Postgraduate Courses in the
School of Computing Sciences

MSc in CAE Techniques

Course Description

The term Computer Aided Engineering (CAE) covers the use of computers in all activities from the design through to the manufacture of a product which can be distinguished from the organisational processes of running a manufacturing operation. It is difficult to draw precise boundaries, but at present, CAE may be said to encompass all those activities carried out once per product or once per batch of products. These include various stages of design, analysis and the planning of the manufacturing process as regards both overall strategy and detailed production methods.

CAE is a fast-moving subject of crucial importance to industry. Its main focus of attention is on the generation of computer understandable product descriptions (which will eventually come to replace the traditional engineering drawing) and their applications in a manufacturing environment. Looking to the future, CAE will increasingly become embedded in a larger organisational framework, the resulting synthesis being known as Computer Integrated Manufacturing (CIM). CAE is nonetheless an important subject in its own right and a key element without which CIM would never be possible.

The course described here deals primarily with mathematical methods and software techniques for CAE, though frequent reference is made to the applications of CAE to various branches of engineering and to the wider context of CIM. It has been designed to provide graduates with programming skills in CAE and with a broad range of interdisciplinary skills including the flexibility of approach required for success in the rapidly developing world of industrial automation. The course is continually changed and modified in line with the needs of industry and in particular, the expectations that industry has of modern day Software Engineers.

Introduction to the Course

Over the past decade, European industry has invested heavily in computer aids for engineering design and manufacture. However, there is a wide disparity in approach. Some companies have acquired ready-made turnkey systems which they hope will meet all their requirements; others are developing systems of their own in-house, whilst others are purchasing separate packages covering various aspects of the design and manufacturing process and endeavouring to interface them to each other. The latter modular approach is likely to become fairly standard in the future, since it enables an organization to choose the particular items of software which best suit its specific purposes and to supplement them with modules written in-house to handle particularly specialised functions.

The flexible modular CAE systems of the future will be capable of dealing with all the separate aspects of design, analysis, manufacture and final inspection in a unified manner, with information flow between the various processes handled automatically by computers. The greater the degree of integration, the greater will be the benefits in productive efficiency, since expensive human interaction with the system will be minimized. However, many problems remain to be solved before complete integration can be achieved, particularly those concerned with interfacing the components of a modular system.

It seems very probable that the techniques of artificial intelligence will often be required to handle the transition between the diverse types of data needed in different parts of such a system.

There is currently a major shortage of specialists in the design and implementation of CAE systems and their component sub-systems. This situation will become worse as more effort needs to be applied to solving the problems of CAE system integration. This course is designed to provide a supply of these software specialists. It aims to provide the interdisciplinary knowledge and skills required for the planning and implementation of an integrated CAE system. Graduates of the course are in demand by manufacturing industries, industrial research organisations and commercial software developers.

Entry Qualifications

The normal entry requirement for the MSc course is a good honours degree in a numerate subject such as Computer Science, Engineering, Mathematics or Physics. In some circumstances combinations of other qualifications and practical experience may be regarded as equivalent. The necessary level of mathematics is covered by most degree courses in the subjects mentioned; specifically, a good grasp is required of linear algebra and numerical analysis. Some computing experience is also desirable but intensive courses in Fortran and C/C++ are available for those students with limited programming experience. Above all, prospective students must be capable of the breadth of approach appropriate to the multidisciplinary nature of the course.

Employment of Graduates

Enquiries regarding availability of potential employees are received from many quarters, both in the EU and elsewhere. There is considerable demand for personnel with expertise in CAE and graduates of this course have been particularly successful in finding long-term employment in both the EU and overseas industries and academic institutions. Some graduates go on to register for PhD degrees, many, on the basis of their MSc research project. Some thesis topics are supplied by individual companies on in-company problems with the view to employment of individuals after graduation - an approach which is being actively encouraged by a growing number of industries. De Montfort University has one of the highest records for employment of its graduates and postgraduates of any UK University.

Course Content and Rationale

At the heart of any integrated CAE system must lie the means for generating a computer representation of the product to be manufactured. Traditionally, this representation was in the form of drawings and it has been possible for several years to use computers to help in the generation and storage of 2D graphical information of this type. However, drawings are essentially intended for human rather than machine interpretation and do not lend themselves to use as the basis of automation of other processes such as manufacturing. During the past few years other, more unified and informationally complete forms of product representation have been developed, including 3D wireframe, surface and solid models. This last type of representation provides a complete and unambiguous description of the geometry of a product and current endeavours are directed towards the development of yet more complete product models containing the wide range of technological data required for automated applications.

