Degradation of Mechanical Characteristics of Some Polymeric Mortars due to Aging

RESEARCH SIGNIFICANCE

A comprehensive program including salt fog spray, dry-wet and temperature cycles, and solar radiation, followed by tests on tensile bending strength and stiffness of epoxy and polyester mortars, significantly contributes to the data on the durability of polymeric concrete. Mass gain due to immersion in distilled water and salt water provides insight into the percolation of moisture and eventual chemical reactions that may explain changes of mechanical properties. Texture observation, along the aging process, with a petrographic microscope and scanning electron microscopy are used and, in conjunction with microprobe analytical data, proposed as a future approach for interpretation of the experimental data.

ACCELERATED AGING OF POLYMERIC MORTARS

Polymeric mortars and concrete have been known to civil engineers for a fairly long time, as the review by Santiago et al.1illustrates. Their more extended application, however, is recent and centered on precast panels, tanks, pools, overlays on pavements, and on the strengthening of the anchorage of structural cables. A question of utmost technical and economic interest that requires accurate data before a more widespread use is possible is that of lifetime behavior. In particular, interest in the degradation of the mechanical properties due to aging heightened because civil engineering structures are exposed to successive conditions of dry-wet environment or of maritime splash, as well as cyclic changes of incidence of solar radiation or temperature. The isolated and sometimes synergetic effect of those factors may cause poor performance, lower the characteristic design values, and reduce the lifetime of the application. A systematic study of the affected properties, their sensitivity to changes in the composition of the polymeric mortars, and the interpretation of experimental data is therefore required to close the associated gap of knowledge on durability of polymeric concrete.

Polymeric mortars are essentially composed of a mixture of mineral aggregates, usually of small effective diameter, and a polymeric resin with low porosity. Polymer cement mortar (PCM) replaces the cement hydrate binders of the conventional material with polymeric resins. The aggregates are graded to minimize the quantity of resin required and reducing shrinkage due to curing. Polymeric concretes hardenrapidly and have higher strength, lower permeability, and better resistance to chemical attack than portland cement concretes. Data on their performance and on the time dependence of their properties are still scarce, despite a number of technical publications published after Kardon's review.2 Several challenges to be faced by PCMs were pinpointed, namely: 1) full knowledge of properties and mechanical characteristics of material; 2) adequate communication among researchers and engineers working with building materials; and 3) possible correlation with established knowledge and practice of portland cement concrete.

The influence of porosity, size, and shape of the aggregates has been studied and the importance of those factors on the strength and strain are established. A comparative study of two types of polyester mortars, with isophtalic (U) or ortophtalic (G) resins, has been reported on which the mechanical strength, porosity, and interface-matrix aggregates are qualitatively suggested to be linked.3

Proshin and Vtorov4 showed the influence that the curing temperature and additives have on the performance and durability of epoxy mortars, as well as the importance that high chemical inertness be achieved in the end product. Plasticization of an epoxy mortar mixture is mainly due to the rheologic properties of the resins that change with temperature increases, raising the mobility of the polymeric chain and lowering the viscosity.

Ohama et al.5 examined the changes caused to the compressive strength of a polymeric concrete by several reagents. Mebarkia and Vipulanandan6 analyzed also the changes on the compressive strength due to a 1-month immersion in various chemical solutions, finding that strength decreased with the increase in pH of the solutions. Results showed fairly good behavior of polymeric concrete exposed to tap water, alkali, salts, kerosene, and rapeseed oil. Chawalwala7 studied the use of vinyl and polyester polymer concrete as a wear surface for a bridge deck, concluding that its degradation rate was linked to the performance of the interface of the aggregate with the polymer matrix, which, in turn, is affected by water sorption.

Preliminary results of the effects of exposure to water and H2SO4 and NaCl solutions on the flexural strength of polyester and epoxy concrete have also been described by Ribeiro et al.8Letsch9 addressed the important problem of the evolution of properties during the lifetime of the system. He does that in connection with creep and, based on a relation due to Findley, plots deformation versus time in a log-log scale and estimates future deformation. The generalization of similar procedures for effects of environmental aggression, for instance, is not known.

Literature exists, thus, that contains much valuable information, but a lack of knowledge is still evident on varioustopics and is reflected on the absence of design guidelines.

