Matching Wall Articulation Methods t Expects Ground Movements
D. Lopes
Dip. Geology, A.R.M.I.T. M. Eng. (S.U.T.), Aus I.M.M.
R.I. Brown
Dip. C.E. M.I.E. Aust., C.P. Eng.
Summary: Since the issue of AS2870-1996 the use of vertical articulation joints in exterior walls has become an accepted building practice. This paper discusses the anecdotal evidence gained from the investigation of more than 500 damaged houses in Victoria, Riverina and Southern Queensland since 1975. This evidence indicates that the greatest amount of wall movement occurs within 2.5-4 metre of overt corners in M-H sites and closer in more highly expansive sites (depending on the soil conditions and superstructure). A detailed “Land Form” table and Site classification is presented, articulation joint construction practices are discussed and the table “matching” the soil behavior with footing and superstructure behavior is represented.
INTRODUCTION
The classification system in AS2870-1996 has simplified the footing “design” process and has standardized the footings constructed, however it has created the impression among engineers that soils are fairly uniform and that ground movement is easily predictable. This may be so in certain geological profiles, but, far from the truth, in others. This is especially so in areas which have the combinations of dry climates and highly expansive alkaline, alluvial or volcanic clays.
The presence of sodic and cracking (“self-mulching”) soils and pervious layers between impervious layers also affect the moisture distribution and suction profiles. The idealized suction triangle and various test methods described in AS2870 are useful in calculating the Characteristic Ground Movement, however there are many conditions where these methods do not represent the onsite conditions.
The soil identification method is still the most popular Classification method, however this method relies on experience and a good knowledge of many factors including climate, drainage, geology and weathering processes.
ARTICULATED WALLING
Technical Note 61, published in 1990 by the Cement and Concrete Association of Australia, states that: “This Technical Notes deals with the articulation of masonry walls, both internal and external, built on a concrete slab-on-ground footings”. It also points out that framed walls with timber or metal studs have very little stiffness compared to slabs and/or strip footings and therefore, plasterboard etc. can be built without articulation.
Australian Standard AS2870 refers to Technical Note 61 as a the method of articulating all construction for housing, including 2 storeys and covers all types of roof, walls, floor, etc.
PERFORMANCE REVIEWS
There have been remarkably very little performance reviews relating specifically to articulation. Some work was done by Kay, Mitchell and Clayton “Domestic Footings” – A Basic For Design From Statistical Date” and Mitchell Kay & Nichols “Performance of Engineered House Footings on Expansive Soils in Adelaide”. These 2 papers ere followed up by more work by Kay and Mitchell Headed “Stiffened Raft Design for Houses Using a Probabilistic Format”. Published in Civil Engineering Transactions 32 No. 3 October 1990.
The investigation covered 222 single storey house and was carried out during 1983-84 drought. Most of the footing types examined were raft slabs built between 1972-1982. This survey report states that only the size of cracks were taken into account and that the data was gained from Council record. The survey was carried out in either highly expansive or extremely expansive sites. There appears to have been no review of the opening or closing of control joints, although the previous studies have included this and a description of the onsite soils.
The types of construction surveyed included: solid masonry, articulated masonry, masonry veneer, articulated masonry veneer. In this report only articulated masonry veneer was deemed to be “flexible”. The other 3 groups were ranked as “stiff”. Perhaps the reason why articulated brick was not actually treated as flexible is as follows, and we quote from the paper… “The inspection practices required to ensure that true flexibility is achieved in solid brick houses are difficult and Council offices have reported deceptive practices on the part of a few unscrupulous contractors”.
With all of those variables in place it became obvious in this work that masonry veneer construction worked very well, even on the extremely expansive clays. The normal crack width found to be generally less than 4mm in highly expansive clays and 4-5mm on extremely expansive clays. According to Appendix C, AS2870-1996 Table C1, this is Category 2 damage only.