The course is concerned with the software techniques and the engineering applications of CAE. Software techniques for CAE discusses the mathematical and computational methods which form the basis of modules entitled Geometric and Solid Modelling, Finite Element Analysis, Modal Analysis and Multi-Body Simulation. These modules cover the application of CAE algorithms for the generation and analysis of product models used to represent the geometry of such complex artefacts as car bodies and aircraft. The modules are designed to provide students with the programming techniques required to write CAE software. This is done by encouraging students to answer problem sheets whose solutions are compounded in the design, coding and testing of various CAE algorithms. In this way, students design and build their own library of CAE functions which, together with a written examination, is assessed at the end of the course. This library is used as a tool for the MSc project undertaken in the third Semester and can be a valuable asset to future employment

The material covers the utilization of CAE software packages for design and analysis. This includes product engineering, product modelling, types of geometric (solid) models, engineering analysis and manufacturing. Stuents are instructed on techniques for modelling complex 3D geometries that often arise in industrial design work and use a range of software for this purpose on both high and low level platforms. This work is the basis of a module entitled CAE Applications and Programming.

The core modules are supported by lectures on various aspects of computing and technology. Relevant material from numerical analysis is introduced to provide the computational basis for some of the techniques used. Solution and visualization packages are available for this purpose. Computer science topics taught include Data Processing Systems and Networks and Human-Computer Interfaces for CAE.

All this material is important in the development of large-scale interactive graphical systems for computer aided design and manufacturing. Computer communications and data exchange is also fast becoming a key aspect of CAE and various aspects of this topic are explored.

The module on Geometric and Solid Modelling discusses the computational techniques used to represent product models as they occur in the real world, i.e. with finite volumes, surface areas and densities.

The module on Finite Element Analysis is provided to teach students the computational background of the computer packages used throughout the year to conduct physical analysis (stress analysis for example) on various product models. This module includes instruction on other methods of analysis such as Finite Difference Analysis, Boundary and Volume Element Techniques.

A new technology which is starting to have a major impact on CAE is Artificial Intelligence. This technology effects both the design process (by providing specialised advice to the designer on many matters as the design proceeds) and subsequent applications (such as the automatic generation of mesh models for finite element analysis or for manufacturing process plans).

Weekly seminars and demonstrations on topics related to CAE are given by speakers from industry and academia. These help to reinforce the material taught and widen its areas of application.

Course Structure

Lectures are confined to the first two academic Semesters, starting in early September, and students are initially assessed by a combination of examinations, course-work and minor project-work. The first Semester will take place at the Fachhochschule Bielefeld (FB). The second Semester will take place at De Montfort University Leicester. The project undertaken in the third Semester will normally take place at FB.

Course Summary

Seminars

Students are additionally required to attend weekly seminars given by invited speakers from academia and industry throughout the duration of the course. The material discussed at these seminars is not formally assessed but forms an important contribution to the general education of students. The principal aim of the seminar programme is to introduce students to the kind of engineering problems and CAE techniques being used in industry and/or current research interests based in both the industrial and the university sector.

Project Skills

In the second Semester students are required to attend a project skills unit which is designed to prepare them for the major project undertaken in the third Semester. The aim of this unit is to:

1. Consider the steps required in defining a project and establishing the work plans necessary for its efficient completion and time management.
2. Techniques for surveying appropriate literature on a research theme including electronic library support, literature searches and the internet.
3. Technical writing as required for interim research reports, research monographs and publications.
4. Preparing the thesis, its presentation, order and content.

Projects are chosen during the second Semester from a list giving the title of the project, a brief summary, the expected deliverables and the name of the supervisor. In the case of externally sponsored students, the project is specified in consultation with the sponsoring company. Full-time supervised work on the project starts in the third Semester and lasts until late August/early September, when theses are submitted. An oral examination is held in late September in the presence of the external examiner when a final assessment of the students performance is given and the course ends. Each student is required to report to their supervisor on a regular basis to discuss the progress being made, indicate any problems encountered and to demonstrate their work. Students are strongly advised to present their thesis to their supervisor for proof reading and general appraisal before submission.

Students are expected to make a commitment to a particular project early in the second Semester and, together with their supervisor, develop an appropriate workplan. The logistics associated with the project such as access to a particular platform, necessery software, making copies of appropriate publications etc. should be finalised during the second Semester so that there is minimal delay in the execution of project over the duration of the third Semester.

Teaching and Learning Strategies

The course adopts a hands-on approach to programming the mathematical functions required by CAE systems. Each set of lectures is accompanied by practical computing laboratory sessions in which live programming demonstrations are held - the lecturer writing a Fortran/C/C++ subroutine/function with the aid of an LCD panel. The items associated with the practical implementation of this approach are as follows:

1. Formal 2 hour lecture given in the morning (10:00 - 12:00 am). Each student will be issued with a set of notes which contain:
   • Complete copy of overhead slides used during lectures
   • Set of supplementary notes prepared by the lecturer
   • Set of problem sheets to be attempted by students
   • Set of previous examination papers (where appropriate)
   • Example software (where necessary)
2. The software written by students for a module and/or the development work undertaken will be assessed and graded accordingly. Each student will be required to adhere to a set pattern of function design and testing.

