Monday, July 29, 2013

Ferro-Cement Prep

We began preparing for building the roof in ferro-cement on July 24th. In order to make the switch from tiles to ferro we had to buy some additional rebar as well as welders mesh, chicken mesh and some wood poles for additional support and reinforcement. The prep took only three days.

first we run rebars along are string lines


then anchor the roof structure to the wall

next we start to lay welders mesh (structural) over the 
whole structure and reinforce with wood poles


binding mesh to rebars


Gaby is all smiles

taking shape


full shape with the welders mesh

then we stabilize the structure with wood poles
until Mike can climb on top of the welders mesh structure

and then so two guys can climb on top

next, the chicken mesh - much finer - is laid over the top

binding and weaving the chicken mesh with wire

in very little time the whole structure is covered

Move to Ferro-Cement

Before we arrived on Mfangano Island we were under the assumption that we were building a "test structure" - a full scale but still usable, prototype that would be used to demonstrate the potential of the tile vaulting technique. We agreed that a vault covering a space that was about 10ft x 10ft was a challenge, but manageable despite the other challenges we were taking up. As we began talking with the Ekialo Kiona Radio staff in early June we quickly realized that 10x10 was far too small to accommodate all of the radio operations. During the schematic design phase (which lasted only a few days) the square footage of the building more than tripled. Nevertheless, we pushed forward with full confidence and immediately began working toward this design.

Things ran smoothly up until point where we began working on the tile vault. As was documented in the last blog post, we experienced catastrophic failures a few days into the build. After much deliberation we determined that there were a number of factors that lead to these collapses: lack of experience of the tilers, strength of the tiles versus the strength of the plaster, and the overall size of the vault. Although we have our own academic and research-based goals in mind with this project, the truth is we are also intending to construct a building to be used by a hugely successful community-based radio station that has been in operation for almost a year.

It is our responsibility to provide this community with a building they are able to finish constructing and use safely once it is complete. As a "Plan B" to this project we have known (although never officially said) that ferro-cement was a viable alternative to tile vaulting since the technique has existed and been practiced on Mfangano for a little over four years.

So, a collective decision was made to continue pursuing the design for the roof using ferro-cement. While it is disappointing in terms of the pursuit of research in tile vaulting, we still feel accomplished a lot. We've succeeded in using earth construction in the structural walls, we've managed to pass on new principles of structure and complex geometry that will still be accomplished in the ferro roof. Although we are missing is the tiling technique, there are many tiles that remain that will be used by construction team on Mfangano for practice and experimentation after we're gone.

photo courtesy of OHR

Realities of Experimental Building Projects

Last week we got a real lesson about what it means to take up an experimental building project. If you take a close read at our project abstract, you'll notice many variables and challenges: earth construction, complex geometries, tile vaulting technology transfer, building in the developing world, among others. 

We've moved along rather smoothly through the various phases so far: soil testing, block production, footings/foundations, structural wall construction, and tile vault prototype construction. When it came to building the vault at full scale, it did not go as smoothly.

July 17: we began construction of the vault using the stabilized compressed earth tiles we fabricated with the sub-soil from the site. We focused on building arches over the rebar formwork.




by the end of the first day, we completed arches
across the length of the vault

July 18th: we worked on extruding the arches off of the rebar.




by the end of the second day each arch was now two tiles wide
and a first course of second-layer tiles was laid 
along the base of the vault

July 19th: we worked on closing the second layer of the arches over the rebars and continued to extend the middle arch.




by late afternoon we had a major collapse right where we had 
been extending the vault and building the second layer. 



Since it was Friday afternoon, we cleaned up the wreckage and ended the work week. Over the weekend we continued to see signs of the structure getting weaker and by he following Monday much more of the structure collapsed leaving just the 2-layer arches supported by the rebar formwork. Upon close inspection we realized that due to lack of curing and many other factors, even these arches were very weak and had to be taken down.





Thursday, July 25, 2013

Formwork and Guidework

With the walls complete the team can begin to focus on the roof. We begin by installing our formwork and guide work. We decided to build formwork using rebar for the two end arches and the middle arch - dividing the office from the broadcast room. The end arches are extended outward with welder's mesh to make an overhang for protection against sun and rain. These end arches were given some additional support with wood poles.

binding wooden supports at back arch


front arch - measuring height for wood pole


supports installed in front

Once the end arches are fixed and fully supported we measured the length (around) of each arch and divided them equally. Next, we run bright colored strings that give us a visual guide for the shape of the vault. Just like in our previous prototypes we get complex curvature by offsetting the strings. To better visualize the shape of the vault we ran the strings in both directions and weaved them together to create a sort of saddle vault fishnet.