Even allowing for some conservatism, the findings seem to indicate that for Adelaide, which has a drier climatic than most of capital cities, a concrete raft with 800mm deep beams is required to give a 2% probability of a 5mm crack. 1,200mm deep beams are required t get the probability below 0.1%. The values recommended by Kay J.N. and Mitchell P.W. for similar structures are 800mm to 1,200mm deep. When one compares 600mm deep beams with 1,200mm deep beams, we have a factor of 8. Therefore, the proposals as put forward by Kay and Mitchell, assuming an uncracked section, have a four or eightfold improvement in deflection control. This clearly means that AS2870 is either very non-conservative or the limits used in the assessment by Kay and Mitchell are too high.
Anecdotal evidence seems to indicate that the general public will mostly accept Category 2 damage, but generally do not accept Category 3 damage and are very aggressive about Category 4 and a lower category damage is demanded from rendered houses. The reaction depends on their personality profile as much as it does on the size of the cracks and the effects on their financial situation.
CURRENT CONDITIONS
Based on anecdotal evidence accumulated by both writers investigating in excess of 300 damaged houses per annum in the past 25 years and many hundred others by members of The Foundation & Footings Society (Vic) (who meet on a regular basis and share information) brickwork cracking and wall movements clearly exceed the 5% probability of cracks developing in highly reactive sites (as accepted by AS2870) Yttrup P. (19930 in “The Performance of Footing Systems” – Report No D8700 – Timber Promotion Council, found that this possibility of damage applies equally to strip footings and raft slabs.
There are many reasons for the incidence of brickwork cracking. The main are:
Current Construction Practice (CCP)
Inadequate Site Investigation (ISI)
Inadequate Design and Detailing (IDD)
Inadequate Drainage (ID)
AS2870 is unconservative in design. (UD)
Current Construction Practice (CCP)
Technical Note 61 requires that articulation joints have a gap of at least 10mm, and full height, starting at the top of the concrete footing and finishing at the tip of the brickwork. It also states that such gaps should be places beside window frames where such are an integral part of the control joint. However, in practice, windows are manufactured to match brick modules, therefore, to create the 12mm articulation joint gap a new window size would be required. This is not in the interests of the window manufacturers and may create confusion for the builder and bricklayer. Alternatively, bricklayers would have to lay their bricks with a wider perp under windows that are going to be incorporated in articulation joints.
In addition, the trend towards rendered housing and the consequent use of cheaper bricks and blocks as the base material has created more problems. Rendering often continues over control joints and is made to adhere to window and door frames, thus creating numerous hairline cracks where the render abuts theses features. The poorer quality brick or block work under the render and poor surface preparation creates other forms of cracking that is sometimes incorrectly interpreted as foundation movement. In wet periods it is difficult t sufficiently dry the brickwork to create a good bond to the render and differential shrinkage of the different materials in the drier months will also induce render cracking.
The use of chemically modified mortar in some states (eg Victoria) instead of a no-bond type of membrane as a moisture barrier (“damp course”) have also caused problems. The chemical addictives increase the strength of the mortar and thus ensure that the brickwork is locked totally onto the concrete slab and/or strip footing, thus when any movement occurs it is immediately transferred into the brickwork by cracking. Horizontal mortar separations above the brick course immediately above the damp course are typical of this movement. In other states, (eg. New South Wales), plastic, aluminium, lead or similar membranes are used as a damp course and as bond breakers. The use of horizontal bond-breakers is imperative in Aerated Autoclaved Concrete construction or its hybrids.
