Structural health monitoring of cultural heritage structures: Applications on Peristyle of Diocletian`s Palace in Split.

Ivan DUVNJAK, Domagoj DAMJANOVIĆ, Joško KROLO

Dpt. Engineering Mechanics (University of Zagreb) Faculty of Civil Engineering – Zagreb, Croatia

iduvnjak@grad.hr, ddomagoj@grad.hr, krolo@grad.hr

Abstract A large number of ancient structures in Croatia are located in the coastal area where the safety of structures is decreased due to aggressive marine environment. There are several cultural heritage structures which are continuously monitored by Faculty of Civil Engineering at the University of Zagreb. One of such application of long-term structural health monitoring system on cultural heritage structures is presented in this paper. This monitoring system is a part of conservation, restoration and reconstruction procedures which were performed during the last ten years on the part of Diocletian’s palace in Split. Installation of the monitoring system was carried out in two phases: first phase of the research included assessment of embedded material, bearing stone and copper clamps and in the second stage monitoring system was installed, which is based on continuous static measurement of displacement, strain and temperature. Data recordings as well as software support are available at remote location via the internet. Aim of monitoring is early detection of structural damage, that timely alert the system and in such way enables prevention of further degradation of the structure.

Keywords: structural health monitoring, static measurement, laboratory testing, strain, displacement.

1 INTRODUCTION

During the period of life structures may experience extreme loading conditions like earthquakes, strong winds, structural deterioration, external environmental influences and random affects which can cause structural damage. Inspections of structures are extremely important in order to detect damages at early stage of their occurrence. In such a way, structures remain safe and reliable for use. Damages on historical structures mainly relate to cracks, foundation settlements, material degradation and structural deformations. There are many techniques which are capable to detect and locate damages even if they are not visible on the surface of the structure [1].

This paper presents continuous monitoring system which incorporates measurement of displacement, strain and temperature at Peristyle which is a part of historic core of Diocletian`s Palace in Split (Fig. 1&2). The objective of this investigation is to detect damage of natural stone masonry structure at the early stage of its occurrence. Monitoring system is a part of the conservation, restoration and reconstruction procedures which were performed during the last fifteen years on the structure of the Peristyle. Installation of the monitoring system was carried out in two phases. First phase of the research included assessment of

8th European Workshop On Structural Health Monitoring (EWSHM 2016), 5-8 July 2016, Spain, Bilbao

www.ndt.net/app.EWSHM2016M

ore

info

abo

ut th

is a

rtic

le: h

ttp://

ww

w.n

dt.n

et/?

id=

2018

7

2

embedded material, bearing stone and copper clamps, which was conducted during 2012. At the end of 2012, through the second phase of research, structural health monitoring system was installed. Since the historical significance of Diocletian`s Palace is enormous as an monument structure special attention is devoted to this experimental research. The investigations are carried out by the Structural Testing Laboratory at Faculty of Civil Engineering of the University of Zagreb.

Figure 1: View on East colonnade, Protiron and West colonade

The restoration and rehabilitation of the Peristyle of Diocletian's Palace is described as one of the most significant and the most extensive projects in the historic core of Split since its inscription on the UNESCO's World Heritage List 30 years ago. Started in 2004 as a stone cleaning operation, after a two-year survey of the as-found condition of the Roman colonnade and the surrounding buildings from later periods, it developed into a complex project. This project included archaeological, geophysical and geo-mechanical research, consolidation of foundations and upper structures, cleaning and conservation of stone, plaster and other materials, lighting and presentation of this multi-layered monument. [2]

Figure 2: View on Chapel St. Roche

3

2 MECHANICAL PROPRETIES OF MATERIALS

First stage of the research, which included assessment of mechanical properties of imbedded materials, was used as the base for the installation of long-term structural health monitoring (SHM) system. In this part mechanical properties of the copper clamps were determined by tensile test in the laboratory. On the site, residual strain measurement was performed in order to determine the actual force in the clamps using hole-drilling method. Furthermore, some basic mechanical properties of stone were determined by static compressive tests which were performed in laboratory.

2.1. Residual strain measurement

The most widely used technique for measuring residual stress is the hole-drilling strain-gage method of stress relaxation. This method involves attaching single strain gauge to the surface, drilling a hole in the vicinity of the gauge and then measuring the residual strains (Fig. 3) [3][4].

Figure 3: View on vault of Protiron and copper clamp

The residual strain measurement was performed in order to determine the actual force in the copper clamp. The hole drilling method was used on twelve measuring points, nine of them at the front and three at the inside of the Protiron (Fig. 4). The clamps are subjected to a uniaxial tensile or compression stress therefore for released strain measurement we used only single strain gauge.

Front side

Rear side Figure 4: Measuring points of Protiron

2.2. Measurement results

An example of continuous records of residual stresses is illustrated on the Fig.5. Using the magnitude of residual stresses the force [kN] is calculated and results are presented in the Fig.6.

