Volcanic Ash in Ancient and Modern Construction

The ability of the deposits in situ to accommodate underground shelters, to support loads and to stand vertically is reviewed. Erosion patterns and some essential chemical and physical properties are presented.

INTRODUCTION

Volcanic ash and other related deposits, such as tuff and pumice, constitute valuable materials for construction. Knowledge of the possible uses of the materials increases our understanding of the technical possibilities that were open to the Minoans of Akrotiri. And although it is not certain which of these possibilities were realized prior to the catastrophic eruption, the interpretation of archaeological findings may gain from this understanding.

A further advantage of knowing the possible uses and properties of volcanic ash is in planning restoration works at the archaeological site(s). Also, as time passes, the number of visitors that must be received will increase and more facilities will have to be constructed. An appreciation of the engineering behaviour of the deposits in situ will contribute to the planning of future works.

CEMENT 

It has been known for millennia that the mixture of volcanic ash or pulverized tuff (siliceous), with lime, produces hydraulic (i.e. water-resistant) cement. An examination of anient Greek and Roman structures provides ample evidence of the effectiveness and durability of this cement. The lining of a cistern in Kamiros, Rhodes (230 km east of Santorini), dating from the 6th or 7th century BC, is still in existence (Mehta 1987; Malinowski 1979). In Roman times the volcanic silicious component of the cement became known as pozzolana, the name deriving from the port of Pozzuoli, 11 km west of Naples and 27 km west of the volcano Vesuvious, Italy, where it was first used extensively. Pozzolanic natural cement was for millennia the only available material for lining cisterns and aqueducts and binding the brick and stone of water-front structures and monumental buildings (including Rome's Coliseum and Pantheon, built during the 1st century AD). It was only in the latter part of the 19th century that pozzolanic cement was gradually replaced by portland cement.

The links between volcanic ash and the archaeology of Akrotiri are varied and several. It was in connection with a quarrying operation in Therasia, to provide pozzolan for the Suez Canal (1867), that the Bronze Age settlements were first studied systematically and attracted international interest (Fouqué 1879).

Quite appropriately, pozzolan from Santorini is called Santorini earth: large quantities were obtained (Fig. 1) to about 1980, when further quarrying was discontinued for environmental preservation. Pozzolan is still being used in various countries like Greece (Efstathiadis 1978), Italy, Germany, Mexico and China (Mehta 1987), because it reduces cost and improves quality and durability of concrete. The chemical and mineralogical properties of typical pozzolans are shown in Tables 1 and 2 (both from Mehta 1987). The improvement produced by the addition of Santorini earth to portland cement is exemplified by the drastic reduction in alkali-silica expansion (Fig. 2): this type of expansion, if left unchecked, may cause the deterioration of concrete. Furthermore, as shown in Fig. 3, the addition of Santorini earth to portland cement does not reduce its strength.

CONCRETE  

As there is evidence for ancient cement, there is also evidence for even more ancient concrete. According to Davidovits (1987), Farkas (1985) and Starr (1983), the stone blocks of the pyramids and the Sphinx at Giza, as well as other monuments built during Egypt's Old Kingdom, around the 27th century BC, are, in fact, of concrete. This theory explains how, for example in the pyramid of Cheops, millions of blocks, weighing up to 30 tonnes each, were lifted to heights reaching 145 m in the relatively short construction period of the pharaoh's life and also, how there is a near-perfect fit (as close as 0.05 mm) between individual blocks. There were no individual items that were so heavy that they could not be carried by man or animal but only raw materials that could be transported by ordinary means. As for the close fit, it was the result of match casting, that is, using the sides of previously cast blocks as the moulds for the new ones. The evidence in support of this theory includes also significant differences between the chemistry of the pyramid blocks and that of the limestone in the quarries. The counter arguments, first, of the bas-relief in the tomb of Djehutihotep showing a giant statue being hauled on a sledge by pulling on ropes and, second, of Herodotus's process (II, 124) of raising the blocks in steps by crane, could be answered by noting that the first lived 9 and the second 22 centuries after the event and, also, that neither method is supported by material evidence (Edwards 1947). Furthermore the bas-relief shows transport on flat ground rather than on a steep slope, and Herodotus's description is based, by his own assertion, on information by the priests, who may have not told the whole truth.

But if (or since) concrete was known to the Egyptians more than 1000 years before the catastrophic eruption, it may have been known (or inventable by) the Minoans of Akrotiri: both the volcanic ash and the limestone for making lime were available to them. This possibility may be worth considering when interpreting constructional finds in excavations.

