3.5. Compressive mechanical properties
To explore the effect of CB incorporation on the mechanical properties, uniaxial unconfined compression test was done on swollen samples in distilled water at 50 % strain (
Figure 7). It was obvious from
Figure 7 that the nanocomposite hydrogels showed reversible behavior (returned to original position) during compression test. The initial elastic modulus (stiffness) was calculated from the initial linear regions of the stress-strain curves (3 to 10 % strain) and the stress at 50 % strain was considered at strength. All the hydrogels were compressed in two directions (parallel and perpendicular) to examine the effect of pore morphology on the mechanical stiffness and strength, and the average values of results were summarized in
Table 2.
Figure 8a-c shows the stress-strain curves of nanocompostie hydrogels obtained from mechanical data when samples were subjected to compression both in parallel and perpendicular direction to the ice crystal growth. It was noticed that the anisotropic mechanical phenomenon became negligible increasing the CB amount in the nanocompostie hydrogels. Clearly, G/PEG/CH (CB-1) nanocomposite hydrogel showed a significant anisotropic mechanical property displaying an anisotropic ratio - ratio of modulus on parallel direction to perpendicular direction - of 3.70
(Figure 8a). Noteworthy, anisotropic mechanical response significantly reduced with 15 % CB content (G/PEG/CH (CB-2),
Figure 8b) while as, completely diminished with 20 % CB content (G/PEG/CH (CB-3),
Figure 8c). The pore morphology of the nanocomposite hydrogels imparted these direction-dependent mechanical responses. Notably, G/PEG/CH (CB-1) nanocomposite hydrogel showed around a four-fold (0.063 MPa to 0.230 MPa) increase in stiffness when compress in parallel direction to macropore channels compared to perpendicular direction (
Table 2). The two best plausible explanations for the increased stiffness when compressed in parallel direction in contrast to perpendicular compression for G/PEG/CH (CB-1) nanocomposite hydrogel might be due to: (i) the large lamellae which act as pillars reinforcing the hydrogel and (ii) the presence of more entrapped pressurized water between two compressive plates so that water could not be easily squeezed out [
29].
The direction-dependent stiffness of anisotropic hydrogel was correlated with their anisotropic pore structures. As a possible confirmation of this, we could notice that a greater extent of macroporous channels along the direction of freezing in G/PEG/CH (CB-1) nanocomposite hydrogel made it an anisotropic hydrogel. On the other hand, a dimensional reduction of the walls of macrochannels in the hydrogel with 15 % of CB corresponded to a loss of the anisotropic behaviour. In G/PEG/CH (CB-2) nanocomposite hydrogel, the difference between directional-dependent stiffness and strength was so little that we could treat it as an intermediate hydrogel in terms of mechanical anisotropy. The typical channel-like morphology disappeared in G/PEG/CH (CB-3) nanocomposite hydrogel, and the more compact pore structure probably allowed the water to squeeze out in a similar way in both directions making it a perfectly isotropic hydrogel. Figure 9 demonstrates the trend of compressive stiffness and strength variation with CB content. Figure 9 and Table 2 showed a significant improvement of stiffness and strength upon addition of CB into hydrogel system. For example, three-fold increase in both stiffness and strength was observed with 20 % CB content. However, nanocomposite hydrogel containing 10 % CB displayed similar stiffness and strength values of G/PEG/CH alone in both directions.
In addition to investigating the effect of CB content on the mechanical stability of the nanocomposite hydrogels, cyclic compression tests with ten loading-unloading cycles were also performed.
Figure 10 shows cyclic compressive stress-strain curves for up to 10 consecutive cycles at 50 % maximum strain without waiting time in parallel and perpendicular direction to the macroporous channels of the G/PEG/CH (CB-1) nanocomposite hydrogel in wet condition.
Figure 11 presents cyclic compressive stress-strain curves of G/PEG/CH (CB-2) and G/PEG/CH (CB-3) nanocomposite hydrogels as well as the corresponding stress-time curves.
