·Home ·Table of Contents ·Conservation and Restoration in Art and Architecture

Monitoring of Fissures on Construction of Saint Duje Cathedral in Split, Croatia Prof. dr. sc. Tanja Roje Bonacci, faculty of Civil Engineering, University of Split, Croatia; Mr.sc.Goran Niksić, Ministry of Culture, State Department for protection of cultural and natural heritage, Split, Croatia; Dott. Ing. Dario Almesberger, Universita di Trieste, SRR.CO.TEC.Trieste, Italy; Doc. dr. sc. predrag P.Miscevic, faculty of Civil Engineering, University of Split, Croatia; Contact

1. INTRODUCTION

Saint Duje cathedral in Split is composed of two parts of distinct age. The main part, Emperor Diocletian's mausoleum, dates at the end of the 3^nd century A.D. In the 17^st century Chatolic church added the "chorus"at the eastern side of the mausoleum. The eastern wall of the mausoleum was taken down in order to unify the two chambers. In the "chorus" chamber multiply repaired fissures were noticed. These fissures are still existent. As it was decided that the construction needs a major repair, the works were peceeded by instrumental monitoring of most obvious fissures. At the same time, air temperature and humidity within and outside the construction were measured. Devices used enable permanent reading of the fissure parameters, but the readings are taken only with period of one day or less. Devices were set up by the italian Company SER.CO.EC. of Trieste. Monitoring was organized and conducted by regional office of the Croatian ministry of culture in Split. Positions of the monitoring devices are shown Fig. 1. At first observations were recorded daily, but as it was found out that this was too frequent, this frequency was reduced. Initially, eleven sensors were used. Observations on measuring location P2 and P3 have shown almost no change of fissure width until 18.09.1996., so that these sensors were then repositioned to points P1 and P11 with intention to suplement the horizontal fissure width measurements with vertical ones. These two measuring locations are positioned on the contact of the construction and mausoleum at core window height. Locations P2 and P3 were abandoned.

Fig 1: Locations of measuring devices

2. DATA ANALYISIS

Fissure monitoring initiated on the 13^th April, 1996 at eleven different locations. These data were supplemented by the air temperature and humidity measurements, both within and outside the cathedral. This paper presents the results from data obtained until 25^th February, 1999. This ammounts to 285 sets of values for measured parameters.

In order to run the statistical analysis, it was necessary to distribute data on time-scale. If this should not have been done, and due to the variable frequency of the readings, wrong picture of the fissure dynamics would have been obtained. Figure 2 was obtained after this has been done, with time on x-axis and absolute sizes of variation of all fissure monitored on the y-axis. This enables the comparison of conduct of fissures. Sensors P2 and P3 were displaced to P1v and P11v when it became obvious those variations at aforementioned are insignificant. Data for P1v and P11v have also shown very small displacements; P11v oscillates around the initial value, while P1v displays a slight increasing trend of an order of 0,15 mm along with the temperature-dependent oscillations. No further analisys has been done for these two fissures.

Fig 2: Time evolution of fissure width

Taking a look at a Figure 2 it is notable that all the fissures monitored can be grouped according to their behaviour. Hence fissures P1, P5, P6 and P11 all have variations in excess of 0,2 mm. The second group contains P4, P7, P8 and P9 and P10 which have variations inferior to 0.2 mm, with a notable increase for P10 towards the end of monitoring period (in future, P10 should be given a special attention).

Figure 3 presents the variations of P4, P7, P8, P9 and P10 with a smaller scale on the y-axis. From it, the behaviour of these fissures can be analised in more detail. The behaviour of these fissures can be decomposed into a constant increase along, with the periodical oscillations either proportional or inversely proportional to the temperature. Fissure P10 shows a distinct increase. As it is notable from Figure 2, P1 and P11 have utterly different behaviour compared to P5 and P6. As the first premise of the analysis was that behaviour of P1 and P11 depends of climatic conditions, so they were analised first.

