Get our free book (in Spanish or English) on rainwater now - To Catch the Rain.
4th and F St. Cob Bench
Original Bench Construction
Built in the fall of 2003, this cob bench located at 4th and F St in Arcata, California, was completed by Humboldt State student Patrick McAuley. The cob bench was constructed in the garden of Arcata resident Lisa Jackson and is viewable from Samoa Blvd.The project was organized through an appropriate technology class (Engineering 305) at Humboldt State University and served as a final project for McAuley. Project leaders were natural builder Kirsten Thompson, appropriate technology course instructor Lonny Grafman and Patrick McAuley.
McAuley and crew gathered materials for the construction of the bench around the city of Arcata. Reclaimed urbanite (broken chunks of sidewalk) was collected from HSU with the intention of being used as the foundation (see fig. 1). Materials for the construction of the cob and plaster were collected locally and kept in the yard of Lisa Jackson (see fig. 2).
The cob materials consist of clay, river sand and straw. River sand is preferred to beach sand as it is more coarse and better for building and straw is preferable to hay as there are no seed heads on the straw that can sprout.
Natural building is rather simple, but it requires some time and effort. Patrick and his team had to assemble this cob bench from scratch, which required lots of time mixing cob and designing the dimensions and shape of the bench. This project also required them to make their own foundation in order to protect the bench from water.
Laying the Foundation
The first thing the cob bench group needed to do before they could build their bench was put down a foundation.
The foundation is essential for a cob structure as it protects it from ground water. Cob is susceptible to taking heavy damage from water. With a foundation to lift the cob structure slightly off the ground it helps prevent water from soaking into the cob, which can destroy it. It also gives the bench a sturdy and level layer to lie on.
To build the foundation McAuley and company first dug a shallow ditch that matched the dimensions of the foundation they wanted. They then filled the ditch with river rock (see fig. 1).
After the group laid down river rock they placed pieces of urbanite on top. The gaps between the urbanite were filled with more river rock (see fig. 3). The group ended up applying two layers of urbanite and this served as their foundation.
After the foundation was laid for the bench the group was able to enter the cob making phase of their project. Cob is a relatively simple natural building material to make. It consists of a mixture of clay, river sand and straw. It is advisable to first run tests on your cob before you begin building with it. It is not known what tests the original group used on their cob, but it is recommended to first make a series of test ratios of clay, river sand and straw (i.e. 1:1:1 or 2:1:1) and then test their resistance to weight/pressure, scratches, and water contact. Since clay can differ drastically depending on region, etc, there is no universal ratio for making cob, hence it is advisable to run new tests each time you use different clay or river sand (the straw doesn’t affect the necessary ratio as much as the clay or river sand). Once tests like these are performed to decide the best ratio for your cob building with it can begin.
McAuely and the rest of his cob group used the standard method of making cob, foot power. They laid down river sand and then the clay on a tarp, added water to the mix to saturate it and began to mix it all together by foot (see fig. 4). It helps to lay the sand down first as it acts like a flour on the tarp and prevents the clay from sticking to it. During the mixing it helps the process move faster if two ends of the tarp are occasionally lifted together. This folds the cob like a burrito and allows for better mixing.
Once the sand and clay stick together sufficiently straw is added to the mix. A finished cob mix consistency should be such that if it is folded it will stick together (see fig. 5). An easy consistency test to perform is to roll a small ball out of the mixture and hold it in your hand. Then, without pinching your fingers together, pull your hands apart. If the mixture does not break apart easily it is of the proper consistency. You can also drop the mixture from shoulder height and let it hit the ground. If the mixture flattens it requires more sand, if it cracks, it requires more clay.
The original cob group led by Patrick McAuely appears to have followed this method when making their cob. After they mixed the cob to the proper consistency they rolled the cob into balls the size of melons (see figs 5 and 6) and transported them by hand to the foundation (see fig 7).
After successfully laying a foundation and then making a batch of cob the bench group began to apply the cob to their foundation.
When applying cob directly to the foundation it is best to slam the cob ball down on the foundation as opposed to slapping it down with your hand. This prevents cob ingredients from moving to the top of the mix (see figure8).
A full layer of cob was applied to the foundation of the bench. This first layer served as the seat (see figure 9). After the first layer was applied more urbanite was placed over it to create the general form of the bench (see figure 10). After the general form of the bench was decided the seats were tested for comfort as after the seat form is decided it is difficult to change (see figure 11).
