Institution news
For full details about the Simulation and Modelling 2022 conference, please visit the event website.
Please briefly explain your current role and involvement in simulation and modelling technologies
Ricardo Paiva (RP): As a Design Systems Specialist at Rolls-Royce, I work on the development of software used to model the behaviour of real aircraft engine systems and components. From implementation of physics based algorithms to data centric models, I have delved in multiple domains ranging from aerodynamics to structural analysis in order to either automate or make these types of analyses computationally possible in the first place. The scope of the design systems we create can be focused on a single component like a fan blade, a compressor sub-system, or even a completely novel engine concept as a whole.
What is the number one challenge holding back the integration of simulation and modelling?
RP: Deployment to the end user in useful timescales. Agile software development processes have helped a lot in battling this in the last 10 years, but there is still much to be done. To have a customer wait for a more refined version of the software rather than provide a minimum viable product as soon as possible is a delicate balancing act. The employment of agile development frameworks teach us that it is almost always the best choice to deploy as soon as able and build up from there. However, defining what constitutes this minimum viable product is a big challenge, especially in tight project timescales, within which the software requirements specification, development, validation, and finally deployment, have to fit.
What are the most common gripes you hear from engineers in the deployment of digital twins?
RP: Digital twins have the uncanny ability of making an organisation more self-aware and honest with itself: when deployed they quickly make bare all the limitations and deficiencies of the physical modelling and techniques thus far employed. A classical example is that when a new parametric model of a system or component is embedded in an automated analysis chain and then wrapped in a DoE/optimisation loop, the latter will quickly find configurations within the design space where results do not appear to be physically sound or where the simulation fails outright. This is something that many see with suspicion, when what is actually happening is that the digital twin that was built and explored in this manner is simply highlighting the deficiencies in the modelling and simulation techniques used thus far (and doing so at a much higher pace than these limitations would have been found otherwise). As such, this ability should be embraced and harnessed to motivate further development of these or make them more robust in the future.
What key topics are you excited to discuss at this year's conference?
RP: I have a keen interest in topics on optimisation and machine learning techniques and their applications in the world of simulation. Also, I would like to learn more about novel modelling techniques for the analysis of structures manufactured via less standardised manufacturing processes (additive layer manufacturing, 3D printing, etc.).
What can engineers expect from your presentation at the event?
RP: The presentation will cover a range of topics around the development of simulation software at Rolls-Royce, in particular that of integrated design systems – software based workflows that chain together multiple tools in order to perform system/sub-system/component level design tasks. An example of such a tool – SDIM (System Design Integrated Modelling) will be given. SDIM is a tool that is used to perform engine level design iterations and engine life predictions.
Who else are you most interested in hearing from on the programme?
RP: There are three:
- Digital twin or digital cousin, IPG Automotive Ltd.
- The class of 2022 – bouncing cars and what it means in terms of digital twins and the validation of engineering simulation, Abercus
- Optimising high-performance electric cars with CFD simulations, Williams Advanced Engineering
What key development/s in simulation and modelling are you most interested in for the future and why?
RP: One aspect of software/model development that really needs streamlining is the structured validation of these using real-life data as much as possible as this is an onerous and usually not very well standardised process. New developments in embedding this in the software development process via the now ubiquitous tools for cloud based software configuration management and deployment using, for instance, Bayesian inference to assess and recalibrate software and models in the presence of new data, would be very welcome.
Why is it important for engineers to join this event?
RP: Events such as these are the lifeblood of the general Engineering Community as they promote sharing of ideas across organisational and technical field boundaries. What is now becoming commonplace for a particular field may become the new revolutionary application in another, and events such as this are one of the best ways for the relevant actors to come together to allow this to take place.
The Simulation and Modelling 2022 conference will take place on 20-21 September 2022 at the Manufacturing Technology Centre in Coventry
Join us in 2022 to hear how thought-leaders and senior engineers are using new technologies and techniques for digital twinning, design, validation, fault detection and process optimisation. Attendees will have the opportunity to engage with colleagues and peers across multiple industries and take back fresh perspectives of how these technologies are being used to optimise design, development, testing, manufacturing and operational engineering: for full details about the Simulation and Modelling 2022 conference, please visit the event website.