Elsevier

Cement and Concrete Composites

Volume 52, September 2014, Pages 34-41
Cement and Concrete Composites

UV modification of tire rubber for use in cementitious composites

https://doi.org/10.1016/j.cemconcomp.2014.04.004Get rights and content

Abstract

The use of recycled rubber as a possible aggregate in concrete is known to result in a reduction of compressive and flexural strength. This paper summarises the results of initial studies on the effect of surface-treating rubber crumb (obtained from discarded tires) with ultraviolet (UV) radiation, with the aim of mitigating such losses. Investigation focussed on changing the surface energy, and therefore the bond strength, between cement and rubber. To identify the most effective UV wavelength for this purpose, a water retention test method was utilized, resulting in the selection of the UV-C wavelength range for treatment. Additionally, specimens containing rubber, treated for different time periods, were subjected to flexural testing. As expected, the addition of untreated crumb rubber resulted in a degradation of flexural strength, however exposure to UV-C generated, at best, values only 6% weaker than those of rubberless specimens, indicating the benefits of the investigated surface treatment.

Introduction

As concern for the environment becomes increasingly important, greater emphasis has been given to recycling materials. The difficulty of recycling discarded rubber tires, however, is a major issue and derives mainly from the vulcanisation process which, on one hand, improves the durability of the tire rubber while, on the other, cross-links the elastomer, making remelting and reshaping impossible.

Presently the construction industry is a significant user of discarded rubber tires, primarily for production of asphalt based composites, as well as for heat and sound insulation in buildings and as a fuel for cement kilns. As far as the authors are aware, rubber has never been used as an aggregate for commercial-based cement composites. Its incorporation in cement deteriorates the mechanical properties of the rubber–cement composite (RCC) so created, which greatly limit the purposes for which it may be used. Eldin and Senouci [1] have quantified these losses to about 30% reduction in both the tensile and compressive strength for a 25 vol.% substitution of fine aggregate with rubber particles.

In previous work, researchers examined the simple mechanical addition of rubber to concrete, considering shape, size and volume fraction of the rubber as the main variables. Khaloo et al. [2], Turatsinze and Garros [3] and others, all reported dramatic reductions in strength, upon incorporation of rubber, which could not be substantially mitigated by an optimisation of the variables they studied. Such outcomes continue to preclude waste rubber aggregates from concrete for structural or high strength applications, although the ability of such composites to consume vast quantities of waste rubber remains attractive.

In identifying the reasons for the dramatic reduction in strength of rubberised concrete, compared to standard concrete, most of the authors cited here agreed that the development of stress concentrations around the rubber particles, due to the lack of bonding between cement and these particles, was a significant factor. Poorly bonded rubber particles would not be able to support any load transfer from the cement matrix and therefore would perform no function other than a space filler. Significantly however, the presence of (in effect) voids in the matrix could result in a reduction in the mechanical properties, such as flexural strength, not only via the reduction in cross-sectional area of the load bearing matrix, but also via the introduction of stress concentrations. Although the latter factor might appear to only relate to flexural strength (and other tensile scenarios), compressive strength would also be expected to be affected by the presence of stress concentrations. The reasons for this are discussed later, but the published experimental work significantly supports this assertion, and all workers have reported a reduction in compressive strength [1], [2], [3], [4], [5].

Given the above, if the bond between rubber and matrix could be improved, a degree of load transfer to the rubber could be achieved and, in theory, the reduction in strength (both tensile and compressive) mitigated. This has been the driving force for investigation by a number of workers (as cited below) for applying surface pretreatments to rubber, and the present study had similar objectives, although via a novel approach.

An early attempt to surface treat rubber particles was by Segre and Joekes [4], whose work was later reproduced by Albano et al. [5] with modifications. For both, treatment consisted of the immersion of rubber chips in sodium hydroxide solution (NaOH) in order to improve the rubber’s hydrophilicity, so as to achieve better adhesion with the hydrated cement paste. Albano et al. modified this by using an additional silane treatment. These initial experiments showed a minor beneficial effect in raising the strength of rubberised concrete.

Later Chou [6] exposed rubber crumb to organosulfur compounds and saw, through analysis with an atomic force microscope, that interactions between rubber and cement hydration products nearly doubled, although the work did not progress to preparing and testing bulk rubberised concrete specimens.

Pelisser et al. [7] reported to have found a more effective treatment through the alkaline activation of rubber with NaOH and the use of silica fume. Unfortunately, in considering the final cost of their treatment, they dismissed this method as it would not have been viable for use in industry. It is clear that any experimental process that succeeds in improving strength must also be capable of application, both practically and economically, for it to be industrially applied. Keeping this in mind, the work presented here sought to employ ultraviolet (UV) radiation as a method of surface pre-treatment. Its effect was investigated initially via a simple water retention protocol (described below) and ultimately by a programme of bend tests, to obtain flexural strength data.