As displayed by
Figure 10a, the stress-strain curves of G/PEG/CH (CB-1) nanocomposite hydrogel clearly showed two different pathways during loading and unloading cycles, resulting in hysteresis loops. With regard to cyclic compression in a direction parallel to the macroporous channels, a pronounced deviation in stress-strain loading-unloading curves was shown between 1
st and 2
nd cycle, which indicated the occurring of irreversible damage during 1
st cycle. It has been hypothesized that those smaller lamellae bridges connecting larger lamellae might be fractured during first cycle; furthermore, the characteristic plateau region was affected by these major micro fractures in the hydrogel involving a more collapsed structure and reduced buckling resistance of the lamellae for the consecutive cycles [
29,
30]. However, a slighter reduction in slope and maximal stress at 50 % strain was observed at each compressive cycle from 2
nd to 5
th consecutive loading-unloading cycles in the material, due to the occurring of other minor micro fractures. After 5
th cycle, all following hysteresis loops were closely overlapped indicating that the hydrogel achieved a mechanically stable structure. It has been noticed that all unloading curves returned to 0 % strain, indicating full shape recovery of the hydrogels even at 50 % deformation. The stress responses (
Figure 10c) of the G/PEG/CH (CB-1) nanocomposite hydrogel compressed at a constant strain level of 50 % was shown as a function of time during the 10 cycles. It was observed that the induced stress exhibited a transition phenomenon: during 1
st cycle the induced stress was 0.064 MPa but, during subsequent cycles, it was gradually decreased and became stationary, reaching a constant value of 0.056 MPa after multiple cycles. This mechanical response is the outcome of a stress softening tendency described as Mullins’ effect, which is characterized by a lower resulting stress for the same applied strain. It is often reported in filled and non-filled rubber-like materials, and such behaviour is known as “preconditioning” [
33].
Figure 10b shows the stress-strain curves resulting from the cyclic loading unloading curves compressed perpendicular direction to the macroporous channels. In this case, stress-strain curves represented very small variation between 1st and 2nd cycle curves, indicating some minor irreversible micro fractures during 1st loading. However, 2nd cycle, and all sequent hysteresis loops were overlapped suggesting that the hydrogel had a good reversible behaviour. The corresponding stress-time plot (Figure 10d) also confirmed the mechanical stability. The maximal stress slightly reduced during the 2nd cycle, however, rapidly reached a stationary value after subsequent cycle. The mechanical responses observed by G/PEG/CH (CB-2) and G/PEG/CH (CB-3) nanocomposite hydrogels during compression testing (Figure 11) were similar. During the 1st loading some sorts of micro fracture occurred, however, 2nd cycle on, and subsequent cycles nearly identical hysteresis loops were observed. They exhibited a preconditioning behaviour up to the 5th cycle, as result of stress softening effect just explained in the previous paragraph.
Cyclic stress softening can be characterized by the amplitude of normalized stress decrease and by the number of cycles needed to reach a stabilized state. This effect was also evaluated by calculating the ratio of the maximum stress of every cycle to maximum stress of the first cycle (normalized stress). As observed in
Figure 12, the normalized stress over 10 cycles preserved at least 85 % of the maximal stress reached in the first cycle for G/PEG/CH (CB-1) in parallel direction, G/PEG/CH (CB-2) and G/PEG/CH (CB-3). A lower stress softening was observed particularly for G/PEG/CH (CB-1) in perpendicular direction, showing an amplitude of normalized stress decrease of 5 %, confirming what discussed above.
In summary, during the first cycles, the hydrogels showed a Mullins effect, with the reduction of stiffness and stress at every cycle. After few cycles, the material behaviour was stabilized, and the hydrogels were able to sustain a compressive strain of 50 % with full strain recovery. All the nanocomposite hydrogels maintained their original shape and their load bearing capability up to such a high level of deformation. It might be possible that many microscopic flaws were created within the hydrogels, but no macroscopic cracks propagated; as a result, the hydrogels were not fractured at a macroscopic level.
