ArticlePDF Available

SAN CARLO DEI BARNABITI: RESTORATION AND REINFORCEMENT OF THE ROOFING OF A FLORENTINE BAROQUE MASTERPIECE

Authors:

Abstract

The church of San Carlo dei Barnabiti houses one of the most important examples of Baroque illusionistic painting in Florence. Given the level of deterioration of the vault, the preliminary investigations required a 3D survey of the current geometry to detect the mainly damaged areas. A point cloud obtained by laser scanning is employed to investigate the deformations and the state of conservation of the vault. The acquired data also revealed some unknown aspects concerning the construction of the structure and its pictorial decoration useful to design the restoration project.
SAN CARLO DEI BARNABITI: RESTORATION AND REINFORCEMENT OF THE
ROOFING OF A FLORENTINE BAROQUE MASTERPIECE
Claudio Mastrodicasa a, Simone Montecchi a, Grazia Tucci b*, Alessandro Conti b, Lidia Fiorini b
a Department of Technical services, Municipality of Florence - direz.servizi.tecnici@comune.fi.it
b GECO Lab., University of Florence - DICEA Dept., Via P.A. Micheli, 8, Florence - grazia.tucci@unifi.it
KEY WORDS: San Carlo dei Barnabiti, 3D modelling, Displacement maps, Basalt fibre reinforcement, Truss end replacement, Lath
and plaster.
ABSTRACT:
The church of San Carlo dei Barnabiti houses one of the most important examples of Baroque illusionistic painting in Florence.
Given the level of deterioration of the vault, the preliminary investigations required a 3D survey of the current geometry to detect the
mainly damaged areas. A point cloud obtained by laser scanning is employed to investigate the deformations and the state of
conservation of the vault. The acquired data also revealed some unknown aspects concerning the construction of the structure and its
pictorial decoration useful to design the restoration project.
* Corresponding author
1. INTRODUCTION
The widespread use of 3D digital models of architectural
complexes for different purposes show their virtually limitless
potentialities and applications. Even though some consolidated
workflows are already in use to produce and process 3D spatial
data, it is however essential to identify on a case by case basis
the most suitable procedures to enhance the peculiarities of
every site. In this case study, the point cloud generated by the
laser scanner is employed to investigate the deformations and
the state of conservation of the vault during the various stages
leading to its restoration. The acquired data also revealed some
unknown aspects concerning the construction of the structure
and its pictorial decoration.
2. HISTORICAL NOTES
The building complex of San Carlo dei Barnabiti, made up of
the church having the same name and the convent, is located in
Via Sant’Agostino, in the centre of Florence, in the block
between Via Maffia and Via dei Serragli. This religious
building was built in 1636 by Gherardo Silvani, who is
considered one of the most prolific architects of the 15th-century
Florentine scene, as well as the father of “Florentine Baroque”,
i.e. a less bizarre and eye-catching version of Baroque, the style
that conquered Rome and the South of Italy in those days.
The work commissioned to Silvani by the Clerics Regular of
Saint Paul (Chierici regolari di San Paolo), also known as
Barnabite Fathers, was accomplished in 1640 with the
demolition of two row houses and part of a third one next to the
ancient little oratory that the Barnabites already owned (Bertani
et al. 1995).
The plan of the church (approx. 16 x 10 m) is still visible in the
ancient built fabric. This is probably the reason why it was
impossible to build the chamber orthogonal to Via di
Sant’Agostino; consequently, the façade and counter-façade
were built within a trapezoid that included the access stairs to
the large window.
Famous artists like Filippo Brilli, Bernardo Ciurini and
Giuseppe Zocchi worked at this construction site. In 1747 the
latter painted the dome and the spandrels.
As far as the nave vault is concerned, according to the historical
researches the church was probably covered by a truss roof. The
documents show that in 1720 a barrel vault with side groins
made of plastered reed mat was built. The following year
Sigismondo Betti painted the “Gloria di San Carlo” in its centre.
The groins were infilled in 1742, while the architectural
perspective was drawn only in 1757-58 by Domenico Stagi,
specialised in “quadrature”.
With the Leopoldine suppression the whole complex became
private, then passed to the Scolopi Fathers and finally in 1866
to the Municipality of Florence, which initially used it as a gym
and then as a place for cultural activities.
3. THE SURVEY OF THE VAULT
The church of San Carlo dei Barnabiti houses one of the most
important examples of Baroque illusionistic painting in
Florence, which resulted from the collaboration of some of the
main 18th-century experts of this artistic form.
The paintings of the nave cover a false “keel” vault. This
structure, widely spread between the 17th and 19th century, was
supported by ribs formed by boards coupled and nailed broken
bonded. The ribs were linked by means of thin rafters and some
plastered reed mats were nailed to them. As these are very
degradable materials, in case of water leakage the mat can be
easily deformed and the painted film, the plaster and the mat
itself can fall off.