Over time the group built up the bench arms and backs (see figure 12). When applying new layers of cob it is best to wet down the previous layer. This allows for the cob layers to form a better bond. It is also advisable to place small holes in the cob in order to increase dry time (see figure 13).
After much hard work the group eventually had a completed cob bench that awaited plaster (see figure 14).
Once the main structure was completed, plaster was applied to the bench. This portion of the project required far fewer people than the construction of the foundation and the application of the cob. Patrick McAuely and Kirsten Thompson were the main two plasterers.
McAuely and Thompson decided to use a 3:2:1, clay to grated manure to sand ratio for their plaster. Grated manure took the place of straw in their mixing recipe, however they could have also used straw for their plaster if they had chosen to do so. If straw is used in plaster it should be chopped up to lengths less than one inch. Placing the straw in a bucket and chopping it with a weed wacker can speed up this process. Since McAuely and Thompson opted to use manure instead of straw for their plaster they grated the manure using a trashcan lid with a window screen placed over it (see fig. 15).
Since the plaster batches were smaller than cob batches they mixed the plaster in a wheelbarrow (see fig. 16) and then applied it to the bench using a trowel (see fig. 17). Before a layer of plaster is applied to the cob bench it is advised to wet down the cob surface and then smear it with some clay slip. Clay slip is simply super saturated clay (see fig. 18). Applying water and clay slip to the cob bench before applying plaster helps the clay and plaster bond together. To give the plaster a smooth finish McAuely and Thompson applied their plaster in an up direction. They smeared plaster on the wall near the bottom and smeared it in an upward direction (see fig. 17).
Like the application of the cob, the application of plaster takes time. The group slowly but surely saw the bench beautify as they applied plaster to it (see figs. 19, 20 and 21).
After the group successfully applied plaster to the bench they opted to apply some reclaimed tiles to the benches top (see fig. 22 and 22.1) and then they applied boiled linseed oil to the bench to protect it against the weather. Once the tiles were placed on top the group’s hard work was rewarded in the form of a beautiful cob bench (see fig. 23)
Bench Update: April 2007
Starting in February of 2007, repairs were made to the cob bench located at 4th and F St. This was done as a solo project for an Appropriate Technology class at Humboldt State University. The project consisted of repairing and providing updates to other natural building structures in Arcata (see Blue Ox Earthen Oven and Strawbale Earthen Plaster). Fixing the bench at 4th and F St. was the largest fraction of the project. The project took from February of 2007 to April of 2007 to complete and consumed around a total of fifty hours.
A Repair Plan
In order to perform repairs on the bench efficiently I followed the following repair structure:
- Assessment of Damage – This consisted of taking pictures of the damaged areas, some minor measurements of the damage areas and then assessing the extent of the repairs needed to repair the damaged zones.
- Measurements – This consisted of measuring the surface area and volume of the bench so that the proper amount of clay, river sand and straw could be gathered to make cob and plaster to repair the bench.
- Gather Supplies – This consisted of gathering clay, river sand, straw, tarps, buckets and trowels to complete the repairs
- Test cob samples – Test samples of cob were made to decide the best cob mix/ratio
- Make new cob to fix major damage such as holes and depreciations in structure – New cob was made to repair major damage such as holes on the bench
- Make new plaster to replace deteriorating plaster – For this stage the whole bench was stripped of its old plaster and then recovered with fresh plaster
- Apply a sealant to protect against rain – This was the final stage of the repairs. New boiled linseed oil sealant was applied to protect the bench from rain
- Make final tests – A few final tests were done on the completed bench to judge the effectiveness of the final plaster and repairs.
The above is just one example of a general repair plan for a cob structure. This is not necessary but it does help to follow a similar plan when fixing a cob structure. It helps the process run more smoothly and it helps keep things organized.
Assessment of Damage
In February of 2007 the bench was in bad shape. The bench had suffered from the elements and was in a dilapidated condition (see Figs. 1a, 1b). A previous group attempted to repair the bench in 2006. They removed the original tiles placed at the top of the bench but they never replaced them. Leaving the tiles off the top of the bench exposed unprotected cob to the elements and by February 2007 the top of the bench had suffered major erosion (see Fig. 2a).