PAVEMENTS

The aerodrome of Santorini, for which we are all grateful, is founded on volcanic ash. Except for the upper 0.20 m the rest of the runway is made of compacted ash (Fig. 4). Also, part of the aerodrome of the island of Kos (150 km NE E of Santorini) was built on volcanic ash. Fig. 5 shows that the portions of the runway that are founded on ash have a thickness of only 0.20 m, whereas those that are on other soils have a thickness of 0.45 m. This difference illustrates the superiority of ash as foundation material for pavements. In the case of Fig. 5b, 0.25 m of compacted ash were interposed for improved strength.

GEOTECHNICAL PROPERTIES

The volcanic ashes of Santorini and Kos were investigated for the design of the aerodromes on the two islands. To civil engineers, ash, and also tuff, tephra etc. are classes of soil or soft rock on which and with which constructions are made. Therefore, the rules of performing and evaluating analyses and tests are the same as for other soils. Summaries of the results of laboratory tests performed according to the standard methods of the American Society for Testing Materials (ASTM 1986) are presented in Fig. 6 and 7, for Santorini and Kos respectively.

Fig. 6 includes particle size, specific gravity, natural water content and compaction, according to ASTM designations D 442, D 854, D 2216 and D 1557, respectively. The shaded area of Fig. 6a is formed by the results of 29 particle-size analyses. In some cases there is an abundance (up to 55% by weight) of particles coarser than 4 mm, belonging by definition to the gravel class; by a strict application of the rules of classification these particles are pumice rather than ash but, still, it is the particles from dust size to 4 mm, properly called ash, that determine engineering properties like strength. The particles which are finer than 75 μm, amounting to 2-40%, are non-plastic in the sense of ASTM D 4318.

The specific gravity shown in Fig. 6b is based on the weight of the solid phase of the ash and the combined volume of the solid phase and internal pores (the voids completely within the particles that to not communicate with the atmosphere). If it were not for the internal pores the specific gravity of ash would be approximately equal to that of SiO₂, Al₂O₃, etc. (Table 1), or 2.65: the magnitude of the difference from this value reflects the proportion of internal pores.

The natural water content (Fig. 6c) which is the ratio of the weight of water to the weight of the ash dried at 105^o C, varies between 2 and 14%. The compaction tests are reported in Fig. 6d as plots of the dry density against the water content, from which the maximum density can be determined for each test; it varies between 1.06 and 1.66 t/m³, the difference between these values and the density of the solid phase (approximately euqal to 2.65 t/m³) is due to pores both interparticle and internal.

The strength, or capacity, of compacted soils to sustain loads of an overlying pavement, is expressed in terms of the CBR (California Bearing Ratio) and is measured by laboratory method ASTM D 1883. A CBR value of 100% representes the prototype maximum, whereas 0% extreme weakness; compacted clays have values of the order of 5% and sands 20%. In the case of Santorini ash, values were 53 to 91%, for a density equal to 95% of the maximum obtained by the compaction test.

In the case of the volcanic ash of Kos, some of the test results are summarized in Fig. 7. On the other hand, CBR tests gave values from 59% to over 100%. The result of the standard penetration test (ASTM D 1586) indicated refusal, which means that the material in situ was too strong to allow the apparatus to penetrate inside it. Fig. 8 shows index properties of volcanic ash of various origins, for comparison. Results of strength tests are presented in Fig. 20, in connection with slope stability.

EXCAVATED CAVERNS

Volcanic ash can be easily excavated to form caverns that have served such diverse purposes as shelters for people or animals (Fig. 9), covert gatherings (Fig. 10), and water supply (Fig. 11) or storage. Even when the ash is consolidated to the point of becoming tuff, it is still excavated, without particular difficulty, for caverns. In Guatemala near-horizontal tunnels are opened on the sides of tuffaceous hills to intercept the water table (Fig. 12, Wolofsky pers. comm.). This method of obtaining water is based on the curved shape of the phreatic line in mounts.

But the sides of the excavated caverns are not stable for ever. Time causes deterioration and scaling. There is a study under way to stabilize the sides of the catacombs of Fig. 10 with chemical grouting. Also, the tunnel of Fig. 11, built in the 7th century BC, appears to have been lined in Roman times (Malinowski 1979), presumably because deterioration had to be controlled.

BUILDING STONE

When volcanic ash develops internal cementation it is transformed into a soft rock called tuff. In spite of its inferior qualities when compared with other stone (lower strength and resistance to erosion), tuff is often quarried and used as building stone. For example, tuff stone masonry was commonly used in Naples, Italy (Fig. 13, Croce 1985) and is still popular in Nairobi, Kenya. 