Additionally, dissipation energy calculated for each cycle also showed sharply reduced absorbed energy after 1
st cycle, but nearly constant for subsequent cycles (
Figure 13a). In the case of G/PEG/CH (CB-1) nanocomposite hydrogel when compressed parallel to the macroporous channels, the calculated dissipation energy was double if compared to those of perpendicular direction (
Figure 13a). Furthermore, the slope of this decreasing is more marked along the parallel direction before the tested samples reached a constant value after a few cycles for both compression directions. This might be explained by a more stable structure in perpendicular direction as depicted before. Additionally, as shown in
Figure 13b, the percentage dissipated energy was the highest at 1
st cycle for all hydrogels. However, the percentage of dissipation energy was found to be similar after 5
th to subsequent 10
th cycle further confirming the achieved mechanical stability after few cycles. Excitingly, all nanocomposite hydrogels maintained their original shapes after undergoing such a high level of deformation.
Moreover, nanocomposite hydrogels were subjected to cyclic compression with increasing maximum strain ranging from 20 to 6 0% and stress-strain curves were represented in
Figure 14 (a-d). In addition, energy dissipation was calculated from the hysteresis loop with different strain levels (
Figure 15).
Figure 14 and
15 highlighted that hysteresis became apparent above 30 % strains and sharply increased with increasing strain levels indicating increased absorption of energy during higher deformation. For all cases, the hysteresis loops area became larger with increasing maximum strain. Furthermore, as presented in
Figure 15a, G/PEG/CH (CB-1) nanocomposite hydrogel showed exponentially increasing trend of dissipation energies with the increasing of maximum strain, suggesting capability of hydrogel to effectively dissipate energy at larger deformation, also indicating a gradual fracture process of the gel network while reaching higher deformation level respect to the previous cycle. The slightly higher dissipated energies for parallel compression of G/PEG/CH (CB-1) with respect to the perpendicular compression might be due to the migration of more pressurized water throughout the porous network. The exponential growth of dissipated energies at higher deformation levels became more apparent when the CB amount was increased to 15 % and 20 %, achieving values of 12 kJ/m
3 and 16 kJ/m
3 for 60 % strain, respectively, which are twice, and three times more than the energy dissipated at 10 % CB content.
Percentage dissipation energy is often used to quantify the energy dissipation ability of a tough hydrogel. Higher level of deformation involves more dissipation due to the more stress applied to the material (higher friction). Applying a greater force, a higher loading energy was conferred to the material and the dissipation energy increased for the water movement. As shown in Figure 15b, G/PEG/CH (CB-1) and G/PEG/CH (CB-2) nanocomposite hydrogels exhibited similar dissipated energy capacities, which increased slightly with higher deformations. At the same time, G/PEG/CH (CB-3) demonstrated lower dissipation ability, but still increased strain up to 60 %, suggesting a less pronounced bond rupture capacity, increased elastic properties, and more deformation tolerance. It is noteworthy that with 20 %CB the percentage dissipation energy is more constant, and it probably be able to sustain higher deformation that led to higher values of percentage dissipation energy (not necessarily linear increasing over 60% of deformation). This CB nanofiller reinforced the hydrogel by giving a greater effect on strength than modulus. This involved high compression energy during loading; as the deformation applied to the material increases, the dissipation energy increases but the compression energy increases relatively more (due to the reinforcement effect). These results clearly suggested that the nanocomposite hydrogel possessed CB concentration dependent and strain-dependent energy dissipation behaviour, and effective energy dissipation occurred at the higher strain deformation.
It is generally accepted that the stiffness and toughness of conventional materials are two opposite mechanical parameters. Interestingly, our G/PEG/CH (CB) nanocomposite hydrogel can simultaneously increase both stiffness and toughness with increasing CB concentration. This could be attributed to hybrid physical and chemical crosslinking in the nanocomposite hydrogel. The enhancement in the elastic modulus (indicator of the stiffness) of G/PEG/CH(CB) nanocomposite hydrogels is attributed to the increase in elastically effective G/PEG/CH chains and the reinforcing effect of the carbon black acting as a filler. The rupture (or peeling) of physically adsorbed G/PEG/CH chains from CB is accompanied by energy dissipation, which consequently improves the crack resistance of the hydrogels. In addition, as more G/PEG/CH chains are adsorbed on CB particles, more energy is dissipated. Therefore, the toughness of G/PEG/CH (CB) nanocomposite hydrogels continue to increase with CB concentration. Taken together, slight decrease of stress and returning and no residuary strain during the consecutive loading-unloading cycles demonstrated elasticity, excellent shape-recovery property and good mechanical stability of the nanocomposite hydrogels.