*3.1. Compressive Strength*

The compressive strength of concrete was tested by using a block of concrete with a compressive strength of 28 MPa. Then, each new concrete block was cast from standard cylinder-shaped concrete formworks with a 10 cm. diameter and a 20 cm. in height. After 24 h, the concrete formworks were removed and the concrete blocks were aged in clean water for 7 days, 14 days, and 28 days respectively. The concrete blocks were subsequently tested for compressive strength. Concrete samples with four different coarse aggregates were collected: natural aggregate concrete, laboratory waste concrete (L-RCA), precast concrete waste (P-RCA), and building demolition waste (B-RCA). The natural aggregate concrete was replaced by L-RCA, P-RCA, B-RCA with the proportion of 30 percent, 60 percent, and 100 percent respectively.

From Figure 8a–c the decreasing trend in compressive strength corresponded with the increasing amount of recycled aggregate mixed in the concrete. Three types of recycled aggregate concrete with a 30 percent replacement rate showed negligible differences in terms of compressive strength compared to the other two replacement rates. As for the recycled aggregate concrete blocks with a 30 percent replacement rate, the vertical crack propagation occurred, and the cracks split apart into two pieces. This stemmed from the fact that the mortar waste could handle less compressive strength than the coarse aggregate. On the other hand, the coarse aggregate and recycled aggregate themselves did not suffer any damages or cracks. As for the 60 percent replacement rate, the vertical crack propagation occurred and reached half of the concrete's height at a 70-degree angle. The crack resulted from the shear force coming from the bond strength of composite and internal friction. As for the 100 percent replacement rate, the vertical crack propagation occurred and reached one-thirds of the concrete's height, but the crack did not cover all the cross-section concrete areas. Moreover, some of the cracks split into tiny pieces due to the dramatic differences in compressive strength of each recycled aggregate, shear force coming from the bond strength of composite and internal friction. However, their failure tended to be the lowest and could handle the least compressive strength compared to the other two.

**Figure 8.** Compressive strength of concrete at 0%, 30%, 60% and 100% replacement rate at: (**a**) 7 days, (**b**) 14 days, and (**c**) 21 days.

#### *3.2. Chloride Penetration Resistance*

From Figure 9, the result showed total charge passed at 28 days of natural coarse aggregate and various types of recycled aggregate by rapid chloride test. It obviously seen that natural aggregate concrete (NCA) was the only type that had a moderate chloride ion permeability rate (2000–4000 coulombs) in accordance with ASTM C1202. On the other hand, all the proportions of recycled aggregate concrete were considered to have a high chloride ion permeability. Concrete made from laboratory and building demolition aggregate are slightly higher than NCA based value about 25% at any replacement ratio. There is also an increase in the recycled aggregate to natural aggregate ratio of these two types of samples do not increase the chloride permeability much. It is worth noting that concrete made from scraps of prefabricated concrete slabs. Instead, the permeability value of the chloride is very high. In one way it could be said that at 30% displacement, the permeability rate is close to that of natural aggregate concrete. But increasing the percentage of renewable aggregates to 60% or 100% increased the permeability significantly, over 71% and 106% compared to the 30% recycle aggregate samples, respectively.

**Figure 9.** The total charge passed at 28 days of natural coarse aggregate and various types of recycled aggregate.

### *3.3. Chloride Penetration and Compressive Strength*

The relationship between chloride penetration and compressive strength of 28-day recycled aggregate concrete is shown in Figure 10. The decreasing penetration rate of L-RCA (Circle), P-RCA (Square), and B-RCA RCA (Triangle) corresponded with their increasing compressive strength. The effect of replacement percentage, it has a high effect on the concrete mixed with P-RCA aggregate when the ratio is increased it will increase the compressive strength and chloride permeability as well. However, for the L-RCA and B-RCA samples, it was found that the increase in the aggregate's substitution ratio had a large effect on the compressive strength and chloride permeability. The two relationships between L-RCA and P-RCA could be explained by R2 which was well fit up to 95 percent and 99 percent respectively. Based on this result, it can be confirmed that linear regression relationship between compressive strength and chloride penetration. However, as for aggregate from the precast slab waste, the chloride penetration rate tended to be very high, which was indirectly proportional to the compressive strength. On the other hand, when considering the water absorption of P-RCA aggregates, it is found that higher water absorption rates than other aggregates have a reliable effect on the absorption values. The passivity of chloride was also higher. This highlighted the sensitivity to a change of compressive strength which significantly affected the porosity of this type of recycled aggregate.

