The roof of Tuntenhausen is just one of the starting points of the development of the baroque roof in Bavaria. Obviously, it did not work quite as well as it should have. The reasons for the failure of the structural concept at Tuntenhausen and the challenge of similar situations elsewhere are quite obvious: Baroque roofs in southern Germany typically exhibit a pitch of around 45°. Contemporary carpenter's treatises call this pitch the "German" pitch. It is a pitch that is considerably lower than the 60° pitch of typical gothic roofs. The low pitch aggravates problems with thrust, particularly in situations where continuous tie-beams are not feasible, and the traditional carpentry techniques that had been developed over the centuries in the area north of the Alps for the high-pitch gothic roofs were ill adapted to the new challenge of low-pitch roofs, as evidenced by many 17th century roofs, including the one of the Hofkirche at Neuburg on the Danube (1615), the lowest of the Bavarian roofs of the time.
Furthermore, brick barrel vaults with small lunettes became the standard vaulting solution in Bavarian baroque architecture. Unlike the adjacent regions in southwestern Germany, Bavarian baroque builders used timber "lattice" vaults only with reluctance, with only very few exceptions, mostly cupolas. In addition, virtually all the brick vaults of south-eastern Bavarian baroque churches rise high enough into the attic space to prevent continuous tie-beams.
At the same time, even minor church buildings of the 17th and 18th century easily reach the 16.50 m span which is a typical upper threshold on late gothic church nave widths (Augsburg, St. Ulrich, 1499).
Technically even more demanding is the church type developed in Bavaria during the 18th century, namely, churches which embed an octogon for the congregation in a rectangular outer shape, the vault being a cupola which is hidden beneath an ordinary ridge roof. These cupolas easily reach 20 m spans (e.g., parish church of Murnau, see below), pushing the required roof span to the limit of traditional carpentry.
The authors of the present paper have conducted, during the last two years, an extensive survey of baroque roofs in southeastern Germany (Bavaria), with the goal of identifying typical solution strategies followed by carpenters in that region to cope with the problem of large span and the challenges of fairly low pitch and missing tie-beam. Around 40 church roofs have been inspected, covering the entire period from 1600 to 1800. Our study complements the survey conducted by Sachse in the region directly adjacent to our area in the west (Sachse 1975). The full results of the survey will be published in a dedicated monograph (Holzer & Kock 2008). In the remainder of the present paper, we present typical constructive strategies encountered during the survey, and we address the question how much information the contemporary carpenters could possibly have gained from the comparatively rich production of carpenter's treatises published in German language. Also, we discuss issues of the mechanical behaviour of the typical structures encountered.
As late as 1814 (!) Franz Sax, of Vienna, Austria, published a section of a barrel-vaulted church with a roof truss in his treatise on Bau-Technologie undBau-Oekonomie (Sax 1814, cf. Figure 4). The lower storey of the roof structure is constituted by a high "liegender Stuhl". The only precaution taken to carry the thrust of the roof is a raised tie-beam, connected to the braces of the "liegender Stuhl" by means of iron clamps.
From a modern perspective, it seems unlikely that someone would rely on such a construction. However, a roof corresponding exactly to Franz Sax' layout was completed in 1721 in Altenerding (Fig. 5). As opposed to the roof depicted by Franz Sax, the carpenter at Altenerding opted for a tenon-and-mortise connection of the raised tie-beam. The tenon traverses the entire rafter and is connected to its outer face by a long iron strap nailed to the top of the tie-beam. The vault at Altenerding is a very shallow, basket-handle shaped barrel vault.
Quite obviously, this system could not be expected to perform well, particularly with a view to the total absence of outer abutments, side chapels, or any other structure capable of carrying the outward thrust of the vault. As a consequence, the roof has undergone very heavy restoration work in the 20th century, modifying the entire structure into a modern truss by members added between the raised tie-beam and the collar beam.
Franz Sax also illustrated a different setup for roofs with discontinuous tie-beam (cf. Fig. 6). Here, two diagonal braces meet at a king post. The details of the joinery at this critical point remain unclear. Roofs of a similar construction were actually built in Bavaria as early as the last decade of the 17th century. One example is given by the former Augustinian abbey church of Weyarn, within 15 km from Tuntenhausen (cf. Fig. 7). The church is a typical example of the Bavarian "wall-pillar church", i.e. a church with a nave vaulted by a longitudinal barrel vault, accompanied on either side by chapels with transverse barrel vaults. In such a church, the transverse barrel vaults typically spring at the same level as the main barrel vault, but their rise is much lower than that of the longitudinal vault due to the larger span of the latter. The difference is accommodated by rising, conical lunettes. Provided the side
chapels are deep enough, the scheme provides optimal abutment to the thrust of the main vault and thus leaves only the problem of roof thrust unaddressed. At Weyarn, the diagonal braces appear in combination with a low first story of the roof truss, whereas the main "liegender Stuhl" of the roof rises on top of some kind of elevated tie-beam. Therefore, the diagonal ties only have to carry the thrust of the lower portion of the roof. Nevertheless, the joint between the king post and the diagonal ties turns out to be the most critical part of the whole structure. In Weyarn, these joints are lap joints secured by a wooden pin. The pins have failed, but nevertheless the structure still stands largely undamaged, probably due to the fact that the thrust of the roof could be transmitted to the abutments through the rudimentary tie-beams by friction. In this respect, the actual design at Weyarn is much better suited to the problem than the one published by Sax, where the lower storey of the roof is also constituted by a "liegender Stuhl".
