A Cracking Update

Today we rammed our last wall on the house. It was not as exciting as we had hoped but perhaps we have passed the novelty stage of exposing rammed earth walls to the elements and the whole process is now becoming fairly routine. Nevertheless, it was a milestone and we are generally very pleased with the overall result. We are fairly regularly reminded of what we've achieved by the awestruck expressions of any new visitor. So are convinced that we have managed to do something fairly special - although the road has been fairly long.

We have learnt many lessons and still don't have all the answers to some of the questions. The biggest outstanding question is how we should properly 'cure' a newly rammed wall to prevent it cracking. Some walls cracked, others didn't. One wall began cracking and then stopped after a few inches while another cracked severely from the top to the bottom. The reduction in lime content certainly improved matters, but how much is hard to say. The final solution will be a combination of several things including (and probably most importantly), the drying rate. However, to put things in perspective, we have rammed 19 panels one of which cracked badly, one cracked slightly (a 4 inch hairline crack), one has cracked in one of our expansion cuts and is not visible and the rest have been fine.

The wall in the above picture sprung a small hairline crack very quickly but we began wetting it regularly and this seemed to make it disappear. Now we can't find the original crack and we continue to keep it fully hydrated. Instead of wrapping the walls in plastic, we have now placed a thin layer of sponge on top of the wall held down with two fibre cement barge boards that protect the chamfered edges. We then covered the entire wall with shade netting and this we soak twice a day. The sponge retains quite a bit of water and the shade netting lets our water reach the wall yet provides it some protection against the full force of the spray. This is one of our theories and it has worked so far. All the cracks begin at the top of the wall and work their way down. The top of the wall obviously dries out the quickest, with three faces exposed to the worst effects of the sun and the wind. By keeping it wet, it remains fairly plastic giving the rest of the wall time to dry out slowly and develop in strength. By the time the top is allowed to dry out, most of the shrinkage has taken place without a crack developing. However this is still a theory and perhaps I will have a definitive answer by the time we complete our next project - the utility building.

We treated our other end-wall similarly. For several days we saturated the wall but kept it covered with plastic rather than shade netting. Finally we sprayed it with the newly recommended sealant. It sucked up around 50 litres of sealant over three days. When we now wet it, it still absorbes water but retains a permanent wet look - which is a much darker shade. We are unsure whether the wet look is due to locked in moisture or whether it is the effect of the sealant. However it has not yet cracked (it may still do so) but we are unsure of how it will develop. Obviously we would like all the walls to look more or less the same colour and we wait patiently while it decides what it is going to do. I would obviously prefer the darker shade than an unsightly crack and time will tell.

We attempted a patch job on the wall that had the severe crack. We cut out the 1.2m panel where the crack had developed and then rammed a new panel in its place. The difficulty is trying to get the coloured lines to flow seamlessly into one another which would help obsure the fact it was a patch. We managed to do an acceptable job, however we expect some shrinkage and hopefully this will manifest itself as a vertical crack at the joint - something we can hopefully hide beneath an 'expansion joint'.

Something else that we have learnt over these past few weeks is the effect of rain. The walls have withstood unbelievable amounts of rain. We have had two storms where 50mm of rain was dropped on them in a few hours. We have also had 3 days of solid light rain which flooded the rain gauge - so was at least 100mm. And the walls have survived this bearing the full brunt of these deluges. But not wishing to tempt fate, we have recently covered just the top of the walls with plastic held down with bricks. However, our nicely sieved sand did get wet despite it being covered. The winds were often too much for our plastic to bear and either ripped it or removed it. This significantly raised the moisture content of the soil, which made it sticky. When mixed in the pan mixer, this created a high number of small round balls which when dropped into the form always sought the outside edge/face of the wall. This is also a result of how the soil is spread in the form as when it's spread, the balls of soil all roll to the lowest points which are invariably against the shutterboard. Consequently during the wet phase, many of the walls have an unusually high percentage of roughly finished surfaces as in the picture above. The individual balls are clearly visible. This can be compared to the finish we were after in the second picture, which has the line through it. However, even the rough finish adds character to the house and certainly every wall has a story to tell.

