BRE Testing

40 Year Accelerated Moisture / Temperature Cycling:
Tests by Building Research Establishment (UK)

CINTEC wall ties were subjected to accelerated moisture/temperature cycling to model the insitu conditions of the ties in traditional cavity construction. The tested wall tie was a standard 8mm diameter x 1 mm CHS stainless steel section in a nominal 16mm diameter drill hole. The tie had a conventional polyester sock and the standard PRESSTEC grout was pressure injected in the usual manner. Clay facing bricks of 212mm x 100mm x 65mm size were used as the test parent material.

The test programme assumed that the insitu cavity construction would be fully saturated by rain-water at least once a year. It was established by trials that a half hour soak in a water tank, followed by a minimum of 2 days drying in an electric oven heated to 40C(±2C) to constant weight, would satisfactorily model insitu conditions.

Five pull out tests on the brick anchor specimens were undertaken seven days after construction, then at 10, 20, and 40 cycles of wetting/drying of the specimens. The tests were undertaken in a Universal Testing machine, calibrated to BS1610: 1985 Grade 2. A side load of 3.5N/mm² pressure was applied to the bed faces to simulate conditions of confinement of the brick insitu.

The full saturation value after 24 hour immersion of the brick in water was 17.5% compared to a water absorption of 15% achieved after the test soak period of 30 minutes. The nominal brick compressive strength was 43.3N/mm². The test pull out values were as follows:

A one way analysis of variance showed the affect on the pull out performance was not significant. Regression analysis (linear as well as polynomial) confirmed this lack of significance.

The general conclusions were:

1. The pull-out performance of the test/anchor clay brick combination would not be adversely affected in any significant manner in conditions of exposure to rain simulated in the test.
2. Failure of the specimens was typically by pull-out of the steel tube, for steel strength primarily governed the capacity of the anchor.
3. Pull-out performance of the anchor/brick system appeared to be directly proportional to the length of embedment.

The full report is available on request.

Fire Test for Wall Ties

by Mr D Chehal & Dr R.C. de Vekey
Technical Director, Centre for Masonry Construction, Construction Division

Unpublished paper from BMS meeting November 1993

A FIRE TEST FOR WALL TIES

D Chehal and Dr. R C de Vekey

SUMMARY

A diverse range of connectors, termed wall ties, restraint ties or cavity connectors, are used in industry to link cladding masonry to either inner leaves of load bearing masonry or to frames of timber, concrete or steel. Their function is to support the cladding and transfer loads arising from wind, impacts, seismic events etc. to the main structure of the building. Many of these connectors have fixing mechanisms or structural components that are made from heat sensitive materials such as resins, plastics and low-melting alloys. Other products use mechanical devices that might be affected by thermal expansion of the components. However, until now, no widely publicised tests have been carried out on the performance of masonry cavity connectors exposed to fire conditions. Under the terms of the EC Construction Products Directive, CEN standards are being drafted for the specification of cavity connectors and resistance to fire is one of the essential requirements for which performance tests are required. Eventually fire performance data will be necessary in order to design in accordance with the forthcoming CEN Code of Practice. The successful application of performance evaluated products will reduce the risks to the public attempting to escape from burning buildings, and to the fire fighting services dealing with the fires. Therefore, BRE has initiated such tests to assess the behaviour of cavity connectors under the effects of fire. In the future it is hoped that the test methodology described in this paper can be extended to other products such as general fixings, support angles, and hangers and straps.

© British Crown Copyright 1993 - Building Research Establishment

To be presented at the Autumn 1993 meeting of the British Masonry Society, at British Ceramic Research Ltd. Stoke on Trent, 9th - 10th November 1993

Fire Test for Wall Ties

D Chehal and Dr. R C de Vekey

INTRODUCTION

Conventional fire tests of panel walls in accordance with BS476⁽¹⁾ or, more recently, ISO⁽²⁾ and CEN⁽³⁾ standards are designed to establish whether the walls remain structurally intact and, in the case of the separating wall between two parts of a building, whether the wall prevents transmission of fire by conduction or via holes or cracks. The criteria are that the wall shall not collapse or become unable to support a service load, the mean temperature of the unexposed surface of the specimen should not increase by more than 140°C above the initial temperature and that no cracks or other openings exist through which flame or hot gases can pass which would cause flaming of a cotton wool pad. This work concentrates on the structural behaviour thus, while wall ties would be expected to affect the performance of such walls by transmitting differential movements and by heat conduction along metal tie bodies, it is not the primary aim of this work to investigate these aspects.

