PSI - Issue 10

M. Papachristoforou et al. / Procedia Structural Integrity 10 (2018) 155–162 M. Papachristoforou et al. / Structural Integrity Procedia 00 (2018) 000 – 000

159

5

men) and 1 % (right specimen) superplasticizer by weight of binder can be seen. It is obvious that the mixture with higher quantity of superplasticizer, even though it fulfills criteria 1,2 and 3, does not meet criteria 4 regarding build ability, since height ratio of 1 st layer versus 5 th is 0.3, so it is characterized as not printable. On the other hand, the right specimen fulfills all 4 criteria and is characterized as printable. Mixture with limestone as aggregate and cement as binder can be seen in Fig.3. By adjusting water and superpla sticizer amount, three different workability levels were achieved, high, moderate and low, however all of them where considered as not printable. Mixture with high workability does not meet criteria 3 and 4. Mixture with moderate workability does not meet criteria 4. Mixture with low workability does not meet criteria 2, since the final layer is printed with voids.

a

b

c

Fig. 3. Limestone and cement mixture with (a) high (b) moderate and (c) low workability level.

By applying the criteria of Table 2 on all the printed mixtures, upper and lower limits of workability parameters that were tested by four different tests and characterize concrete as printable were obtained and results are presented in Table 3. Expansion measured according to flow table test was found to be more consistent than Yield stress or Vicat value, since the printability lower limit of 18 cm and upper limit of 24 cm applied to all the mixtures that were tested and printed. On the contrary, threshold values of Yield stress and Vicat for accepting concrete mixture as printable were not so clear. For example, some mixtures with Vicat value lower than 1mm were found to be printable according to criteria of Table 2, while others with 10mm were not printable. The above can also be seen in Figs. 4 and 5 where expansion is correlated with yield stress and Vicat respectively. In Fig.4. It can be seen that for flow table value of 18 cm, yield stress ranges from 200 to 6000 Pa. Even higher fluctuations are observed in Fig.5, for example for 22 cm expansion, Vicat value was measured 0.5 mm in one mixture and 30 mm in another one. Power consumption of the electric motor rotating the screw extruder was also considered as a workability parameter. Unfortunately, it could only be measured on two mixtures with river sand, one with 500 kg/m³ and the other with 830 kg/m³ cement as binder . Threshold values of power consumption that were obtained are also given in Table 3, while in Fig.6, expansion versus power is given. It can be seen that regardless the quantity of cement, for the upper limit of expansion (24 cm), the limit of power consumption is the same for the two mixtures (630 W). On the other hand, for the lower limit of expansion (18 cm), the mixture with higher quan tity of cement (830 kg/m³) has lower power consumption than the mixture with 500 kg/m³ . This can be attributed to the fact that lower quantity of cement renders to higher quantity of aggregate in a given volume of concrete and consequently, more friction induced by the aggregates in the moving parts of the screw extruder, especially in the case of low workability mixture. Another important parameter of workability in 3D printing concrete is the loss of workability with time. Expansion of mixtures with the three different aggregates used was measured 0, 15 and 30 minutes after mixing and results are

Table 3. Threshold values for accepting concrete mixture as printable. Parameter tested

Lower limit Upper limit

Expansion (cm) Yield stress (Pa) Vicat value (mm)

18

24

500 ±300

1800 ±500

2 ±1.5

20 ± 10

Power consumption of electric motor rotating screw extruder (W)

750 1 /670 2

630 1 /630 2

1: mixture with river sand and 500 kg/m³ cement 2: mixture with river sand and 830 kg/m³ cement