Crafts and Trades
Characterisation of the State of compression of Pietra D'Istria elements by Non Destructive Ultrasonic Technique
[Reprinted from: http://www.ndt.net/article/wcndt00/papers/idn173/idn173.htm.]
ABSTRACTAbout the 80% of the Venetian monuments and edifices have been built up, decorated or covered by limestone coming from quarries located in the Istria peninsula.
Nowadays, these pits are still exploited for the extraction of Pietra d'Istria stone, and their veins have maintained the main characteristics unchanged.
This means that through the studying of the properties of limestone produced by these quarries in our days it could be possible to gain some new insight also into the material widely used in the Venetian architecture during past centuries.
Starting from these assumptions, the purpose of this work was to characterise the material extracted from the above mentioned limestone quarries so to obtain a complete understanding of its properties and therefore to achieve a better comprehension of the issue concerning the anamnesis, diagnostic, monitoring, conservation and restoration of numerous edifices in Venice.
In order to organise the working program, the collaboration with the Kamen Pazin quarries has been essential for the supply of Pietra d'Istria samples.
After an initial evaluation of material main physical characteristics, the dependence of ultrasonic velocity on the compression state has been verified using different waves frequencies (55 and 120 kHz).
Velocity and the correspondent oscillogram have been recorded every 10 MPa during compression tests till fracture.
Variations in ultrasonic waves velocity and shape have been analysed
and studied so to identify any possible correlation with the load borne
by the samples.
1. INTRODUCTION"Pietra d'Istria" is the material mostly used in the Venetian architecture for its great applicability to a wide range of different functional solutions; this lithotype is used both as structural element, thanks to its relevant mechanical strength, and with decorative function.
This sedimentary rock behaviour has always been studied even if most of the surveys performed was limited to an evaluation of the state of degradation. In fact, most of the time, these studies have been conducted pushed by the precise need to perform a structural diagnosis preliminary to a project of conservative restoration. This implies the adoption of non destructive techniques, to be used "in situ", in order not to damage monuments which artistic importance is often out of discussion. Moreover, the number of samples which can be extracted is usually limited by the necessity to preserve structure integrity.
The approach adopted for this work has been completely different: a serious historic research conducted to determine the original quarries used by Venetian during past centuries, has identified in the Orsera pit, one of the most exploited sites. Thanks to the collaboration with Kamen Pazin, company owner of that quarry, it has been possible to have a good number of samples without which, this research would have been impossible.
2. EXPERIMENTAL METHODSThe development of efficient techniques of ultrasonic surveying for a specific class of materials, presupposes the acquaintance of its main physical characteristics. For this reason, the first part of this work has been dedicated to the characterisation of samples.
2.1 Material characterisation
Their weight, dimensions, density, ultrasonic velocity (55 and 120 kHz) and compressive strength are given in Table1 and Table 2.
In particular, the ultrasonic speed results, recorded in the following tables, have been calculated, for each sample, as mean values of 6 different measurements performed on both transversal sides of the specimens. Ultrasonic wave measurements have been made using a CCT 6 tester connected with HAMEG HM 205-2 oscilloscope and HM 8148 graphic printer (Fig. 1).
2.2 Ultrasonic characterisation
The commonplace practice of processing ultrasonic signals in terms
of wave frequency and amplitude attenuation, as applied in petroleum geophysics,
is rarely exploited for the purposes of civil engineering.
The ultrasonic velocity has been calculated starting from the measures of crossing times and parallelepipeds lengths so to avoid the potential error due to samples enlargement during compression. Lengths have been determined with a vernier on unload samples, and electrical transducers have been applied to estimate their enlargements while subjected to increasing pressures (see Fig. 2).
As during the static conditions tests the transducers working at 120
kHz had demonstrated more sensibility, this frequency has been also chosen
to perform the dynamic investigation. Measurements of the travel time of
the 120kHz source pulse have been taken along the axis of the core sample.
Vaseline has been used as a coupling medium to improve the acoustic contact
between the sample and the transducers. The instrument was calibrated by
contact of the transducers.