Wall ties used in new masonry, built in the traditional way, are normally metal plates or wires with a loop, fishtail or other gripping feature which are bedded in cement mortar as the wall is built. Metal ties bedded in inorganic mortar are inherently fire resistant and appeared to give no problems in early fire tests ^(4,5) or in practice. For a number of situations, however, a range of alternative materials and fixing methods have been introduced. In the UK the problem of corroding steel ties has led to the use of many types of remedial wall tie systems ⁽⁶⁾ and the use of non-corroding plastic ties for some domestic housing ⁽⁷⁾. In continental Europe thin-joint mortars have become popular, of which some are based on organic binders. Additionally, the difficulties of aligning course heights when using thin joint mortars in one leaf and normal mortars in the other have led to the use of ties which use remedial style fixing methods to the inner leaf. Thus there are a number of situations where fixings are made using organic polymers as either the tie body or, in the form of resin glues, expanding plugs or screw plugs. Such ties are not inherently fire resistant and could fail and shorten the life of a cavity wall in a fire, or lead to the collapse of cladding during the course of the fire resulting in danger to escaping occupants and the fire services. The only published work known are a few tests in a CIRIA report ⁽⁸⁾. The CEN standard for ties ⁽⁹⁾ is performance based and there is a requirement in the work plan for a fire performance test. There is a need, therefore, to devise an acceptable test method which can establish a time to failure for such products so they can be specified appropriately for different fire classifications.

The method under development is to install the test ties in a piece of typical cavity walling but to mechanically isolate the bricks to which the ties are fixed in the outer leaf from the rest of the wall. This arrangement makes possible the application of a notional serviceability load to each individual tie during the test. This could lead to minor changes in the heat flow but trials using thermocouples have not indicated a serious problem. The serviceability load can either be the measured room temperature characteristic ultimate resistance of the tie divided by the factors of safety ?? and ?? or can be calculated as a 50 year return wind load on the tie at the standard spacing and for a typical building type and exposure condition. For most tie types tested to date a load of 1.3kN has been used which would be adequate for most applications in UK walls and gives comparability between products.

EXPERIMENTAL METHOD

Specimens

Figure 1 shows a side elevation of the specimen in the rig and Figure 2 is a orthogonal projection of the outer leaf in the rig.

Figure 1 Side view of the test rig and the test wall

The block inner leaf wall specimens are built in steel frames, 5 courses high and 2½ units long, giving a nominal area of 1.4 m². The nominal dimensions of the blocks are 40 mm x 210 mm x 100 mm. A 1:1:6, cement, lime, sand mortar designation (iii), to BS 5628 has been adopted. All tests are carried out after at least 28 days of conditioning in a dry laboratory. Early tests were undertaken with air dry walls at 28 days but they still had a higher than equilibrium moisture content. A better solution was to either dry the walls in a kiln or to condition them for 3 months at room temperature.

Figure 2. 3-D view of the frame and outer leaf

The outer leaf brick wall is made up of Ibstock multibuff rustic solid bricks, nominal dimensions of 215 mm x 100 mm x 65 mm laid in the same mortar as the block leaf. The brick has a water absorption of 15.9 % and a compressive strength of 24.8N/mm². As the outer leaf is re-usable after each test, three gaps have been left in the finished wall, into which, the bricks containing the ties are slotted. The exact locations of these slots are shown in Figure 2. The configuration of 3 ties in the wall meets the requirements of 2.5 ties/rn² specified in BS 5628 and CP3: Chapter 5: Part 2. The spacing of 0.9m x 0.45m has also been adhered to.

Figure 3. Positions of transducers for each specimen

The cavity is a conventional 50mm air cavity but it is intended that limited trials of filled cavities may be carried out in future tests.

