Mortarless Walls and Seismic Resilience in Inca Architecture
Inca polygonal walls are built without mortar of any kind. Stones are held in place by their own weight, by gravity, and by the precise matching of their surfaces, and the joints between adjacent blocks are so close that a thin blade cannot be inserted into them. This mortarless construction has a striking visual effect: the walls appear to be made of a single piece of stone, with the joints visible only as fine lines. The technique is, however, more than an aesthetic statement. By avoiding mortar, the Inca created walls that are capable of significant deformation under stress without losing integrity, and this characteristic has become especially clear in the major earthquakes of the twentieth and twenty-first centuries, in which Inca walls have generally performed better than the colonial and modern buildings built above, beside, or in place of them.
The 1950 Cusco earthquake
The most important demonstration of the seismic resilience of Inca walls was the earthquake of 21 May 1950, which measured approximately 6.0 on the Richter scale (modern estimates vary between about 6.0 and 6.9) and whose epicenter was located in the Cusco region. The earthquake destroyed or seriously damaged many of the principal colonial buildings of Cusco, including large sections of the cathedral, the church of La Compañía, and the church of Santo Domingo, which had been built on top of the Coricancha following the Spanish conquest.
The pattern of damage was striking. The colonial buildings, which were constructed of fired brick bonded with lime mortar, cracked and partially collapsed, while the underlying Inca walls, built of polygonal andesite and diorite blocks, generally remained intact. At the church of Santo Domingo, the colonial walls cracked along lines that followed the original Inca doorways and niches, while the Inca foundations remained structurally sound. At the church of San Cristóbal, built on Inca foundations in the upper city, the colonial structure failed completely, but the underlying Inca walls were still in place. The same pattern was observed at the fortress of Sacsayhuamán, where the lower cyclopean walls and the upper polygonal courses survived the shaking with little damage, and at the walls of the Twelve-Angled Stone on Calle Hatunrumiyoc, where the joints remained tight and the wall structure was preserved.
The 1950 earthquake was not the first seismic event to demonstrate the resilience of Inca walls. The chronicler Bernabé Cobo, writing in the 1650s, recorded that the 1650 earthquake, which destroyed large sections of colonial Cusco, had caused the Spanish to take note of the way in which the Inca walls had withstood the shaking, and that the colonial authorities had been instructed to preserve the Inca walls as foundations for new buildings wherever possible. The 1986 Cusco earthquake, although of smaller magnitude (5.7), produced a similar pattern, and the 2007 Pisco earthquake, which measured 8.0 and was centered in southern Peru, again demonstrated the relative resilience of dry-stone Inca walls in comparison with adobe and colonial buildings.
The mechanical principles
The mechanical behavior of Inca walls under seismic loading has been the subject of detailed study by engineers, including those associated with the Pontificia Universidad Católica del Perú (PUCP), the Getty Conservation Institute, and a number of Japanese and Italian institutions. The studies have identified three principal features that contribute to the seismic resilience of the walls.
The first is the polygonal geometry of the individual stones. Each block is cut with multiple convex and concave faces that interlock with the neighboring blocks, distributing loads in three dimensions and preventing the wall from failing along a single joint. The second is the inward batter of the wall face, which lowers the center of gravity of the wall and gives it additional stability against overturning. The third is the rubble and mud core between the two facings, which does not fully fill the gap and which allows the wall to absorb small motions without losing integrity.
The key feature of the system is that the joints are not rigid. The stones are not bonded to one another but are held in place by gravity and by the polygonal fit, and they are free to rock and to slide relative to one another in a small, controlled way when the wall is loaded. Under seismic motion, the wall deforms as a whole, with the stones moving on their joints and then resettling into their original positions. The wall is, in effect, a dry-stone rocking structure, in the same mechanical family as the wooden temples of Japan, which are designed to rock on their foundations during earthquakes.
Laboratory and field studies
The mechanical behavior of Inca walls has been studied in the laboratory using small-scale models and full-scale replicas. The Italian architect Enrico Cancian, working with the University of Padua, has tested a series of wall models under simulated seismic loading and has demonstrated that polygonal walls are capable of withstanding accelerations several times greater than those of the 1950 earthquake without losing integrity. The work has also shown that the walls perform best when the polygonal blocks are fitted with the kind of close joint characteristic of the imperial style, and that wider joints reduce the resilience of the wall.
Field studies have included the photogrammetric analysis of the walls of the Coricancha, Sacsayhuamán, and Machu Picchu, and the study of post-earthquake damage to Inca buildings in the Cusco region. The post-earthquake surveys have shown that Inca walls tend to fail by the displacement of individual blocks rather than by the fracture of the wall as a whole, and that the displaced blocks can be reset in their original positions without the need for new stone. This characteristic is a major advantage in a seismic environment, since the wall can be repaired relatively easily after each significant event.
The corbel vault and roofing
The seismic resilience of Inca architecture is not limited to walls. The corbel vault, which the Inca used as an alternative to the pitched wooden roof in some of the most important imperial buildings, has also proved to be a remarkably stable structure under seismic loading. The corbel vault is built of thin, carefully cut stone slabs laid on a curve, with each slab projecting slightly over the one below, and the resulting structure is held in place by the same principles of gravity and dry fit as the polygonal walls.
The most famous surviving corbel vault of the Inca state is the long hall of the Coricancha, sometimes called the “Hall of Serpents,” which has stood for more than 500 years despite repeated seismic events. The vault is also found at the Sun Temple of Ollantaytambo, at the Temple of the Sun at Machu Picchu, and in the smaller temples of several other sites. The corbel vault has been studied by engineers, including the Peruvian structural engineer Julio Vargas Neumann, who has shown that the vault distributes lateral loads through the curving surfaces of the stone slabs and that it is capable of significant deformation without losing integrity.