EXPERIMENT 1
Aim:
a. To experiment working with PoP as a mortar to lay the first layer of tiles using guide strings to obtain the right curve.
b. To learn the challenges of such a technique which is unfamiliar to our masons, getting them used to building in thin air
c. To see if we can successfully lay the first layer such that it stands stable under its self weight without the support of centering
Step 1:
Drawing the curve and fixing centering
a. A catenary arch with our design span (4.47m) and rise (447mm) was traced by hanging a chain on a wall.
b. 2 templates were made by bending steel rods to the shape of the traced curve. These templates were connected at corresponding points with guide strings to help lay the tiles in the right shape.
c. A strip of bent plywood was fixed to the template to serve as centering for the first tile.
d. The two templates were held together with horizontal supports in between to make them stable.
Step 2:
Laying first tile line in PoP
a. First line of tiles is laid resting on the formwork . Ends rest on the support walls at an angle.
b. It took us a few trials and errors to get the right consistency of PoP mix. The biggest challenge was that it sets so fast that we had little time to align the tile according to the guides. However, we gradually came to form a system with one person mixing the PoP and applying it on the tile and another holding it in place while also checking whether it is in line with the guide string.
The entire line is completed from support to support. 2nd, 3rd, 4th, 5th tile lines laid, cantilevering from preceding tile line and with staggering joints.
Laying tiles in PoP mortar
Note how each tile takes approximately 30 seconds from applying mortar to fixing and adjusting in place to initial setting of PoP.
Laying the edge tiles
The tiles are NOT supported on the horizontal ties between the templates. Only the first line of tiles is supported on the formwork. The other tile lines cantilever from the preceding tile line
Step 3:
Laying successive lines in 1st course
Each successive line is laid only after the preceding line is completed from support to support to form its own arch. Once an arch is complete, it bears its own weight to the support walls and is, in effect, no longer cantilevering from the previous tile line. This prevents any unnecessary loading on the weak PoP joints and avoids cracking.
Step 4:
Finishing laying the 1st course
14 lines are laid and we stop just before the plywood strip on the other side (this piece was installed earlier for a previous trial that failed)
While the layer seems to be stable – standing mostly without support except for the first tile line, we were starting to feel apprehensive that we had allowed the curve to sag on one of the portion which was touching the horizontal ties.
Step 5:
Laying plaster
We laid a thin layer of plaster, 1:3 (cement : sand), before starting the second tile line. However, at this point, our apprehensions started to grow that the vault was resting on one of the horizontal ties and was therefore not entirely in compression. Any further work could be futile if that was the case because the vault would fail on removing the centering.
The vault failed while removing the center support of the centering.
Exp1 collapse_2
Step 6 :
Removal of centering
We decided to remove the centering to check if the vault stands by itself.
POSSIBLE REASONS FOR FAILURE:
a. As already mentioned, we were already doubtful about the stability of the vault because we feared that one portion of it was resting on one of the horizontal ties of the formwork support. Hence the vault was never fully in compression and failed when the centering was removed
b. Our centering was poorly designed- while removing it, the vault curve was disturbed. Note that we had only constructed the first layer which is fragile because of its thinness and the PoP joints are extremely weak.
c. We rushed in removing the centering and did not allow the 3 layers of the vault to get completed which would allow the vault to achieve the thickness and strength required to stand without centering.
d. It is also possible that the water from the plaster (note that our plaster was also slightly runny to enable a thin layer) also weakened the PoP joints which could not take the movements while removing the centering.
LESSONS:
a. Our centering needs improvement- it needs to be designed such that it is free from the vault and can be dismantled without disturbing the vault or causing unnecessary movements in the structure.
b. Instead of laying the entire layer 1 first before starting layer 2, it might be better to finish a few lines of Layer 1, move to layer 2 and build the two layers back and forth but together.
c. We need to be more careful maintaining the curve. In Exp 1, our curve was slightly off which is why one portion rested (or seemed to rest) on the horizontal tie.
d. It might be better to avoid the plaster layer between layers 1 and 2 and instead use a rich mortar (perhaps pure cement) to lay layer 2 on 1.
Our first set of trial and errors, understanding working with PoP and its challenges. Learning how to build a vault in air using guides and no shuttering.
Experiment to check whether an arch can stand stable under self weight without centering.
Specifications of materials used in the experiment with quantities and approximate costs to extrapolate the economics of such construction.