EROSION

Volcanic ash and tuff are vulnerable to both wind and moving water. Central to the erosion process are the endogenous fissures (Fig. 14), running in verticalor near-vertical directions; spacing of fissures can be 1 to 10 m. On the other hand, waves affect adversely coastal cliffs (Fig. 15) either by causing the gradual fall of blocks (Fig. 16), or by first forming caves (Fig. 17) and then enlarging them to the point that their roofs collapse (Kotzias and Stamatopoulos 1984).

STRENGTH AND SLOPE STABILITY

The high strength of volcanic ash is evident from the great height at which steep slopes can stand (Fig. 1, 15, 18), sometimes reaching 30 m or more. A uniform continuous material standing vertically at 30 m, with unit weight 1.3 t/m³ (12.7 kN/m³), must have an average shear strength of at least 10.1 t/m² (99.4 kN/m²), if it is to resist downward displacement along a cylindrical slip surface such as that of Fig. 19 (Lambe and Whitman 1969). The results of direct shear tests, performed according to ASTM D 3080, shon in Fig. 20, indicate strengths of at least as much. The specimens on which these tests were performed were from the vicinity of the cliff of Fig. 15.

Still, the failure of steep slopes of volcanic ash or tuff does not conform to the slip surface configuration (Fig. 19), which is common in other soils. Observation shows that the pattern proposed by O'Rourke and Crespo (1988) is nearer reality (Fig. 21). This pattern was studied in connection with the volcaniclastic Cangahua formation in the Andes of Ecuador and Colombia. The characteristic properties of this formation are: dry unit weight 1.12-1.47 t/m³ (11.0-14.4 kN/m³), specific gravity 2.58, compressive and tensile strength at average unit weight, 50 and 12 t/m² (490 and 118 kN/m²), respectively, and angle of internal friction 39-40°; the particle-size distribution is shown in Fig. 8. Steep slopes of Cangahua have failed during earthquakes. Also, earthquake horizontal accelerations may induce tensile stresses that hasten the process of Fig. 21.

OTHER ENGINEERING ASPECTS

The strength mentioned above for Milos and Cangahua, would lead to a conventionally computed maximum soil-bearing pressure of footings(Terzaghi and Peck 1967) of the order of 800 t/m² (7840 kN/m²). Yet, in actual practice, footings are designed for bearing pressures of only 30-50 t/m² (294-490 kN/m²), to safeguard against inhomogeneities, fissures, settlements, etc.

Other aspects of in situ ash and tuff that must be considered in the planning of buildings, bridges, canals. harbours, etc, is their variability, even within short distances, and their high permeability to water. Variability may cause problems in foundations and permeability may necessitate sealing of the surface.

Further applications include the use of pumice (particles between 4 and 75 mm) for light-weight concrete, for insulating concrete and for thermal insulation by itself. Concrete made with pumice weighs about 1.0 t/m³ (9.8 kN/m³) which is only 40% of what ordinary concrete weighs (Neville 1973a). Pumice concrete has a coefficient of thermal conductivity of 0.15 Btu/ft h ^oF, whereas that of conventional concrete is 6 times as much (Neville 1973b).

Addendum

The subject and approach of this paper relate to civil engineering; the authors are practising civil engineers, engaged in the particular speciality of geotechnical engineering. The inclusion of this discipline for the first time in the Thera Congress will no doubt enhance its multi-disciplinary character. Indeed, there are many recent instances where geotechnical engineering and archaeology have co-existed, for example, Session 10 of the First Hellenic Conference on Geotechnical Engineers (Athens, February 1988) dealt with 'Geotechnical Engineering Related to Hellenic Historical Monuments' (Stamatopoulos 1988).

As both geotechnical engineering and archaeology are concerned with the upper few metres of the subsurface, they inevitably mingle. To a geotechnical engineer, potsherds may signal the danger of founding on loose soil. To an archaeologist, the lower end of a deposit containing potsherds or other remains may signal no further need to excavate.

Some methods of investigation are now common to both disciplines: pits or trenches, geophysical soundings, aerial photographs, boreholes, etc.

The important subject of earthquakes is also of common interest. By observing ancient cities or monuments, it is possible to infer the magnitude and type of earthquakes thousands of years ago; this is of value to engineers. On the other hand, the knowledge of the same engineers can and is being used in protecting the monuments from future quakes.

----------------------------------------------------

----------------------------------------------