The seats were suffering from erosion as well (see Figs. 3a, 4a) and the original plaster coat was dilapidated and falling off of the wall (Figs 5a, 6a). The original design plan for repairing this bench was to save most of the plaster that was still on the bench and to patch its cracks with new cob. However, upon further analysis of the existing plaster a decision was made to remove the majority of the old plaster and to plaster the bench from scratch. The existing plaster had been infested with bugs and would easily fall off the bench if gently tapped when wet. It was not suitable to be kept which is part of the reason it was decided to take off and redo the plaster.
There were a few sections of the bench in good condition. These sections were left untouched and only received plaster to fill the cracks. The rest of the bench was plastered from scratch.
Before making the repairs I took some quick measurements of the bench to determine its surface area and volume. I used a standard tape measurer to measure the length, width and height of the bench. I did not subtract the volume of the seats themselves and instead measured the bench as if it were a solid cube structure. My decision to do this was driven both by ease of measurement taking and because I wanted the final measurements to be larger so that I would end up with more materials than were needed to fix the bench. My philosophy here was that it was better to overestimate the bench dynamics than to underestimate them.
I measured the length of the bench to be 10 feet, its height to be 2 feet and its width to be 3 feet. Using these measurements I calculated the surface area of the bench to be 112 square feet and the volume to be 60 cubic feet.
I also measured the dimensions (length, width, height) of the significantly damaged areas. After taking these measurements I combined dimensions of the damaged areas to discover the combined damage length, width and height. Doing this I got a length of about 94 inches, a width of 10 inches and a height of 40 inches. I then calculated the volume of the combined damaged sections to be 21.8 cubic feet.
I used the bench measurements and the damage measurements to determine the amount of materials I needed to fix the bench. I took the total bench volume and decided to use 1/3 of the volume number when gathering clay, river sand and straw. I estimated 1/3 the total amount of the volume to be a reasonable volume for gathering clay, sand and straw as the total volume of the damage appeared to close to this number.
Materials and Costs
The following are the materials and the costs of this project:
|Rice Straw||1 bale||$6.25|
|Boiled Linseed Oil||1 gallon||$16.00|
|10 Gallon Bucket||5||FREE|
|River Sand||1/3 yard||FREE|
I attempted to gather all of the necessary materials for this project in the Arcata community. The goal was to find everything for free, but in the end some of the materials were paid for. I needed clay, sand and straw to make the cob and plaster and I also needed ten-gallon buckets to take measurements and to hold sand, clay and water. A couple of tarps were needed to cover the clay and sand and to mix the cob on. Other materials included a plaster trowel to apply plaster to the bench and linseed oil to cover and protect the bench from rain.
I decided to redo the plaster on the majority of the bench so I took the surface area and volume measurements and used them to decide how much straw, clay and river sand I would need to create enough cob and plaster to cover the bench. Since I did not need materials in excess of the total volume of the bench, I opted to gather about 1/3 the amount of the volume of the bench in clay, sand and straw. I ended up gathering 1/2-yard of clay, a bale of straw (200 sq. ft) and about 1/3-yard of sand
I could not find the clay for free so I purchased 1/2 yard of clay from Wes Greens Contractor Supplies in Arcata for ten dollars (Fig. 7a). I paid six dollars for rice straw at Three G’s Hay and Grain, also located in Arcata. The river sand that was used for the project was gathered directly from the Mad River and shoveled into buckets (Fig. 8a). The trowel I ended up using was purchased for 6 dollars at Ace Hardware, which is where I also purchased a gallon of boiled linseed oil for 16 dollars. Two tarps were purchased for 3 dollars at a local Harbor Freight tools. Luckily the buckets were found on site or were gathered for free elsewhere.
TestingA series of seven test samples of cob and plaster were made prior to making repairs. The mix for the cob/plaster consisted of clay, river sand and rice straw. Cob Sample Ratios were as follows (clay/sand/straw):
The different tests composed on the cob samples were as follows:
1. Dry/crack test: The crack test is designed to determine how much each cob sample cracks after it has been given ample time to try at room temperature. An excellent test sample has no noticeable cracks while a poor test sample has many cracks after drying. (See fig. 9a *note - samples are in order from left to right. Sample 1 is in the upper left of the picture and sample 7 is in the lower right)
2. Scratch Test: The samples are scratched with a wire brush for 100 up/down strokes at varied intensity in order to determine how readily the sample deteriorates when abraded by a rough surface. An excellent case withstands 100 or more strokes and does not fully deteriorate while a poor case noticeably deteriorates under 50 strokes. Samples were also scratched with the edge of a putty knife. A good test with the putty knife survives thirty or more strokes while a poor test survives 15 or less. (see fig. 10a)
3. Water Resistance Test: The samples are placed under running water in order to determine how long it takes for the water to deteriorate the sample. Flow rate of water was10 gallons per minute.. An excellent case lasts for 30 or more seconds, while a poor case lasts for 15 or less seconds.
