Analysis of historic rammed earth construction

PhD Thesis Paul Jaquin January 2008

If you would like a copy of all or part of the thesis, please contact me.

This thesis has resulted in a number of publications, see this page for details.


Rammed earth is an ancient construction technique which has recently become popular for sustainable building. Soil is compacted in removable formwork to make a homogeneous wall. A lack of experimental evidence and a poor fundamental understanding means that current design guidelines are highly conservative and inappropriate for the analysis of historic rammed earth buildings. This thesis shows that rammed earth can be viewed in a geotechnical engineering framework and that doing so helps to explain many aspects of the material behaviour. Rammed earth walls were built and tested in the laboratory then modelled using techniques available to practising engineers. Unsaturated soil mechanics was considered useful in explaining much of the behaviour of rammed earth. This was investigated through a series of uniaxial compression tests and the results are explained using unsaturated soil mechanics. Visits to Spain and India were made to investigate rammed earth in the field. Historic construction techniques, modes of failure and repair strategies were studied. The unsaturated nature of rammed earth is used to explain modes of failure and to suggest the most appropriate repair strategies.

Chapter 1 – Introduction and Literature Review

This section presents a review of literature relating to rammed earth construction. It is argued that while some research has been carried out, and guidelines exist for the construction of new rammed earth buildings, there is a lack of understanding of many of the concepts which underpin earth building, and this means that many basic questions about the nature of rammed earth remain unanswered. While these questions have been posed, and it has been argued that rammed earth be treated in a geotechnical engineering framework, it is shown that work to date has not achieved this.

Chapter 2 – Classical testing and modelling

rammed earth testing

Rammed earth wall. Point load applied at the centre

The current design guidelines are unsatisfactory for use with historic rammed earth. This is because they are based heavily on masonry or concrete guidelines and have been adapted for use in modern rammed earth. The lack of definitive testing and these adapted guidelines mean that large factors of safety must be used. When investigating a historic building possibility close to collapse, such safety factors cannot be used. Therefore improved modelling strategies must be developed. This chapter uses simple physical modelling to inform numerical modelling. Physical modelling involved the construction of five walls in the laboratory. These walls were differently shaped and loaded in different ways because construction of the next wall was informed by the behaviour of the previous wall. The tests were intentionally simple and only limited instrumentation was used. These walls improved our understanding of the construction technique and provided insights into the types of failure which occur when loading rammed earth. The layered nature of rammed earth is highlighted.

rammed earth analysis

Finite element model of a rammed earth wall. Red areas have failed.

Simple numerical models, available to practising engineers were used to represent rammed earth behaviour. Finite elements models of some of the physical walls were developed. The layered nature, highlighted in physical modelling was investigated first. A layered rammed earth model was developed based on the Mohr-Coulomb yield criteria, with the layers being assigned different properties to the body of the rammed earth. A second approach looked at the compaction of rammed earth using an elastoplastic hardening approach also making use of the Mohr-Coulomb yield criteria. In both cases a parametric study of Mohr-Coulomb parameters was carried out.

The first part of this chapter discusses soil from a geotechnical engineering perspective. A range of constitutive models are then described. The construction and testing of walls in the laboratory is then explained, and this leads to the development of the numerical modelling based on the constitutive models.

Related publication

Jaquin, P.A., Analysis of Historic Rammed Earth Construction9th Young Geotechnical Engineers Symposium, Belfast, September 2006

(1.12MB)  Analysis of Historic Rammed Earth Construction

Chapter 3 – Advanced testing and modelling

Classical soil mechanics is well suited to a wide range of geotechnical engineering problems. The use of simple formulae and experience based safety factors leads to efficient design of geotechnical structures and can be used to predict the behaviour of a wide range of soil types under different loading conditions.

rammed earth testing

Testing a rammed earth cylinder in uniaxial compression

However there are situations where classical soil mechanics cannot accurately predict the behaviour of a body of soil (as described by Jennings and Burland 1962). This usually occurs when the soil becomes unsaturated, that is there is a mixture of air and water within the pores of the soil, and the pore water within the soil is not continuous.

