R&D Activities and Discussions

Prof. Allstair Borthwick led the presentations to theWTEC visiting team. He indicated that Oxford’s Department of Engineering Science teaches computational methods to undergraduates. It usually takes about 6 months to train graduate students to use simulation and modeling software (typically by their supervisors). They have excellent computer lab facilities; students take the lab course (e.g., MATLAB) for one term, about 90 minutes per week. The department needs more next-generation computing capabilities. About two-thirds of the students are trained in computer imaging, especially in the bio-med areas. Overseas students are mostly from China and India.

The following is a summary of the presentations and discussions.

Allstair Borthwick

• 
  River co-sustainability: Yellow River does not have flow to the sea for 200 days a year
  
• 
  Simulation of dyke break, solitary waves; some collaboration with UCLA and Delft
  
• 
  Urban flooding—Thamesmead
  
• 
  Chaotic advection; lead-zinc mines in Canada
  
• 
  Tsunami simulation
  
• 
  Waves up the beach
  
• 
  Ice mechanics
  

40 years of CFD (computational fluid dynamics)

• 
  Interactions with offshore structures (Question: Ekofisk offshore design in North Sea)
  
• 
  Resonances in gaps between vessels, liquid natural gas tanker mechanics.
  
• 
  Wave energy
  
• 
  Fluid structure interaction (BP-funded)
  

• 
  Coastal engineering
  
• 
  Wave diffraction; develop software—DIFFRACT
  
• 
  Wave hitting a structure; can go up to 35 meters
  
• 
  Offshore platforms (with BP)
  
• 
  Extreme waves on deep water
  
• 
  Raleigh tail study
  
• 
  Energy and physics studies
  
• 
  Boussinesq wave modeling
  

• 
  Power grid; airline hubs simulation, etc.
  
• 
  NY garment industry
  
• 
  Data by UNITE, 7000 firms to 300 firms
  
• 
  Defects spread, affected by 40% of network
  
• 
  Simulated 1000 times for random input; same result
  
• 
  Engineering networks simulations: computers/telecoms, electricity, etc; fungal networks vs. minimum spanning tree
  
• 
  Manufacturing systems simulation: reconfigurable, make to order, cycle time, etc.
  

• 
  Computational turbomachinery applications (funded by Rolls Royce)
  
• 
  Turbine blades, flutter, optimization of blades
  

• 
  Biomechanics
  
• 
  Clinical driven collaboration with Princeton, Harvard, and others in the United States
  
• 
  Patient-specific aneurysms; modeling and curing
  
  
  

Prof. Graham Richards

Dr. William Barford

Dr. Jonathan Doye

Dr. Peter Grout

Prof. David Manolopoulos

Dr. Mark Wilson

The University of Oxford has a long tradition in theoretical chemistry, going back to the 1956 Nobel Prize-winning work of Hinshelwood on chemical kinetics and the work of Coulon’s group on the theory of valence. Today the Theoretical Chemistry Group has 10 academic staff and approximately 30 students and postdocs studying with them. Prof. David Logan is the head of the group, and serves as the Coulson Professor.

During the WTEC visiting team’s two-hour visit to the Oxford Theorgy Group, our hosts first briefly introduced their research interests. Prof. Logan focuses on the study of strongly correlated electrons in condensed matter; Dr. Grout studies defects in solids. Prof. Richards is a widely recognized pioneer in biological simulations. Dr. Barford studies electronic properties of conjugated polymers. Dr. Doye works on simulations of soft condensed matter with a focus on biological systems, and Dr. Wilson also studies condensed matter systems including low-dimensional crystals, glasses and networks, and self-assembly. Finally, Prof. Manolopoulos is interested in dynamics and has recently developed tractable semiclassical methods for treating condensed matter systems.

Different parts of the research depend to very different degrees on high-performance computing. The soft condensed matter groups (Doye and Wilson) are in the process of installing a large local cluster to meet a significant part of their computing needs. Oxford University itself is in the process of moving its supercomputer facility to a new site that will accommodate a 15 TFlop machine (at a cost of approximately $6 million). Additionally, the Oxford campus as a whole is experimenting with grid computing through the creation of the OxGrid facility, which uses spare cycles from across the campus. The new Doye/Wilson cluster received local funding support that was conditional on it becoming a part of OxGrid.