4. Drop test. A 2’ X 2’ sample of the cob is dropped from a height of 25 feet on to concrete to test its durability. An excellent case survives the drop with minimal cracking and retains its form, while a poor test breaks apart into fragments.
5. Heat Test: Cob samples are placed in an oven set to 600 degrees Fahrenheit and are baked for an hour. Samples are then noted for any damages retained during the heating, such as cracks, brittleness, etc.
6. Soak Test: The samples of cob were completely immersed in water for a total of 15 minuets to test how well the cob mixture held together when soaked by water. A good test consists of a sample of cob that holds together when picked up out of the water, a bad test is one that dissolves in the water (see figs. 11a, 12a).
Results of Tests
After the testing was complete the 1:1:1 ratio (sample 1) came out victorious (see fig 13a). The sample passed most every test with an excellent score. The samples that contained more straw performed the worst and the more balanced or clay heavy samples performed the best.
If I were to go back and do these tests again I would scrap the heat test, as it didn’t give me any valuable information about the quality of the cob. The other tests, particularly the soak, scrape and water pressure tests, were more valuable in discerning the quality of the cob.
After all the tests were completed repairs began on the bench.
Making the Repairs
Once materials were gathered and the cob bench was finished being assessed, measured and tested it was time to make the repairs. Fixing the bench consisted of three simple steps:
- Make new cob to patch noticeable deterioration and holes in the bench
- Make new plaster to cover replace and cover cracks in deteriorating plaster
- Apply a sealant (i.e. boiled linseed oil) to help protect against rain and other elements.
Step 1: New Cob
The first step after testing was completed was to make new cob to repair the holes in the bench created by erosion. The ratio for a cob batch consisted of one five-gallon bucket of clay, one five-gallon bucket of river sand and about one five-gallon bucket of straw. This mixture was thrown directly onto a tarp and then water was applied to it (see figure 14a). While I placed the clay on the tarp first and then the sand I later found that it is better to place the sand down first as the sand acts like flour and prevents the clay from becoming stuck to the tarp.
The Cob was mixed on the tarp and with bare feet (see fig. 15a). Occasionally two sides of the tarp were lifted together to bring the bottom sections of the mixture to the top so they could be mixed. After the sand and clay were thoroughly mixed together I added straw to the mixture. The batch was considered done once the mixture would fold into a burrito form when two sides of the tarp were lifted up. For one person this process took about an hour to complete.
Once the cob was finished it was rolled into balls (see fig. 16a). The cob balls were made to be about the size of a grapefruit. A simple test was done to determine if the texture of the cob balls was correct. This method was explained earlier in the “mixing cob” section of this article. The method consists of grabbing the cob ball with both hands, gripping the ball firmly and then pulling your hands apart without pinching the ball. If the ball does not separate easily, the texture is probably right.
The finished cob balls were placed on the damaged areas of the bench (see fig.17a). I ended up building up the front of the damaged seat with the cob and a few patches were placed throughout the bench (see fig. 18a). Before I placed the cob onto the damaged sections the area was thoroughly moistened and then smeared with a layer of clay slip. It is not as necessary to apply slip when working with cob as it is when working with plaster, but the slip helped to promote better bonding with the existing cob bench. As was also noted earlier in this article, clay slip is simply supersaturated clay. Once water and clay slip was applied to the damaged sections the cob balls were placed on the damaged area and then shaped to match the existing shape of the bench. Small thumb sized holes were then placed around the new cob ball in order to promote dry time.
The new cob bonded well with the bench. I did not sift the clay before mixing the cob and this did not cause any complications until I made my last batch, which had a few unmixed clay chunks left in it. I ended up remixing the sections that had clay chunks and there were no further complications. It is not necessary to sift clay before mixing, though it does help ensure that the clay and sand will mix together easily. The dry time of the applied cob seemed to be about one to two days, depending on the available sunshine.