Other than the concerns about existing and future cracks, the walls do look stunning. To remind us that progress is being made, the rib and block roof arrived today and just in time for the builders holiday! We have been through several roof designs and finally decided on the steel rib and polystyrene block arrangement. This reduced the weight of the roof and also improves the insulation qualities. Since this roof will be the last thing standing in this town, the polystyrene is likely to remain in place for several hundred years so I am quite OK with this decision - for the benefit of my conscience that has this product high on its list of things never to use. At least it is not one of those single use polystyrene coffee cups that fill up landfill sites!

The food forest is also beginning to establish itself. The water collects behind the swale just as Geoff Lawtons Permaculture video shows it should. The trees have been well watered and some have been the unfortunate victims of drowning - the ground is so saturated and high in clay that the trees just seem to spend their time swimming. The rest have survived almost tornado speed winds, had their leaves blown off and are now recovering back to where they should be. Some have decided not to grow leaves at all this year in view of the weather. The marula's are just not performing at all. They look like planted sticks with name tags attached to them. The fluted milkwoods are battling with the winds. One has lost all of its leaves and the others look tired from fighting. The acacia's are fine as are most of the citrus. Several pawpaw's have drowned but one or two seem to be returning from the dead. The rest are sprouting nicely. The bees are having a fantastic time and a fairly strong hive has recently taken residence on top of one of our test walls. It's obviously the karma of the wall that so energises them! The birds are obviously happy with the trees and a local fiscal shrike has taken to impaling his food on the thorns of some exhausted plant.

Below is a video slideshow showing the disassembly of the shutterboard - the daily reveal!

Exacting Standards

After we had discovered that 5% lime behaves very differently to 4% lime, we began improving our standards. To establish the water content of the soil, we first weighed a sample and then saturated it in alcohol. This we set alight and stirred while it burnt itself out and in the process evaporated off all the moisture. We then weighed the sample again and the difference was the water content. The next step was to weigh a standard wheelbarrow of soil and using a bit of simple arithmetic, we calculated the dry mass of a standard wheelbarrow of soil. We then recalculated the mass of the lime based on the dry mass of the soil. This required us to drop the amount of lime added and now we are operating at far more exacting standards. To what extent this will improve the shrinkage of the soil is anyone's guess but at least we now know that we have minimised the possibility of stabilisation cracks.

Our window boxes have improved significantly with experience. The amount of kicking is now minimal and the openings are perfectly straight - whereas before we always had small kick somewhere on the face. Now using lots of wedges and bracing, the end panels remain true. On the window openings, we have been flying blind and living in hope to some extent. We know the window sizes, but there was a degree of uncertainty of the windows and their dimensions that would finally arrive from China. Well they did arrive yesterday and there were many panels and each one was surprisingly heavy. I ended up taking two panels to the site and we were pleasantly surprised. Johan was simply grateful that they were actually made to size and fitted comfortably into the spaces we had created for them. This made his day. I was really pleased with the colour and the neat internal blinds that gracefully slide between the double glazing. So between showing off the fact that they fit and that they have fancy little features, there was relief all around on another gamble that had paid off handsomely.



The steel reinforcing for the roof arrived and now needs to be assembled. We finally decided to go with a rib and block roof structure, where the ribs are steel channels and the blocks are made of polystyrene. Due to the significant spans and the fact that we are going to have a heavy (planted) roof, the roof needs to be unusually strong. As a consequence, we have quite a bit of reinforcing. To avoid using steel reinforcing mesh, we have decided on a fibre additive that is added to the concrete that equally boosts the strength. Trevor the concrete specialist advised us to use waterproof concrete. He said that the torch on bitumen product always fails - not that it is a poor product, but it is only as good as the application, and it is the application that always leads to problems later. So we have asked for waterproof concrete which is stronger and is guaranteed by the supplier. Presuming that it is 100% effective, I am now revisiting how I will lay the planted roof. Before I was planning to have a waterproof layer, then a drainage layer with special plastic drainage cups, then a geotech fabric and on top of that the planting medium. Now with a genuinely waterproofed roof that has a 7.5degree fall for drainage, I'm inclined to place a course gravel layer directly onto the concrete slab. Over this can go the geotech fabric and then above this a layer of riversand that will aid in the filtration of the rainwater and finally the planting medium. This will be significantly cheaper, longer lasting and probably as equally effective.

