2014 MS&E Innovation Summit - Save the Date

Speaker Abstracts

Day 1

Presenter: Dr.Krishnan Viswanath , Global Solutions Executive - Chemicals & Petroleum, IBM
Presentation: Smart Operations - Making Smarter Decisions Faster

Challenges such as varying feedstocks, market variations and scarce expertise lead to tough questions being asked of operations.  Decisions about process conditions and targets affect the yields, conversions, energy usage, qualities and throughputs, which directly impact the margins of the business. Economics and planning decisions about operations' targets and inventory levels affect margins and multiple cost elements such as working capital. Decisions about asset maintenance affect overall equipment life, reliability, and maintenance costs. The time spent gathering data to make operational decisions affect productivity. In the past, steady operations and reliable workflows were considered as being sufficient for success. Moving forward, Information Visibility, Predictive Analytics, and Sustainability have become some of the key focus areas for the value chain.

Therefore, Smart Operations, enabled through a set of work processes, technologies and skills, is essential to enable making ‘Smarter Decisions Faster’

Presenter: Michael Elliott, CEO, Atrium
Presentation: New Paths for ELN

The market for ELN is quickly separating into two as it expands beyond its traditional biopharmaceutical base.  Capabilities once desired in a monolithic platform are increasingly being viewed as restrictive and complex.  This talk will examine the transformation of the market and will offer a perspective on the future application of the technology.

Presenter: Bryan Billings, Senior Tech Support Engineer, Accelrys CISPro
Presentation: EH&S Compliance through Electronic Inventory Management

Laboratory safety and insight is vital for environments where accidents or oversights within a lab can have dire consequences.  Regulatory compliance is becoming a more critical and difficult task for organizations to comply due to new and continually evolving requirements.  Explore the main challenges for Environmental Health and Safety (EH&S) and learn how CISPro, the world’s leading laboratory inventory management solution, can provide a foundation for inventory control and EH&S reporting.

Presenter: Ian Stott, Informatics & Maths Science Leader, Unilever
Presentation: Using Pipeline Pilot to run 1 million Genetic Algorithms on the Daresbury Supercomputer

Genetic Algorithms have been widely used for many years as a method for solving optimisation problems. However, their conceptual simplicity has proven to be both a blessing and a curse. Their blessing is that they are a simple and powerful method that can be applied to a vast range of optimisation problems. Their curse is that this simplicity means that very many different genetic algorithm variants have evolved over the years – which poses the non-trivial problem to a user of Genetic Algorithms: which variant should I use for a particular optimisation problem?
This talk will describe one approach to this problem. A generic genetic algorithm component has been developed within Pipeline Pilot. By installing and running Pipeline Pilot on the STFC supercomputer at Daresbury, this has enabled a multitude of different variants of the genetic algorithm to be compared. This talk will describe the implementation of a generic genetic algorithm component within Pipeline Pilot and the details behind running Pipeline Pilot protocols on massively parallel hardware.

Presenter: René de Ruiter, Catalyst Development and Support, Shell Products & Technology
Presentation: Getting the Best Out of Experimental Research. An Improved Data Workflow Approach

Data collection around complex experimental workflows has always been a challenge. An improved data workflow approach is developed to address this complexity.  The presentation will address what has been the ratio behind this approach, the key workflow requirements and some results are shared.

Day 2

Presenter: Reza Sadeghi, Materials Science & Engineering Managing Director, Accelrys
Presentation: Predictive Materials Science Today and Tomorrow

Predictive science is today playing a key role in helping researchers to understand the link between molecular structure and material properties in order to optimize product performance.  This talk will cover Accelrys’ future direction for predictive science, and how it will grow and evolve to become more impactful in the full product development lifecycle by supporting the discovery, application, and production of materials as a part of product systems.

Presenter: Matthew Jackson, Scientist, Cytec
Presentation: Understanding Reaction Induced Phase Separation via Atomistic Simulations

Thermoset epoxies are strong and stiff, but are notoriously brittle.  Toughness can be realized through blending with certain thermoplastics.  However, toughness enhancement without compromise of other properties requires control of the morphology of the blend.  The term “Reaction induced phase separation” is used to describe the circumstance where the thermoplastic is initially soluble in the monomers of the epoxy.  As the epoxy “cures” to form a network structure through reaction of the monomers, the thermoplastic phase separates.   Controlling when and how the phase separation occurs are vital to obtain the desired morphology and properties.  Atomistic simulations within Materials Studio 7.0 are being investigated to determine whether these methods can be applied to predict the phase separation behavior of this class of materials. 