Fig 3: Time evolution of fissure width at locations P4, P7, P8, P9 and P10

Locations P5 and P6 are positioned on the same fissure, but at different heights. Position P4 is also located on this fissure. All these positions are within the construction on its eastern wall. Position P9 is found on the same wall at height of 9.50 m a.s.l. but on the outer side. Analysis of these data will be presented in a separate chapter. (Fig. 1)

Device locations P1 and P11
The behaviour of fissures at P1 and P11 is completely different from other ones. It is oscillatory; i.e. their width oscillates around fixed average value. Amplitude of these oscillations is quite large, and amounts to 0.79 mm and 1.03 mm respectively. As it can be clearly seen from Fig. 2 and fig. 4 their dynamics seems to be correlated to air temperature outside the construction. As shown on Figure 5, for both fissures correlation coefficient is superior to 90%. All this indicates that these two fissures formed as a consequence of different thermal responses of the two adjacent walls belonging to two buildings of differing age. From Figure 4 it can also be noted that on locations P1 and P11 average width of these two fissures slightly decreased during monitoring period.

Fig 4: Time evolution of fissure width at P1 and P11

Fig 5: Correlation between temperature and fissure movements on the measure points P1 and P11

The correlation analysis of fissure width at P1 and P11 and humidity has shown a null result, and henceforth humidity was not taken into account.

Locations P4, P5, P6 and P9
Locations P4, P5 and P6 are positioned along the same fissure on inner side of construction. Heights of these locations is: P4 at 11.60 m a.s.l, P5 at 14.53 m a.s.l. and P6 at 16.50 m a.s.l.respectively. Initial and final widths of fissure on these locations are given in Table 1.

	Relative minimum	Relative maximum	13. IV 1996	25. II 1999	Total opening in 1048 days
Devices position no.	mm	mm	mm	mm	mm
P4	-0,01	0,09	11,71	11,76	0,06
P5	-0,04	0,66	12,26	12,82	0,56
P6	-0,06	0,78	15,16	15,94	0,78
Table 1:Fissure widths on the beginning and on February 1999

Figure 6 displays the behaviour of fissure on P4, P5 and P6 (0,19 mm p.a., 0,22 mm p.a., 0,25 mm p.a. respectively). Moving average analysis for these locations extracts fissure width temperature dependence. Fissure dynamics on all these locations is dominated by steady widening. Superimposed on this widening is minor, but clearly notable temperature-dependent annually periodical behaviour.

Fig 6: Time evolution of fissure width at P4, P5 and P6

Location P9 is positioned on the outer side of construction's eastern wall, about 1 m below the construction floor. Figure 7 presents behaviour of fissure at this location. As for three previously analysed locations, along with the annually periodical temperature-dependent behaviour, steady widening of the fissure on this location is obvious. In this case, dominant component is temperature-dependent annually periodical behaviour (Figure 7, amplitude of order of 0,06 mm), with widening (0,06 mm p.a.) being somewhat slower than on P4, P5 and P6.

Fig 7: Time evolution of fissure at P9

Fig 8: fissures width's seasonal variation at P4, P5 and P6 from 13 April 1996 to 24 February 1999 made with moving average method

Fig 9: Evolution of fissure on the eastern construction wall (measuring locations P9, P4, P5 and P6). Distance from the construction floor is only approximate

Detailed analysis of dynamics of this fissure should enable us to establish the global dynamics of construction's north-eastern wall. Other walls are presumed immobile, as the adjacent buildings support them.

Table 2 contains data for fissure widening speeds in degrees and millimetres. Rotation axis is located approximately at the floor level.

speed	degree pro year (mean)	0.003564
	device No	P4	P5	P6
speed	[mm/god]	0.020841	0.229246	0.270927
speed	[mm/10god]	0.21	2.29	2.71
Table 2: Widening speeds for fissures on the eastern construction wall

At the beginning of this century, northern wall of the construction has been released by demolition of adjacent building. Archaeological excavations deep under construction floor along the northern wall followed. Foundation soil under the construction is uncontrolled embankment of diverse low-quality materials. Vertical cut-off in this soil, along the construction's northern wall, enables its free deformation influenced by horizontal soil pressure. Roof damages are also enhancing the widening of this fissure. All these reasons called for necessary operations during reconstruction of subfoundation soil undertaken in autumn 1999.

3. CONCLUSION

Non-destructive damage analysis methods are very useful when damages are not so large as to endanger the building's stability, since they demand long-period data gathering. Automatic monitoring and data recording, as well as permanent statistical data analysis is welcome. This paper demonstrated that monitoring shorter than a year is almost of no use in this sort of analysis.

Correct interpretation of the obtained data, based on statistical analysis, is also of major importance. This paper presented just a few possible statistical methods.

© AIPnD , created by NDT.net

|Home|    |Top|