In the meantime, builder's holiday begins on the 11th December and everything (despite the recession) closes down for a month. We were hoping to get the roof up by then but were concerned that we would not make it. We are still concerned about this timeframe however we have now found a subcontractor who specialises in the erection of these roofs and has all the tools to do it. He is confident that it can be done in time but this is obviously reliant on the electrician and plumber installing their services in time. The benefit of this is significant - concrete takes 28 days to reach its full strength and after the concrete is poured, it must remain supported for around 3 weeks before the props can be removed. By letting the roof cure over the holiday, we are saving a significant amount of time.

Finally we have begun the trenching for all the plumbing and electrical reticulation. We've managed not to confuse the plumber with all the groundwater drainage and the rainwater, greywater and blackwater reticulation needs. It seems that houses were build like this in a previous era and the plumber adapted fairly easily to the requirements when he managed to relate the layout to one that exists in his mental archives.

It Never Rains but it Pours

We had a bit of rain the other day. I was sitting with Stijn who is about to take occupation of the house I sold him when it started to drizzle. After a few minutes it picked up and then it began to flood. Every gutter was turned into a weir and there was water everywhere. Since the house was built in 1924, I was silently praying that there would not be the telling drip drip of a leak, especially as he was about to move in. Of course I told him it was an unusual rain and that the gutters don't always behave that badly. Fortunately a day or two later the local newspapers headlines claimed it was the worst storm in 22 years. And it was bad. Upmarket houses lost their roofs, trees were bowled over onto cars and signposts were flattened. We had Allover Mzansi's builder's toilet blown 300m down the valley and now is quite twisted. We now use it to explain to people the dangers of smoking in a pit toilet since it looks as if it's been through lift-off at Cape Canaveral. Other than that, we only lost the Clingfilm cladding that was used to seal the walls whilst curing, but on closer inspection I lost my sense of humour also.

The walls had developed a speckled look which was hardly noticeable and just added to the character. It took us a while to work out that it was the calling card left by the hail. And then we came across the crack. And the crack continues to get worse, but it was developing under the cover of the Clingfilm and by the time we saw it, there was little we could do. I was stupid not to realise that this was likely to happen, especially over a long expanse of wall that had no expansion joints. Also the bend in the wall created from differential drying rates mentioned in an earlier post should have set the alarm bells ringing. But it didn't and only after the appearance of the crack have we begun cutting expansion joints in any section of wall longer than 1.2m. This suits us as the panels are 1.2m wide and they leave a line at the joint between panels which is unsightly. Now we cut along the line and this is just as unsightly but necessary to avoid cracks developing where we don't want them.

Mike the engineer thinks that they may be stabilisation cracks. Apparently these are quite common in roads (which are also rammed earth structures except they use massive road machinery rather than hand tampers) especially if the lime content is slightly too high. So the lime content is a critical component (!!!) and Mike took off some samples to have them tested. Whether they are shrinkage or stabilisation cracks, I really don't want them and wish they had chosen someone else's wall.

I'm now beginning to believe that Johan's quote "this is not a building site, it is a school" deserved a lot more credit than I first gave it.

Going Backwards?

We stripped Saturday's compacted wall on Monday and it looked great except for two lifts that were around 18cm deep and barely compacted. They were very soft - so much so that it was possible to make a significant indentation with the press of a finger. It was obvious that the wall had to come down. This was a pity as the rest of the wall was fine. One of the lifts was placed in the confusion after the giraffe accident so has a valid excuse for being soft, but the other lift did not. However we tried to turn it into an educational experience and the entire labour force was shown the failure and the reasons why it was unacceptable. Then they set about breaking it down which took a surprising amount of effort. We thought that we could just push it over but that turned out to be impossible. Finally we used picks to hack our way through it which was an uplifting experience. These walls are here to stay!


During the past two days we finished off preparing the remaining stem walls and now we are ready to proceed with the compacting until all the walls are up. Fortunately Deon managed to get the giraffe's bucket repaired and this is back up to speed. Thanks Deon!!