Materials and tests

The distribution of the grain size of the sand used in the mixing of the mortars is briefly characterized in Table 1 and their chemical composition is defined in Table 2. The sand is siliceous with grains of quasi-uniform size with an average diameter d50 = 342 m.

Mortars with either epoxy or polyester resins were mixed. In both types, the resin was used in a 20% proportion by weight. Mixing of the mortars was made in accordance with RILEM PC2 (1995) standard. The main properties of epoxy and polyester resins, reported elsewhere,10 are quoted in Table 2. The orthophthalic polyester resin was preaccelerated by methyl ethyl ketone peroxide catalyst with 2% by mass. The low-viscosity epoxy resin was based on a diglycidyl ether of bisphenol mixed with an aliphatic amine hardener on a proportion of 2/3 resin and 1/3 hardener. The mixtures of polyester and epoxy mortars were the same as used in a previous European research project.10

Batches of 18 prisms, each of 160 x 40 x 40 mm, were made for tests of bending and compressive strength conducted throughout the process of environmental aging. Specimens were cured for 1 day at room temperature and then at 80 °C for 3 hours (post-cure treatment).

Sorption, or water uptake, is an important factor of change of mechanical strength and can be due to absorption and adsorption. Absorption relates to capillary uptake by pores existing in the mortar, whose grain size distribution is also known to affect mechanical strength, and does not plasticize the matrix of composites. Adsorption relates to the process of formation of a solution, causes a temperature increase and swelling of the composite, and is a major cause of water uptake if the polymer is free of voids and air bubbles. The designation of sorption or water uptake combines both phenomena.11

The gain of mass due to moisture sorption was studied because moisture was seen as a major cause of possible mechanical degradation. For those studies, prisms were initially dried in an oven at 35 to 40 °C until their mass stabilized and afterward were kept submerged in distilled water and in salt water at 18 °C, in closed containers, making measurements of gain of mass at regular time intervals.

EXPERIMENTAL WORK

Diffusion

Curves of mass gain were obtained for both mortars. Figure 1 shows curves of mass gain for the polymeric mortars corresponding to immersion in distilled water and in salt water. The estimate of the coefficient of diffusivity D in distilled water, for both mortars and the initial period of sorption,when mass gain is proportional to time1/2, led to Dpoly = 0.936 × 10-4 mm2/second for the polyester mortar and to Depoxy = 0.643 × 10-4 mm2/second for the epoxy mortar. The submersion in salt water led to higher values, especially for the epoxy mortar, 0.892 mm2/second, but still showed a higher value of diffusivity, 0.949 mm2/second, in the case of the polyester mortar.

The saturation for salt water was reached at lower values of mass gain, probably due to a smaller nucleation and growth of pores subsequently occupied by water. Sorption of salt water shows a plateau approximately between 350 and 1000 hours of immersion and a prolonged second plateau later on, while no such plateau is evident for a sample immersed in distilled water in the first 10,000 hours. The coefficient of diffusion or diffusivity of the epoxy matrix increased from distilled water to salt water, for the epoxy used by the authors, by a factor of 1.96, whereas the increase in the epoxy mortar was 1.48.

A linear analytical model of sorption of water by standard 160 x 40 x 40 mm prisms immersed in water led to qualitative conclusions, but requires further improvement.

The concentration at points on the central line of the prisms is compared at 1000 and 2000 hours for the two mortars. The results obtained from the mathematical model, for the batch selected, indicated that the polyester mortar almost attained the maximum concentration after 1000 hours,whereas it took more than 2000 hours to reach the same condition in the epoxy mortar.

Comparing those computed maxima with the experimental data, they take longer to be reached in the prototype, that is, the linear model with constant diffusivity does not accurately represent the water sorption phenomena.

Distilled water absorption took place at a higher rate and reached higher values than those found for the salt water with the prescribed salinity. Saturation for immersion in distilled water was also reached at higher level than for the sodium chloride solution.

Desorption was also studied based on the mass gain of cylindrical specimens initially saturated with distilled or salt water and then dried at 22 °C. Figure 2(a) and (b) summarize the results that show a more rapid initial loss that becomes very slow around 1000 hours and that is slightly more pronounced for the polyester mortar than for the epoxy mortar. Comparison of desorption for the same type of mortar showed a more rapid loss of mass for the specimens saturated with salt water.