During the construction of articulation joints, mortar is often permitted to drop between the abutting brickwork, thus making it impossible for the joint to close. This even occurs when the bricklayer uses the BCA method of construction, ie. the brick veneer is tied to the timber frame and there are no ties in the joint. It is also usual to construct window openings so that the brick sills and the adjoining brick walls make a tight and neat fit thus preventing any articulation action. It is common practice to use angle iron lintels to support the brickwork. These lintels nearly always go through the control joints, thus eliminating any chance of a control joint closing. On the inside of the house, plasterboard is generally cut too tight against door and window frames, thus creating the common diagonal and horizontal cracks above and beside door and window frames. Good practice requires a minimum 12mm gap between plasterboard and the window/door frames to enable movement without stress being applied to the plasterboard. This situation became a major problem from the 1970’s inwards as the door frame trimmers used were thinner and cheaper. In some cases the joints between plasterboard and door or window frames are filled with plaster and do not allow movement to occur. In “H” sites the size of gaps in and around doors and windows relative to the timber frame are, in most cases, too narrow to cope with an expected nominated movement of 30mm over 3 metre or 15mm over 1.5-1.8 metre in the cantilever mode. It is not surprising therefore that such differential movements induces plaster cracking and door and window jamming.
The incorrect placement of articulation joints also contributes to wall cracking. The evolution of articulation joints owes it origins to the use of brickwork control joints, which were in common use to avoid the damage due to brick growth. Prior to the use of well-controlled furnaces, bricks were often improperly baked and absorbed moisture in use. The best location for these joints was clearly close to corners. In recent times poorly baked bricks are rare but perhaps less rare during boom building periods. It has taken many years to convince engineers, builders and brick layers that articulation joints serve a different purpose and that they are best placed further away from concerns than control joints but at a maximum of 3 metres.
Inadequate Site Investigation
AS2870 understates the importance of a skilled site investigation and in commentary does not encourage soil testing. In some cases the soil conditions are difficult to assess and are often misunderstood by inexperienced or unqualified investigators. Also some matters raised in Section 2 of AS2870 are still being ignored in many states. The most commonly overlooked Clause is 1.3.3. which refers to “Abnormal Moisture Conditions” and we summarise:
“Recent removal of construction that has influenced soil conditions, influences of drains, channels, ponds, dams, tanks, etc. recent removal of trees, proximity of trees, excessive or irregular watering (changes in watering patterns) lack of maintenance of site drainage, failure to repair plumbing leaks and other sill affect the Classification of a site.”
Many investigators have noted that a significant number of wall cracks and other superstructure movements occur within the first 2 years of the houses being occupied. During the recent 1997-2001 drought, this incidence has increased greatly. The use of numerous sprinkler outlets along the house walls on a desiccated building site is a recipe for damage.
Inadequate Design and Detailing
The most common cause of failure resulting from incorrect design (IDD) is the overlooking of excessive soil water or tree problems. The down-sizing of building blocks and the popularity of large native trees creates “boundary problems” which are difficult to overcome, as a consequence, articulation of walls where they are near land boundaries are particularly important. These matters are still being ignored in Victoria.
Another (IDD) that is very important, is the lack of place allowed beneath timber floors on highly and moderately reactive sites to facilitate the re-levelling of floors. It is inevitable that the ground under a timber floor dries after construction. This is particularly so in temperate and cold climates where under-floor ducted heating created a year round summer condition. Heave of the perimeter footings founded in clay always occurs after a drought breaks. Sadly, the first reaction by builders and owners when this floor settlement occurs, is to start shaving the doors, taking them off and re-aligning them, re-doing window frames. When eventually re-levelling of the floor is though of as an option, re-working all the windows and doors is required thus greatly increasing the remedial costs.
Another (IDD) is the improper detailing of plasterboard and its fitting, particularly at corners and to ceilings and joists. In the typical edge heave where the outer walls go up and thus take the trusses with them, the result is the development of high stresses between the ceiling and the middle walls of the house. The solution may be to avoid the use of large spans in large open areas by using a central footing and a load bearing wall to support smaller spans. Alternatively, free up the ceiling by articulating it relative to the walls, ie. shadow-line or similar ceiling/wall treatments. None of these methods are currently invoked at construction stage, although they are being used in the repair of housing that have experienced these modes of movement.
