Assessment of Concrete Compressive Strength by Ultrasonic NonDestructive Test

. This paper has carried out an experimental program to establish a relatively accurate relation between the ultrasonic pulse velocity (UPV) and the concrete compressive strength. The program involved testing concrete cubes of (100) mm and prisms of (100×100×300) cast with specified test variables. The samples are tested by using ultrasonic test equipment with two methods, direct ultrasonic pulse (DUPV) and surface (indirect) ultrasonic pulse (SUPV) for each sample. The obtained results were used as input data in the statistical program (SPSS) to predict the best equation representing the relation between the compressive strength and the ultrasonic pulse velocity. In this research 383 specimens were tested, and an exponential equation is proposed for this purpose. The statistical program has been used to prove which type of UPV is more suitable, the (SUPV) test or the (DUPV) test, to represent the relation between the ultrasonic pulse velocity and the concrete compressive strength. In this paper, the effect of salt content on the connection between the ultrasonic pulse velocity and the concrete compressive strength has also been studied.


Introduction
The (UPV) method estimates the compressive strength of concrete by measuring the (UPV) from the pulse passing the time between the transmitter and receiver at certain distances in a concrete structure. The ultrasonic pulse velocity can evaluate concrete quality factors such as the compressive strength by obtaining a correlation between the (UPV) and the compressive strength. This study aims to correlate the compressive strength of the tested cubes and the results of the non-destructive tests (UPV) for the prism cast from the same concrete mix by using statistical methods.
Several studies have been made to develop the relationship between the ultrasonic pulse velocity and the compressive strength;

Experimental Results
The study covers 383 test results on 172 prisms and nearly 900 concrete cubes of 100 mm [8]. These cubes were a product from the mixtures designed for this research using ordinary Portland cement compatible with the Iraqi standard (No.5) and sulfate resisting Portland cement with different curing conditions. The mixing properties and the experimental results are shown in Table 1. For this research, different graded and maximum size coarse aggregate are prepared to satisfy the grading requirements. The coarse aggregate grading and characteristics are given in Table 2.

Discussion of Experimental Work Results
The ultrasonic pulse velocity (UPV) had been measured in two ways, the first one was (SUPV) and the second was (DUPV) in order to find the best relation which can be correlated to the compressive strength. A comparison had been made to find the best form of the (UPV) type: surface ultrasonic pulse velocity (SUPV) or direct ultrasonic pulse velocity (DUPV) to represent the relation between UPV and the compressive strength. To investigate the SUPV and DUPV, data have been drawn with the compressive strength for all the samples subjected to normal curing. Two exponential curves were found, as shown in Figures 1 and 2. The correlation factor and R 2 were found for the two curves, as shown in Table 3.     Type 1  Type 2  Type 3  Type 4  Type 5  37.5  100  100  100  100  100  20  70  100  100  100  100  14  40  70  100  100  100  10  10  40  50  100  0  5 0 0 0 0 0  Figures 1 and 2 show that the rate variations of the DUPV are less than the variations of SUPV for the same variation of the compressive strength. Thus, an increase in the DUPV happens at a lesser rate than increasing the compressive strength. The SUPV seems to be more sensitive to the increase in compressive strength. This tendency is noticed since the propagation of surface waves is restricted to a region near the boundaries of the free external surface of the material. The depth of the penetration is on the order of one wavelength thickness. The cement paste content of this layer is greater than the average paste content inside the concrete due to the so-called wall effect. Therefore, the velocity of a surface wave SUPV is influenced more by the paste properties than that of the direct waves DUPV that travel through the whole mass of the concrete. Since the concrete strength is also controlled by the strength of the hardened cement paste, SUPV may be a better indicator of the concrete strength than DUPV. Figure 3 shows that for pulse velocity less than (4.5 km/s), the (DUPV) was more significant than the (SUPV) for the same compressive strength. For pulse velocity greater than (4.5 km/s), the (DUPV) was less than the (SUPV) for the same compressive strength. This happens because, at low pulse velocity (less than 4.5 km/s), the ultrasonic wave passing throws the coarse aggregate and gives us high (DUPV) when the compressive strength low. The (SUPV) wave passing throw the cement mortar and will represent the compressive strength more accurately. Besides that, the correlation factor of SUPV was more significant than (DUPV), so using (SUPV) is better than using the (DUPV) to represent the relationship with the compressive strength.

Results of Statistical Analysis
The statistical methods can explain the test results and the prediction of concrete strength, in the case of the test that was carried out in a satisfactory way and standard tools. The statistical methods were proved to have good values. The goal of a regression method is to fit a line through points (results) so that the squared deviations of the observed points from that line are minimized. In statistical modeling, the overall objective is to develop a predictive equation relating a criterion variable to one or more predictor variables. In this research, the criterion variables include the compressive strength, the direct ultrasonic wave, and the surface (indirect) waves. The summary of descriptive statistics of all the variables is shown in Table 4, and the correlation matrix for the data are shown in Table 5. The terms shown in Table 4 and 5 are defined as follows: C= Compressive strength (MPa), D=Direct ultrasonic wave velocity (km/s), S= Surface (indirect) ultrasonic wave velocity (km/s) The correlation of the proposed equation is (0.84726), and R 2 is (0.7883). Verification of the proposed equation and previous equations depends on separated data taken from a different source. The data adopted for the study in this part was taken from ref erences [9,10], as shown in Table 6. By using the direct ultrasonic velocity illustrated in Table 6, the compressive strength estimated from some of the previous equations and also estimated from the proposed Eq. (7), which depending on the SUPV. The equivalent surface SUPV was taken from Eq. (8), built only for verification purposes from the research data. S = 1.405 D _ 1.406 (8) where: S is a surface ultrasonic velocity and D is a direct ultrasonic velocity in km/sec From Figure 4 one can notice that no equation can satisfy all the points entirely. The proposed equation could be considered the nearest one to the points taken from reference [9], and that agrees with the correlation factor appear in Table 7, which equals (0.9611).  Figure 4. Relation between compressive strength and ultrasonic pulse velocity for proposed and previous equations.

Effect of The Salts on The Proposed
Relation. The salt content in fine aggregate affects the relationship between the compressive strength and the (SUPV). Suppose the inspector suspects that there's salt in the tested concrete and to be on the safe side, the following correction of the (SUPV) reading should be used before using Eq. (7). This correction was found from separating the normal curing data for concrete with salt from the data of concrete with no salt. These two groups of data are interred in the SPSS program, and with two equations are found one for the case of high salt content, and the other represents the data for concrete with a bit of salt, as mentioned earlier = .
. ′ By equating Eq. (7) with Eq. (8), the relation between (SUPV) for concrete with high salt and concrete with a little salt is found as shown in Eq. (10).

Conclusions
 The SUPV is more sensitive than the DUPV for assessment of compressive strength where R 2 value equal to (0.7883) and correlation factor equal to (0.84726) for equation depending on surface reading while R 2 value equal to (0.6504) and correlation factor equal to (0.7389) for equation depend on direct reading.  SO3 content in fine aggregate effect decreases compression strength while the SUPV reading does not affect the same rate. For using the general proposed equation, the proposed correction of SUPV reading must be used.