Apparatus

Figure 4. Loading Connection

The fire test rig was designed for use in the measurement of the performance of wall ties in a fire situation while subjected to a mechanical load which might be a result of wind suction or fire-induced thermal movement. The  designer had to pay particular attention to the arrangement of the measuring devices to track the movement of the wall leaves and the ties themselves. For this purpose, linear displacement transducers were used which were output to a data logger. As illustrated in Figure 3, a pair of transducers are used to monitor the movement of the inner leaf and the outer brick and a single fifth unit tracks the tie. The positions of the transducers are similar for each brick. The voltages from the transducers could then be converted to millimetre displacements. A graphical display of displacement of the ties and wall leaves against time can be obtained for each test. Figure 4 indicates the method of connecting the load to the brick and hence to the tie itself. The load applied to the ties has, in most cases been based on a manufacturer’s recommended characteristic load capacity for the tie divided by the factors of safety (giving 1.3kN for most products). This capacity will cope with 50 year return period wind forces for most situations in the UK at the standard density of 2.5 ties/m² using the model calculation in BS DD14O:Part2 ⁽⁷⁾. Since the combined lateral load of the three ties on the wall is not inconsiderable a further check was carried out in accordance with BS5628:Part 1. Clause 36.4 ⁽¹⁰⁾. This indicated that the wall with four-sided support and a modest applied stress from the furnace lid, should have adequate capacity for the load. The test rig is attached to a furnace at the Fire Research Station in Borehamwood, which simulates fire conditions on the interior face of the backup/ inner leaf. The rig also applies static tensile serviceability loads to the cavity connectors during the fire using weights. The temperature of the furnace is controlled, by an experienced operator, to vary with time as closely as possible in accordance with the heating conditions laid down in BS 476:Part 20: 1987. For each cavity wall considered three wall ties were inserted in the positions indicated in Figure 2.

Fire tests

There is a step-by-step procedure involved when inserting the ties and proceeding with the test. Care must be taken so as not to damage the ties before testing and every effort is made to insert the ties simulating site conditions. That is, insertion is carried out in situ. Figure 5 shows the rig before the block wall is hoisted into the furnace and before the remedial ties are inserted. The slots into which the bricks containing the ties will be placed are visible on the left and the gas firing system can be seen on the right. Figure 6 shows the bricks in their respective slots, the transducers to measure relative displacements of the inner and outer leaf and the loading wires and rollers. Any voids between the furnace and the inner leaf are filled with ceramic fibre thus reducing heat loss to the cavity during a test.

Once the ties have been inserted and the cavity depth checked, the reinforced concrete roof platform is set down to apply a uniform downward load onto the blockwall inner leaf and the transducers are checked to ensure adequate displacements can be measured. The load is then slowly applied by stacking slotted steel plates onto a large plate on the end of the loading wire and when checks are complete, the furnace, datalogger and computer are all set running simultaneously. The Orion datalogger which is used to control the rate of measurement and to transfer the measurements to a computer is visible in the foreground of Figure 5 and stands on the transducer power unit.

RESULTS

Up to October 1993, tests have been carried out on examples of several generic types of remedial wall tie including resin/resin, metal expander/metal expander, metal expander/resin, pvc expander/pvc expander and cement grout/cement grout. Several other types are still to be tested, including traditional mortared ties, screw-in ties, plastic-bodied ties and frame connectors. The results to date are listed in Table 1.

Figure 7. Temperature profile along tie after 2hrs

The following teething problems have been encountered:

1. Walls cured in the laboratory even for quite long periods still have very large amounts of moisture in them. This is released early on and holds the temperature of most of the inner leaf at 100°C for long periods. It was concluded that the walls should be either oven dried or stored for much longer periods so to bring them to a moisture content more representative of the inner leaf of a cavity wall. Even oven dried walls emit steam for around 30 minutes.
   
2. The temperature of the lower tie, No. 3, is significantly lower than the upper pair although there is no evidences from the results, that this tie is performing any better as a consequence. Figure 7 shows the temperature of the tie body for positions along the length. (From test 12 onwards the three ties were all situated near the top to even out the temperature.)
   
