Software-based engineering optimisation methods have been around in the back rooms of engineering and design firms for more than 20 years. Despite the best efforts of advocates and a suite of convincing case studies, most engineers have not embraced the idea of letting a computer do the heavy lifting during the design process.
Techniques such as topology optimisation are used mainly as a way of reducing the weight of parts. The software algorithms take a lump of space where an engineer is creating a part and decide on the best shape, depending on the loads input by the engineer. Some versions of the software are freely available in open-source forms but, even so, this has remained a niche area. Martin Gambling, managing director of GRM Consulting, says one reason for it not being more widespread is that topology optimisation has not been integrated into the everyday CAD tools that people use. Furthermore, there is scepticism about the technique, he adds. “There is a human fear that it replaces the engineer. But it doesn’t – it adds to their capabilities. You can see what the code is guiding as a design for load paths and structures, add that to your own understanding, and design something brilliant.”
However, software-based engineering optimisation does have limits. A computer program will do only exactly what it’s told – nothing else. It won’t account for other, more subtle load cases. It also can’t substitute for an engineer’s experience and common sense. “It won’t replace good engineering – it needs that to be added to go further,” he says. “But it can be put into your design process. If you fully engage with it, and combine it with good engineering, it will become more useful.”
Laminates shape-up
As well as helping firms to optimise structures using topology optimisation, GRM works with composites. The firm’s engineers use a process called topometry to develop the optimal design for ply laminates in a structure, working out the best angles and shapes of the plies with a software tool. The user inputs the requirements for stiffness and strength into the software, which then evolves the shape and angle of the plies the design should use. Every Formula One team that has won the Driver’s Championship for the past 10 years has used GRM’s software.
GRM and firms such as SolidThinking in the US, part of software company Altair, have carved successful niches developing these bespoke optimisation tools and providing consultancy services alongside them.
A drive to reduce the weight of parts across almost all sectors has led to a steady growth in interest in the area, even though many of the tools are “semi-black box”, as Gambling terms them.
This trend to make optimisation more mainstream has seen CAD vendors become involved in the technology, and has been further reinforced by increases in computing power and 3D CAD packages maturing and expanding their feature set. GRM has recently developed topology optimisation tools that embed in the main CAD packages. Its Truform tool works in both Solidworks and Creo. “Topology optimisation has been around for more than 25 years, but it’s been quite an exclusive club. We want to make it more accessible to the everyday designers and engineers,” says Gambling.
Inspiring performance
Altair’s SolidThinking Inspire is a topology optimisation tool that integrates with most of the main CAD packages, and generates concepts using code based on Altair’s Optistruct structural analysis solver. As with GRM, SolidThinking aims to democratise the use of topology optimisation. The latest release of its Inspire software can consider new variables, such as temperature loads, velocity, acceleration, G-loads, and enforced displacement, to simulate precise loading conditions. It also features functionality that allows users to wrap topology optimisation results with geometry to, it is claimed, create manufacturable designs more quickly.
Andy Bartels, programme manager for SolidThinking, says: “Our new PolyNURBS toolset is a game changer, allowing users to create geometry from optimised results much faster than traditional CAD modelling. The result is a robust tool that has not only expanded the use cases for Inspire, but also accelerates the path to cost-effective manufacturing.”
GRM used Innovate UK funding to develop its Truform tool, and it is also part of Innovate UK projects with larger firms, such as Jaguar Land Rover and Ford, to develop more advanced optimisation methods using Genesis, its main optimisation software package. A project with JLR, called Ultran, completed last month, aimed to optimise powertrains. Previously, optimisation had focused almost entirely on lightweighting vehicle bodies. GRM’s optimisation work helped inform the design of the lightweight rear drive unit, which was carried out by Ricardo, and the unit is now being tested and manufactured.
Jon Wheals, senior technologist at Ricardo Innovations, says: “The unit can be adapted for almost any front, rear or all-wheel driveline configuration, providing a means of achieving a significant mass reduction – proving that even highly scrutinised components can be improved significantly.
