Tara Green, Zoe Roettger, and Clark Glymph
Introduction
Our experiments involved testing out full scale, working models of pendulum saws, evidence of which is found in masonry remains primarily in Mycenae, during the height palatial period (1370-1190 BC). Modeled after the three types of cuts left on these masonry remains—cuts in the ground, horizontal ones, and vertical ones—saws were designed and tested to see how effective the three coordinating pendulum saws could be.
Methods
This lab occurred over four hours. During this time, each saw had two 4-person lab groups assigned to it who were responsible for sawing as deep as possible into one stone or more.
The questions that this data is concerned with involves the effectiveness/efficiency of the pendulum saw, which was tested by comparing the workability of a fulcrum based, vertical, and horizontal saw designs. Each of the saws’ blades were made of bronze and were about the same size and shape, confining the variables to those pertaining to the saws’ designs.
To cut the limestone, emery dust and water was applied to the rock, which then transferred onto the dull blade. Since emery is harder than limestone, this abrasive dust allowed the saw to cut through the rock.
Original Pendulum Saw
Two groups worked with the design by Nick Blackwell (Indiana University), an upright pendulum saw with a triangular frame and an adjustable beam that the fulcrum hung from. The fulcrum was attached to the bronze blade, which had a radius of 17 cm and an arc length of 40 cm (top of blade) and 49 cm (bottom of blade). In order to cut the item underneath the saw, the fulcrum was pulled left-to-right by opposite ropes, looped eight times around and positioned 35 cm above the blade. (Shortly after we began sawing, however, we realized that it was more beneficial to have the rope positioned lower and closer to the saw as well as coming out on either side, which are adjustments that we made accordingly.)
The fulcrum of our saw was designed to have an extra long shaft hole so that we could fit woodchips into the excess space and take them out as needed, allowing us to lower the saw as we continued cutting, rather than taking the center beam out entirely. In practice, however, the woodchips fell out often, and we thought that the saw did a good job of keeping contact with the stone without us needing to remove the woodchips—partially due to how slow our progress was. We also found it helpful to stabilize the fulcrum by keeping it in place with ropes wrapped around both ends of the center bean as well as to attach a bucket that was loaded with sand, which provided additional stability.
We began with a large piece of limestone, with bases of 30 cm and 44.5 cm, and a height of 23 cm. We did not chisel an initial line into due to our ability to effectively gouge a deep enough line into the rock. As we did not have a chisel line for a guide, we ended up sawing multiple grooves into the rock over the course of roughly 1 hour, all of which were less than 0.5 cm deep.
When it became apparent that it would be nearly impossible to establish one groove after multiple shallow ones had been created, we started again with a new rock, which was a 30 cm by 30 cm square, with a height of 2.5 cm. After deciding to cut along the diagonal so as to maximize the length of the cut and, thus, the rock-to-saw contact time per swing, we chiseled a line in the middle of the rock. We chose to only chisel in the middle as we had learned that this is where the pendulum saw makes first contact and, seeing as our ability to make a straight line with a chisel (or a line at all) was rather poor, we only did the requisite amount of chiseling to set up our cutting line.
To establish a single groove, we did our best to keep the saw solely on the rock as we found that, when the saw lost contact with the rock prior to a groove being formed, it did not stay in one place. In order to do this, we had to make fast and tight pulls on the rope, and most people found this pace tiring, especially when combined with the rope burn produced by quickness of the sawing. Once a groove had been established, however, we were able to make longer strokes.
Unfortunately, because we were sawing on the ground, our saw seemed to press the rock into the dirt, which had become wet and muddy from all the water and sludge that was produced by the sawing and had been washed off the rock. So, we propped the rock up on two other stones, which both brought it closer to the blade and provided a foundation that allowed for the rock, rather than the earth, to take the majority of the force from the saw and thus make our sawing more efficient.