This chapter first outlines the concepts of tensile strength of water and of surface tension. The notion of suction is outlined by explaining the phenomenon of capillary rise which is then linked to the relative humidity of the surrounding air. Suction in soil mechanics is then explained and the concept of equilibrium pore radius defined. The idea of a liquid bridge is outlined and the attractive force across the liquid described. The volume of water held within a soil due to suction is then described through the idea of a Soil Water Characteristic Curve.

The differences between the behaviour of saturated and unsaturated soils; and ideas for the conceptualisation of unsaturated soils are described, highlighting the double structure theory and evidence of liquid bridges. Constitutive models which distinguish between the behaviour of saturated and unsaturated soils are then outlined.

The origins of strength in rammed earth are discussed. Rammed earth practitioners disagree on the reasons for strength in the construction material (for example King 1997; Norton 1997 and Houben and Avrami 2000), and attempts are being made in the field of unsaturated soil mechanics to quantify the magnitudes of the different actions.

Electrostatic actions such as van der Waals forces and Double Layer attraction are combined as DLVO theory, and it is shown that it is suction rather than DLVO forces which provide additional strength to highlight unsaturated soils such as rammed earth. Cementing is briefly introduced to later explain why there is an optimum cement content for stabilised rammed earth building.

rammed earth testing

Failed rammed earth cylinder sample

A short series of unconfined compression tests were carried out, where suction was measured to establish a link between sample strength and suction. Tensiometers were introduced as a way of measuring suction. The testing procedure, and the way in which issues in the experimentation were resolved are described.

As a result of the experimentation is it possible to describe the relationship between a number of the measured parameters such as water content and strength. A change in suction on loading; and the increased strength and brittleness of drier samples was observed. The results are then explained within a double structure framework and it is argued that this concept may be of use in explaining the behaviour of earthen structures.

The nature of water in rammed earth is then discussed. It is argued that the concepts outlined previously may also be used to explain the infiltration to and evaporation from rammed earth structures. Finally further work is suggested and implications for modern rammed earth buildings are highlighted. The work in this chapter allows a much fuller understanding of the failure of and repair to rammed earth structures which are discussed in Chapters 5 and 6.

Related publication

Jaquin, P. A., Augarde C.E and Legrande, L. Unsaturated characteristics of rammed earthFirst European Conference on Unsaturated Soils, Durham, July 2008

(160KB)  Unsaturated characteristics of rammed earth

Chapter 4 – Typologies of rammed earth construction

18th century rammed earth barn in the Tapial con Lunetos style. Villafeliche, Spain

This chapter investigates different types of historic rammed earth construction, looking at techniques and material used in construction and the resulting rammed earth walls. A comprehension of both the methods of construction and the composition of historic walls is vital for the understanding of failure and repair of rammed earth. However, until recently typological rammed earth descriptions have ‘raised little interest when compared to ornamental and spatial studies’ (Graciani García and Tabales Rodríguez 2003).

This chapter looks at descriptions of rammed earth in Seville, Spain proposed by Graciani García and Tabales Rodríguez (2003) and aims to extend and improve their framework based on field visits discussed in Appendices B, C and D.

Related publication

Jaquin, P.A., Augarde, C. E. and Gerrard, C.M. Historic Rammed Earth Structures in Spain.International Symposium on Earthen Structures, Bangalore, August 2007

(1.40MB)  Historic Rammed Earth Structures in Spain

Chapter 5 – Failures in rammed earth structures

Rammed earth, being constructed from soil, is perceived by many to be a very delicate material, requiring a sympathetic climate and frequent maintenance to preserve its appearance and structural integrity.

crack in rammed earth

Crack between rammed earth infill and brick columns. Villafeliche, Spain

This chapter aims to determine the ways in which rammed earth structures fail. This will improve the ability of engineers to assess whether rammed earth is able to remain in a viable state over a long period of time, and if there are any limiting factors to the survival of a rammed earth structure.

There have been many specific case studies regarding the repair of historic earthen architecture, often looking at the failure mechanisms before proposing and implementing repair strategies. A number of authors have presented general failure mechanisms for earthen architecture, but few take account of the geotechnical nature of earth building, and none take account of its unsaturated nature outlined in Chapter 3.

