Two themes dominated this year's Meeting of Teachers of Geotechnical Subjects (MTGS 2000) hosted by University of Birmingham - the teaching of Eurocode 7 and use of the parameters c9 and f9.
Teaching and EC7 Dr Brian Simpson of Ove Arup and UK representative on Eurocode 7's drafting committee outlined why Eurocode 7 should be used.
Apart from necessity - because EC7 will replace national geotechnical design codes - Simpson said that in principle it will make it easier to work in other European countries. It should also lead to greater economy, however in the short term this benefit may be slight as it is 'difficult to break down national boundaries' A significant aspect for practising engineers is that EC7 achieves consistency with structural codes without compromising geotechnical principles, said Simpson. It defines characteristic and design parameters and provides a framework for calculations. It is good for complex problems and is structured in a way that should prevent designers overlooking potential design cases.
Simpson ended by highlighting the practical aspects of EC7 that teachers need to focus on: namely selection of parameters and characteristic values; the process of checking more than one set of factors; applying factors at source, not later in the calculation; and finally emphasising that the code provides a geotechnical design framework consistent with structural engineering codes.
Dr Trevor Orr, from Trinity College, Dublin, explained how objectives for geotechnical education could be incorporated into teaching based around EC7.The underlying principles of geotechnical education he said, were:
to develop good understanding of soil and its behaviour;
to show how the soil profile and the properties of soil strata may be determined from appropriate investigations and tests;
to teach appropriate models for analysing soil behaviour and predicting the behaviour of different types of geotechnical structures, such as spread foundations, piles, retaining structures and slopes;
to provide guidance on the design and construction of common geotechnical design; to encourage an inquisitive approach to soil mechanics;
and finally to impart a liking and enthusiasm for geotechnical engineering.
He said EC7 provides a good set of principles and a rational and consistent framework for teaching geotechnical design.
It clarifies aspects that can cause confusion among students by focusing attention onto the distinction between loads (actions) and resistances, and the selection of parameter values to be used in design. Furthermore it provides a clear distinction between ultimate limit state and serviceability limit state.
In short it moves geotechnical design from being a 'black art'with apparently arbitrary factors of safety that are different for each design situation, to being more consistent and scientific with the same partial factors for all design situations, said Orr.
He also cited Professor John Burland's 1987 Nash lecture, which defined ground profile, soil behaviour, applied mechanics, and empiricism (precedent) coupled to well-winnowed experience as the four most important aspects that need to be emphasised to students of soils mechanics: These basic educational objectives can be applied to EC7, he explained, by using a series of examples in the form of lectures, eg 'lecture on geotechnical investigations'and 'lecture on the determination of parameter values used for design'(Orr and Farrell,1999).
He also showed the Eurocode provisions related to Burland's four aspects of soil mechanics (Tab le 1).
Robert Haslem in his talk on characteristic values, factors and risk, outlined a general limitation of Eurocodes which assume parameters for variable actions fit a normal (Gaussian) distribution (Ground Engineering November 2000).
Haslem illustrated a shortcoming in this assumption using data from a pipe jacking case histories. In this the observed forces on the pipes conform to a normal distribution, but pipe deviation angles fitted a lognormal distribution. When force is considered independent of deviation angle the combined risk produced is much lower than if force is assumed to be dependent on deviation angle.
Haslem said engineers should be aware that variable actions do not necessarily conform to a normal distribution and should be aware of the consequences.
Discussion on EC7 centred on the issue of characteristic values and how to teach parameter selection. Use of 'real data'to illustrate the variability of conditions should be introduced at an early stage, before teaching basic mechanics using 'simplified parameters'.
Once these concepts have been embraced teachers could reintroduce real ground conditions and show how to obtain appropriate parameters.
Delegates discussed students' resistance to anything that does not have an exact answer. This could be overcome by introducing gradually the concept that soils are variable, leading to idea that there is never a single correct answer, and that experience plays an important role in geotechnical design.
Delegates also believed that European engineers seemed to be more comfortable with statistical approaches than engineers in the UK. However, the important point that came across was to make students aware of the variability of soil data and show how they might approach the problem of parameter (characteristic value) selection.
Geotechnical heresy and the f and c words Dr Andrew Schofield, retired professor at Cambridge University, demonstrated why he believes we should not be teaching traditional theory based on stress circles. Instead we should teach 'correct'continuum mechanics theory based on q = Mp9.
This is more accurate and a less restrictive relationship than Rankine's condition of limiting stress obliquity on conjugate planes. Critical state soil mechanics offers a scheme of geotechnical teaching based on plasticity theory, said Schofield.
'Terzaghi was wrong to apply the Mohr Coulomb equation to states 'on the wet side of critical' as Hvorslev only had data of peak strength in states 'on the dry side'' In addition he stated that there are no peaks in the strength data at low effective pressure either.
