Friday, January 29, 2010

Rammed Earth Testing - Large Sample

The success of the small rammed earth samples meant it was time to build a larger mold for a larger sample. Some leftover block-board, a couple of trips to the wood shop and 2m of 20mm threaded rod later, we had our large sample mold. It consists of four panels and produces a 25x25cm column with 45⁰ chamfered corners.

Out on the site, we dug a fresh hole approximately where the earth pit would be, so as to test the actual soil that we hope to use. After digging about a meter to get bellow the organic top soil, we hit nice orange sub-soil.  friendly neighbor came over to investigate, and ended up helping out (he is a mason and is now part of the team working on the foundations). The soil had about the right moisture content as it came from the ground, so we used it as is. We laid the freshly dug soil in 4” courses and rammed it down with our improvised rammer.

There was a bit of a struggle releasing the nuts due to the lack of a wrench, but after a couple of well placed blows the threaded rods came lose. This is defiantly something that will need to be addressed in the design of the full sized formwork.

The surface finish and strength were both better than I had hoped for – the sides were incredibly smooth and free of defects. The layering was subtle but is apparent enough to add some visual interest to any walls that we leave unfinished – which will be the default.

After a week sitting unprotected in the field and a number of heavy rainstorms, the Large Sample showed little sign of damage. The top surface had some minor pitting from the rain, but the sides were untouched. Like the small samples the block has cured into something resembling sandstone and is reassuringly solid. The sample suggests that, assuming good coverage of the tops of the walls and window sills, the walls made from this earth could survive the rains without the need for stabilizers.

Rammed Earth Testing - Small Samples

The decision on using Rammed earth for our house was dependant on the suitably of the soil on site (see material selection for the reasoning behind this). I had a strong suspicion that the soil could work from seeing the way the footpaths in the around here set into firm smooth surfaces with the passage of people and animals. The paths do not erode easily nor do they swell of crack with the cycle of wet and dry.  It is one of the qualities that makes mountain biking so enjoyable here in Rwink.   Further down the valley the soils are clay heavy and get stick when they are wet, then crack if they dry too quickly.

Earth Sample with metal tube rammer, showing the wood plug blocking the end, and the black pvc tube as a mold.

To investigate the suitability of the soil, I followed a few of the usual tests, many of which are well outlined in David Eatons Rammed Earth House. To begin with, a bucket-sized sample was taken from an existing hole on site and a simple sample mold was constructed from 50mm pvc pipe and a steel tube. The initial samples showed promise, but the mold was difficult to release. Paper and plastic sleeves were both tried as aids to remove the samples but as you can see from the image, there was a lot of damage.  After finding some pipe clamps in Kigali, the mold now has a split and can be used to produce full height samples.

From Left to Right: Hand molded ball, the first rammed sample showing extensive deformation from getting stuck, and then progressing toward the samples that were rammed using sleeves to help with releasing from the mold.

Even in their damaged state, the samples developed great strength after only a day of curing, and by the end of a week they were as hard as Sydney sandstone. This was good enough to try again with a larger sample.  Rammed Earth Testing - Large Sample

Thursday, January 28, 2010

Early Design Phases - The Plans 2

The winning concept that came out of the redesigning process is a single unit with open living areas on one side and sleeping and bathing on the other. The building’s front and back elevations get progressively more closed as it moves towards the private areas; creating a pavilion-like airiness in the living area and maintaining a balance between fenestration and privacy in the bedrooms and bathrooms.

Valley side elevation, showing screened in porch, master bed, and master bath. note that the stairs down to the valley side garden are not shown in the model
Road side elevation closeup, showing entrance porch, which will be screened,  and the living, dinning and kitchen area.
Plan view: living areas on left with porches front and back, master bed and bath front right, kids/guest bath and bedrooms back right.

This new design uses the same roof truss at an even 2m spacing across the structure. It incorporate more storage, and brings the private areas together. the gabled roof protects the earth walls better than the shed roof, and provides better shading from the western sun. The integrated porches will be screened to provide bug-free outdoor space and cross ventilation to the living area, which is critical because we’re building in a malaria zone.  while this design lacks some of the showmanship of the previous design, it does create a much more private and protect space with an open airy feel.

Early Design Phases - The Plans 1

The design of the house evolved over the three months between the purchase of the land and our return to Rwanda. Following the ideas as laid out in The Brief and Material Selection we experimented with different layouts and feels for the house. The front runner for a lot of the process was a two part house that consisted of a shed roofed front house containing the living areas and master suite, and a gabled roof back house with the remaining bedrooms and storage.

