Abstract
Background
A round robin test program to evaluate whether the ASTM-procedure are able to produce consistent results for nominally identical panels tested at different laboratories with different
measuring systems has been carried out. The tests are performed at four different laboratories.
The four laboratories are:
Catholic University of Leuven (KUL)
Belgian Building Research Institute (BBRI)
Norwegian University of Science and Technology (NTNU) together with SINTEF Building and Infrastructure
Norwegian Public Roads Administration (NPRA)
In total 96 round concrete panels are tested according to ASTM C 1550. The test program consists of 6 different concretes, involving steel fibre reinforced concrete, synthetic fibre
reinforced concrete and concrete without fibres. Details of the fibres and fibre dosages are as follows:
Blanco concrete without fibres
Synthetic fibre reinforced concrete with fibre content equal to 4,5 kg/m3
Synthetic fibre reinforced concrete with fibre content equal to 9 kg/m3
Metallic fibre reinforced concrete with fibre content equal to 20 kg/m3
Metallic fibre reinforced concrete with fibre content equal to 40 kg/m3
Metallic fibre reinforced concrete with fibre content equal to 60 kg/m3
From every series, four panels were tested at each laboratory. To avoid variation in results due to differences in calculation method, the results from every laboratory are calculated by
the same person. All calculations are done at NTNU/SINTEF.
Calculation
In general, the calculations are performed according to the description in ASTM C 1550, which means to calculate the absorbed energy up to 40mm of deflection. In addition, the
absorbed energy up to a deflection of 5mm, 10mm and 20mm is calculated. At BBRI and KUL, there were mounted LVDT’s to measure not only the deflection but also the crack
opening. Another subject that is investigated is the maximum load, and the corresponding deflection.
In general, there will always be differences in the results when tests are performed at different laboratories. And further, it will always be differences in the results when test are
performed several times at one lab as well. To investigate whether these differences are real, or just an expected variation, statistical calculation may be performed. In this test program, a statistical calculation based on a two sample Welch t-test is performed to compare the results from one laboratory with the results from another laboratory. This means that all laboratories are compared with each other. In the statistical calculation, the /2 is set to be 0.025, which means that the results from the statistical calculation are at the 95% significance level.
Main findings
Even though the energy absorption seems to be dependent on where the panels are tested, the statistical calculation shows that all panels with identical fibre content may be considered to be within one series. This means that the ASTM-procedure in fact is able to produce consistent results for nominally identical panels tested at different laboratories with different measuring systems.
Regarding maximum load and corresponding deflection, it seems like both are dependent on the fibre type and the fibre dosage. Furthermore, it seems like the different laboratories do
not necessarily gives equal results. This subject is not properly discussed because it is not within the scope of this report.