4

Figure 5: Recordings of residual stresses at measuring points 3 and 12

Front side

Rear side Figure 6: Results of measured residual force (kN)

Mechanical properties of the copper clamp were determined using tensile test (E=65 GPa Fig. 7 left) [5]. Standard compressive test on stone cubes (σM=65 MPa) was performed for determination of strength (Fig. 7 right) [6], while additional tests in which strain was measured were used for determination of elasticity modulus (E=48,4 MPa). Mechanical properties of materials were determined using electro mechanical testing machine Zwick Z600E in laboratory.

Figure 7: Recordings of tensile test of copper clamps and compressive strength of stone

3 THE CONTINIOUS MONITORING SYSTEM

It is known for a long time that service loads, environmental and accidental actions may cause damage to the structural systems. In this process, long life maintenance plans play an important role. Regular inspections and condition assessments of engineering structures can

-20

-15

-10

-5

0

5

0 50 100 150 200 250 300 350

Res

idua

l str

ess

(M

Pa)

Time (sec)

Measuring point 3

-5

0

5

10

15

20

0 50 100 150 200 250 300 350

Res

idua

l str

ess

(M

Pa)

Time (sec)

Measuring point 12

0

20

40

60

80

100

120

140

0 1 2 3 4 5 6 7 8 9

Str

ess

(M

Pa)

Strain (%)

Tensile test of copper clamp

0

10

20

30

40

50

60

70

0 0,5 1 1,5 2 2,5 3 3,5 4

Str

ess

(M

Pa)

Displacement (mm)

Compressive test of stone

5

allow programmed repair works and avoid undesired economic, cultural and life losses. In case of historical constructions these aspects are highlighted due to the high level of the structures importance [7].

Long-term structural health monitoring system was installed in January 2013 and it is based on LabVIEW software and data acquisition system NI CompaqDAQ. The system is a high speed mulit-channel digital data acquisition system with 24-bit resolution. During monitoring the data were collected on total of seventeen measuring points at sampling interval of 30 minutes. The Labview software is used for data acquisition and signal processing. For the monitoring system a new program was designed to monitor data from sensors. The monitoring program consists of graphic interface (front panel window) on which important measuring values are visualized (Fig. 8). Acquired data as well as software support are available at remote location via the internet.

Figure 8: Front panel window of software

The objective of this SHM system is to detect anomalies and to perform checks of assigned thresholds. Many different sensors can be used in SHM system. For this purpose system was designed to measure displacements, strain and temperature (Fig.9 & 10) at discrete nodes of a structure and to detect sudden movements. Monitoring consists of high precision sensors, such as linear variable differential transformers (LVDT) and strain gauges.

SG1

SG2SG3

SG4

L2L1

Strain gauge - SG

LVDT - L

Legend:

L3

LVDT - L

Legend:

Figure 9: Disposition of the measuring points at Chapel St. Roche and East colonnade

6

SG7

L6

Strain gauge - SG

LVDT - L

Temperature - TEMP

TEMP

Legend:

L5

L4

L8

L9

L7

SG5

Strain gauge - SG

Legend:

SG6

Figure 10: Disposition of the measuring points at Protiron and West colonnade

On the base of visual inspection of the structure, crack disposition was determined and relevant measuring locations were selected. Transducers for displacement measurement (LVDT) were installed to measure relative displacement between the stone blocks and for crack opening measurements. In total nine displacement transducers were installed in monitoring system: two at Chapel of St. Roche (L1 and L2, Fig.9&11 left) one at East colonnade (L3, Fig.9&11 right) and six at Protiron (from L4 to L9, Fig.10.).

Figure 11: Displacement measuring points at Chapel St. Roche and East colonnade

Figure 12: Strain measuring points at Chapel St. Roche

L3

L1

L2

SG1

SG2

SG3

SG4

7

Strain gauges measure the strain state at discrete locations of the masonry and vaults. Two pairs of strain gauges (SG1&2 and SG3&4) are mounted at Chapel St. Roche in horizontal and vertical direction. These four strain gauges are mounted to detect crack propagation of the bearing stone (Fig.12.). Another two strain gauges (SG5 and SG6) in two directions are mounted at West colonnade at keystone of the vault (Fig.13. right). The last one strain gauge (SG7) is mounted on springer of the vault at Protiron (Fig.13. left). Simultaneously, temperature and its effects on the structure are measured.

Figure 13: Measuring points at Protiron and West colonnade

3.1 Measurement results

The measurement results are represented in time domain which provides information related to how the instrumented structural elements behave during monitoring period. The results of displacement, strain and temperature are illustrated over three year period, from January 2013 to March 2016.

Figure 14: Time series of displacements

Measurement results of displacement do not indicate substantial changes in the period of three years (Fig. 14), although there are some shifts in measured results. The maximum value of displacement of approximately 0.1 mm was detected at LVDT 2, measuring relative displacement between stone blocks. Behaviour trend of measured displacements is in accordance with ambient temperature (Fig. 15).