The effect of UV on polymers is well documented [8]. UV has the ability to modify the surface of polymers through a range of mechanisms. Some, like photo oxidation, generate free radicals in the presence of air (via oxygen or ozone) which can cross-link polymer chains, thereby raising stiffness and reducing toughness. Chain scission can also occur in some systems. The presence of free radicals species, polar groups, and dangling bonds, together with changes in surface free energy [9] is thought to be the reason why UV exposure is effective in also improving adhesive bonding to such surfaces. The degradation of rubber via exposure to UV radiation is well known [8] and usually manifests itself as an increase in brittleness due to further crosslinking of the elastomer structure. It was therefore hoped that UV could alter the adhesion of rubber to aqueous based systems, by changing its polar surface energy. Other surface treatments have been previously employed in order to modify adhesion to polymers in general, including electrical discharge [10] and flame-based methods, although none of these have enjoyed adoption with regards to commercial RCC materials. Assuming a beneficial outcome, UV pretreatment would provide a cost-effective method for incorporation of rubber into RCC, largely due to its clean, contactless, and easily implemented characteristics.

The authors are not aware of UV being used previously in this application and have investigated its potential over a number of years. The study described here utlized rubber in a cement matrix (i.e. as an additive) as opposed to rubber as an aggregate substitute, in concrete. This differed from the cited literature which almost universally utilizes the latter. The justification for this was simply that the focus was on interfacial bond strength (between rubber and matrix) as the weak link in the resultant composite. Given that there is no expectation that rubber would bond directly to aggregate, it seemed sensible to eliminate this non-influential variable from the study. The authors recognise that this may make direct comparisons with other studies difficult but it should not invalidate any benefits accrued. In any case, overall rubber volume fractions were chosen to be broadly similar (see Section 4).

The following sections describe the preliminary work undertaken to investigate the benefit of UV treatment on mechanical properties, and in particular the effect of UV wavelength and time of exposure. Further aspects of the treatment will be covered in forthcoming publications. Section 2 outlines the main features of the UV treatment process. Sections 3 and 4 describe the initial testing protocols used to evaluate the effect of UV irradiation. Section 5 constitutes a discussion of the results obtained and consideration of possible underlying mechanisms together with analysis of further tests carried out to characterize the nature of the cement–rubber interface.

Section snippets

Materials, apparatus and treatment

The cement used for practical experimentation was sourced from Cemex Ltd. (Cemex Rugby + Premium Cement). This is a Portland cement containing added lime, which raises the workability. Its oxide composition by mass is given in Table 1 [11].

The rubber was sourced from Crumb-Rubber UK in the form of fine, high quality, rubber crumb. This originates from used car tires and buffings from truck tires created during retreading. Tires are reduced to crumb through a process of initial shredding,

Water retention test

It is well known that physical bonding (e.g. via van der Waals interactions) is at the heart of most adhesion mechanisms [18]. Chemical bonds formed by reaction can occur in some adhesive–adherent systems, but are generally not the norm. Polar interactions can often be enhanced by surface pretreatments. The level of polar interaction can often be established by characterizing the degree of wetting that is achievable between adhesive–adherent pairs [18]. In the case of cement bonding to rubber,

Mechanical testing

Flexural testing of RCC specimens was chosen as the most appropriate means of characterising the effect of rubber incorporation and pretreatment. A compression test would have been more conventional, however the quantity of treated rubber required (given the number of specimens needed for adequate statistical rigour) would have been prohibitive. A flexural test, however, generates failure via tensile stress states, and may at first seem inappropriate for a material principally used in

Discussion

From the mechanical testing data gathered, it can be seen that the result of the addition of untreated rubber to the cement mixture causes an immediate loss in flexural strength, as was expected from the published literature. This behaviour supports the hypothesis that untreated rubber acts as a void inside the cement and therefore creates stress concentrations at the rubber–cement interface. As treatment time increases, the flexural strength clearly improves until reaching a plateau at

Conclusions

The work presented here has confirmed the previously observed detrimental effect upon strength, generated by the insertion of rubber in a cement matrix, however, using a UV based pre-treatment on the rubber crumb has been shown to reduce the majority of these strength losses. This is largely thought to be due to an improvement in interfacial bonding between the cement matrix and the rubber. By means of optimising the UV treatment, either by changes in the wavelength employed, and/or by

Acknowledgement

The authors would like to thank M. Aitken of Crumb-Rubber UK for providing them with the rubber crumb used in this study.

References (27)

  • N.N. Eldin et al.

    Rubber-tire particles as concrete aggregate

    J Mater Civil Eng

    (1993)
  • L.H. Chou

    On improvements of rubberized concrete with organosulfur compounds

    Adv Mater Res

    (2011)
  • J.F. Rabek

    Photodegradation of polymers: physical characteristics and applications

    (1996)
  • Cited by (74)

    View all citing articles on Scopus
    View full text