Masonry roofs

 If you have been to the Wits Club or Alumni Offices at the University of the Witwatersrand in the last couple of years, you may have seen this unusual structure. It is a recent interpretation of a structural type that is extremely ancient, with surviving evidence of examples from Ancient Egypt and the most ancient cultures of the Middle East. The curved roof forms are amazingly efficient, such that they can be built from materials that have very little structural strength, such as sundried earth and cement-stabilised earth, which is the material used for the roofs in the Wits building.

The Wits masonry vaults
http://www.claisse.info/2019%20papers/5021.pdf

My interest in this material goes back to the mid-2000s, when I was exploring labour-intensive approaches as part of my PhD. I was already aware of stabilised earth, as several of my colleagues in the School of Civil and Environmental Engineering had been researching it, and we already had a diesel and a fully manual block press. My excitement was motivated by the fact that the building components could be manufactured by local, relatively unskilled, workers with predominantly locally quarried materials. This would save enormously on the cost of "imported" materials (a very small fraction of cement is needed as a stabiliser) and at the same time the cost of transportation of materials to more rural sites could be circumvented. This would leave more of the project cost to be available for wages for local workers and investment into local businesses in the vulnerable communities that were the target of the labour-intensive interventions.

The block press is designed to make walling components, larger than bricks, and in the case of the diesel press, interlocking so that no mortar was needed, with the blocks fitting together horizontally and vertically. However, my cost analysis for my PhD showed that much pf the project value was still being used up by the roof, if conventional methods such as profiled metal sheeting on timber trusses were used. This prompted an exploration into shell structures, which I discovered were already being studied by my colleagues at Wits and a parallel study was underway at MIT in the States. With some minor (and reversible) modifications to the hand-press, we managed to make thin tiles of cement-stabilised earth, allowing a very versatile type of structure, the timbrel vault, to be built.

This led to a number of experimental small-scale structures, testing out various configurations, two of which were built on the roof of our School's building. These were subject to several years of testing to help us understand the structural performance and the thermal properties.

Two of the 1/4 scale roofs showing the extraordinary strength 
https://scielo.org.za/scielo.php?script=sci_arttext&pid=S1021-20192013000300011

Woven into this story of research and innovation was the highly successful Mapungubwe Interpretation Centre that I have explored in a previous blog, a wonderful example of the migration of ideas across several continents and many centuries of discovery and application.

To come back to the building near the Wits Club, this was designed and built by one of my colleagues, Dr Ryan Bradley, a structural engineer whose own PhD was focused on developing alternatives to low-cost housing that could be more affordable and would have much better thermal performance than most of what was being built on the government's RDP housing programmes. The building has a kitchen, bathroom and living area on the ground floor, and space for bedrooms on the upper floor. 

It has been extensively monitored over the years since it was completed in 2017, particularly to identify cracking in the roof vaults, a common problem with masonry roof structures. Many of the world's most famous curved or vaulted structures have extensive cracking, such as the Pantheon in Rome, however, in most cases these do not compromise the strength or stability of the building. The cracks are "self-sealing" through the main compressive action of the structural forces, which prevent distortion, further cracking or collapse, so if you ever see a vault with visible cracks, this is usually not cause for concern!

The thermal performance has likewise been measured and analysed for a number of years, with some interesting modelling having been done by the Maths in Industry Study Group, which I have covered in a blog earlier this year. What is particularly interesting is how this thread of research has drawn together so many important themes around environmental sustainability: carbon footprint, thermal performance, economic sustainability through job creation and small business development, and social sustainability through striving for greater comfort for inhabitants that does not require artificial heating or cooling.