DM308 Additive Manufacturing Lecture 3 PDF
Document Details
Uploaded by SafeDiscernment
University of Strathclyde
Vassili Vorontsov
Tags
Summary
This is a lecture about additive manufacturing (AM), focusing on the various problems in additive layer manufacturing (ALM) processes and their solutions. It also reviews current trends and approaches in additive manufacturing, and the different types of ALM.
Full Transcript
DM308 Production Techniques 2 Lecture 3 – Additive Manufacturing (Part 3) Dr. Vassili Vorontsov Acknowledgements: Prof. Jonathan Corney, Mr. Colin Andrews Department of Design, Manufacture and Engineering management, Faculty of Engineering, University of Strathclyde [email protected] To...
DM308 Production Techniques 2 Lecture 3 – Additive Manufacturing (Part 3) Dr. Vassili Vorontsov Acknowledgements: Prof. Jonathan Corney, Mr. Colin Andrews Department of Design, Manufacture and Engineering management, Faculty of Engineering, University of Strathclyde [email protected] Today’s lecture Additive Manufacture (AM) Lecture 3 1)Additive manufacturing problems and Booth’s checklist 2)Trends in additive manufacturing – Design for AM, WAAM & RepRap Problems in ALM 3D printing is not “automatic” 3D printing is not “automatic” like 2D printing Lots of stuff can go wrong! 1) Cavities and trapped volumes 2) Distortion, shrinkage & warping 3) Feature damage & size 4) Overhangs 5) Surface finish 6) Volume 7) Internal heterogeneity 8) For metals process sensitivity …here are some illustrations of these errors Cavities and trapped volumes Cavities: Internal cavities are hard to form since it may be difficult to remove the support material from internal regions. Overhangs and bridges Overhangs: Overhanging features may affect the surface flatness (distortion). Large overhanging features may require supports, which may affect the part quality. Overhangs and bridges in FDM In FDM, overhangs with an angle of greater that approximately 45 degrees will droop if printed without support structures. This will vary between different machines and materials. Overhangs and bridges in FDM The distance that can be spanned without support structures is also limited. This will also vary between different FDM machines and materials. Cooling of the print using a direct stream of air can help print longer bridges. Distortion, shrinkage & warping RP processes that involve solidification can give rise to residual stresses. The size and effect of the associated strains are dictated by the geometry of the components. “Wet” shape Plastic Dry shape Distortion, shrinkage & warping Distortion, shrinkage and warping also occur in metal ALM. Feature size Feature size: Features must not be too small, too closely spaced or require accuracy beyond the technology’s capabilities (e.g. 0.5mm 3D printing). For example the wall of the hole below was closer than 0.5mm to the edge of the block. Typical ALM feature size limits www.3dhubs.com Surface finish Surface roughness is defined as the absolute value of the deviation of the fabricated surface from the desired surface. Zooming in on one of the terraces gives us: If the layer thickness is t, then Rmax, the maximum roughness is: Slice generation Which orientation will produce the best results? Problems can depend on build direction Vertical build direction gives the best quality hole (i.e. the roundest profile) Other problems are volume related Volume: The maximum size of the component is defined by the capability of the individual additive layer system. Typical Build Sizes for midrange ALM systems Overcoming build volume limitations Limitations of insufficient build volumes are overcome by assembling large structures from smaller ones. But these assemblies will need to be designed, adding costs. Problems of process sensitivity Ex ce ssi ve he at (p ud d les ) Insu Inco fficien t n (wa sistent heat vy s urfa melt ce) Ex ce ssi ve v in olu su me rfa ce (ripp ) le s Ins uffi cie nt vo lum e( sp utt er) All metal ALM processes have outcomes that are very dependent on the raw material quality and the process parameters (e.g. laser power, speed, scan pattern). Checking for anticipated problems Booth’s design for 3D Printing worksheet (see MyPlace for detailed pdf), Booth’s ALM worksheet Booth’s ALM worksheet Booth’s ALM worksheet Post processing and ALM We’ve discussed the Economics of ALM systems in Lecture 2, but I is not just the machines, materials and operation that cost time and money. There is often post processing required. Laminated Object Manufacture (LOM) Selective Laser Sintering (SLS) Fused Deposition Modeling (FDM) Power Based methods Post processing and ALM All ALM methods produce models which require “finishing” steps. Post processing and ALM Finishing is labour intensive and can take as long as the ALM manufacture itself. It is all done by hand and is hence expensive! Expensive, but finishing adds value Trends in ALM Trend 1: Less is more Design for additive layer manufacture FEA driven topological optimisation FE stress analysis is slowly changing from the job of analysing a given geometry to generating an optimised geometry. Topological optimisation examples Topological optimisation examples Topological optimisation used to make an ALM mould for a cast concrete slab. Topological optimisation not only for ALM Topological optimisation can also be used with conventional manufacturing (e.g. machining, casting, etc.) Design for additive manufacture 1. Minimise volume: because that drives both raw material and machine costs. 2. Open structures to allow easy powder/support structure removal. 3. Thin walls to avoid warping: Thicker volume of material solidify and cool at different rates to other parts of the structure causing residual stress and warping of the final shape. Design for AM replaces thick sections with honeycomb/cellular structures so they are stiff but light! 4. “Organic” forms generated by FEA driven topological optimization: Rather than using FEA to calculate the stresses on a shape, these systems use FEA to generate the “optimum” shape (i.e. lowest stress or strain). 