PSI - Issue 13

Zoi S. Metaxa et al. / Procedia Structural Integrity 13 (2018) 2011–2016 Z.S. Metaxa and S.K. Kourkoulis / Structural Integrity Procedia 00 (2018) 000 – 000

2015

5

tions are seen in Fig. 2b. All samples were tested at the age of 28 days. It is observed that, in accordance with the electrical resistivity results, the nanocomposites demonstrate higher strength compared to the reference paste. A 25% increase is observed when dispersion is performed using a superplasticizer concentration of 0.7% by weight of cement. The results for the mechanical and electrical properties demonstrate that, under the test conditions employed in the present protocol, to achieve a homogeneous GnPs distribution a superplasticizer concentration of approximately 0.7% by weight of cement should be employed. At higher concentrations the dispersing agent amount is possibly quite high. According to the literature, the dispersing agent molecules interact with each other resulting to their flocculation (Konsta-Gdoutos (2010), Rastogi et al. (2008) and Yu et al. (2007)). At lower concentrations, the amount of the dispersing agent is not sufficient so as to repel the strong Van der Waals forces between the GnPs (Yu et al. (2007)). 3.2. Effect of ultrasonication energy application To further optimize the GnPs dispersion in the Type II cementitious matrix, the effect of ultrasonic energy was investigated. According to the previous results, the GnPs (0.1% by weight of cement) were dispersed in an aqueous solution containing the mixing water and the superplasticizer at the optimum concentration of 0.7 % by weight of cement. The resulting suspensions were ultrasonicated by applying different energies (i.e., 100, 300, 400, 500, 600 kJ). Fig. 3 demonstrates both the results of the electrical resistivity and normalized flexural strength. It is observed that the nanocomposites dispersed using ultrasonic energy exhibit lower electrical resistivity and higher flexural strength compared to the ones without sonication energy (0 kJ). Specifically, the samples dispersed by applying an ultrasonic energy of 400 kJ demonstrate the lowest average electrical resistivity and at the same time the highest increase in the average flexural strength. The nanocomposites that were subjected to lower or higher ultrasonic energies demonstrate lesser improvements at both their electrical and mechanical properties. The application of ultrasonic energy at aqueous suspensions incorporating carbon nanomaterials has two main effec ts a) it disentangles the carbon nanomaterials’ helping with their dispersion and b) it breaks the individual nanomaterials reducing their aspect ratio (Yang et al. (2006)). The application of a low ultrasonic energy is not effective enough to uniformly disperse all carbon nanomaterials. On the other hand, the application of quite high ultrasonic energies results to GnPs rapture reducing their reinforcing efficiency. According to the mechanical and electrical properties results, it can be concluded that, under the test conditions employed in the research reported here, the optimal ultrasonic energy for the efficient distribution of GnPs having a lateral size of 8 μm is close to 400 kJ.

1,4

1,4

Cement paste (CEM II, w/c=0.3) + 0,1wt% GnPs 8 μm

Cement paste (CEM II, w/c=0.3) + 0,1wt% GnPs 8 μm

1,2

1,2

1,0

1,0

0,8

0,8

0,6

0,6

0,4

0,4

0,2

0,2

Normalized flexural strength (-)

Electrical resistivity (MOhm*cm)

0,0

0,0

0 100 200 300 400 500 600

0 100 200 300 400 500 600

(a)

(b)

Ultrasonic energy (kJ)

Ultrasonic energy (kJ)

Fig. 3. Effect of ultrasonic energy application of type II/GnP cementitious nanocomposites on the (a) electrical resistivity and (b) flexural strength normalized over the respective value of the nanocomposites dispersed without application of ultrasonic energy.

4. Conclusions

The dispersion of graphene nanoplatelets reinforcing a cementitious matrix made of Type II cement was experi mentally investigated. A 3 rd generation polycarboxylate ether-based superplasticizer, fully compatible with the ce mentitious matrix, was found to be suitable for dispersing the GnPs. To achieve a homogeneous GnPs distribution a