Page 77 - Riem-Vol6_nº3

422 IBRACON Structures and Materials Journal • 2013 • vol. 6 • nº 3
Monitoring of crack opening in masonry and concrete using mobile platform
the automatical method independent of the amount of measures
accomplished. The pachymeter presented a repeatability error
between 4.2% and 4.9%. In particular, these errors were higher
during the beginning of the only crack opening observed, on each
the crack dimensions are close to the pachymeter resolution.
Starting from the 9th measuring with the pachymeter there is an in-
crease on the repeatability error due to a bigger index of cracks on
Figure 11 – Measurement comparison
between the proposed method, conventional
method with graduated scale and digital
pachymeter in 22 measurements in the specimen
Figure 12 – Repeatability errors of crack
measurements in the specimen. Total were
performed 22 comparisons between the
digital pachymeter and the proposed method
Figure 13 – Measurements time during 22 stages of
the crack opening for a total of seven observed cracks
the evaluated specimen. On the other hand, the proposed method
gathered better results due to its capacity to detect and count minors
quantities of pixels for the same phase of increase of the cracks on
the specimen, presenting a 0.42% and 0.77% repeatability error.
A performance evaluation for the proposed method (Figure 13)
shows that it was necessary nine seconds to capture and process
an image with up to seven cracks on the specimen against an av-
erage of 235 seconds for the conventional method. Those dura-
tions correspond to 22 measures collected on the same specimen.
It is observed in Figure 13 that the time demanded by the conven-
tional measuring until the 8th measure is contrary to the increase
of the only crack opening observed on the specimen. It is also
observed that the time demanded by the measuring of the 9th and
22th measures is directly proportional to the increase of the index
of cracks identified on the specimen.
The comparison of the repeatability error in relation to the dimen-
sions of the cracks is illustrated in Figure 14. It is noticed that the
errors presented by the digital pachymeter outweigh the ones
found in the automatical method. The outcomes in Figure 14 indi-
cate that, as the only observed crack grows in dimension until the
8th measure, smaller are the errors registered by the conventional
method. That behavior occurred due to the higher facility to view
the crack increase, along the compression test.
In Figure 14, as the new cracks arise in initial phase, the error
becomes higher. In these cases, the error for the conventional
method increased from 4.2% to 4.9% between the 9th and 22th
measures. On the other hand, the error exhibited by the proposed
method increased slightly at the end of measures, due to the high-
er quantity of pixels that the crack occupied. However, that impreci-
sion remained below 1% during all measures.
6. Conclusions
The proposed monitoring method in this work identifies the per-
centual index of the region of the crack on masonry or concrete,