Given the level of deterioration of the church, the preliminary
investigations leading to the restoration work required a 3D
survey of the current geometry of the vault to detect the mainly
damaged areas. According to the severity of the deformations,
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-5/W1, 2017
GEOMATICS & RESTORATION – Conservation of Cultural Heritage in the Digital Era, 22–24 May 2017, Florence, Italy
This contribution has been peer-reviewed.
doi:10.5194/isprs-archives-XLII-5-W1-515-2017
515
restorers will identify the most effective techniques on a point-
by-point basis.
As complete scaffolding was already present to perform other
preliminary examinations and restoration work, we decided to
employ a laser scanner following the already consolidated
workflow for this type of surveys. Due to the height of the
scaffolding, we had to place the scanner very close to the
surface; for this reason, every scan inevitably has an area with
very high sampling and big poor quality portions, because the
angle between the incident beam and the acquired surface was
very acute. Consequently, to obtain uniform sampling we
needed to perform 15 scans by selecting the useful parts from
each of them.
The result was a point model made up of more than 560 million
points, which generated a surface model of more than 76
million triangles describing in detail the nave vault and the
dome with spandrels in the presbytery, made of masonry (Tucci
and Bonora 2011).
4. GEOMETRY OF THE VAULT AND
DISPLACEMENT MAP
The method generally used to assess and quantify deformations
employs displacement maps obtained by comparing a point or
surface model of the current condition to a reference model of
the non-deformed condition. If this is not possible, an
elementary surface or plane can be used as a reference model.
Some software offers specific tools to build horizontal or
vertical planes or best fitting planes to a set of points selected
by the user. In some cases, it is also possible to build reference
surfaces starting from curves with a known equation.
In literature, this type of vaults is called “polycentric”, as the
ribs have the configuration of arcs of a circle, a geometric
construction that can be easily reproduced in a construction site.
So, we initially tried to rebuild a similar generating curve
starting from a series of transverse sections in order to measure,
if possible, the displacement of the real curves compared to it.
It turned out that it was impossible to rebuild analytically the
theoretical design as the sections cannot be compared to a
combination of arcs of a circle or to any other elementary curve.
Moreover, the plan is slightly irregular and the height of the
vault is variable, so the sections are all different. Anyway,
regardless of the design, it is no wonder that such a light and
non-rigid structure has progressively settled under its weight
since the construction phase.
The reference surface can therefore be suitably modelled only
starting from the real generating curves identified during the
survey, after having eliminated the deformations due to the
subsequent deterioration processes. To this aim, we assumed
that the mat was more firmly nailed close to the ribs, resulting
in less severe deformations.
However, at least during the initial phase, it was impossible to
access the roof space to verify the position of the ribs. So, this
element was obtained by examining the created surface model,
which unveils some aspects of the vault that cannot be seen to
the naked eye. The mesh shows four large triangular areas on
each side where the surface of the plaster is more regular,
corresponding to the infilling of the pre-existing groins
according to the archival documents (Fig. 1). Based on the
knowledge concerning the construction of this type of vaults,
the main ribs should inevitably be between the groins.
The five sections obtained in the detected areas were
approximated with flat spline curves and then a spline surface
was built by means of a loft function (Fig. 2). This was used as
a reference surface to measure the distance of the point model to
produce a displacement map with the algorithm M3C2 (Lague
et al. 2013) implemented in (CloudCompare, 2017).
Figure 2. The spline surface obtained by five cross sections of
the vault
Given the empirical system used to determine the reference
surface, the less severe anomalies can be ignored, as they are
due to a kind of modelling not sufficiently close to reality or to
the uneven thickness of the plaster. However, the number of
areas in which the geometry deviates significantly from the
reference surface is high; so, they should be immediately
examined by the restorers to check for possible damage. The
maximum deformation is approx. 12 cm.
Figure 1. Mesh model of the vault of San Carlo dei Barnabiti showing the infilled groins
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-5/W1, 2017
GEOMATICS & RESTORATION – Conservation of Cultural Heritage in the Digital Era, 22–24 May 2017, Florence, Italy
This contribution has been peer-reviewed.
doi:10.5194/isprs-archives-XLII-5-W1-515-2017
516
The substantial validity of the adopted process has been
confirmed by the fact that the displacement map is consistent
with the areas where the deterioration is macroscopic and by the
visual inspections subsequently carried out in the roof space,
which confirmed the position of the ribs and also showed that
the wooden structure of the infilled groins is still present.
During the restoration work, the survey could be integrated with
some acquisitions of the roof space to relate the overhanging
structures with what has already been surveyed.