During the compression tests, good indications have come from the oscillographs printed every 10 MPa. In fact, in most of the cases, the state of cracking and also the imminent fracture were anticipated by the shape of the wave (Fig. 3). Sample IVA has been the only exception: in this case, the specimen has reacted as a unique block without showing any evident forewarning signal (Fig. 4).
RESULTS AND CONCLUSIONS
The experimental results presented in this paper evidence the discontinuous nature of this kind of rock: different values have been obtain not only from one sample to the other but also in the same sample.
However, since from the static conditions tests, it has been possible to point out a dependence of ultrasonic velocity from ultimate compressive strength. In fact, an increase in Rc, is characterised by higher values of ultrasonic speed. Unfortunately, at least for the frequency used in this laboratory work, these differences are not so significant to allow a precise correlation between the compression strength and ultrasonic signal velocity in Pietra d'Istria. Even if Graph. 1 demonstrates similar angular coefficient for trend curve derived from measurements done using the same frequency, it is not possible to achieve a sort of calibration curve applicable during an "in situ" investigation.
Moreover, it can be observed that different frequencies of ultrasonic signal have determined different ultrasonic velocities on the same sample. From the results obtained it has also been possible to appreciate the greater sensibility that can be achieved using 120 kHz probes. In fact, considering that the defect dimension is determinant for the disturb that can be created on the wave, the higher frequency should detect smaller defects. This consideration is even more evident considering the type C sample which has demonstrated an ultimate compressive strength of 98 MPa. 2 different ultrasonic velocities have been measured with 55 kHz (6108 m/s) and 120 kHz (5736 m/s): the value determined with the higher frequency is considerably lower than the other due to a defect located in one third of the sample which has conditioned the final value. In the case of 120 kHz, the wave have been so much disturbed by that defect that 1 value on 6 (those used to calculate the mean value) has been sufficient to produce a result so far from the others. With the frequency of 55 kHz, the signal has not been scattered so strongly in the correspondence of that defect to point out such a difference.
For these reasons, it has been decided to monitor in real time the ultrasonic velocity in the samples during compression tests with the 120 kHz probes.
During the dynamic tests, type A and type C samples have demonstrated a different behaviour: in the case A, before the final collapse, cracks develop starting from the veining and are emphasised by velocity decreasing, while in the case C, samples react as a bulk, showing, after destruction, a shape quite similar to the typical double pyramids.
On the whole, the samples cut perpendicular to the veining (type C) have demonstrated similar behaviours, showing the first damage effects around the 90% of the ultimate load achieved. Type A samples have not had an unique answer to loading. In particular, sample IVA has shown a complete different feature (Graph. 3 and Fig. 4): while loading it with growing pressures, velocity increased without evidencing any damage effect till final collapse. This behaviour is probably due to a different composition of the veins present in that sample which did not work as defects but simply as a continuous with the bulk. This hypothesis is still under investigation with SEM, X-ray and DTA and TG surveys.
Also in the case of the dynamic conditions tests, it is not possible to identify a general correlation applicable to a non-destructive investigation on Pietra d'Istria. Moreover, in the case of an "in situ" application it should be necessary to reproduce not only the loading conditions, but also other variables which, for the Venetian case, should be the phenomenon of "acqua alta", capillary rise, humidity, etc.
Conversely, elastic wave propagation measurements can be successfully conducted for monitoring changes in limestone properties and for detection of discrete events (for example the initiation of dilation and the onset of macroscopic fracturing).
In any case, non-destructive ultrasonic technique has shown the potentiality in the diagnosis of the state of conservation of materials such as Pietra d'Istria structural elements. Further information would be achieved by the elaboration of a digital signal coming from the oscilloscope through Fourier transform analysis and from the study of the attenuation.
The laboratory results presented here have provided a more complete picture of the properties necessary for the interpretation of the data collected during a monitoring.
The authors express their thanks to Professor S. Meriani, Chairman
of the Department of Materials Engineering of the University of Trieste,
for valuable discussions during the course of this work.