This chapter combines those failure mechanisms reported by previous authors with those identified during field visits by the author described in Appendixes B, C and D. These failure mechanisms are explained, and in a number of cases the previously ignored unsaturated nature of the material is explored. A number of case study structures are then discussed, charting combinations of the different failure mechanisms which combine to precipitate major problems in the structure. Finally, conclusions are drawn as to the most common and most problematic failure mechanisms, drawing out pertinent advice on the determination of problems in rammed earth structures.

Related publication

Jaquin, P. A. Study of historic rammed earth structures in Spain and IndiaThe Structural Engineer,

(1.52MB) Study of historic rammed earth structures in Spain and India

Chapter 6 – Rammed earth repair

Chapter 5 introduced failures which may beset historic rammed earth buildings. This chapter looks at repair methods which could be adopted following such failures.

Stitched crack in rammed earth, Basgo, Ladakh

It is important that the cause of any damage be determined before repair is carried out. In many instances, if the structural integrity of the building is not at risk, then stopping the cause of the damage may be all that is required to ensure the safety of the building. Warren (1993) notes that property owners may rush to repair damage for aesthetic reasons when it may not be necessary.

The principles of repair are discussed and it is argued that western ideals of historic building conservation may be at odds with other repair philosophies. Although no specific document relating to the repair of historic rammed earth buildings exists, a small number of authors have discussed repair strategies for earth buildings and their applicability to historic rammed earth buildings is discussed.

This chapter presents methods for dealing with ground movement, highlighting those which have been suggested for historic earth buildings. As many solutions are independent of building type, this section is intentionally brief.

Issues with structural elements were highlighted in Section 5.4, and while many of these impact on rammed earth structures, the elements themselves are not rammed earth, and their repair requires further specialist knowledge (for example for the repair of timber roof members). Where the structural member issues were integral to rammed earth (for example the joint between rammed earth blocks) it was considered that this issue would be brought about by a further problem, (for example ground movement) and strategies for the alleviation of these problems are presented. Both soft crack stitching and hard wall tying techniques are given.

In Chapter 5 it was argued that excess water in a rammed earth wall is undesirable, so ways to reduce the amount of water entering a structure and methods for increasing evaporation are presented here. Where erosion has occurred a range of techniques to replace missing material are given. Finally the techniques presented are evaluated and those considered most effective are recommended.

Appendix A – Historic rammed earth distribution

Rammed earth is a construction technique where soil is taken from the ground and compacted to form structures. Removable formwork is installed, and the soil compacted within it. The technique is widespread but the distribution of rammed earth across the world and its development over time has not previously been fully documented. Many sources quote the same examples of the Potala Palace in Lhasa, parts of the Great Wall of China, and the Alhambra in Granada. The distribution of rammed earth is however more complex than usually portrayed, appearing to spread over the world in a number of temporal waves, each precipitated by a different set of needs.

In this appendix, a strict definition of the rammed earth technique is first presented, identifying it by name in different languages. It will be argued that the rammed earth technique appears to have developed independently in China and around the Mediterranean. The technique then spread with the movement of peoples to different parts of the world. Rammed earth has continually been reinvented as a building material. At times it has been used as a quick technique for the building of fortifications, a cheap way a man can build his own home, and a sustainable construction technique using only what is available on site.

Related publication

Jaquin, P.A. Augarde, C.E. Gerrard, C.M. Historic rammed earth distributionInternational Journal of Architectural Heritage

Appendix B – North Spain

A field visit to northern Spain was carried out in October 2007, with assistance from the Institution of Structural Engineers Rowen Travel Award. Dr Charles Augarde and Dr Chris Gerrard were present for the first three days, and for the following week I was accompanied on a number of days by Mr Nick Watson.

A large number of sites were visited, not all of which were found to be rammed earth, and the four sites described in this appendix relate to those mentioned in the body of thesis.

Appendix C – South Spain

Seventeen locations in southern Spain were visited in January 2006. Over a two week period I travelled from Murcia in western Spain to Seville in southern Spain. The seventeen locations have been further split into individual sites which range from whole castle complexes through individual to individual walls or parts of city walls.

Appendix D –  North India

A short field trip to northern India was undertaken prior to a conference in the region in late October and early November 2006. A number of historic sites were visited, of which three rammed earth ones (shown in Figure D.1 and Figure D.2) are described in this appendix.