3. The walls have proved more fragile than expected and blocks have been pulled out in some tests. This is partly due to damage caused by the need to transport the walls between laboratories. Standard practice from wall 8 onwards has been to build near the furnace.
   
4. Some torque specifications for metal expander ties seem to be too low and result in premature slip. Some were thus overtorqued and then performed better.
   
5. The standard static load of 1.3kN is too high for the screwed helical tie as it is close to the ultimate load at room temperature. The test needs to be repeated with a reduced load.

As was expected there is a tendency for resin or polymer based systems to fail at around 30 minutes due to creep of the softened plastic. One product, however, lasted a lot longer and deserves further investigation. Surprisingly three examples failed in the relatively cooler outer leaf connection.

Most resin and cement based grout systems do depend on a degree of ‘re-entrancy' or shape of both the tie and the hole to give a reliable performance. If very smooth holes are cut in solid materials or ties which are essentially cylinders with very little ‘shape’ are used then the tie or the grout plug is able to slide out if the adhesive bond fails. In resin systems this occurs when the resin softens and in cementitious systems when the bond fails under the action of differential expansion and steam. With socked systems in perforated units or in units with naturally occurring large pores this will not be a problem. If the drilling tool cuts a very clean smooth hole it would be advisable to roughen it by use of a second tool.

It was also expected that metal-based expansion anchors might fail if differential expansion of the components relieved the gripping force. This was indeed observed in some cases and may lead to a revision of recommended installation torque. Well-torqued metal expanders lasted for up to 3 hours but earlier failure was observed in many cases.

CONCLUSIONS

1. No generic type tested so far has been completely reliable or infallible but all types would give on average the ½ hour resistance required for domestic dwellings provided they are installed correctly.
   
2. Metal expanders are capable of longer periods of resistance of 1hr+ but can fail prematurely if they are not sufficiently torqued up.
   
3. Cement grouted ties are capable of at least 2 hours resistance.
   
4. PVC expanders, as expected, give about ½ hour resistance.
   
5. Polyester resin ties generally only give ½ hour resistance in the inner leaf but one product managed more than 1 hour. The failure mode in one case suggested that resin expansion may be one factor. More research is needed on the mechanisms of failure.
   
6. Polyester resin fixings are likely to give at least one hour’s resistance when installed only in the outer leaf.
   
7. All types of tie are likely to perform better in rougher holes and best in under-reamed holes. Additionally, resin and cement - grouted ties will perform better with the maximum amount of roughness (or shaping) of the tie body.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the considerable amount of development work on the apparatus by Steven Osmond, a former member of the section, technical input and guidance by Alan Ferguson and assistance with the specimen preparation from Pat Scott.

REFERENCES

1. BS476 Part 20: 1987, Fire tests on building materials and structures. British Standards Institute, London, 1987.
2. ISO/DIS 834-1 Fire resistance tests - Elements of building construction. Part 1:
   General requirements for fire resistance testing; Part 2: Fire resistance testing of non-loadbearing elements; Part 3: Fire resistance testing of loadbearing elements
3. CEN prEN YYY1, prEN YYY2, prEN YYY3 (Renumbered equivalents of ISO drafts).
4. Davey, N. and Ashton, L.A., Fire tests on structural elements, National Building Studies Res.Pap (12), London, HMSO, 1953.
5. Malhotra, H.L., Fire resistance of brick and block walls, FRS Fire Note No.6, London, HMSO, 1966
6. BRE Digest 329, Installing ties in existing construction, revised, BRE, Feb., 1988
7. DD14O: Part 2: 1987. Draft for development. Wall ties. Recommendations for design of wall ties. British Standards Institute, London, 1987.
8. CIRTA Report 117, Replacement ties in cavity walls, CIRIA, 1988
9. CEN prEN 845-1 Soecification for ancillary components for masonry: Part 1: Ties, straps, hangers, brackets and support angles.
10. BS 5628: Part 1: 1992, Code of Practice for the use of Masonry: part 1. Structural use of unreinforced masonry.

Previous Page | The Anchor Menu | Home | Applications Menu

CINTEC, CINTEC MC Systems, Presstec and Archtec are all Registered Trade Marks, Copyright ©1996-2002