When on any plane the normal effective stress component falls to zero the aggregate ceases to be an effective solid and continuum grains fall apart; with a high hydraulic gradient an aggregate fails by channelling, boiling, cracking, or piping; behaviour we call liquefaction. He made a passionate plea for teachers of geotechnical subjects to teach correct mechanics (Ground Engineering November 2000).
Dr Mike Keedwell, retired lecturer at Coventry University discussed how M and f9 parameters are dependent on the rate at which load is applied to samples in the triaxial test, which he demonstrated by means of two-dimensional analogues. He showed this effect for both strain and stress controlled triaxial tests. The consolidation behaviour of soils is dependent on the rate of loading, which has an effect on the values of M and f9.Clearly the amount of secondary consolidation allowed to occur during the consolidation stage of a triaxial test will affect the failure stress obtained in the subsequent axial loading stage and is consequently another factor affecting the value of M or f9.He referred those attending to his book on Rheology and Soil Mechanics (Keedwell,1984).
A show of hands from the floor indicated that most teachers do teach critical state soil mechanics at some stage in their courses, but mainly in later stages of an undergraduate course and/or at masters level. However, most teachers use traditional approaches in their introductory soil mechanics courses because they involve 'simpler' concepts, and use methods students are more likely to meet in practice and common text books. However, as Schofield retorted, these are not 'real' reasons for not teaching correct theories from the start.
John Greenwood, from Nottingham Trent University presented data on the teaching of geosynthetics in UK universities (Greenwood, 2000). From 21 questionnaire responses, he found that most universities taught typically two to four hours of geosynthetic related material mainly in third and fourth year undergraduate modules. He thought it would be desirable to start teaching geosynthetics earlier in degree programmes and devote more hours to it, including some simple laboratory sessions.
Some of those attending felt there was little capacity to add more into already overcrowded civil engineering degree programmes. And what could be taken out to allow geosynthetics more coverage? In addition, the point was raised that few lecturers have the specialist knowledge required to teach the subject to sufficient depth. It was suggested the International Geosynthetics Society could organise seminars for academics on geosynthetics and provide case histories, including photographs for use in lectures.
Dr Minna Karstunen from the University of Glasgow discussed how collaboration between universities can alleviate growing pressures associated with training postgraduate students.
Fewer students and staff, increasing specialisation and increasing demands from industry are the main drivers of collaboration, she said. Advantages of collaboration include 'complimentarity' critical mass and economy of scale, while the chief disadvantages are logistics and cost.
The key to successful collaboration is working with partners close by and winning external funding (from either national or European schemes).
Collaboration can work on a variety of different levels. A number of joint masters courses exist already. Activities associated with formal training networks (ie involving external funding) include EPSRC funded networks, European training networks, CPD modules and MSc classes. Informal training network (ie involving no external funding) initiatives could include activities such as full utilisation of research seminars, visiting speakers, joint 'away days' and brain-storming sessions.
Karstunen described the Glasgow/Herriot-Watt joint master's courses in which teaching is shared and students spend one day a week at each institution. Other areas of collaboration at a national level include research training networks (Soft Clay Modelling for Engineering Practice) and the Scottish Universities Geotechnical Network.
She concluded that collaboration and networking is certainly worth the effort, but you need to work with people who are genuinely interested. Setting up the network/collaboration and securing external funding can be difficult.
Professor Arnold Verruijt from Technical University of Delft in the Netherlands, added there is considerable potential for developing networks for teaching soil mechanics subjects across Europe. Engineering soil mechanics has greater potential than other branches of engineering such as mechanics and fluid mechanics in which teaching levels vary considerably from country to country.
Professor Verruijt outlined a computer based teaching resource developed to help test students and encourage learning.
He said there had been only a marginal improvement in standard since the introduction of the computer based exercises.
However one delegate described his experience of integrating some of the exercises (available from www.geo.citg.tudelt.nl) into his course, which resulted in a noticeable improvement in the pass rate. Feedback suggested students enjoyed the computer-based exercises.
Simpson B and Driscoll R (1998).Eurocode 7: a commentary, BRE, DETR and Arup Publication.
Burland J (1987).The teaching of soil mechanics - a personal view, Nash Lecture, Proceedings of the IX ECSMFE,3,1427-1447, Balkema.
Haslem R (2000).Eurocode 7: characteristic values, factors of safety and risk, Ground Engineering, November, Vol.33 Number 11, p40.
Orr TLL and Farrell ER (1999).Geotechnical design to Eurocode 7, Springer.
Greenwood J R (2000).Teaching of geosynthetics in UK universities, Presented at the Second European Geosynthetics Conference, Bologna, Italy, October 2000.
Keedwell MJ (1984).Rheology and soil mechanics, Elsevier Applied Science Publishers.
Schofield A (2000).Rankine's earth pressure fallacy, Ground Engineering, November, Vol.33 Number 11, p39.