5. Lower factor of safety than traditional designs: Because the material has more homogeneity (than say casting or forging) , involves less fabrication (e.g. welding or bolted assembly) and has been precisely stressed there is significantly less uncertainty about the parts composition and behaviour under load. Consequently the factor of safety can be reduced. How much is it worth? A real example Gearbox housing Traditional design (milled from solid) - 1716kg of stainless steel - £3800 cost - 58 hours manufacturing lead time Optimised design (ALM from titanium alloy) - 281kg - £800 - 8 hours 79% reduction in cost, 86% reduction in manufacturing lead time 83% reduction in weight Generative design Generative design is different to topological optimisation. Topological optimisation is a tool for refining the design produced by a human. In generative design, the human sets the required parameters and the computer does the design iteratively. The human judges the design and adjusts the input to the computer. Generative design requires implementation of artificial intelligence. Generative design may use topological optimisation algorithms too. Generative design Generative design allows a wide variety of possibilities to be explored and compared. (e.g. price vs. performance). Big picture comment It is often combinations of technologies that disrupted established ways of doing things: PCs + Speadsheets YouTube + Cheap video recorders iPod + iTunes Word Processors + Laser Printer High Speed Internet + Cloud Computing Internet + Web Browsers Additive Layer Manufacture + Topological Optimization? Trend 2: Multi-material ALM Are multi material printers the next big thing? E.g. 3D printed electronics If one material is an electrical insulator (e.g. plastic) and the other material is electrically conducting (e.g. metal). E.g. 3D printed electronics 3D electronics? Are smartphones flat because that is the best shape? Trend 3: Wire arc additive manufacturing How does Wire Arc Additive Manufacturing work? Unlike the more common metal powder AM processes, Wire Arc Additive Manufacturing works by melting metal wire using an electric arc as the heat source. The wire, when melted, is then extruded in the form of beads on the substrate. As the beads stick together, they create a layer of metal material. The process is then repeated, layer by layer, with a robotic arm, until the metal part is completed. WAAM can work with a wide range of metals, provided they are in wire form. This list includes stainless steel, nickel-based alloys, titanium alloys and aluminium alloys. WAAM in action WAAM equipment Limitations of WAAM Not fully commercialised: Although a number of companies are developing WAAM technology for the production of metal parts, currently there is no commercially available system. However, collaborators such as Kuka Systems, Airbus Defence and Space, FMC Technologies and other companies are currently working to develop a systematic methodology for WAAM. Heat management is another challenge associated with WAAM. The printing process involves high temperatures, causing build-up of residual stress — a problem commonly faced by metal 3D printing. As residual stress can often lead to deformation in a component, cooling must be factored into the process. Shield gas: When using certain materials, like titanium, shielding is necessary to create an inert atmosphere to ensure the right building conditions, meaning that the process has to take place in an inert gas chamber. However, the inert gas chamber limits the size of parts that can be produced with this technology and installing such chamber will increase the cost of the equipment. Trend 4: Crazy fringe And finally for Trend 4 we have to mention the RepRap movement and the “Crazy Fringe” of ALM - lowering prices and pushing boundaries. 23 March 2005 The most important current trend in ALM started! Dr. Adrian Bowyer, an academic at the University of Bath had and idea! The idea was... Suppose we made a machine that: Self-replicated, but didn’t self-assemble (like a virus) Existed symbiotically with people, giving them goods in return for being helped to replicate (like flowers)… The Replicating Rapid Prototyper Project The above remained a dream, but in order to work towards it Dr. Bowyer and team decided that he had to create an additive layer manufacturing system made mostly from plastic. The idea attracted a group of enthusiasts Who collaborated on-line: Nuffield Foundation EPSRC Bath University IMRC Rapid-prototyped FDM write head A – geared motor B – screw drive C – heated extruder D - electronics B Then they made a DIY 3D printer Test-bed machine 13 September 2006 Then a rapid prototyped Cartesian robot Where all the plastic bits could be printed on the printer. Open source electronics followed The robot could be controlled by an opensource Arduino project development and education board. The board is programmed using a C++ based programming language which is compiled on free and open source software. This allowed even more enthusiasts to participate in the RepRap project. October 2009 A more sophisticated robot had been developed. Want to build one yourself? https://reprap.org/wiki/Build_A_RepRap Commercialisation? Unexpectedly cheap commercial versions of the RepRap Design started to appear. These were sold as DIY kits. £900 (RapMan) $1200 (Makerbot) Commercialisation? Originally sold as kits these printers are now mass produced in China. “Print the Legend” is a 2014 documentary film and Netflix Original focused on the 3D printing revolution. It delves into the growth of the 3D printing industry, with focus on start-up companies MakerBot and Formlabs, established companies Stratasys, PrintForm and 3D Systems, and figures of controversy in the industry such as Cody Wilson. So what is the point of all this? The success of the RepRap project has inspired a constant stream of open source innovation in this area. Open source AM systems being developed for metal, composites, mixed material its a trend that lowers the bar for everyone.