The WTEC team exchanged views with our Oxford hosts on a variety of issues regarding opportunities and challenges in research and training in simulation-based engineering and science. Some of the perspectives expressed by the hosts included the following points:

• 
  At Oxford, high-quality students and postdocs are readily available and comprise one of the most important strengths of the theory program. The hosts felt that members of their group were able to become adept quite quickly at developing appropriate tools, whether they be formal models or computational methodologies.
  
• 
  Tie-ins to industry can be very strong, as exemplified by Prof. Richard’s personal involvement in the founding of several companies. However this remains the exception rather than the rule. Most of the Group members operate in the traditional academic manner of pursuing curiosity-driven research.
  
• 
  Towards the physical end of theoretical chemistry, funding is a challenge, and the funding models have changed in several important respects. First, the Group no longer receives significant numbers of studentships directly to redistribute to its academic staff. Second, scientific grants to researchers are now funded based on full cost accounting; it is not yet clear whether this will lead to increased opportunities or not. National and international collaborations of RICS researchers are increasing significantly. Third, there is now a national Materials Modeling Consortium (started about 3 years ago) that brings together teams and provides some additional funding (overall at the rate of about $2 million/year).
  
• 
  Towards the biological end of theoretical chemistry, the funding situation is considerably better, due primarily to the availability of the Welcome Trust, which sponsors research relevant to human and animal welfare.
  
• 
  Funding for the new Oxford supercomputer facility is for hardware only.
  
  
  

Dr. Philip Biggin, P.I., Computational Studies of Receptors

For the past fifty years, the Department of Biochemistry at the University of Oxford has been one of the preeminent institutions in the SBES field with pioneering contributions in metabolism, immunology, protein structure and function, microbiology, cell biology, and genetics. In 2008, the department will move to a new, state-of-the-art building. Within the department, three faculty members are the most active in simulation-based research. Of the three, Prof. Mark Sansom is the most senior, and his research group (Structural Bioinformatics and Computational Biochemistry Unit), composed of 26 researchers (13 postdocs and 13 graduate students), is the largest. Dr. Philip Biggin and Dr. Bela Novak are more recent arrivals to the department, and their newer research groups (each circa 5 researchers), are expected to grow over time.

As noted during our earlier visit to the University of Oxford, the university is in the process of moving its supercomputer facility to a new site that will accommodate a $6 million, 15 TFlop machine. Larger-scale national facilities (HECToR) are also important for compuational simulations of biological macromolecules.

In lieu of formal presentations, our two hosts engaged the WTEC visiting team in a free-format discussion on issues pertaining to SBES activities (computational biochemistry) in their research groups, in their department, in their interdepartmental collaborations, and in the biochemistry landscape in the UK. In the first part of this discussion and after our overview of the goals of the WTEC-SBES study, we were briefed on the research interests of two of our hosts, Prof. Sansom and Dr. Biggin (Dr. Novak was on travel and not on campus during our visit).

Prof. Sansom’s group is working on topics ranging from the dynamics of water in nanopores to large-scale MD simulations of bacterial membranes. The group’s computational studies of membrane proteins range from molecular simulations of channels and transporters, to computational bionanoscience and membrane protein folding and stability. In his research description, Prof. Sansom notes that an estimated 50% of potential new drug targets are protein membranes (see the group’s research website http://sbcb.bioch.ox.ac.uk/index.php). Also of relevance to our report’s life sciences chapter, Dr. Biggin’s simulation-based research activities are centered around receptor dynamics and ligand binding. Dr. Bela Novak’s interests are in computational and mathematical modeling of regulatory networks.

In the second part of our discussion, our hosts considered some of the broader SBES perspectives aligned with our list of questions. The following points were particularly memorable:

• 
  Prof. Sansom sees the growth of computational biology along two axes: one axis is the connection with experiments and closer collaborations with other disciplines; and the other is the build-up of the systems biology framework.
  
• 
  High-quality graduate students and postdocs drawn from the global pool are readily available and interested in joining their research groups (the strong reputation of Oxford and its effect on student recruitment was a recurring theme throughout our visit to the four departments at Oxford). For large-scale simulations, e.g., of protein membranes, the students do not develop new codes but use the standard MD packages on the major computational facilities.
  
• 
  Consistent with the WTEC team’s findings throughout the week, the biologically oriented research groups at Oxford biochemistry are well funded, and scarcity of research funding does not appear to be an issue. In addition to the BBSRC (Biotechnology & Biological Sciences Research Council, http://www.bbsrc.ac.uk) the funding stream emanating from the Wellcome Trust (the world’s largest medical research foundation funding research on human and animal health, http://www.wellcome.ac.uk) provides a major boost for high-quality research in the biological sciences, including SBES activities therein.
  
