We start from some observations made during an intensive inspection of the southern Bavarian pilgrimage church of Tuntenhausen, some 30 km south-east of Munich. The church was erected in 1628, during the ThirtyYears War. It replaced an older, gothic church, of which only the tower in the west is persisting - a strange twin tower, consisting of two towers built wall-to-wall. The church is one of the earliest "baroque" churches in the area. However, its layout resembles very much the layout of a gothic church. The church has three parallel naves with three bays. The main nave is in fact somewhat higher than the aisles, but the whole keeps the general impression of a "hall" church with three naves of roughly equal height (cf. Fig. 1). The main arches of the church are supported by two pairs of slender pillars. In the east, the church features a chancel with an ambulatory, encircling the partly preserved old gothic polygonal apse with the principal altar.
When the authors of the present paper inspected the church, they were confronted with preparations for restauration work, and their opinion concerning wide cracks in the vaults was asked for. It was not clear whether the damage was caused by defects in the roof structure or rather by ongoing settlements. Investigations on the stucco surfaces indicated that the cracks were not new, but had developed with a decreasing rate since the 17th century (Rosenheim 2007).
It turned out that the present state of the structure could not be explained without a thorough understanding of its building process, which could be deciphered from observations in the attic space.
Quite like the church itself, the roof structure resembles gothic constructions very much (cf. Fig. 2).
The main girders of the roof are constituted by two superimposed trussed frames. The lower one is characterized by two rows of timber columns rising directly above the main arches of the church. This part is well stiffened by various braces rising parallel to the rafters. The upper storey of the roof, by contrast, makes use of a "liegender Stuhl" structure, the most characteristic element of 17th and 18th century roofs in Germany. The roof accommodates the rising barrel vault of the central nave by simply leaving out the tie-beam in the central section.
However, in the intervals between the lunettes of the main vault, projecting beam ends suggest that, originally, the principal transverse trusses of the roof in fact did have continuous tie-beams (cf. Fig. 3). There is no immediately obvious explanation for this feature. However, it turns out that this tiny peculiarity - easily overlooked - is intimately connected to the construction process of the vault and roof, and combines well with the following observations on the vaults, as we shall see:
- The longitudinal barrel vault of the central nave is slightly pointed in cross-section
- The lunettes of the central barrel vault rise considerably (1 m circa) above the capstone of the main arches
- There is no transverse arch, neither in the nave nor in the aisles, neither on the extrados or intrados of the vault
- The vaults of the aisles are rounded transverse barrel vaults, connected to each other by conical lunettes in longitudinal direction
- In cross-section, the profile of the vaults over the aisles is visibly distorted where the adjacent conical lunettes meet each other.
All these observations can be accommodated in the following hypothetical account of the vault erection: First, the outer walls of the church and the two arcaded walls separating the naves were built. At this point, there was no intermediate transverse connection or stiffening between the free-standing, slender arcades and the outer walls. Next, the roof was erected. The roof made use of continuous tie-beams as anchoring beams in the main trusses. This ensured temporal stabilization of the free-standing, very slender masonry structures (even the outer walls are very slender structures, exhibiting huge windows, but no external or internal buttresses). After roof erection, the vaulting was started with the transverse barrel vaults of the aisles, the anchoring beams of the roof still being in place. After completion of the transverse barrel vaults, the builders obviously trusted them to carry the thrust of both the roof truss and the central barrel vault, and dared to cut the anchoring beams. Then, finally, the central barrel vault was built. One might speculate whether the unusual shape of the pointed vault (unique in 17th century Bavarian architecture) was already a reaction to immediate damage or not. However, it is sure that considerable deformation set in immediately after the completion of the construction. This is testified by the deformed shapes of the aisle vaults, the outward tilt of the circumferential walls, as well as a mess of historical rehabilitation structures above the vault (cf. Figs 2 and 3). These include scissor braces intended to carry the thrust of the roof, as well as another pair ofscissors anchored to the circumferential walls, obviously meant to carry the vault thrust. From a modern point of view, it is obvious that the slowly hardening historical mortar would have prevented the aisle vaults from performing their intended function as abutments, and that movements would slowly come to a standstill only as the mortar reached its final stiffness. Most of the thrust is now carried by the conical lunettes. On the other hand, the scissor braces presumably did not do a great service towards stopping the process of deformation.
Anyhow, a large percentage of the present-day deformation must have occured immediately after completion, since the stucco work of the vault already
takes account of the deformed vault shapes, and is also fastened to the vault in such a way as to effectively hide already existing cracks. Therefore, the authors felt entitled to conclude that the present-day cracks present no hazard for the structural stability of the church, even though some of the cracks are up to 10 mm wide.
As a remark concluding the case study, we would like to note that the timber columns of the roof are all tilted. However, this effect seems to be mostly independent from the problems with the vaults. Rather, the upper storey of the roof, the "liegender Stuhl", leans towards the east due to a lack of stiffening. The stiff construction of the lower storey prevented the upper part from toppling, even though the construction of the hip roof over the ambulatory is very defective. Furthermore, it is worth noting that the horizontal beam at mid-elevation of the lower storey of the roof (cf. figure 2) cannot be considered as a raised tie-beam, being connected to the columns only by tenon-and-mortise connections.
Was this article helpful?