Over the course of the 1.5 hours that we spent sawing the second stone, we used about 6–7 cups of emery dust and 7–8 cups of water. Those who were sawing reported that they were able to feel when the stone needed one or the other—when the sawing felt smooth, more emery dust was needed; when the blade felt almost stuck, more water was needed. The sawing of rock also produced a distinctive sound and made the sludge produced by the emery-water combination more grey (due to the color of the rock), so those who were not sawing were still able to tell when the saw was working effectively.
In the last 30 minutes of the lab, we changed our handhold on the rope to one that is akin to rowing by wrapping the rope around a sturdy object, such as a hammer. This made the physical labor much less demanding and allowed us to pull even more quickly as well as synchronize the pace of the two people pulling. These changes led to an increase in sawing efficiency, as shown by the table below.
The following day, we measured the depth of the cut on the second stone at various points in order to give us a full grasp of the shape of the curved cut. From these measurements, the pendulum was around 1.575 m.
Horizontal Pendulum Saw
The vertical saw design was another design proposed by Nick Blackwell to make saw cuts horizontally in the stone, possibly to create thresholds. For example, at the Treasury of Minyas at Orchomenos, the threshold of the rooms has a horizontal threshold carved out of the bedrock that is only a couple inches above the ground. Nick proposed a couple of designs for the horizontal saw that are shown below.
When the groups initially tried to use the horizontal saw, they began by setting up the entire saw on the mulch and tried to hold the limestone block in place using their feet. Soon though, they realized that the saw was unable to cut the stone because it kept moving so they placed it against a wall to give it stability and stop it from moving while being cut. Unfortunately there was still not enough weight on top of the stone, leading to someone always having to use their body weight to hold the stone down. Eventually, the group moved the entire saw onto flatter ground, since the saw was making two grooves in the stone, and they hoped that flatter and sturdier ground would fix the issue.
The design originally had a vertical pendulum beam which would have been pulled back and forth via a rope. The horizontal arm, which had the blade at the end, had a wooden brace on the other end which fit around the vertical beam. Thus when the vertical pendulum beam was moved back and forth, the horizontal arm and the saw would also move back and forth. However, after only 5 back-and-forth swings (about 8 seconds), the wooden brace connecting the two pieces snapped, so the groups adapted by tying a rope to the horizontal arm and pulling it back and forth manually. Since the brace broke, the vertical pendulum arm was no longer useful, nor did the group have the gravitational advantage that was essential to the design.
We worked with one stone for the entire lab period. While we did not initially chisel a groove into the stone, after 74 swings (140 seconds) we realized that we were not making a mark and chiseled a guide line into the stone. We sawed the stone for 3951 seconds (01:05:51), for a total of 1,785 back-and-forth swings. That averaged out to about 2.2 back-and-forth swings per second. However that amount varied greatly depending on who was pulling the ropes and where the saw was set up. Emery and water were added periodically, roughly every 40-60 back-and-forth swings, or 90-130 seconds. It was sometimes added by mixing together a paste (about a 3:1 emery to water ratio, sort of a wet-sand texture) and then applying it to the stone, and sometimes by pouring water over the stone and flicking on dry emery. Once we had been cutting for a while, we noticed that some of the wet emery would drip onto the ground below the stone, and we would sometimes just use our hands to put the wet emery on the stone.
After 317 swings (611 seconds) we started to notice a visible cut mark, along with the gray slurry. The saw needed to be moved forward into the groove about every 120 back-and-forth swings (260 seconds). This was done either by using a sledgehammer to inch the whole saw forward or by lifting the saw and moving it forward. The saw blade became loose and needed to be tightened after the first 354 swings, then again after another 460, at which point we replaced the screws with bolts. Unfortunately, the blade continued to be loose, but the groups could not think of another solution that would not take a long time, so we continued to saw with the loose blade.
As mentioned earlier, the saw created two grooves in the stone, instead of the single groove that the other groups created. We measured the depth of the groove the day after the lab was completed. The side of the stone which we cut into was not flat, so we used a yardstick to measure depth from a flat position and then subtracted the width of the yardstick. The two cuts in the stone, Cut A and Cut B, were both 1 cm wide and the depth for each was measured in 0.5 cm increments along the cut.