Complementary to the novel aspects comprising the new axle, this mass reduction project is a great example of how lightweight design, analysis and optimisation can deliver radical mass savings – and consequent reductions in vehicle-based carbon dioxide emissions – for all types of driveline and transmission systems.
From car engines to train seats
GRM’s next Innovate UK project in the automotive sector will look at optimising engines. Meanwhile, other sectors, such as defence, aerospace and rail, are becoming aware of the benefits optimisation can bring to engineering and development. GRM has recently worked with train manufacturer Hitachi Rail to optimise the seats on its Class 331 train, a refurbishment of the Class 321. The carriages have recently gone into service in East Anglia.
The work started because of a change in regulations on train seats, which aimed to improve the survivability of seat occupants. Engineers at GRM used experience from the automotive sector to design a 26kg seat, about 10kg lighter than a standard seat, optimising the structural strength, weight and protection of the occupant.
During sled tests at safety testing facility MIRA (formerly the Motor Industry Research Association), the prototype seat passed first time. It was then developed into a cantilevered design for Hitachi Rail, which subsequently ordered 6,000 of the seats. “You can slam 200kN into the back of it and it won’t break,” says Gambling.
It’s the type of work that he believes the company will do more of as more engineers learn about optimisation. But the other trend driving interest in the area is additive manufacturing (AM). Several CAD firms, and many consultancies, present a technologically intoxicating idea of “generative modelling” – computer algorithms and CAD software working in tandem. A computer designs the optimum part, resulting in some strange item comprised of blobs and splines, which is then printed at the push of a button using AM.
Design for manufacture
“Topology optimisation lends itself well to additive manufacturing, because you can print what it gives you,” says Gambling. “But there are different rules for AM – you need supports and overhangs. The topology tools need to consider the AM rules in the process. So what you get out doesn’t need so many supports and overhangs. It’s taking the human rules and putting them into code.”
Achieving the reality of topologically optimised weird-and-wacky parts and push-button printing from 3D CAD files is a few years off. Gambling believes researchers have progressed about half of the way through moulding topology optimisation towards push-button 3D printing.
One company that is active in the 3D printing space is Altair. It has high-profile partnerships in automotive, aerospace and manufacturing, with companies such as Airbus, RUAG and Scania. The latest version of Hyperworks CAE software included a solution in its Optistruct solver specifically for the optimisation of lattice structures in AM.
“Engineers and programmers are developing and implementing optimisation principles into the everyday design process,” says Gambling. “We’re bringing the entry level in, but the high-end level is also continuing to grow. This is still an evolving technology.
Virtual technique eases the load for tailor-made bikes
Additive manufacturing is an enticing proposition for bicycle makers because it offers the potential for greater customisation. But bringing a commercial, customisable product to the mass market has proved challenging.
Welsh start-up Robot Bike Co set about solving this challenge by developing a carbon-
fibre mountain bike that features titanium “nodes” connecting the various parts. These nodes are manufactured based on an individual rider’s height, weight and riding style.
Altair ProductDesign’s engineering team designed the 12 nodes using topology optimisation – including the head tube, seat post, and chain stay lugs – to make the frame as light and strong as possible. Weighing in at just 3.2kg, the resulting bike is as light as any high-end, custom-made, carbon bike frame.
To achieve this reduction in weight, the team took the designs into the virtual environment and applied the loading data that the bike frame would be required to withstand. This data was used to output a geometry that maximised the efficiency of the material layout while achieving all performance targets. The designs were optimised for AM, including determining the ideal print angle and placement of the supporting structure to avoid the component collapsing during manufacture.
Paul Kirkham from Altair says: “AM is the perfect partner for design optimisation, as it allows us to produce components and systems that are far closer to the ideal balance of weight and performance. Robot Bike Co now has a bike that is truly innovative.”
Ed Haythornthwaite, co-founder of Robot Bike Co, says: “Using Solidthinking Inspire Altair product design helped us further reduce the weight of our frame while ensuring that stresses are kept below a predetermined maximum. This has allowed us to provide a lifetime warranty.