This page has not been updated recently.
We have kept it, within the OnWindows Archive, for your reference.
Simulation at the core of design
20 December 2012
Under pressure to shorten time-to-market, manufacturers are seeking to further expand the use of simulation by taking advantage of it earlier in the design process, says Jan Larsson at Siemens PLM Software
Throughout the design cycle, organisations race to create a design so they can analyse, build and test physical prototypes to confirm intended product functional performance. As a result, many changes occur along the way and design problems are often not found until the testing of the physical prototype. This leads to development schedules slipping, costs skyrocketing and products often falling short of market and business requirements, increasing the risk of customer dissatisfaction and/or recalls.
In many instances, the design team and the simulation analysis team are working almost entirely independently. This is often because the simulation team is only brought in to test something if it fails, or as a final analysis step in the approval process before a product goes to market. Even when bringing simulation into the process earlier, teams are using an array of disparate, specialised tools that are disconnected from each other, which creates redundant data and workflows, and ultimately impedes the speed of simulation.
Not only is this process slow and cumbersome, but it also sets a precedent for ‘good enough’ product design. This ethos is outdated and companies need to move to the next level of product development, integrating simulation more deeply into the design process by using better and more synchronised tools across both the design and analyst teams that enable a concurrent design and analysis process.
Historically, simulation has often been out of step with the design schedule; results have come too late to be useful. This, in turn, has created a lot of the frustration, waste and confusion associated with computer aided engineering (CAE) software. According to a report by the AutoSim Consortium, of the total time needed by engineers to perform a full simulation for a system or subsystem, 80 per cent is devoted to generating the model.
To speed things up, every step in the simulation process must be aligned. The goal is to get to a point where simulation is in sync with design and, in some cases, leading design so analysis results can be fed back into every design decision. Simulation models (such as finite element models) with detail matching the accuracy desired at a particular design stage must be available.
Integrated multi-discipline analysis allows users to more easily evaluate real-world conditions
Today, a wide range of applications are used in product development. Designers have a variety of CAD applications and analysts have a number of CAE software technologies at hand, including multi-physics simulations, electromagnetics, fluid dynamics, structural finite element analyses, fatigue and failure analyses, acoustic predictions and design optimisation. In a simulation-driven design process, engineers have access to powerful geometry editing tools, like direct modelling and dimension-driven design. The most recent advance comes in the form of synchronous technology, a unique capability that combines the speed and flexibility of direct modelling with the precise control of dimension-driven design. These powerful tools allow engineers and analysts to easily edit and obtain the idealised geometry they need without having to wait for designers to perform these tasks. This enables them to respond rapidly to design changes or suggest changes to the design based on simulation results.
Because models and data can be shared easily, this level of integration can provide confidence for a manufacturer’s decision makers, while delivering a more consistent interface and feature set that allows more flexibility between the roles of the designer and the analyst. This is not to say that one person could ever replace the role of the other, but through a centralised hub, it’s possible to deliver tools – skinned versions of the applications that meet the individual requirements of the different teams – which allow designers to run basic simulations, and analysts to make any necessary tweaks to a model’s geometry. This helps garner greater synchronicity and trust between the designers and the analysts, and gives both teams the ability to make alterations without having to go through a complex and time-consuming back and forth process.
By turning to a simulation-driven approach to product development and introducing it right in the concept stage, product developers can reuse existing models and design geometry instead of rebuilding them from scratch. This allows them to explore alternatives, spot flaws and optimise product performance before the physical prototype or detailed design is created. The process allows important decisions to be made on functionality, geometry and materials early in the cycle based on simulation results.
This high level of coordination between designers and analysts can be achieved through the implementation of a complete suite of integrated process automation tools, such as Siemens PLM Software’s NX software. By using NX – which integrates high-end analyst modelling tools with world-class geometry capabilities, and combining it with data management software, such as Siemens PLM Software’s Teamcenter, users are able to develop and share analysis models faster than with traditional CAE workflows.
Changing the culture
The simulation-driven design process represents a significant cultural change for many companies. In addition to new technology, these requirements will also call for dramatic changes in processes and attitudes. It may mean re-organising the way groups work together and it is likely to mean changing old habits.
People that used to hold data for as long as they could to get as much context from other groups and to minimise their chances of revision, for example, have to be encouraged to release preliminary information early to support a more rapid process. Each group working on the design has to learn the needs of the other groups.
This kind of concurrent and collaborative engineering – where 3D models, data and results are shared so everybody is able to see the geometric model in real time – creates workflows that help facilitate design reviews and enables multiple departments to review and approve or reject the design. As an added benefit, by properly synchronising and centrally managing processes, companies can more easily adhere to all regulations around tracking and traceability.
Simulation guides critical trade-off decisions to balance competing product objectives such as reliability, cost and weight requirements. But it also plays a major role in product and process innovation. New ideas have to be tested, qualified and refined before they can be put into practice or introduced into the market.
Usually hundreds of concept alternatives are evaluated before detailed design is begun. In the past this required physical testing, and so very few design alternatives or radical ideas were tested and products evolved slowly.
Advances in technology and processing power, combined with the coherence of a centralised design and analysis hub mean that there is much greater scope for experimentation. This is crucial to innovative product design. If the simulation and feedback process is fast enough, even radical but unfeasible ideas can reveal useful information and increase insight into a design. This is especially true early in the design process when you want to eliminate potential ‘losers’ and focus subsequent efforts on a smaller set of potential ‘winners’.
The message is clear – by putting simulation at the heart of the design process, engineers are better able to understand, predict and improve product performance digitally. More design concepts can be explored, which in turn reduces direct costs associated with expensive physical prototypes and enables faster, more informed decisions.
Jan Larsson is marketing director – NX, EMEA Marketing at Siemens PLM Software