**Figure 10.** The relationship between the total charge passed and compressive strength.

#### **4. Discussion**

#### *4.1. Failure Mode*

The aggregate porosity played a vital role in how the crack occurred, which also directly affected the compressive strength of the concrete. Ref. [26] Normally, the crack path advances through aggregate, mortar, and ITZ, depending on the ability to handle the strength of each part as shown in Figure 11. In natural aggregate concrete (NCA), the crack path went through the mortar and ITZ around the aggregate. In contrast, in laboratory waste (L-RA) and building demolition waste aggregate concrete (B-CA), it was possible for some crack paths to go through the previously existing ITZ, which occurred between some of the previously existing mortar and aggregate, while other crack paths might go through the new ITZ, depending on which ITZ could bear higher strength. This result is explained by research from Li et al. [27] which found that old mortar is the weakest of Recycle Concrete Aggregate. If the parent strength of recycled aggregate is high, the ITZ might be able to handle more strength. Lastly, in precast slab waste aggregate (P-RA), due to high porosity in this type of aggregate, which resulted from the use of small-sized stones and in turn leading to a larger area of ITZ. Thus, the crack path could go through its ITZ much more easily compared to the other two.

**Figure 11.** Crack propagation in composition containing: (**a**) concrete containing NCA, (**b**) concrete containing L-RA and B-RA, and (**c**) concrete containing P-RA.

#### *4.2. Durability Mode*

Photos from an image processing technique showed the area of each phase as shown in Figure 12. In the first row are raw photos which were denoised. The photos in the second row show the phase distribution in the cross-section area of the concrete after using an image processing technique. Finally, in the last row are the percentages of each phase of NCA, B-RA and P-RA. The phase of NCA and B-RA showed a similar distribution, unlike that of P-RA. To elaborate, the phase area of P-RA aggregate was nearly twice that of NCA and B-RA. The larger phase area of P-RA resulted in the increase in ITZ. The length of ITZ was determined by drawing a line in the CAD program as shown in Figure 13. Then, the area of ITZ of concrete cross-section was identified by multiplying the length of ITZ by its thickness 40 μm on average based on the research by Zouaoui et al. [28] which proposed that the ITZ thickness ranged from 30–50 μm.

Figure 14 demonstrates the schematic of chloride penetration through the cross-section of concrete. Chloride could penetrate the mortar to certain extent, but could penetrate the ITZ area better, while chloride could barely penetrate or could not penetrate the stone area at all. According to the research by Silva et al. which investigated the relationship between the total charge passed and chloride migration coefficient based on 120 studies, the relationship between these two is linear as presented in the equation below

y = 0.0034x

$$\begin{array}{ll} \text{If} & \text{y} = \text{Chloride migration coefficient (De)}, & \times 10 \text{--} 12 \text{ m}^2/\text{s} \\\\ \text{x} = \text{Total charge passed}, & \text{Completeness} \end{array}$$

$$\begin{array}{ll} \text{Motor Metrix} \\\\ \begin{array}{ll} \text{I} & \text{Inter-Fasial Transformation Zone (ITZ)} \\\\ \text{Same} & \text{Agarge} \\\\ \text{Choloride Penetration} & \text{S} \end{array} \xrightarrow{\text{Allow Penetration}} \begin{array}{ll} \text{Mept Penetration} \\\\ \text{Zero Penetration} \\\\ \text{Zero Penetration} \end{array} \xrightarrow{\text{KroP Penetration}} \begin{array}{ll} \text{High Pentration} \\\\ \text{High Penetration} \end{array}$$

**Figure 14.** The schematic of chloride penetration through concrete.

The relationship between the ITZ area determined by the calculation above and the total charge passed was exponential. To illustrated, the area of ITZ obtained by an image processing technique played a crucial role in the amount of chloride penetration in exponential relationship which R<sup>2</sup> = 0.9994 (Figure 15). This also directly affects the concrete durability in accordance with the research by Azarsa and Gupta [29].

**Figure 15.** The relationship between the area of Inter-facial Transition Zone and the total charge passed.