For a quicker localisation of the geometric anomalies, the
displacement map has been superimposed to the image of the
pictorial decoration (Fig. 3). As is known, terrestrial laser
scanners link every acquired point to the corresponding
intensity value of the reflected signal, enabling display modes
similar to a photograph. This allows us to immediately locate
the geometric anomalies by referring them to the pictorial
decoration, as well as to point out even small cracks with a
width lower than the adopted sampling pitch, hence not
represented on the surface model (Tucci et al. 2016).
5. RESTORATION WORK
According to the first results of the surveys carried out, we
identified a restoration plan affecting mainly the structures of
the roof and the roof space to avoid further damage to the vaults
of the nave and presbytery. It includes:
1) In situ refurbishment of the end of one of the trusses by
building a specific glulam prosthesis with the following
techniques:
a) resection of the damaged portion;
b) creation of the glulam prosthesis;
c) connection of the two parts by means of carbon fibre
connecting bars.
2) Reinforcement of the extrados of the hemispherical calotte
of the presbytery dome:
a) shoring of the intrados after placing special protections;
b) levelling of the surfaces with milk of lime and/or lime
putty to fill microcracks or fissures;
c) reinforcement by applying basalt fibre ribbons
embedded in lime putty to the dome extrados.
3) Reinforcement of the curved ribs supporting the lath and
plaster of the nave vault:
a) shoring of the decorated vault after placing special
protections;
b) supporting of every single wooden rib with a light metal
structure to stiffen it;
c) longitudinal connection of all the ribs by stretching
stainless steel cables.
6. FURTHER DEVELOPMENTS AND CONCLUSIONS
Additionally, a mesh model representing the real form, freed
from the decorative perspective drawn exactly to hide the
underlying geometry, shows also the unevenness of the plaster
which cannot be seen to the naked eye. Among the unexpected
results of the survey, the surface model reveals sketches of the
architectural perspective scratched on the plaster, sometimes not
congruent with the painting (Fig. 4).
Whether these are evidences of change of mind or traces of a
previous frame around the central scene painted by Betti before
the intervention of Domenico Stagi, they show that the spatial
information undergoing suitable data processing can be used for
other purposes compared to the traditional ones in the field of
architectural survey (Tucci and Bonora 2015). In fact, it can
both provide useful point indications for restoration work and
inspire further researches that can lead to a more detailed
reconstruction of the history and construction of a building,
through suitable archival and diagnostic confirmations.
Figure 3. Displacement map between the reference surface and the point model, superimposed on a orthoimage of the vault
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-5/W1, 2017
GEOMATICS & RESTORATION – Conservation of Cultural Heritage in the Digital Era, 22–24 May 2017, Florence, Italy
This contribution has been peer-reviewed.
doi:10.5194/isprs-archives-XLII-5-W1-515-2017
517
Figure 4. The mesh model showing a sketch scratched on the
plaster different from the painting (image processed with Adobe
Photoshop’s edge enhancement filters)
REFERENCES
Bertani, L., Cagni, G.M., Castellani, E., Trotta, G., 1995. San
Carlo dei Barnabiti a Firenze, I.T.E.A, Anghiari.
CloudCompare (version 2.8) [GPL software]. 2017. Retrieved
from http://www.cloudcompare.org/
Lague, D., Brodu, N. and Leroux, J., 2013. Accurate 3D
comparison of complex topography with terrestrial laser
scanner: application to the Rangitikei canyon (N-Z), ISPRS
journal of Photogrammmetry and Remote Sensing
Tucci, G., Bonora, V., 2011. From real to "real". A review of
geomatic and rapid prototyping techniques for solid modelling
in Cultural Heritage field. International Archives of the
Photogrammetry, Remote Sensing and Spatial Information
Sciences - ISPRS Archives, 38 (5W16), pp. 575-582.
Tucci, G., Bonora, V., 2015. Geomatics and management of at-
risk cultural heritage. Rendiconti Lincei, 26, pp. 105-114.
Tucci, G., Bonora, V., Fiorini, L., Conti, A. 2016. The Florence
baptistery: 3-D Survey as a knowledge tool for historical and
structural investigations. International Archives of the
Photogrammetry, Remote Sensing and Spatial Information
Sciences - ISPRS Archives, 41, pp. 977-984.
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-5/W1, 2017
GEOMATICS & RESTORATION – Conservation of Cultural Heritage in the Digital Era, 22–24 May 2017, Florence, Italy
This contribution has been peer-reviewed.
doi:10.5194/isprs-archives-XLII-5-W1-515-2017
518
... Geomatics, by means of high-resolution sampling, allow to overcome these challenges [9], producing accurate metric drawings, regardless of the complexity of surveyed geometries also dealing with garden architectures [10]. The resulting point model can be used as a replica of the real object (or, more correctly, of some of its properties) to support many aspects of the conservation process [11], like for designing a restoration project [12] or for building models for structural analysis [13] and other simulations and researches [14]. So, the digital model is not only as a visual representation, but a flexible, integrable, transmissible, and shareable 3-D database. ...