Inadequate Roof Drainage
The current practice of building a house is to primarily get it to lock up stage and construct the roof as soon as possible. At this stage thee are no agricultural drains yet constructed and the roof gutters discharge at discrete points around the footings and create localized heave conditions. After heavy rain, it is not uncommon to see a partly built house surrounded by a moat of water. A solution would be to install the agricultural drains and a temporary system of down-pipes, which are angled well away from the house, as soon as the roof is constructed.
Reinforced Brickwork
The reinforcing of brickwork is also another methodology, which deserves a more detailed study. Penetration of reactive clays by deep footings causes as many problems as it solves by allowing water to access more deeply into the more reactive profile and thus creating a greater heaving force. Where a greater superstructure stiffness is required reinforcement of brickwork should be considered, at least where there are not too many large openings in the lower wall sections.
ARTICULATION TABLE
The writers have compiled an articulation spacing table, which considers the type of structure and footings and soil behavior. The location of the articulations is designed to provide flexibility to the walls where they are most prone to movement considering the footing/slab and wall stiffness.
CONCLUSIONS
Clearly, the methods of installing articulation joints in housing needs to be improved. No-bond horizontal articulation should be considered for any footing and is imperative wherever glued masonry walls are used. Timber floor supports built on expansive clays need to be made adjustable.
The consideration of Site Classification, superstructure stiffness and footing flexibility can provide a better Classification method for both footing types and articulation spacing.
However, the evidence would appear fairly strong that even with good articulation solid brick or block work construction, some failures are inevitable in highly reactive soils, therefore this type of construction should be discouraged in such soil conditions.
The distance of articulation joints from corners is critical and is related to ground movement and superstructure stiffness. The cantilevering effect is important. The articulation spacing considered in this paper apply to “standard” AS2870 raft and footing designs. Where “waffle” slabs are used, the designers should consider using slightly larger vertical articulation spacings.
It should be noted that a properly constructed control joint with a soft filling can usually open or close 12mm without the general public raising a question, or in fact being disturbed by them. Whereas properties with hairline plaster cracking or 2-3mm brick cracking causes concern and distress to the owners. Clearly articulation joints are an excellent construction technique at a modest cost. It is not uncommon to hear the public say …..”the problem with this house is that there are not enough articulation joints”. The next phase therefore is to make sure that such flexibility is brought in closer and that junctions between footing and brickwork are also articulated.
Consideration should be given to reducing the span of large roof trusses by supporting them on internal walls, or where that is not practical, allowance made for shadow-line construction between cornices/wall and/or ceilings.
Further research is being conducted by A.W. Page and other at the University of Newcastle. We believe that the funding of further research in increasing the flexibility of housing will be money well spent.
REFERENCES
Brown, R.I., McManus K.J., (1996) – “Rehabilitation of Damaged Houses Founded on Expansive Soils Using Moisture Recharge”, Seventh Australia New Zealand Conference on Geomechanics – Australian Geomechanics Society and the New Zealand Geotechnical Society.
Holland, J.E. (1981) – The Design, Performance and Repair of Housing Foundations” – Swinburne Institute of Technology Publication.
Mitchell, P.W. (1986) – “The Design of Raft Footings on Expansive Soil”, - Institution of Engineering Australia Transactions.
Smith, R.L. (1993) – Analytical Design Methods and Practical Construction Details” – Australian Geomechanics Society, Victoria Group and Institution of Engineers, Australia – Seminar – Extending the Code Beyond Residential Slabs & Footings.
Standards Association of Australia – AS2870.1 (1996) – “Residential Slabs and Footings: AS2870.2 (1988) “Guide to Design by Engineering Principles”, AS2870 “Residential Slabs and Footings”.
Walsh, P.F., (1974) – “The Design of Residential Slabs-on-Ground” Division of Building Research Technical Paper (Second Series) No.5 C.S.I.R.O.
Yttrup, P., (1993) – “The Performance of Footings Systems” – Report No D8700 – Timber Promotion Council.
