEGRADDE Project, LimitState (together with the University of Sheffield and other partners) have been exploring how DED AM process constraints can be integrated into the design optimization process, to ensure that the parts generated are manufacturable.
To achieve this, partners with experience of employing various DED manufacturing processes were consulted. The processes considered were Wire Arc Additive Manufacturing (WAAM) and wire and powder-based Laser Metal Deposition methods (LMD-w and LMD-p respectively).
From these discussions a series of design issues were compiled for use in an optimization.
These were included for WAAM and LMD-w processes:
Restrictions on the achievable work envelope or build volume
Rules outline the relationship between workpiece orientation and the ability to manufacture overhangs
Consideration of methods of generating infill and its effect on the inclusion of part features below a minimum size
Dealing with issues surrounds the permissible radius of fillets
Restrictions on the achievable diameter of lattice / truss elements
And for LMD-p processes:
Necessity for a suitable substrate
Inability to provide support during the build
No “closed volumes” in the part permitted
Resolution of the deposited bead
Maintaining beam and powder deposition head access
Considering stress and material deformations during manufacture
Once the constraints were established, their effect on manufacturability was evaluated. In some cases, it was deemed appropriate to handle a given constraint in a post-processing step, whereas, in others it was necessary to identify ways of including these from the outset. Considering the latter, when using truss layout optimization techniques to design a part using 3-axis DED equipment in which the workpiece remains stationary, elements inclined at large angles from the vertical, that cannot be manufactured, are simply omitted from the list of available elements in the optimization setup. However, this may lead to far from optimal designs. Thus, it was found that much better designs could be generated when 5-axis DED equipment was involved, where movement of the workpiece is possible. In this case a novel approach involving the use of “printing curves” was developed for use in the optimization to ensure manufacturability.
The picture below shows the optimized structures with vertical load on the bottom edge and the corresponding printing plan. (a) Optimized structure from layout optimization; (b) Re-optimized structure with angle constraints corresponds to the printing curves, which are illustrated by the grey lines across the design domain.
Once an optimization has been completed, part manufacture can be simulated virtually to check for nozzle clashes etc. prior to final manufacture. The picture below shows the Cantilever design optimized for AM manufacture (a) Optimized line structure; (b) 3D continuum model; (c) AM slices
In addition, to process-related issues, the feasibility of using a new hybrid optimization methodology involving layout and topology optimization has been explored. Whereas layout optimization can be used to rapidly identify optimal truss structures when the design freedom is high, topology optimization methods, which involve removing low stressed material from an initial volume, work best when there is rather less design freedom. In the hybrid approach developed in INTEGRADDE, each region of design domain is optimized using the most appropriate technique, with a view to getting the best of both. See below a picture of the hybrid optimization of a tip-loaded cantilever.
In summary, the INTEGRADDE Project is providing an invaluable opportunity to use optimization to more fully realise the potential of DED additive manufacturing methods.
LIMITSTATE is a provider of engineering software, design services and innovative solutions to the civil, structural and mechanical/aerospace engineering sectors. The company comprises several skilled engineers and software development staff with many decades of combined experience in successfully overseeing the transformation of novel research methods into step-change tools for use by industry. The company has developed extensive experience of successfully collaborating with academic and industrial partners on both software development and research projects.
LimitState is currently commercialising layout/topology optimisation tools to design lighter and more efficient AM components. This software implements novel optimisation algorithms based on layout and geometry optimisation to determine an optimised CAD model (also compatible with FEA tools). Key benefits are speed and the fact that it obviates the need for time-consuming manual remodelling of the raw optimisation output - invaluable when seeking to automate the design process.