Cements for floor-levelling screeds

Introduction

screed(1) material is a mixture of cement, aggregates up to 5mm, water and, in some cases, admixtures. This is laid in-situ onto a base, bonded or unbonded, or onto an intermediate layer or insulation material, for one or more of the following purposes: To achieve a defined level

To carry the final flooring BS 8204:1999: Part(1), which includes cement-sand levelling screeds, restricts the choice of cement to:

Portland cement to BS (2) Sulfate-resisting Portland cement to BS 4027(3)

Portland blastfurnace cements conforming to BS 146 (4)

Combinations produced in the concrete mixer from Portland cement to BS 12 and ggbs conforming to BS 6699 (5). The proportion and properties, apart from manufacture, conform to BS 146.

The commercially available cements and combinations not currently approved by BSI for use in floor screeds include, for example, Portland limestone cement, Portland pfa cement and enhanced/air-entrained Portland cement. As this lack of approval can be regarded as an obstacle to improved competition within the concrete materials sector, the situation is unsatisfactory. It may operate as an impediment to the wider use of EN197-1: 2000 common cements, particularly those incorporating pfa. (This is a new European Standard for common cements which has been implemented in the UK.)

The objective of the work described in this paper was to establish, by full-scale assessment of floor screeds, whether all commercially available EN197-1: 2000 cements are suitable for such applications(6). The assessment was undertaken by a skilled floor-screed tradesman through performance testing carried out by the BCA and its member companies (see acknowledgements on facing page). The test programme was developed by BCA member companies with guidance from BSI committee B/544/6 Screeds and insitu floorings.

Materials

The sharp sand was chosen to be at the finer end of the grading limits in BS 8204: Part l(1) (75% was graded below 1.18mm). Seven cements were selected for trial, and the standard 28-day mortar compressive strengths(7) of the cements, determined according to BS EN196: Part 1, are given in Table 1.

Practical trials

Seven floor screeds (Panels 1-7) were laid by a skilled floor screed tradesman in an unheated storage building at the Blue Circle Northfleet Works in 4m^sup 2^ wooden forms, together with one 0.75m^sup 2^ floor screed (Panel 8). Three control screeds were used, two using PC and one Portland blastfurnace cement. Screed materials were gauged to give a cement/sand mass ratio of 1:4.5. Enough water was added to produce a semi-dry mix. Screeds were machine-mixed and pumped. The screeds were laid 55mm thick as single-layer unbonded levelling screeds on a polythene sheet, levelled with a screed rail and floatfinished. From each batch of screed material, between two and six 100mm cubes were cast, handtamped and left alongside the screed panels, with the top face exposed or placed inside a polythene bag.

Each screed was left uncovered for 24 hours, after which the wooden form was removed. Half of each panel was then covered with polythene, which was removed after seven days, allowing the panels to dry out naturally. Cubes were demoulded at 24 hours, weighed and wet-cured in polythene bags or in air to 28 days, when the compressive strength was measured.

At 1, 2, 8, 14, and 28 days, the insitu crushing resistance of panels 1 to 7 was determined using the BRE screed tester(1). In this test, a 4kg weight is dropped from a standard height of 1000mm onto a case-hardened anvil which has its 25.2mm diameter cylindrical face in contact with the screed. The crushing resistance is taken as the induced indentation depth, in millimetres. Panel thicknesses at each corner and flatness, as indicated by the level at the midpoint of the panels, were measured one day after formwork removal. Flatness was measured again after six months.

Observations

The following observations were made by the tradesman:

All screed mixes were a little too dry

He recognised that screed panel 8 was the same cement as panel 1 Screed panels 1, 2 and 8 did not stick to the float

Panels 1, 2, 6, 7 and 8 were similar as regards ease of finishing

Panels 4 and, particularly, 3 were the most difficult to finish

The ranking of the screed panels for ease of finishing were: 5 (the best), 1, 2, 6, 7, 8, 4 and 3 (the most difficult).

The observation team considered panel 7 had the finest finishing texture.

Results and discussion

Screed

From eight days onwards, the in-situ crushing strength of all the panel sections, both cured and uncured, was lower than 2.3mm, indicating a Category A floor screed(l). No clear dependence of in-situ crushing strength upon cement type was apparent. Panels 1 and 3 resulted in mean 28-day crushing strengths in the range 0.4-0.8mm, and in other panels of 0.4-1.1mm. The lowest individual cured panel in-situ crushing strength was 1.2mm. and, of the uncured panels, 2.6mm. The change in panel flatness from 1 day to 6 months was in the range 0.2mm-0.9mm, with the Portland, Portland limestone and Portland pfa interground cements giving the highest values.

Cubes

The lowest 28-day compressive strength, as expected, was obtained using the air-entrained screed mix, namely 10.5MPa, while the others gave compressive strengths ranging from 16.5MPa (Panel 6) to 22.6MPa (Panel 5).

Conclusions

The following conclusions are applicable to the particular materials and floor-levelling screed mixes previously discussed:

Ease of placing and finishing floor screeds made with Portland pfa cement, enhanced/air-entrained Portland cement and Portland limestone cement are similar or superior to those for screeds made from cements recommended in BS 8204: 1999: Part 1(1.

The water content giving suitable placing and finishing characteristics was lowest with the Portland pfa cements and highest with the plain Portland cement.

In-situ crushing strength of screeds depends on age but not on cement type or cement strength class (32.5 or 42.5MPa).

From two days onwards, the insitu crushing strength of all the screed panels was less than 3mm, indicating a Category A floorlevelling screed.

Recommendation

These findings indicate that the list of permitted cements for floor-levelling screeds in BS 8204: 1999: Part 1 should be extended to include BS 6588 Portland pfa cements, BS 7583 Portland limestone cements and airentrained BS 12 Portland cements.

Acknowledgements

The author acknowledges the help and assistance of the observation team members, listed below, in the development, implementation and site management of the screed programme. In-situ testing of the screed panels and laboratory testing of cements/screeds was carried out by a Blue Circle Technical Centre Team, led by J. Davies and N. Hollingum on behalf of the BCA. The observation team comprised D. Hobbs (British Cement Association), M. Monk (Blue Circle Cement UK), D. Cole (Construction Confederation/ Federation of Plasterers and Dry Wall Contractors, and representative of B/544/6), G. Barnbrook (Advisory Engineer, The Concrete Society and Committee Member of B/544/6), D. Timlin (Progressive Plastering), and two operatives (Progressive Plastering).