Valley side elevation, showing master suite, lving area, and porch
Overall floor plan with telephone-pole rafters visible over the front house. 
Interior of Living area looking out over the valley

The layout allowed for phased building because the front house could be lived in while the back was being finished and kept the overall size of each roof relatively small. It made good use of the sites valley views, and despite the dramatic valley side elevation, the road side was discrete and fitting. unfortunately this design does have some weakness:
  1. Excessive solar gain -  those large windows pulling in the view, also face NW  so by the time we added enough shading, the dramatic effect of the windows is lost.
  2. Weather Proofing – the shed roof faced towards the weather, and windows here have a tendency to leak. plus the detail between the lip of the roof and the top of the windows also looks like a weak point.
  3. A Little Too Separate – while guest would potentially appreciate their own space in the back house, the reality is, as a house for our future young family, the 2am feeds and nightmare banishing would require trekking out to the other building, an idea that we don’t relish.
  4. Site Gradient – the back house runs up the slope of the site, this mean that for that little skinny building, we would have to climb half a flight of stairs and throw down a lot of foundation to make it level. this became even more of a problem once we took a closer look at the land – an old foundation created a steep bank at the top of the property halfway through where the back house would have been.
We had a lot of solutions for these various issues but eventually our dissatisfaction drove us to try afresh.

Tuesday, January 19, 2010

Early Design Phases - Material Selection

Through a combination of financial and environmental factors we decided that minimising the amount of cement required for construction would be a key factor of the design. This drove an exploration of alternatives to the standard cement block. Burnt Bricks ( wood fired clay bricks ) are cheap and quick, but their manufacture here contributes to deforestation because of the wood used to fire them plus they need excessive mortar use due to their irregular shapes. 

Mud Brick construction as traditionally made in the village using soil dug from the property mixed with water and straw. They are cheap and work, however the soft organic filled bricks are fairly susceptible to insect infestation and erosion unless regularly rendered in cement mortar, which tends to crack and fall off if not well tended. Also the long drying time and handling damage make for a long fabrication time for the bricks. 

One person we know had successfully built a house with unstabilized Compressed Earth Block (CEB) using earth from nearby. This looked like a very promising material and we even located a couple of CEB machines available in Rwanda. There is even an interlocking block model that allows the blocks to be dry-stacked. There is still a need for rendering in cement on the outside to protect the joins, and plastering on the inside if a smooth wall is desired. 

The next idea, and current winner, is Rammed Earth. Monolithic walls, free of joins and mortar, built from the earth on the property using local labour. The thermal mass and insulation qualities will help stabilise the relatively minor fluctuations in temperature. Plus the excellent sound deadening will add some privacy to village life. This material choice would also fulfil a promise, made to Cheryl long ago at a hippy festival, that one day I would build us a house of dirt. Because this technique is so dependant on the quality of the earth, some testing will be required to double check the suitablility of the earth on site

The roofing options come down to: basic gal roofing, fancy metal roofing, simple tiles, and factory tiles. The metal options both produce lightweight, weather resistant roofs that are noisy and hot unless insulated. The tile roofs are pleasant and insulating, but are heavy and more prone to leaking. The basic gal tin roofing with a papyrus mat insulating layer is the current choice, because of cost, visual appropriateness (tin roofs are more common in our valley), lightweight, and forgiving in its installation. Eucalyptus poles harvested from the local managed forests will be used for rafters, ring-beams and ridge-beams.

This was a relatively straight forward decision; the most appropriate option is already in common use in the valley. Large rocks quarried from a ridge nearby are used to make rubble-trench foundations and a rubble infill for the base of the floor. A thin cement layer caps the foundation and and, with a burnished skim coat, finishes the floor.

The basic window module will be simple wooden framing embeded into the rammed earth wall to creat a low-profile mount for 8mm glass. The aim is to play on the thickness of the walls (40cm) and create a feeling that the window is not even there - a kind of less extreme version of the windows in Utzon's house in Majorca (& here)

Early Design Phases - The Brief

Once the decision to build had been finalised with the purchase of land, the designing and dreaming began. While we didn't write a formal brief, the ideas we based the designs on could be roughly distilled to:
  • Low Cost: with no financing available and little chance of a conventional return on investment, the cash cost of the project needs to be kept low.
  • Open: Bring in the natural ambiant light and make the most of the valley views
  • Private: as foreigners in a rural village, we are minor celebrities and exposed to constant curiosity from the other residents, particularly the children. While the interest is mostly friendly it can be wearisome and is not something that we wish to experience in our own home.
  • Visually Appropriate: The house should blend with the landscape and the village and not look ostentatious or fortress-like.
  • Socially Responsible: maximise the potential benefits for the local area through employment and capacity building
  • Environmentally Sound: Emphasise local materials, low carbon footprint, and efficient operation of the building.
  • Semi self sufficent: While there is town water and electricity, they are not very reliable yet, so rain water collection and a simple electrical backup system will be used to cover the gaps in service.

This one time, in Uganda...

This story of this house begins with a group of friends and collogues going for a weekender to Uganda. The eco lodge cabins that we stayed in were without electricity but still managed to provide a lovely hot bath with simple passive solar panels. Half jokingly, we started talking about how doable it would be to create something similar back in Rwink. This resulted in a wine fuelled pact to buy land and build houses together.

I'm not really sure how seriously we all took this idea at the time, but it developed a life of its own, and only two weeks later we were in the sector office handing over a brown paper bag full of Rwandan francs to our new neighbour.

A couple of months of contemplation researching and designing while away visiting our families leads us to where we are now: one fairly solid design, a plot of land, and two and a half months before the rains come.