Humidity cycles at fixed temperature

For three-point bending tests, aging of the standard prisms by humidity cycles took place in a commercial chamber. Coupons were subjected to a continuous sequence of cycles, each of 12 hours under a 20% relative humidity (RH), followed by 12 hours at 90% RH, keeping the temperature at 40 °C. Records were kept up to 10,000 hours of conditioning of the coupons.

The mechanical bending and compression tests were made at fixed times (1000, 3000, 5000, and 10,000 hours) soon after retrieval of groups of three or four coupons from the chamber. The force and midpoint displacement were recorded until failure, in the three-point bending tests. The values of the bending stress at failure as well as the midpoint maximum deflection at the onset of the brittle rupture are shown in Table 3.

The values of the midpoint bending deflection for the three-point bending tests after 5000 and 10,000 hours of moisture cycles are displayed in Fig. 3, for both mortars. It can be seen the higher bending rigidity of the epoxy mortar and a quasi-linear response at 5000 hours of humidity cycles for the epoxy mortar, followed by a curve that approaches a bilinear law after 10,000 hours. The polyester mortar follows a more strongly nonlinear force-displacement response.

Both mortars show degradation of their ultimate bending strengths, more so for the epoxy mortar with a decrease of 30% at 10,000 hours that almost doubles the 18% decrease shown after 5000 hours. The ultimate deflection almost doubles at 10,000 hours, showing an important increase of the ratio deflection versus load at failure, characterizing an increase of bending deformability.

In the case of the polyester mortar, there was some irregularvariation of the ultimate bending strength followed by a slightdecrease, whereas for epoxy mortar there was a steady decline of strength. The effects on bending strength are much less visible and unimportant for polyester mortar, as seen also in Table 3.

Figure 4 shows the ultimate compressive and bending strength of both mortars, from tests made at 1000, 3000, 5000 and 10,000 hours of moisture cycles. Previously mentioned initial increase-decrease of ultimate bending strength of the polyester mortar can be seen, whereas the epoxy mortar curve shows continuous degradation, although very small in the initial segment of the first period. The causes of the fluctuations are not obvious; they may have to do with initial filling of voids prior to chemical reactions, and more studies are planned to establish their connection with physical and chemical changes.

Tables 4 and 5 show the modifications recorded on the compressive failure strength and ultimate strain of the epoxy and polyester mortars, respectively, under the same cycles of moisture. Compressive tests were performed using the three-point bending coupons after their bending rupture, sufficiently away from the sections of failure. The average strength values for epoxy mortar are fairly constant except for a drop of nearly 28% on the compressive strength at 10,000 hours (Table 4). The changes on polyester strength for moisture cycles were much more reduced, as shown in Table 5. These differences on the changes of compressive strength of both mortars can be easily compared in Fig. 4(a).

The comparison of the values obtained for the artificially aged specimens has been reported referring to values found at zero hours. Part of the changes, however, can be due to the normal process of aging superimposed on the artificially accelerated process. To analyze that aspect, tests of specimens naturally aged and of the same batch as those tested after submission to the humidity cycles were made at the end of each predetermined time interval. The results are also shown in Tables 4 and 5. Degradation of measured properties was evaluated by comparing results with those obtained at zero hours of artificial aging. Tables 4 and 5 show that the conclusions drawn remain essentially the same if the reference coupons are tested at the same age as that of the artificially aged, thus validating the procedure of selecting coupons at zero hours to generate reference data.

Temperature effects at fixed relative humidity

Temperature effects, namely low-temperature effects on compressive and bending ultimate strength, were measured on coupons from another batch of epoxy mortar of similar composition. The average bending ultimate stress obtained from standard three-point tests, at a room temperature of 18 °C, was 29.6 MPa, with a corresponding midpoint deflection of 0.83 mm. These values changed, respectively, to (34.3; 0.88) at -15 °C and to (37.1; 0.97) at -30 °C. The compressive ultimate stress and strain at room temperature were 80.6 MPafor a strain 0.016 that changed, at -15 °C, to (88.8; 0.017) and, at -30 °C, to (95.8; 0.018). In this negative range of temperature, the flexural strength of the epoxy mortar increases as the temperature decreases, the same applying to the compressive strength. The midpoint deflection also increased, mainly as a consequence of the increasing value of the applied force required for flexural failure, but at a slower rate, revealing an increase of Young's modulus E. This increase, calculated from the experimental data, showed that E at room temperature, 4.47 GPa, increased 4% and 7%, respectively, for negative temperatures of -15 and -30 °C, to attain average values of 4.67 and 4.79 GPa.

