Fish-Bone Feature in Web Masonry
A feature of the web conical shape that draws little notice is the fish-bone pattern for the masonry courses that make up the vault.
Stone courses, whether serving for load walls or vaults, are preferably and more simply built using stone blocks of a uniform height. Changing the height between courses is not a problem, but changing the height of stone blocks in the same course creates difficulties. In the extreme, if the dimensions of the stone blocks are random, the resulting wall has a ramble finish.
The vault web in each of the four sections of the cross vault is made up of two symmetrical triangular shells that rest on the groin ribs and meet at the crown ridge. The stone blocks that make up the vault web are arranged in horizontal uniform courses, stacking them from spring to crown during construction to fill the triangular shells.
The geometric problem is that filling the shell area with uniform stone courses results in a fish-bone pattern of the courses and grout lines, because of the slight conical shape of the web shell. The problem would not exist if the shell were a perfect cylindrical surface.
This is an effect recognized and amply discussed by Fitchen, especially when the conical shape of the web surface is further distorted, such as changing the base of the cross vault intersection to a rectangle or stilting the transverse arches, figure 18.
Figure 18. Fish-bone pattern in the stone courses of cross vault webs – Fitchen (1961, fig. 44).
The fish-bone result is a cumulative effect that becomes acute at the end of the triangular shell, at the opposite side from where the course laying is started. For obvious reasons, the course laying is started at the bottom of the triangular shell near the spring line. If the course laying is started with perfect horizontal lines, the fish-bone effect accumulates at the crown line, creating the result of a fish-bone pattern in the stone blocks and grout lines at the crown line.
To end up with horizontal grout lines at the crown, the web has to start with highly tilted courses at the spring line, relegating the fish-bone effect to the small area in the lower triangular corner of the web shell. Fitchen indicates the use of a triangular stone block above the horizontal spring line to create and inclined bed plane for the successive web courses, figure 18.
Starting to lay courses at an angle may seem a haphazard beginning, not knowing for sure the angle to start with. In actuality this is not an issue for the masons erecting the webs. The planks that make up the vault formwork discussed above, spanning from groin to the transverse arches, define the conical shape of the vault web. They also serve to define the direction of the masonry courses at every level of the triangular web shell.
Starting at the crown line and laying the planks successively, each at a uniform distance from previous one, maintaining strict parallelism among them, results in the last number of planks close to the spring line to take on the fish-bone pattern. All the masons have to do is lay the courses following the planks lines. Starting with strongly tilted courses near the spring line and laying each course along the plank direction at each point along the hutch of the shell results in the final courses near the web apex that are perfectly aligned with the crown line.
The fish-bone pattern is clearly noticeable in the masonry of the vault webs at Castel del Monte, figure 19.
Figure 19. Fish-bone pattern in the web coursing near the spring line, above the tas-de-charge.
A number of courses just above the tas-de-charge, near the spring line, are strongly inclined even to a casual observation. The course inclination tapers down at about one-third of the wed rise above the spring line, where the courses become essentially parallel to the crown line. The area that would have even more course tilting is actually taken over by the masonry that makes up the tas-de-charge. Indeed, the first web masonry starts above the tas-de-charge, about one third of the vault height above the spring line.
This is an indication that the vault webs at Castel del Monte have a conical shape. The transverse arch crowns must be lower that the groin crown at the cross vault center.
Measurements reported by Shirmer indicate that the transverse arch sections have a circular radius that is the same as the radius of the groin arches. This confirms the conical shape for the vault webs at Castel del Monte deducible from the fish-bone pattern in the grout lines. The transverse arch crown is 4.4% lower in height than the crown of the groin arches, at the keystone, based on the geometry discussed above.