Step 2: New Plaster
As noted in the introduction, the old plaster was removed from the bench before the new plaster was applied. The removal of the old plaster was a fairly easy task as most of it was already falling off. A putty knife was used to take off the majority of the plaster. Some sections of plaster were left untouched as they were in good condition and did not want to be removed easily (see figs 19a, 20a).
After the old plaster was removed and new cob was applied to deteriorated sections of the bench, new plaster was made. The ratio of materials used for the plaster was the same as the cob, except in the plaster recipe the straw was cut into small pieces no longer than three-quarters of an inch in length using hedge clippers. Two and a half batches of plaster were made during the course of the project. Each batch consisted of one five-gallon bucket of clay, one five-gallon bucket of sand and one five-gallon bucket of rice straw.
The plaster was applied in a similar manner to the cob. First, the section of the bench to be plastered was thoroughly wet down with water, then a layer of clay slip was applied to the surface and then a handful of plaster was applied to the wall. The plaster was smeared onto the wall with a masonry trowel and spread from the bottom up to cover the area. Care was taken to make sure that the plaster was smoothed out as much as possible and in most cases the plaster was applied in an upward motion so as to create uniformity in the plasters texture (see the arm in fig. 18a for an example of new plaster being applied to the wall).
I did the plaster in sections at different times. This caused a few noticeable edges in the plaster showing where one section stopped and another began. However these became less noticeable once sanded.
I did the largest sections of the bench all at one time so as to have more uniformity in the plasters texture (see fig. 21a for an example of new plaster being applied to the right side of the bench. This section was covered at the same time.)
Applying the plaster took the majority of the time on this project. It was well worth it though to see the bench take form with a new coat of plaster (see figs. 22a, 23a. The photos depict a re-plastered bench with some plaster sections still drying)
Step 3: Applying Linseed Oil
Once the new plaster was finished and dried, boiled linseed oil was applied to the bench to protect it against weather damage. I first sanded down the bench with fine sandpaper to make it smoother and take out the little imperfections. After I finished sanding I applied the oil directly to the bench with a standard paintbrush (see figure 24a). This is a very easy process to complete and it goes relatively quickly. I simply poured some of the oil into an old plastic container and brushed the oil into the bench in an up/down motion. I tried to saturate the bench as much as I could, but not oversaturate it so as to conserve oil and increase dry time. It takes a few days in the sun for the oil to dry properly and it leaves a smell for a while after it is dried. This isn’t anything to worry about for structures out in the open that receive plenty of fresh air.
Boiled linseed oil (as opposed to regular linseed oil) is great to apply to cob structures, as it is natural oil that comes from the flax seed, it has a fast dry time and it is good at protecting structures from water damage. It’s better to use boiled linseed oil as it dries faster than non-boiled linseed oil.
A coat of the oil was applied to the cob bench and left to dry. Some slight color change appears to occur with the addition of linseed oil. The original color of the bench was a light brown, but after application of the oil the bench darkened to a medium brown (see figs. 25a, 26a). After the linseed oil was applied the bench was officially finished and ready for use.
The cob and plaster coats have bonded to the wall successfully. Some immediate concerns come from some erosion that has occurred in small places of the wall during recent rains. The bench was covered with a tarp yet moisture still penetrated into the bench and caused minor erosion in some locations in the plaster. This is before any weather protecting linseed oil has been applied. These small bits of damage have since been repaired.
The effect of rain on the bench is the greatest concern at this point. The solidity of the bench is great. The plaster has hardened and handles scratch and contact tests effectively.
Although this project was a repair job a considerable amount of time was required to fix it. The condition of the bench at the start of the project was not deplorable, however it required numerous patches and new plaster in order to re-beautify it.
Part of the reason this bench didn’t last longer in its original form is probably due to exposure to the rain. There is currently no cover for the bench and in the past the bench was only occasionally covered with a tarp during the rainy season. Cob does not fare well when it receives extensive contact with rain or any other form of water.
The damage that can be done to a cob structure by rain can easily be avoided if the structure is either covered or if it receives routine maintenance. One way to protect a cob bench is to cover it with protective oil, such as the boiled linseed oil that was applied to the bench at 4th and F St. If every so often new coats of boiled linseed oil or another protective oil are applied to the bench it should greatly extend the benches health.
Now that the cob bench at 4th and F St. is refurbished it is hoped that it will last longer than the first version of the bench and not require more serious maintenance work again for some time.
After many years of weather and use, the bench is now mostly decayed. It has lost its usable form as a bench and is returning back to earth.