Article
Full-text available
In the past, it was difficult to exhaustively document gardens and artificial grottoes with traditional manual survey techniques and maybe also for this reason the restoration Charters used to pass over the need of an accurate survey. Currently, high resolution sampling allows to obtain 3-D data acquisition and realistic and metrically accurate representations. However, the quality of a digital model is not validated by automatic measures or likelihood but following rigorous acquisition and processing workflows, which must be described in metadata and paradata. The new survey reveals some aspects of the historic water supply system for a better understanding of the construction of this monument and will be extended to the narrow tunnels included in the retaining wall around the grotto.
Article
Full-text available
The Baptistery of San Giovanni is one of the most important pieces of architecture in Florence. It is an octagonal building, encrusted with marble both internally and externally (including the pyramidal roof) and covered inside by a magnificent dome with sparkling gold mosaics. During Dante’s time, it appeared much older than the other monuments, so its origins were considered as hailing straight from Florence’s most remote and mythical history. Even though we have much more data now, scholars still disagree over the interpretations on the origin and construction sequence of the monument. Survey has always been considered a main instrument for understanding historical architecture, mostly from constructional and structural points of view. During the last century, the Baptistery was surveyed using both traditional techniques and the most up-to-date instruments available at the time, such as topography, close-range photogrammetry and laser scanning. So, a review of those early applications, even if partial or isolated, can significantly attest to the state of the art and evolution of survey techniques. During recent years, the Opera di Santa Maria del Fiore promoted new research and a wide range of diagnostic investigations aimed at acquiring greater knowledge of the monument in anticipation of the cleaning and restoration of the outer wall surfaces during 2015. Among this research, GeCo Lab carried out a new systematic and complete laser scanner survey of the whole Baptistery, acquiring data for the more inaccessible parts that were given little attention during other survey campaigns. First of all, the paper analyses recent contributions given by instrumental surveys in advancing knowledge of the building, with references to the cutting-edge techniques and measurement tools used at the time. Then, it describes the new survey campaign, illustrating the approach followed in the planning, data acquisition and data elaboration phases; finally, it gives examples of some interpretations of the structure stemming from the new acquisitions.
Article
The paper starts off by defining cultural heritage and how the concept has extended over time. Then, after attesting to the fragility of our heritage, it continues with the awareness of the need to enable its transmission to posterity. It pinpoints a change in the philosophy of preservation practices, with the preference of prevention over intervention. Then it goes on to define the contribution of geomatics to knowledge, seen as a prerequisite to any decision-making, and to outline the models produced by the techniques classified as part of geomatics: discrete and continuous models, with different balances between the accuracy and efficacy of the descriptions. Some considerations follow on the geomatics tools and techniques now available to document the cultural heritage, with particular reference to the possibility of recording high-resolution data and integrating information from different sensors. The problems presented are illustrated with some examples of applications and the paper concludes by identifying some current research topics.
Article
Surveying techniques such as Terrestrial Laser Scanner have recently been used to measure surface changes via 3D point cloud (PC) comparison. Two types of approaches have been pursued: 3D tracking of homologous parts of the surface to compute a displacement field, and distance calculation between two point clouds when homologous parts cannot be defined. This study deals with the second approach, typical of natural surfaces altered by erosion, sedimentation or vegetation between surveys. Current comparison methods are based on a closest point distance or require at least one of the PC to be meshed with severe limitations when surfaces present roughness elements at all scales. We introduce a new algorithm performing a direct comparison of point clouds in 3D. Surface normals are first estimated in 3D at a scale consistent with the local surface roughness. The measurement of the mean change along the normal direction is then performed with an explicit calculation of a confidence interval. Comparison with existing methods demonstrates the higher accuracy of our approach, as well as an easier workflow due to the absence of surface meshing or DEM generation. Application of the method in a rapidly eroding meandering bedrock river (Rangitikei river canyon) illustrates its ability to handle 3D differences in complex situations (flat and vertical surfaces on the same scene), to reduce uncertainty related to point cloud roughness by local averaging and to generate 3D maps of uncertainty levels. Combined with mm-range local georeferencing of the point clouds, levels of detection down to 6 mm can be routinely attained in situ over ranges of 50 m. We provide evidence for the self-affine behavior of different surfaces. We show how this impacts the calculation of normal vectors and demonstrate the scaling behavior of the level of change detection.
From real to⋯ "real". A review of geomatic and rapid prototyping techniques for solid modelling in Cultural Heritage field. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences-ISPRS Archives
  • G Tucci
  • V Bonora
Tucci, G., Bonora, V., 2011. From real to⋯ "real". A review of geomatic and rapid prototyping techniques for solid modelling in Cultural Heritage field. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences-ISPRS Archives, 38 (5W16), pp. 575-582.