Presenter: Benoît Schnell, Fellow, Michelin
Presentation: Multi-scale studies of polymers in a tire tread

A tire tread is a nanocomposite in a particular shape, which has to ensure good grip for safety without neglecting other performances such as wear or rolling resistance, the latter being actually directly related to the fuel consumption of the vehicles. It is well known that the material processing has a great impact on the properties of treads and more quantitative relationships are still under investigation. One possibility is to achieve improvements by considering the process/product as a whole, but another possible approach is to design an optimized structure for one or several properties by coupling theory / simulation / structure analysis, and then develop the process which would allow reaching the desired target. The presentation will focus on the step “design of an optimized structure”.

Several properties of the tire tread materials are mainly governed by the polymer dynamics. As the dynamic properties of a polymer melt involve a large interval of relaxation times, a multi-scale approach is necessary for its study. From atomistic simulations (Molecular Dynamics done with Materials Studio from Accelrys), we can only obtain information about the high frequency / high temperature polymers behavior. However, from the obtained atomistic configurations, we are also able to develop for different polymers the so-called “realistic mesoscopic interaction potentials” (usable in DPD or more general CG simulations) with an inverse iterative Boltzmann method [1], in particular deriving a method for the NPT ensemble [2]. These mesoscopic potentials allow to obtain good static properties such as the squared end-to-end vector Re2 or the rubbery plateau modulus Gp, which cannot be reached with classical MD simulations. Work is still in progress to also obtain the polymer dynamics from chemically realistic DPD simulations. Optimizing the polymer network topology and the effect of confinement on polymer chains is also under investigation through chemically realistic DPD simulations, similarly as what has been done in references 3 and 4 in order to understand respectively the non-affinity during swelling and the modification of the entanglement density near a wall.

[1] Lahmar F. et al., Onset of entanglements revisited. Topological analysis, Macromolecules, 2009, 42, pp 7474 – 7484.
[2]    Maurel G., Malfreyt P. and Schnell B., to be published.
[3]  Saalwächter K. et al., Swelling Heterogeneities in end-linked model networks : a combined proton multiple-quantum NMR and computer simulation study, Macromolecules, 2004, 37, pp 8556 – 8568.
[4] Vladkov M. and Barrat J.-L., Local dynamics and primitive path analysis for a model polymer near a surface, Macromolecules, 2007, pp 3797 – 3804.

Presenter: Jérôme Gomar, Research Scentist, L’Oréal
Presentation: Exploreal: Alerting on the danger of explosion

Keeping the safety at the work place at the highest possible level, is an important issue. In order to improve the risk analysis in chemical synthesis, The Exploreal tool, based on Pipeline Pilot collections has been developed, and transversally integrated within the L’Oreal Advanced research division. This presentation will give an overview of the development story of this application.

Presenter: Misbah Sarwar, Principal Scientist, Johnson Matthey Technology Centre
Presentation: Molecular Modelling : An Industrial Perspective

Atomistic modelling is playing an increasingly important role in understanding material-property relationships. This talk will give an overview of how Materials Studio is being used at Johnson Matthey to understand how our catalysts work and guide catalyst design. Case studies of where atomistic modelling using the GULP, Forcite and QMERA modules has been used to aid our understanding of emissions control catalysts and where CASTEP is being used to predict new alloys for our fuel cell catalysts and to help characterise some of the materials we make will be presented.

Presenter: Stephen Todd, Senior Product Manager, Accelrys
Presentation: Materials Studio 7.0 - "More Science. More Applications. More People."

With new features and user enhancements, Materials Studio 7.0 extends its leadership in modeling and simulation software as the easiest to use and most complete software for materials innovation.
Learn how new capabilities in solubility property prediction, force field enhancements, performance and visualization enhancements and prediction of electron transport properties will enable modelers to do more.

Presenter: Björn Winkler, Professor, University of Frankfurt
Presentation: Interpreting and complementing experimentally determined structure-property relations by DFT-based model calculations

Any interpretation of spectroscopic measurements beyond a phenomenological comparison requires the use of predictive and accurate models. For most crystalline compounds, atomistic model calculations based on the density functional theory-based codes CASTEP and DMOL3 provide a mature and reliable approach to compute a large number of physical properties and spectroscopic signatures. I will discuss the application of DFT-based calculations for the interpretation of infrared, Raman and inelastic x-ray scattering spectra, of electron energy loss spectra, and of ultrasound
spectra. The model calculations allow us to interpret the spectra quantitatively and to predict the effect of phase transitions. For some experiments, model calculations are required to plan the
experiments, e.g. by predicting the dynamical structure factor. Model calculations also allow us to complement experiments by computing structure-property relations for conditions which are difficult to achieve in experiments, or to systematically search for compounds with a specific desired property, and I will discuss examples for such applications.

Presenter: Marc Meunier, Principle Field Apps Scientist, Accelrys
Presentation: Automating Workflows for Property Prediction

With the recent releases of Materials Studio, the capability to connect directly to Pipeline Pilot has changed the way modelling can be used in an organization. This presentation will give an overview of how customers are automating tasks and talk about the advantages of using Pipeline Pilot in collaboration with Materials Studio.