While the bucket was being repaired, we began erecting the final layer of formwork on our first north wall panel. The difference now is that we are dealing with a 3.3m high wall and not a 2.3m one. And this is significant. It also requires a second level of formwork to make up the height. When we had just installed the last panel, I realised that the giraffe might not be able to reach over the additional height created by the stem wall. This was a moment of panic as it will not be practical to load these forms by hand. So we moved the giraffe into position and raised the bucket to confirm our worst fears. Well it fits like a glove and this was one of those perfect moments where all the disconnected planning worked better than expected. What a beauty!

By the end of the day everything was in place including the window VDB. We will be ramming the wall in three sections: the walls on each side of the window will be rammed separately and then the window sill will be rammed after these are up. This means that tomorrow is going to be rushed in an attempt to get both walls up and to complete the extra height. I just hope that we will get both walls done in one go and then on Tuesday we can see the fruits of our labour. We can then ram the window sill at our leisure. This is probably the most exciting point as the tall walls will show off the majesty of the rammed earth and we want to be stunned by its beauty.

Break and B@*$#@

Things have been going fantastically well for a bunch of amateur rammed earth builders. That was up until today. We were busy with the last wall on the south side and were almost finished. In fact we had around six lifts to do. Our progress had been quicker than ever before. Twice the hoist motor stopped and we think it was due to rate at which it was working. It literally did not stop moving up or down and it was quite hot as a result. The electrician had told us that the emergency stop was likely to burn out due to the way it had been wired by the manufacturer. However, after a short break it continued to work normally and we decided that there was a temperature protection circuit in the motor.

On the final lift (where it was moving smoothly as we have been quite specific about lubrication) it let off a crack and then bucket the flew down its guide rail and crashed at the foot of the hoist. Fortunately we have ensured that everyone is kept clear of the path of the bucket at all times so there was no chance of an injury, but the bucket took a massive blow on impact. It is now bent rather badly and brought the entire operation to a halt. It is however repairable and we hope to get it repaired fairly soon.

The bucket is latched by a hook connected to the winch pulley. The hook and pulley assembly is an integral unit and are designed to match the hoist. The shaft that attaches the hook to the pulley broke. This should have been one of the last things to break as it must at least be able to withstand the maximum torque of the motor - which it clearly did not. Throughout the process the bucket has always moved very easily when loaded and never gave any hint of breaking. But it did.

Now we need to define an inspection program for all moving parts to prevent anything like this happening again. The delay is not an issue, but the continued safety of everyone on the site certainly is.

Ramming Update #1

We have managed to ram a section of wall per day. Today, we rested the ramming crew and instead, started preparing all the waterproofing and concrete chamfers on the remaining stem walls. Tomorrow we will ram the last wall on the south side of the building. Johan being a traditional builder has calculated that our progress is equal to the build rate in brick. The difference is that we don’t have to plaster or paint so there is probably one day’s saving in that; however we have spent quite a bit of time manufacturing the formwork. By the end of the project, the build rate is probably equal to brick.

There have been some issues that have arisen. The major issue being the forces at play on compaction. Besides having to dramatically increase the number of walers we use and the huge increase on endboard support to prevent movement, we have since discovered a curing issue. We braced the endboards on our last attempt where we rammed an unsupported wall, with steel jacks. Despite this, there was still movement between the bracing points. Now we realise that we need to brace the endboards at around 300mm intervals. The traditional support of wood with wooden wedges has proved to be the simples and probably most effective. The curing issue is another story.

Our first attempt at a window opening was not great due to the significant movements in the VDB. We attempted a repair and it looks like a repair. It is no longer an option to attempt repairs of the rammed earth wall and we now realise that our preparation needs to be significantly better in order to prevent issues later. Also, we had a fairly slender section that was adjacent to a column. As this dried out, it curved away from the wall. The movement was around 10mm and I think it had to do with differential drying rates. The section of the wall next to the concrete column did not dry out, but it did on the opposite face. Due to the fairly high clay content, it shrinks as it dries. This would have caused the one face to shrink significantly more than the other face which resulted in the slender section bowing away from the concrete wall. When we pulled it back in line, it cracked. Since this all is around our repaired window, we will either cut out a large section and redo it or alternatively redo the whole wall. Now we have decided to brace all slender sections. Also, the moment we remove the formwork, we wrap the wall in cellophane or ‘cling film’. This is quite amazing stuff as it bonds with itself and creates a watertight plastic layer which is an ideal way to control the rates of evaporation and consequently the moisture content. The most important effect of wrapping the wall in cling film is that it stops us from attempting repairs on the wall. When the face is exposed, the inclination to fix little imperfections is overwhelming. Again, every repair looks like a repair and by wrapping it in cling film keeps our grubby hands off the slight imperfections until they’ve had time to cure.