• 
  The near completion of the new biochemistry building and the pending move to these state-of-the-art facilities is a major boost to the research efforts of the department, including those of our hosts.
  

Pfaffenwaldring 9

G. Lewison, P. Westmoreland

Rainer Helmig, Civil Engineering

Barbara Wohlmuth, Mathematics and Physics

Wulfgang Ehlers, Civil Engineering

Haus Hasse, Mechanical Engineering

Thomas Ertl, Computer Science

The University of Stuttgart took its present form in 1967; its present rector is Prof. Dr. -Ing. W. Ressel. The university has 20,000 students (approximately 5000 are international) distributed in 10 faculties (departments), and 5000 staff. It is a research university with a focus on engineering and the natural sciences. Relevant faculties (colleges or departments) in these areas are Civil and Environmental Engineering; Chemistry; Energy Technology, Process Engineering, and Biological Engineering; Computer Science, Electrical Engineering and Information Technology; Aerospace Engineering and Geodesy; Engineering Design, Production Engineering, and Automotive Engineering; and Mathematics and Physics. Key research areas are modeling and simulation, complex systems, communications, materials, technology concepts and assessment, energy and the environment, mobility, construction and living, integrated product and product design.

The university has an international range of study programs, including seven master’s courses taught in English. In addition, is has a number of collaboration programs. It has an inter-faculty (inter-departmental) research structure, with the different faculties linked through transfer and research centers interacting with national institutions (Max Planck Institutes, Fraunhofer Institutes, German Aerospace Center), international institutions, and industry. (The university sits in one of Europe’s strongest economic regions: Bosch, Fischer, Daimler, HP, IBM, Festo, Pilz Deutschland, Porsche, Trumpf, Stihl, Züblin, and BASF all have facilities in the area).

Stuttgart has a number of superlative technical and computing facilities and projects. These include Europe’s fastest vector computer (12.7 TFLOPS computing power); Gauss Centre for Supercomputing (Europe’s most powerful high-performance computing alliance among Jülich, Munich, and Stuttgart); VISUS (visualization Research Center, one of the leading facilities in the region); ASCS (Automotive Simulation Centre Stuttgart, in close cooperation with industry since 2007); Archi-Neering (Bangkok’s new airport is the result of close cooperation between Civil Engineering and Architecture); Baden-Württemberg Astronautics Center (opening in 2010); IZKT (International Center for Cultural and Technological Studies); SOFIA (joint U.S.-German project, a Boeing 747SP, equipped with a high-performance mirror telescope); wind tunnel (tests of aerodynamic and aeroacoustic properties of vehicles, up to 265km/hour); and VEGAS (subsurface remediation facility for the simulation of contamination processes).

DFG has provided support for collaborative research centers (SBF), transregion projects (TR), transfer units (TBF), research units (FOR), Priority programs, and “Excellence Initiatives.” Many of these (listed below) are entirely based on or have major components in simulation-based engineering and science:

• 
  DFG Collaborative Research Centers (SBF) and Transregion Projects (TR): Spatial World Models for Mobile Context-aware Applications; Selective Catalytic Oxidation of C-H Bonds with Molecular Oxygen; Dynamic Simulation of Systems with Large Particle Numbers; Incremental Specification in Context; Control of Quantum Correlations in Tailored Matter (TR involving Stuttgart/Tübingen/Ulm). The budget of these centers is about €2–3 million/year for 12 years, in three 4 year terms.
  
• 
  DFG Transfer Units (TFB): Simulation and Active Control of Hydroacoustics in Flexible Piping Systems; Development of a Regenerative Reactor System for Autothermal Operation of Endothermic High-Temperature Syntheses; Transformability in Multivariant Serial Production; Rapid Prototyping; Computer Aided Modelling and Simulation for Analysis, and Synthesis and Operation in Process Engineering. These units execute transfer from university to industry.
  
• 
  DFG Research Units (FOR): Nondestructive Evaluation of Concrete Structures Using Acoustic and Electro-magnetic Echo Methods; Development of Concepts and Methods for the Determination of Reliability of Mechatronics Systems in Early Stages of Development; Noise Generation in Turbulent Flow; Multiscale Methods in Computational Mechanics; Specific Predictive Maintenance of Machine Tools by Automated Condition Monitoring; Positioning of Single Nanostructures—Single Quantum Devices. These units are smaller than the centers, generally consisting of 4–5 researchers.
  