For completeness of the presentation, a brief reference is made to the effects of aging due to cycles of temperature, already mentioned in Reference 12 for batches of similar composition. The cycles consisted of 12 hours at 20 °C, followed by 12 hours at 50 °C, with RH kept at 80%. The results are summarized in Table 6 and Fig. 5.

The time evolution of the ultimate compressive stress of polyester mortar differs from that of epoxy mortar. For the polyester mortar, the average compressive strength decreased 13.6% by 10,000 hours, after remaining almost constant at the previous times of control, whereas for the epoxy mortar, it showed alternating changes until 5000 hours and a subsequent marked decrease, over 30%, at 10,000 hours.

The bending strength of the polyester mortar was scarcely affected until 5000 hours, actually with a small recovery from 3000 to 5000 hours, and then declined approximately 22% at 10,000 hours. For the epoxy mortar, the average tensile bending strength showed a marked decline from 0 to 1000 hours, remained fairly constant afterward until 3000 hours, and increased slightly until 5000 hours and showed a strong reduction of approximately 40% at 10,000 hours.

Cycles of salt spray

Salt fog spray, with salinity imposed at 50 g of NaCl per L of water, was cyclically imposed on prisms of 160 x 40 x 40 mm. Cycles were of 8 hours of salt fog (98% humidity) followed by 16 hours of drying at 35 °C and lasted for 10,000 hours. Table 7 shows the effects on the strength, evaluated by standard three-point bending tests and compression. Bending effects are graphically shown in Fig. 6.

The bending strength of polyester mortar shows an initial gain, followed by a decrease until 10,000 hours, when the deterioration exceeds 30%. The initial gain of strength confirms data known8 and a better understanding of the causes requires analytical data yet to be obtained.

The epoxy mortar exhibited a monotonic decrease of strength that reached more than 50% at 10,000 hours. A change of composition of epoxy mortar for structural applications submitted to this type of weathering seems advisable, perhaps changing the resin and the aggregates.

The main results found after aging by salt fog cycles can be summarized as follows:

1) Decrease of bending strength, reaching 52% in epoxy and 32% in polyester, at 10,000 hours;

2) Apparently anomalous initial behavior of the polyester mortar, showing higher strength after enduring 1000 hours of salt fog spray cycles. Equivalent behavior has been detected in comparable situations by different authors, for example, Reference 10; and

3) Higher strength and lower bending deformability of the epoxy mortar specimens when compared with those of polyester mortar. The values found consistently show a more reduced effect on polyester mortar than on the epoxy specimens.

Microscopic observation of salt fog effects

Different techniques were used to observe changes on the surface of the specimens subject to salt fog, whereas analytical exams had to be postponed due to difficulties with equipment. Observations made with hand lenses and a stereo microscope showed the shapes of the pores and their spatial distribution along the saline fog cycles. Under the petrographic microscope, quartz fine grains (<0.1 mm) and very few ferromagnesian minerals are identified in the mortar matrix, which behaves as an isotropic substance. Scanning electron microscope (SEM) observations were also made-a more in-depth interpretation waiting for analytical data.

In general, the mineral grains (essentially very fine quartz), the resin, and the pores are well defined. The grains of very fine sand are angular, making a striking contrast with the pores, which are round and quite spherical in polyester mortar. Pore sizes in the polyester mortar vary from <0.1 to 0.5 mm in diameter and show slight changes in the contour lines of the borders at different cycles; they are well defined at zero state, become less pronounced at 1000 hours, are absent at 5000 hours, and return after 10,000 hours when the contour lines are particularly visible as a transition thin line, between the cavities and the matrix, as observed on Fig. 7.