I decided to add a little bit extra lime to the wall thinking it would be a good idea. However, I have just learned that the extra lime may result in hairline cracking. So this is another unintended consequence of erring on overdoing something. Another lesson is how to calculate the moisture content. We have been advised to purchase a scale that can weigh 500g and is accurate to 0.1gram. With this, we are supposed to weigh out 500grams of soil. Then we add some methylated spirits to the soil and set it alight. While it’s burning we are supposed to stir it and this burns off all the moisture – as alcohol would. We then weigh the dried soil and by doing this, we can calculate the moisture content of the soil. From this we can calculate precisely how much water we should add to get to our 9.4% moisture content. Apparently if the moisture content varies in our mixing, it may affect the colour of the cured wall. However, it must be said at this stage that I’m grateful to have a rammed earth wall of any colour and I’m sure a slight variation in colour will just add to the character.

So now we are going to leave the walls under plastic for a week and then we will remove the plastic for a day. By then we are hoping that the moisture content of the exposed surfaces will have reduced by around 50%. Then we will spray on a water repelling sealant. This will have two purposes, one to protect the wall from rain and other sources of moisture and the second is to prevent carbonation. Carbonation occurs where carbon dioxide in the air reacts with the bonding process of the lime. This results in the wall being much weaker and layers of the wall can literally be rubbed off. It comes off as a powder/dust. We discovered this occurring with one of our test walls before we sealed it. However, once it is properly sealed, the wall will cure to rock and look stunning to boot.

The Wall

We took off the shutters this morning to discover a stunningly beautiful wall. Beside the expected drama with the window box that did kick and pull, we were very happy with our first effort, especially the coloured veins that had a natural look about them.

The red stem wall bricks and the red vein complement each other well. I'm hoping that the windows will do the same with the charcoal grey vein but this we'll only find out later in the month.

Now we're at the 'patch 'n hide' school of rammed earth. We had to cut and fill the window cavity to get it properly square & plumb and making it match the existing finish is testing our abilities of improvisation - but we're learning something new every step of the way. In the meantime, we began ramming the second wall and made far better progress than we did with our first attempt. We doubled the number of walers around the midsection and this seemed to provide the additional rigidity that we were missing yesterday. Without a window or door volume displacement box, there was no fiddling around and the tamping team progressed easily through all the lifts.

Tomorrow, with two walls up we will begin to get a better appreciation of the final product. And if we continue at this pace, most of the walls on the south side will be up by next week. Then we will have something stunningly beautiful to remind us of what is possible when you set your mind to it.

And finally, the weather deserves the greatest thanks: overcast, nicely cool and dry.

Finally

Despite the warnings of thunder storms, today was perfectly cool and dry. Only as we were wrapping up did the clouds darken and spit out some rain. The timing could not have been more perfect.

We rammed our first wall. Finally. It was a fantastic experience. Progress seemed frustratingly slow but the pan mixer and giraffe worked better than expected. The ramming crew were exhausted and we rotated the whole crew after a while. Working at the bottom of the form where you are closed in and hot obviously saps energy but as soon as their heads poked above the forms where there is a lot of fresh air, the crew's output increased quite dramatically. It was great.

We were expecting the forms to come under significant amount of pressure and we did as much bracing as we felt reasonable. Nevertheless, we were very surprised by the amount the forms kicked out. In some places it was probably as much as 15mm which is more than we wanted and had planned for. As a result, our window endboard (or volume displacement box) was hanging from its top section that was securely clamped to the formwork. Where the wall sections had kicked out, the endboard was not in contact with anything except earth and shifted rather worryingly. I just hope that when we strip off the formwork that our window opening is square and wide enough for a window. Obviously, our next wall will have to have significantly more rigidity.

