Multi-scale Modelling Platform (MMP): Design of LED applications through distributed simulations

A.W.J. Gielen, F.O. Valega Mackenzie (Corresponding author), B. Patzák, V. Smilauer, O. Tapaninen, P. Myöhänen, Mikko Majanen, Aila Sitomaniemi, V. Hildenbrand

    Research output: Chapter in Book/Report/Conference proceedingConference abstract in proceedingsScientific


    The high-tech industry strives to increase overall functionality and quality of products by the application of nano-enabled materials and devices. The development of such products significantly benefits from a thorough understanding of multi-scale and multi-physics phenomena and adequate numerical tools to guide nano-enabled design. Multiscale modelling and therewith multiscale design will considerably reduce development costs, decrease time to market and improve process yield and device functionality. These approaches contain scientific and organizational challenges: - The main scientific challenge for building this platform lies in a proper definition of scale transitions and the associated information exchange between the relevant scales - As nano-engineering is intrinsically strongly multidisciplinary, the expertise and simulation resources are distributed over different companies, research institutes, and academic groups. The MMP project is developing an integrated modelling platform MuPIF (, especially equipped to target multiscale and multi-physics engineering problems. The innovation of this platform lies in its generic, modular, and distributed concept, supported by data standardization and proper definition of application and data interfaces. This allows integration of simulation software, either academic, open source, or propriety, and data repositories as plug-in components, without any necessity to have all software in one computer or even in one network. This eases the cooperation between parties as no sensitive data or models need to be handed over, as well as development of the different (sub)models and simulation software can be handled independently. Although MuPIF can handle quite complex data streams and simulation control over multiple simulation servers, the IT security policies of companies in general prohibits these kind of complex connections. A key aspect of the IT security policies is that only out-bound connections are allowed. Therefore we implemented a nameserver+hub approach (Figure 1) to conform to the company IT security policies, but harnessing the power of MuPIF. We will demonstrate the use of the models and platform in the current poster / presentation by assessing the performance of phosphor light conversion in LEDs in a distributed simulation chain. The opto-thermal simulation chain consists of four models: particle level scattering model, device level ray tracing model, and microstructural and device level thermal models. We will present particulars on the models and summarize the requirements for the application programming interfaces (API) to connect any software to the MuPIF platform. We intend to demonstrate the platform from the Plugfest location, using the simulation servers at VTT and TNO. The main results of this project discussed here are two-fold: (1) We created a state-of-the-art simulation model for the opto-thermal behavior of LED's, and (2) we can run the simulation chain over the internet, with direct and immediate data transfer between the different models, including remote control of the simulation chain.
    Original languageEnglish
    Title of host publicationMultiscale Simulation
    Subtitle of host publicationFrom Materials through to Industrial Usage
    PublisherUniversity College Dublin
    Publication statusPublished - 2016
    EventMultiscale Simulation: from Materials through to Industrial Usage - Dublin, Ireland
    Duration: 5 Sept 20167 Sept 2016


    ConferenceMultiscale Simulation: from Materials through to Industrial Usage


    Dive into the research topics of 'Multi-scale Modelling Platform (MMP): Design of LED applications through distributed simulations'. Together they form a unique fingerprint.

    Cite this