Module Specifications

Semester 1

CAE Applications and Programming

This will be followed by the manufacturing process of prototypes, called rapid prototyping, and a comparison of the many different ways of producing prototypes. After the necessary modification of the product (often arising from the prototype), the manufacturing process and the simulation of this process on the computer will be considered. The aspects of simulation of this process on the computer will be considered together with the simultaneous engineering and the data exchange between different CAE systems. Finally, reverse engineering (i.e. the automatic generation of a CAE representation from an existing component) and its growing importance in CAE will be discussed.

Syllabus

Computer Aided Design (CAD)

• Basic 2D and 3D techniques (wireframe, surface, solid and hybrid models, layers, macros, parametrics)
• Basic geometric data processing (numerical descriptions of 2D and 3D structures)

Computer Aided Engineering (CAE)

• Different kinds of simulations
• Optimization (kinematic/dynamic; FEM, BEM)
• Optimization
• Rapid Prototyping
• Simultaneous engineering
• Data exchange
• Data management control systems
• Knowledge-based systems
• Reverse engineering
• Surface reconstruction

Assessment: Examination and course work

Syllabus

• Single-degree-of-freedom
• Modal parameters in the systems frequency domain
• Frequency response function
• Hysteretical and viscous damping
• Phase resonance and phase separation
• Normal frequency and damping ratios
• Observability and controllability
• Multi-degree-of-freedom systems
• Frequency response analysis
• Equations of motion
• Modal partial fractions representation
• Eigenvalue/eigenvector calculation
• Simulation of structural responses
• Modal mass, stiffness and damping matrices
• Modal parameters in the time domain
• Normalization and orthogonality
• Generalized coordinates
• Time domain procedures
• Proportional and non-proportional damping
• Eigenvalue/eigenvector determination
• Real and complex eigenvectors
• Coupling finite element results with measured normal modes parameters

Assessment: Examination and course work

Syllabus

• Modelling and analysis of kinematics
• Plain kinematics
• Absolute and relative constraints
• Differential equations of motion
• Transitional and rotational joints of the centres of gravity and mass moments of inertia
• Absolute and relative drives
• Locations, velocities and accelerations
• Dynamics of plain systems
• Variation equations of Newtonian mechanics
• Kinematic analysis
• Numerical methods
• Lagrange multipliers
• Inverse dynamics
• Organization of and conditions of equilibrium
• Dynamic assembly of the system

Assessment: Examination and course work

Syllabus

• Network analysis
• Network design
• Switching systems and oscillators
• Logic synthesis
• Digital systems design
• Sequential logic design
• Digital simulation
• Computer-aided test technology
• Digital signal processing
• Microprocessor engineering
• Computer-aided design and VLSI (very large scale integration)

Assessment: Examination and course work

Semester 2

Solid Modelling is a branch of geometric modelling that looks at methods for creating complete representations of physical solid objects. As such they enable one to determine all geometric properties of the object bounded by the surface.

This course provides a comprehensive introduction to the subject and its practical implementation. Emphasis will be placed on the use of the boundary representations for solid models, the associated Euler operators for their low-level manipulation and the higher level modelling operations that are built on top. Algorithms are included for some of the techniques covered.

Syllabus

• Different types of geometric model: Graphical, Wireframe, Surface, Solid
• Decomposition, CSG, B-rep, Hybrid
• Introduction to topology
• 2 manifolds, plane models, Euler characteristic
• Geometric modelling
• Manipulation of boundary models, Euler operators
• Implementation - Euler operators
• Basic modelling algorithms: sweeping, gluing
• CSG type manipulations - splitting and Boolean operations

Assessment: Examination and course work

The stability of solutions is investigated as is the approach required when considering coordinate geometries other than a Cartesian system. After an introduction to the historical background of the Finite Element Method as used to solve problems in structural analysis, the course goes on to show how these methods can be used to solve continuous field problems. It investigates the use of Finite Element Analysis as a general approximation method for the numerical solution of physical problems described by field equations in continuous media, actually containing many of the finite difference schemes as a special case.

Syllabus

• Basic concepts
• Applications to differential equations
• Finite differencing schemes
• Stability analysis
• Variational calculus
• Interpolation functions
• The Galerkin method
• Numerical integration
• Weighted residuals
• Conservation laws
• Assembly and solution problems
• Banded formulation
• The frontal method
• Substructures
• Element types
• Alternative formulation methods
• Dynamic problems
• Pre- and post-processors
• Mesh generators
• Boundary element methods
• Finite volume methods
• Spectral methods

Assessment: Examination and course work

Syllabus

• Mathematical deduction
• Theorem proving
• AI in automatic programming
• AI in engineering design
• AI in process planning
• AI in scheduling and robotics
• Programming languages for AI
• Expert systems
• Knowledge engineering
• Neural networks

Assessment: Examination and course work

Syllabus

• Processing Methodologies
• Process Scheduling
• System Calls in DOS
• System Calls in UNIX
• Interprocess Messaging
• The X-Windows System
• Motif programming
• Top-Down Design
• Graphics Standards
• X-designer
• Visual C++ systems
• J++ systems

Assessment: Examination and course work