Cut A:
Cut B:
Vertical Pendulum Saw
The vertical saw design was another proposed design by Nick Blackwell to make saw cuts vertically against stone possibly for decorative designs. For example on the Treasury of Atreus facade Nick proposes some of the images below as possibilities of how this vertical saw would have been used.
Using the vertical saw two of our lab groups attached a stone to a large tree using ratchet straps. These straps proved to be difficult being so close to the sawed part of the stone that they were cut through eventually.
These lab groups also attached a bucket of sand to the post that hangs down to help with a possible “mechanical advantage” but it didn’t seem to work that well. At the front of the saw a wooden frame moved the saw blade up and down but proved to be a problem as the lab went on and thus the groups decided to move it to an upright position. In the upright position the frame did not touch the base of the saw which with the bucket of sand strained the beam holding the saw leading to the beam and base to split in several places. The group noted that if we were to do this again they would want to remove the movable front frame and have a shorter distance from the saw blade to the fulcrum. This group believed the effort they were putting in was more than other saw groups from observation (though groups did not switch to different saws so this may be unknown). They believed this excess effort was due to the poor design of having a person on the front pushing the blade up and then someone in the back pushing the blade back down causing an odd mode of motion for both parties.
| Time (h:m) | Cut Depth (cm) |
| 1:25 | 0.3 |
| 1:39 | 0.4 |
| 1:59 | 0.7 |
| 2:19 | 0.9 |
| 2:27 | 0.9 |
After the first hour and twenty five minutes the vertical saw had cut 0.3 cm and then they timed themselves for fifteen minutes and got another 0.1 cm. Afterwards they decided to do a hard push instead of a more leisurely pace and were able to get 0.3 cm in fifteen minutes. After these pushes it was noted that the bronze saw blade was hot to the touch. Their average work rate would be 1 cm per hour. Their stone was 5 cm thick so to cut through the whole stone it would take 5 hours. However they noted that the initial depth data was affected by the grey sludge that was a byproduct of cutting the stone.
After cleaning the stone of any byproduct they measured the depth of the whole cut and concluded that the actual cut was actually 1.761 cm deep making their original calculation 0.8 cm off. With this new depth their revised work rate is 0.7188 cm/hr over the course of the full two hours. It’s important to note that their pace sped up after the first hour and a half so this work rate is less than what it likely was. In order to get a good idea of the work rate there must be more experimentation. It also could be that the stone was more difficult to initially cut into but easier once a groove had been carved out or as users got more comfortable with the equipment.
Depth of saw cut, measured at each ½ cm. after a 2 ½ hour period of cutting, fully cleaned and dried.
All data points for the depth of cut
This group also decided to interview a few passing spectators including students, alumni, professors, and wedding planners?! All of them were intrigued by the saws and wanted to know what our class was doing. No one said they were grand but a few said that if they needed to make an outdoor patio and it was between these saws or a handsaw they would pick these saws. Noting these saws could possibly be an aesthetic or powerful appeal Also, one person said he thought we were mining for ore. The wedding planner said it was the most interesting thing he’d seen done in front of Anderson in his time at Carleton. A geology professor said it was interesting and not what he was expecting.
Conclusions
All groups reported that they struggled with establishing one groove by keeping the saw blade in the same place every time that they cut. Additionally, all groups were surprised by how slow their saw’s progress was and by how shallow their cut was after four hours of work. These qualitative observations, supported by quantitative data, reveal that the mechanical benefit of using a pendulum saw does not, at least from these experiments, seem as advantageous as we all had assumed. However, when one cut was established, it was incredibly straight and precise, which might have been preferred (by the elites in the Bronze Age who could afford the delays) to the inexact cuts produced by a handsaw.
It is important to note that the saw blades that we were working with were at least twice as thick as those that would have been used during the Bronze Age, and this width likely slowed down our sawing progress. However, our progress was quite slow regardless!
0 thoughts on “Week 9 Lab Summary: Bronze Age Saws”