In the macroscopic observation, epoxy mortar shows pores of irregular contour (not well rounded). As the salt fog cycles proceed, the contour border lines become more intense (Fig. 8(b)), practically vanish at 5000 hours (Fig. 8(c)), and become again very distinct at 10,000 hours (Fig. 8(d)).

An interesting comparison is also possible from the scanning electronic microscope, as shown in Fig. 9, where the images contrast the aspects captured at 0, 5000, and 10,000 hours, both for polyester and epoxy mortars. The different roundedness of the pores and the evolution of the boundary between matrix and pores are clearly seen.

The images show that there are changes of the matrix, in the first stages, attributed to plasticization caused by sorption of moisture. This plasticization and increase of viscosity of the resin promotes better stress distribution among the mortar components and compensates for the adverse effects of salt water sorption until a time threshold of around 2000 hours. Along this same period, the temperature in the chamber speeds up the completion of the cure of the resin, contributing also to the delay of some effects of the degradation.

Past this stage of post-curing and plasticization, the continued diffusion of salted moisture and concurrent chemical phenomena, crystallization and microcracking progress, and the mechanical strength of the mortar degrades rapidly.

This pattern is more clearly evident on polyester mortar (Table 7).

Ultraviolet radiation

A new batch of prismatic coupons of the standard dimensionsof 160 x 40 x 40 mm were subjected to filtered arc xenon rays from a bulb that created an average radiation of 550 w/m2 for a period of 3000 hours. The incidence of the rays was mostly over one lateral 160 x 40 mm face, without rotation of the coupons, simulating the usual condition of a member subjectedto solar rays solely over a part of its external surface.

The results obtained are depicted in Table 8. It can be seen that the tensile bending strength hardly changed and, in fact, changes contributed for higher strength, perhaps due to further cure of the resin and natural aging of the prisms. Midpoint deflection increased along with time of exposure, denoting some increase of the ductility of the specimens, more so for the polyester mortar. Figure 10 shows that the coupons underwent severe color change during the tests, especially those of epoxy mortar. The intense coloring caused by solar radiation adds to the prudence required when using some mortars on conservation, especially epoxy mortars.

In addition, the increase of superficial permeability due to solar radiation causes higher moisture sorption, indicating that those synergetic effects cannot be neglected.

FINAL REMARKS

The study led to results and conclusions that can be succinctly stated:

1. Comparatively, accelerated environmental actions had a more adverse effect on the strength of epoxy mortar than on the strength of polyester mortar;

2. A substantial decrease of strength of epoxy mortar was found after humidity cycles reaching 30% at 10,000 hours, and its average ultimate deflection almost doubled;

3. Salt fog spraying cycles caused the most severe effects on strength reduction, especially on the epoxy mortar;

4. Changes of physical texture, as detected in microscopic observations, do not seem explanatory of the modification of mechanical properties. Preliminary observations, however, indicate that thermal cycles affect the texture of the mortars and cause coalescence of pores, especially in epoxy mortar;

5. The initial increase of compressive strength, especially inpolyester mortar, may be linked to further curing of the resin-an effect that disappears with the continuation of the cycles;

6. Effects of solar radiation on strength were minor, while an increase of bending deformability was identified;

7. The effects of solar radiation are essentially present in the vicinity of the exposed surface and the change on the color of the surface is very intense, especially for epoxy mortar, showing red tones that may not respect architectural principles on conservation;

8. Simultaneous ultraviolet action and moisture cycles are likely to be detrimental to the strength of the mortars;

9. For the mixtures used, values of diffusivity are larger for polyester mortar than for epoxy mortar; and

10. The saturation of the prisms due to immersion is reached at lower values of mass gain in salt water than in distilled water.

Further studies are necessary to generate more data and further the interpretation of the results, including with different amplitudes and mean values for the cycles, as well as imposing accelerated aging under stress.

The influence of the composition of the aggregates and of their grain size distribution, as well as the importance of the type of resins used also require additional attention.

Correlating degradation of mechanical properties with physical and chemical changes is an important step for civil engineering and the study of analytical data coupled with microscopic observations appears as a promising path to follow.

ACKNOWLEDGMENTS

The authors thank C. Ribeiro, who prepared the coupons; former students R. Marreiros, A. S. Cruz, and M. Salgueiro, who performed part of the tests; and J. Simão for his help obtaining the SEM images.