• 
  DFG Priority Programs: Molecular Modeling in Chemical Process Engineering; Nanowires and Nanotubes: From Controlled Synthesis to Function. These programs bring €1.2–2 million/year (spread over 10–20 projects). Ten new priority programs are opened per year. The call involves preproposals, out of which 80 are invited for full proposals to compete for 10 awards. Reviewers are academics and convene in panels. Generally, the PI meets the reviewers.
  
• 
  Excellence Initiative: The goal of the Excellence Initiative, planned for an initial period of 5 years, is to foster excellence in science and research and to raise the profile of top performers in the academic and research community by means of three lines of funding: strategies for the future; excellence clusters, and graduate schools. The University of Stuttgart has been successful in two of the Excellence Initiative’s three lines of funding. Funding has been granted to the Advanced Manufacturing Engineering Graduate School and the Simulation Technology Excellence Cluster.
  

The Simulation Technology Excellence Cluster is a University of Stuttgart DFG Excellence Initiative award. The SimTech Excellence Cluster is coordinated by Dr.-Ing. Wolfgang Ehlers, and the initiative covers topics from isolated numerical approaches to integrative systems science. Interestingly, in order to obtain funding, the group gave arguments partly based on the U.S. NSF Blue Ribbon Panel Report of February 2006: “… [C]hallenges in SBES … involve … multi-scale and multi-physics modelling, real-time integration of simulation methods with measurement systems, model validation and verification, handling large data, and visualisation.” “… [O]ne of those challenges is education of the next generation of engineers and scientists in the theory and practices of SBES.” DFG has agreed wholeheartedly with this report and provided funding accordingly. The SimTech Excellence cluster brings €7 million/year for 5 years.

The SimTech cluster ultimately aims at linking three types of interactions:

• 
  the interactive handling of multiscale, multiphysics systems under consideration of uncertainty by simulation in order to gain a qualitative and quantitative understanding, to predict the system behavior, and to prepare decisions
  
• 
  the interactive situation-adapted or context-adapted optimization of systems that are influenced by sensor data streams in real time
  
• 
  the interaction between developers and systems for evaluating the consequences of system behavior for management, optimization, control, and automation
  

• 
  from Empirical Material Description towards Computational Material Design
  
• 
  towards Integrative Virtual Prototyping
  
• 
  towards Interactive Environmental Engineering
  
• 
  from Classical Biology to Systems Biology
  
• 
  from Biomechanics towards the Overall Human Model
  

The only way to achieve this integrative, simulation-based vision is through a long-term sustained research agenda in the following research areas:

1. 
   Molecular and Particle Simulations, serving for nano and micro simulations, and bridging scales by coupling particle dynamics with continuum mechanical approaches
   
2. 
   Advanced Mechanics of Multiscale and Multifield Problems, exhibiting the basic scientific key for the description of complex problems in almost all branches of engineering simulation and design by bridging scales, coupling physics, and linking domains
   
3. 
   Systems Analysis and Inverse Problems, focusing on model validation, parameter identification and model reduction, as well as dynamical system analysis, control, aspects of autonomy, automation and hierarchical or networked structures of systems
   
4. 
   Numerical and Computational Mathematics, guaranteeing advances towards multiscale and multiphysics numerical models, including the quantification of uncertainty and a self-adaptive choice of scales and physics in order to simulate dynamic and coupled processes in complex real-world problems
   
5. 
   Integrated Data Management and Interactive Visualization, mastering the explosion of information in simulation technology through human-system interfaces for model setup, real-time simulation, control, and interactive visualization, and through design of sensor networks for real-time control simulations
   
6. 
   Hybrid High-Performance Computing Systems and Simulation Software Engineering, with the basic idea of harnessing the power of large-scale systems through advanced simulation software technology to solve grand challenge problems
   
7. 
   Integrative Platform of Reflection and Evaluation, which contributes to the above research areas regarding Theory of Science, Philosophy of Technology, Sociology of Technology, and Ethics
   

• 
  Vaporisation of a droplet in an air stream, a fully 3D enhanced visualized phase transition problem (B, E)
  
• 
  Salt water on fresh water in a porous medium, a multifield flow simulation obtained by parallel computing (B, F)
  
• 
  Collision of methane and ethane, a molecular dynamics nanoscale simulation with 3D point-based visualization (A, E)
  
• 
  CO₂ sequestration in a geological formation, systems analysis based on a multiscale and multiphysics simulation (B–F)
  
• 
  Flexion of a lumbar spine, a multifield and multiphysics biomechanical simulation (B–D)