An almost identical roof can be found in the church of Perlach (today part of Munich). It dates to 1728.
As opposed to Weyarn, the builders at Perlach one generation later would not rely on traditional joinery to carry the load in the diagonal braces, but introduced iron straps fastened to the king post by a bolt with an iron wedge (Fig. 8).
Surprisingly, none of the roofs discussed so far made use of a roof truss element that was already well known in the 15th century, namely, scissor braces. During the 15th century, scissor braces became something like a standard solution for roofs where continuous tie-beams were not feasible because of a wooden barrel vault. Medieval wooden barrel vaults are not frequently encountered today in Germany, but traces prove that this kind of construction was once quite popular, not only in town halls, but also in churches (cf., e.g., Cramer & Eissing 1996). In what is today Munich, the church of Ramersdorf (first quarter of the 15th century, see Fig. 9), is an example of such a structure. The roof with its 60° pitch once accommodated a trefoil-shaped wooden barrel vault, as is evidenced by traces of stucco at the gable ends of the roof (see Fig. 9, background; cf. also Hosch 1996).
In Ramersdorf, the wooden vault was soon replaced by a masonry one with a much lower rise. The roof shows scissor ties connecting all pairs of rafters.
A roof of very similar structure is also depicted in Johann Wilhelms carpentry treatise, Architectura Civilis, first published in 1649 (Wilhelm 1649, cf. Fig. 10), the earliest dedicated carpentry treatise to appear in German language. However, it needs to be noted that Wihelm's roof is stiffened by a liegender Stuhl frame, which collects the loads of the rafters in transversal frames, and only these frames are equipped with the scissor ties, rather than every pair of rafters as in the medieval roof. Wilhelm's treatise sparked a rich production of further German carpentry treatises, including the Architectura Civilis published by Caspar Walter, of Augsburg, in 1704 (Walter 1704). Each of these treatises until the end of the 18th century presented at least one example of a roof truss with missing tie-beam, in combination with a wooden or lattice barrel vault (e.g., plates 18-21 in Walter 1704). Also, several examples prove that the technique was well known to contemporary carpenters and actually employed in building practice. The roofs of the 17th and early 18th century protestant churches of Augsburg (St. Ulrich, constructed by Caspar Walter in 1710) and Regensburg (Trinity church) make use of scissor braces.
Under these circumstances, it seems somewhat odd that the scheme was only reluctantly introduced into roofs which had a brick vault rather than a wooden vault. In the case of the brick vault, the added weight causes greatly increased thrusts, so that proper precautions to carry the thrust become a must. Nevertheless, carpenters adhered longtime to traditional structural forms and did not readily transfer solutions from one problem class to another.
A roof equipped with scissor braces between every pair of rafters (not just in the principal trusses) can be
found in the church of Murnau (1717-21). However, this church was a completely singular structure at the time of its building. The "nave" of the church is constituted by a Bavarian variant of the Roman Pantheon, namely, an octagonal space with a diameter of 20 m. A perfect sphere is inscribed into the nave, the principal corniche coinciding with the equator of the sphere. Strangely, the huge hemispherical cupola (one of the largest baroque cupolas in Germany, surpassing the one at Weingarten, Swabia, cf. Kutnyi & Wiesneth 2007, with a diameter of 15 m, and approaching the cupolas at Ettal and Dresden, both with a diameter of approximately 25 m) is not visible from the outside, but hidden under an ordinary ridge roof. Accommodating such a large cupola in the attic posed an unprecedented challenge to the carpenter.
In Murnau, the carpenter resorted to scissor braces connecting all pairs of rafters to each other in the area where the cupola protrudes into the roof space (see Fig. 11). Also, he added a bridge-like truss in the plane of the rafters to carry the load from the "open" part of the roof to the parts with continuous tie-beams. Last not least, both the brick cupola and the "hole" cut into the base of the roof are reinforced by closed rings of tie-beams at various levels. The Murnau roof is one of the most remarkable feats of Bavarian baroque carpentry and would deserve a monographical study in which we cannot give here.
Only by the middle of the 18th century, scissor braces had developed into a commonplace solution for the problem of interrupted tie-beams. Regularly, the south-eastern Bavarian architect Johann Alois Mayr (1723-1771) employed scissor-braces whenever there was a need to use discontinuous tie-beams. His church roofs at Baumburg (1756), Marienberg (1764) and Kirchweidach (completed only after Mayr's death, in 1774) all employ scissor-braces. The case of Baumburg (Fig. 12) is particularly interesting for two reasons. Firstly, Baumburg is a standard "wall-pillar"
church with a longitudinal barrel vault covering the nave; as such it reflects the typical situation of Bavarian baroque. Secondly, the joinery of the roof exhibits a marked "engineering" character: The scissor-braces are not connected to the other timber members by traditional joints, but rather, exclusively by iron bolts. On the other hand, all the less important joints follow tradition, even to such an extent that lap joints with a curved end - a characteristic element from the vernacular architecture of the region - abound in the roof truss.
Was this article helpful?