Based on our initial progress, I thought that we would end up ramming half a wall but in the end we managed to finish it completely and just in the nick of time. We wrapped up the wall in plastic to protect it from the rain and tomorrow, we will remove the forms to see what lies below. It will be dark in colour as it contains quite a bit of moisture but when it dries out, it will lighten and mature into a wonderful natural finish with designer coloured layers at various heights.



Below are some snippets of the activity:

Weather

Looks like we won't be working this week either:

Mon Tues Wed

2nd 3rd 4th November

Rain Delay

Yesterday, it took us quite a bit of time to sort out the few things we had left unfinished from Tuesday. We ground off the sharp edges of our tamping tools, marked out the various details inside our wall cavity, finished installing our chamfers, built a chute, tested and fixed a few problems with the pan mixer, marked out all our buckets with the correct volumes of lime, stone and water according to our recipe and fixed the bend in our walers by installing spacers between opposite walers.

We then started building a 'training' wall where we ran everyone through each of their tasks. We also showed the crew how to tamp the soil to the correct hardness. We tested our Zimbabwean compaction testing tool, but at 2N/mm2 it leaves an impression on most of the test sites. However if it was set to 1.5N/mm2, I am confident it would not. Nevertheless, the depth of the indentation that it makes gives us a clear indication of the quality of the compaction and we will continue using it at the current setting. By the time we were ready with the main event, it was too late and therefore planned to start on it today.

But today looked something like this:


In fact, we had 40mm of rain during the day which meant that the postponement due to rain was likely to last well beyond the weekend. The weather report is not very promising and this is proving to be very frustrating.

Broken Tubes and False Starts?

Today was the final day of preparations for the rammed earth phase.

We assigned a crew to start sieving the large pile of sand and this seems to work fairly well. I was concerned that the aperture size of the sieve was too large (10mm) and instead of completely replacing the wire mesh, we overlaid it with another mesh of a different aperture. This reduces the quantity of large diameter stones and clay balls that make it through the sieve whilst the smaller sizes remain unaffected. Larger pieces still have a way to get through, but their options are significantly reduced.

Several coats of waterproofing have been painted on the stem walls. They are now ready for the rammed earth and the outstanding matter now is the endboards and custom sized forms. The endboards have been fabricated and we are waiting for the sealant to dry. The wet weather has arrived and we have to make doubly sure that the forms are resistant to water in the coming months.

Poen, our millwright made a tool to test the hardness of our compacted earth based on the Zimbabwean rammed earth standards. I thought it would take him a few days as it had a fairly low order of priority in the scheme of things but the novelty of the item seemed to capture his imagination and it was ready the same day. I have just calibrated it but am yet to be convinced of its effectiveness. The rammed earth immediately after ramming is still quite malleable and I do not see it withstanding a force of 2N per sq. mm until after it has cured - which is too late to test the level of compaction. But I'm sure we'll adapt it to our needs and in the end, will get a consistent level of compaction.

We have been anticipating the arrival of the solar panels and today they were delivered. It was a long wait but worth it in the end. It burnt having to pay more to transport the container from Durban to White River than from Shanghai to Durban but this is Africa.

The container was well packed but despite this, 47 of the vacuum tubes were broken. All of the broken units were placed at the top of the container and it seems that they were broken whilst loading them into the fairly confined space. The rest of the consignment seemed fine including the bulk of the vacuum tubes. The photovoltaic panels were in four boxes which were too heavy to pick up by hand so we had to open the boxes and carry them panel by panel into the storeroom. This and sorting out the broken vacuum tubes meant that we took a much longer than anticipated to offload the goods. However, at least it has been done and they are here. Now we just have to finish the building in order to put them to good use and as I have lots of spare tubes, the damaged items do not constitute a crisis.

The apprehension is building for tomorrow and I can see that unless everything is just perfect, we will delay the start until it is. The rammed earth phase is not difficult and we've done everything to make it painless and foolproof. But despite this, we are apprehensive. It's like jumping into a dark pool not knowing what lurks beneath the surface. Let's just hope there are no man-eating crabs.