From the Fig. 16 it can be observed that strain at position SG2 increased significantly over the period of first year up to maximum value of 400 µm/m. Thereafter, strain at SG2 changed the slope and behaved almost linear with time reaching a maximum value of 650 µm/m. At same position strain gauge (SG1) in horizontal direction increased slightly over the period of first eighteen months and thereafter decreased significantly from 330 to 100 µm/m. After

-0,08

-0,06

-0,04

-0,02

0,00

0,02

0,04

0,06

0,08

0,10

0,12

01.13. 04.13. 06.13. 09.13. 12.13. 03.14. 06.14. 09.14. 12.14. 03.15. 06.15. 09.15. 12.15. 03.16.

Dis

plac

em

ent

(m

m)

Time (month.year)

L1 L2 L3 L4

-0,08

-0,06

-0,04

-0,02

0,00

0,02

0,04

0,06

0,08

0,10

0,12

01.13. 04.13. 06.13. 09.13. 12.13. 03.14. 06.14. 09.14. 12.14. 03.15. 06.15. 09.15. 12.15. 03.16.

Dis

plac

em

ent

(m

m)

Time (month.year)

L5 L6 L7 L8 L9

SG5 SG6

SG7

L5

L6

8

sudden decrease strain at SG1 behaved almost linear with time reaching maximum value of 350 µm/m. The measured data at position SG3&4 also show that the strains are almost equal and increased in time without any sudden jumps.

Figure 15: Time series of ambient temperature

Figure 16: Time series of strains

The measured strains of the West colonnade and Protiron (SG5, SG6 & SG7) presented on the right graph in Fig.16. Measured strains at these measuring points are generally stable over time excluding two decreases at position SG6 which could be noticed at same year period at beginning of June 2014 and 2015.

After each sudden decrease of strain visual inspection of structure was performed. Strain gauges at position SG1&2 are located close to the cracks in stone block. Based on visual inspection it was concluded that relatively high strain level measured at these positions is affected by propagation of existing cracks and formation of new cracks.

4 CONCLUSION

This paper presents long-term structural health monitoring of the historical masonry structure. A total of seventeen measuring points were selected to monitor relative displacements between the stone blocks, crack openings, strain and temperature at discrete nodes of structure. The aim of monitoring is early detection of structural anomalies and check of assigned thresholds, that timely alert the system and in that way enable prevention of further degradation of the structure. Structural health monitoring of Peristyle is useful instrument for improving the efficiency of maintenance of this historical structure.

Based on the presented measurement results, after three year period of monitoring and according to previous experience we can conclude that current state of the structure is stable. According to measured displacements, strains and visual inspection it is obvious that some

-5

0

5

10

15

20

25

30

35

40

45

01.13. 04.13. 06.13. 09.13. 12.13. 03.14. 06.14. 09.14. 12.14. 03.15. 06.15. 09.15. 12.15. 03.16.

Tem

pera

ture

( C)

Time (month.year)

-100-50

050

100150200250300350400450500550600650700

01.13. 04.13. 06.13. 09.13. 12.13. 03.14. 06.14. 09.14. 12.14. 03.15. 06.15. 09.15. 12.15. 03.16.

Str

ain

(μm

/m)

Time (month.year)

SG1 SG2 SG3 SG4

-100-50

050

100150200250300350400450500550600650700

01.13. 04.13. 06.13. 09.13. 12.13. 03.14. 06.14. 09.14. 12.14. 03.15. 06.15. 09.15. 12.15. 03.16.

Str

ain

(μm

/m)

Time (month.year)

SG5 SG6 SG7

9

crack are propagating. . Structural health monitoring of historical masonry structures is much more demanding

than the monitoring of modern structures. Behaviour of masonry structures is very difficult to assess in short period of monitoring. Therefore, it is necessary to monitor the condition of structure through longer time period.

REFERENCES

[1] S. K. Verma, S. S. Bhadauria, and S. Akhtar, ‘Review of Nondestructive Testing Methods for Condition Monitoring of Concrete Structures’, J. Constr. Eng., vol. 2013, no. 2008, 2013.

[2] G. Nikšić and S. M. Sunara, ‘Historic core of Split and the Peristyle of Diocletian`s. Economic impact of cultural heritage preservation’, pp. 813–819.

[3] G. S. Schajer, ‘Measurement of Non-Uniform residual stresses using the hole-drilling method.Part 1-Stress calculation procedure’, J. Eng. Technol., vol. 110, pp. 339–349, 1988.

[4] L. Suchanek and I. Zetkova, ‘Evaluation of the surface small holes drilled by unconventional methods’, Energy Procedia, vol. 100, no. C, pp. 1582–1590, 2015.

[5] EN ISO 6892-1 Metallic materials - Tensile testing - Method of test at room temperature. 2009, pp. 1–72.

[6] EN 1926:2006. Natural stone test methods -- Determination of uniaxial compressive strenght. 2006, p. 18.

[7] L. F. Ramos and P. B. Lourenço, ‘Static and dynamic structural monitoring of the Santa Maria of Belém Church , in Lisbon’, pp. 1–20, 2005.