| dc.description.abstract | The ever-evolving concrete construction industry has required stronger concrete, faster turnarounds, and better durability. To achieve this, the concrete mixtures have changed drastically, including a large amount of cementitious materials, lower amount of water relative to the amount of cementitious materials, supplementary cementitious materials, and chemical admixtures. Each of these changes alters the properties and the behaviour of both fresh and hardened concrete. The standards and codes relating to the quality control and assurance of concrete structures were based on research conducted decades ago, which utilized on concrete mixtures which contained none of these changes and would not meet the expectations of modern concrete construction. The parameters of core testing, time, location, and direction of core extraction, the condition of the core between extraction and testing, the diameter, and the length-to-diameter (l/d) ratio of the core, all have effects on the results of the compressive strength test. Due to the property changes caused by the modern concrete mixtures, the effect that these parameters had on the result of the compressive strength test may or may not be true nowadays.
At the request of the Ministry of Transportation of Ontario, an experimental project was conducted to determine the effects that various parameters have on the day 28 compressive strength of concrete samples, and which combination of these parameters would be optimal for quality assurance testing purposes. In total, 8 sets of concrete structures were created and tested, totaling 884 concrete samples, of which 713 underwent compression testing. These sets of concrete samples included beams, wall sections to represent girder webs, large box structures, manhole risers, and standard concrete cylinders which were constructed in the University of Waterloo laboratory and in pre-cast manufacturer facilities. Other tests were conducted, including the bulk resistivity, rapid chloride permeability, and air void systems. These tests are not discussed in this thesis but were discussed elsewhere [1]. The purpose of this project was to determine how different core parameters affected the compressive strength results, and how modern concrete mixtures adhere to the current practices outlined in standards and codes. These parameters are the time of coring (day 3, 7, 14, and 28), location of the core along the length and height of a structure, direction of core extraction (perpendicular or parallel to the direction of casting), the condition of the core between extraction and testing (sealed in plastic or soaked in a saturated calcium hydroxide solution), the diameter of the core (75 mm or 100 mm), and the core l/d ratio (2 or 1.5). A statistical analysis was carried out at a 95% confidence level to determine these effects.
Each sample set varied a parameter to isolate its effect on the compressive strength. Once isolated, the effect of the parameters was determined through statistical comparisons. The time of coring was found to have no significant effect on the day 28 compressive strength, regardless of how the core was conditioned between the coring and testing day. The height and length along a structure was also found to be insignificant, provided the concrete mixture included supplementary cementitious materials (SCMs) which reduce the bleed water. A similar conclusion was found for the direction of coring: provided the concrete mixture includes SCMs which reduce bleed water, there was no significant difference between the two directions of coring: perpendicular and parallel to the casting direction. However, the condition in which the core is stored between coring and testing was found to be significant. The cores which were soaked had a 2.3% lower compressive strength than cores sealed in plastic, as recommended by ASTM C42 [2]. The soaked cores were also half as variable with a coefficient of variation (CoV) of 4.73% compared to 9.31% for the sealed cores. The strength correction factor (SCF) in ACI 214 [3] of 0.917 for equating a sealed core compressive strength to a soaked core was found to be inadequate for the data presented, which found 0.977 to be adequate. Similar for the diameter of the samples: 75 mm diameter samples were found to have a 1.5% higher and 25% more variable compressive strength than 100 mm diameter samples. The SCF found in ACI 214 [3] was found to be adequate for the data presented. Lastly, the l/d ratio models presented in CSA A23.1 [4], ASTM C42/C39 [2, 5], and ACI 214 [3] were found to not adequately represent the data in this project. Instead, a modified version of the ACI 214 model was suggested; however, this model was sensitive to the input data. Another suggestion is to calculate a SCF for each unique concrete mixture and structure type by averaging the compressive strengths of samples with the standard l/d ratio of 2 to the compressive strengths of samples with a non-standard l/d ratio (i.e. F_(l⁄d)=f_(c,l⁄d=2)⁄f_(c,l⁄d≠2) ). This second method provided SCFs on par with the modified ACI 214 model, while compensating for sensitivity in the concrete mixtures and construction techniques.
Once all the effects of the parameters discussed above are combined, the variability of 100 mm diameter samples with an l/d ratio of 1.5 was found to be less than the variability of 75 mm diameter samples with an l/d ratio of 2. To ensure equivalent variability in sets of samples, additional cores should be extracted, depending on the CoV of the sample data and the non-standard parameters. For example, with a CoV of 6.1%, four saturated cores with a 100 mm diameter and an l/d ratio of 1.5 or five saturated cores with a 75 mm diameter and an l/d ratio of 2 would be required to be equivalent to three standard cores. To ensure the same probability of passing quality assurance testing, where the average compressive strength of three cores must be at least 0.85f_c^' with no single value below 0.75f_c^' [3, 4], the limits may be changed to accommodate the increased variability associated with non-standard parameters on the core. For instance, with a CoV of 6.1%, five 75 mm diameter cores with an l/d ratio of 2 having no single value below 0.66f_c^' would be equivalent to the current three 100 mm diameter cores with an l/d ratio of 2 having no single value below 0.75f_c^'.
All of findings above lead to the conclusions that concrete samples from modern concrete mixtures, which include high cementitious contents, low water to cementitious materials ratio, SCMs, and chemical admixtures, are not represented adequately by the current codes and standards. | en |
