03.03 Design

Open full view...

Categories: Buildability

Introduction
Design for buildability requires that the building and cladding designers take account of:

  • Handling and transportation of the cladding
  • Positioning of the cladding
  • Ease of installation
  • Tolerances and fit
  • Appearance of the finished wall

The designers also have to have regard for the requirements of the Construction Design and Management (CDM) regulations.
 


Overall design issues
The form of cladding used will have great implications on the sequence of construction, ease of construction and safety.

  • Stick systems assembled at site are more tolerant of inaccuracies in the building structure than unitised or panelised systems.
  • Stick systems and rainscreen overcladding are easily transported to site and positioned on the building.
  • Unitised and panelised walls require less site work and involve fewer people working at height for a shorter period of time.
  • Unitised and panelised walls require fewer site made joints and can be erected faster than stick walls.

All brackets and fixings should be designed to be adjustable to the required tolerance accomodation.

Ease of construction entails designing for accessibility to fixings, bolts and other components so that the wall can be installed without the need for fixers to use:

  • Three hands
  • Special tools
  • Mirrors to observe hidden components

If a wall is difficult to install it is likely that the installers will not achieve the best result.  From the installers' point of view the best wall is the one that goes together quickly and easily.
 
 


Cladding design for fit
The geometry and scale of a building are significant to the buildability of most forms of cladding in the following ways regarding fit and tolerances:

  • Difficulties of assembly are related to the degree of dimensional control required, that is, the number of critical dimensions:
  • Minor deviations/errors can be magnified by the dimensions or geometry of the component or assembly to create significant positional discrepancies:
  • Although components made in a factory have a higher degree of accuracy than those constructed in-situ, manufacturing tolerances are important in the context of repeated elements in large buildings:
  • Components forming designs of complex geometry have more complex interrelationships and are therefore more difficult to set-out on site and check for errors in position and shape.

The design of the cladding system and its method of fixing must take account of the degree of accuracy that can be achieved in the supporting structure.  Conversely, the dimensional control required of the supporting structure will depend to some extent on the amount of adjustment provided by the cladding fixing and jointing system.  In some instances, this may also limit the level of movement allowed in the structure (ACA, 1991).

Each type of cladding varies in terms of the manufactured and erected accuracy achievable and its freedom to adjust and allow fit.  The table below summarises the aspects of cladding that can aid or hinder satisfactory fit.
 

Aspect 
(variants)
Effect on buildability 
Joint type 
(butt/lap/open)
Inherent provision for adjustment; 
frame dimensions critical to the fit of cladding units, lap joints are more tolerant.
Joint seal type 
(sealant/gasket)
Ability of an individual joint to accommodate deviations.
Unit size 
(component/sub-assembly)
Number of joints present to accommodate deviations; variability increases with size and quantity of components fixed together.
Unit weight 
(light/heavyweight)
Ease of handling - significant to the positioning of components.
Unit material 
(metal/concrete/stone/glass)
Manufacturing tolerances, dimensional changes & flexibility; option of cutting panels to overcome unexpected discrepancies.
Assembly 
(factory/site)
Factory precision (and quality) versus site flexibility.
Fixing design 
(direction/extent)
Designed provision for adjustment.
Fixing location 
(slab edge/column)
Need to accommodate bending deflections and other movements of frame.


 


Dimensional co-ordination/critical dimensions
Problems of dimensional co-ordination afflict assemblies of prefabricated components such as windows and non-traditional forms of cladding and depend on the degree of dimensional control required or the number of critical dimensions.  This emphasises the need to identify locations where dimensions are likely to be critical to fit, to select design details which avoid potential problems and minimise the number of tolerance constraints and to specify reasonable and achievable tolerance targets when it is not possible to eliminate tolerance problems by design (Elliott, 1992).

The dimensional controls required to install windows and cladding are discussed below.
 


Manufactured component into pre-formed space
A window may be installed within an opening pre-formed either:

  • within a masonry wall or cladding panel, or
  • between four panels.

In all cases three degrees of restraint on dimensions are produced.  The critical dimensions are the height and width of the opening and the difference of the diagonals to determine square.  All of these must be controlled if the unit is to be successfully installed.  Obviously, the same condition applies to the window but the lower variability of manufacturing processes means that this is comparatively less critical.  Of course the window and opening must also be square and plane.  Fit of this assembly depends on:

  • selection of the optimum size for the window and opening;
  • design of the joint and seal to take account of the permissible deviations in the joined components.

Disparities in the size and shape of the cladding opening and unit are accommodated at the interface; this is normally designed as a butt-joint but too often is actually sealed as a fillet joint due to the close proximity of the joint faces.  A pvc-u window can vary in size by +/- 3mm (BS 7412) and a brickwork opening by +/- 20mm (BS 5606).  At the interface between the window and the opening the maximum likelihood estimate of the tolerance will be equal to the square root of the sum of the squares of each individual deviation, which in this case is:

Tolerance = (202 + 32)1/2 = 20.2mm, say 20mm.
 

  • Consequently, for two jamb joints with a target joint size of 20mm, the actual joint may vary in width from 10mm to 30mm; which would make proper sealing impossible.  The delay involved with deferring manufacture of the units until measurement of the opening is viable only in refurbishment situations.  Hence, if the variability is considered unacceptable, possible solutions are to:
  • increase accuracy in the cladding - by the use of templates or a more accurate construction method, or by improving quality control/construction skills;
  • increase provision for inaccuracy - by modifying the window surround or cladding edge detail to incorporate a more accommodating lap-joint or by increasing the number of joints.

Clearly if, say, a 6mm joint had been designed, over-size would lead to lack of fit (4mm clash per joint) and the need for remedial action: either the opening has to be cut (by 20mm) to allow the window to be installed and properly sealed or the window replaced and the glass re-cut.
 


System of manufactured components/units enclosing a pre-formed frame
It is very often the case in unitised concrete cladding that mullion and spandrel units need to be precisely positioned to the grid to form the correct size of window opening.  Dimensions that are critical to the fit of the panels are thus column spacings and floor heights.  In the same way, the opening dimensions formed by the four cladding units are critical to the fit of the window.

Cladding systems that are supported by rails/purlins (e.g. metal sheeting) or an intermediate framework of mullions and transoms (i.e. curtain walling) usually run past columns and beams and are therefore not directly related to the space between them; the only dimensions that are critical are the overall length and height of the structure, which allows inaccuracies to be accommodated over many joints rather than a few.

When actual deviations are excessive and impede satisfactory fit of the cladding, the solution may be to either modify the structure, drill some additional fixing positions or supply one-off bespoke brackets.  The latter course of action would involve the cladding contractor, structural engineer (solution alters loads on structure) and the architect (solution generates a different aesthetic effect).
 


Co-ordination with other building elements
The cladding itself can be critical to the fit of other building elements that must subsequently be installed, for example windows within an opening formed by four cladding units.  Cladding may also be required to accurately co-ordinate or ‘tie-up’ with adjoining interior work (e.g. ceiling and floor finishes), building partitions, or mechanical equipment.
 


Geometrical form/magnification of errors
Dimensional and positional inaccuracies of a component can be magnified by its dimensions or the geometry of the assembly to form larger positional errors, potentially leading to lack of fit.  To illustrate, minor errors in the squareness of the ends of a column or in the erection of the member will place the opposite end out-of-position by an amount equal to the initial error magnified by the ratio of length to width.  Similarly, misalignment of cast-in sockets for connection of the cladding are equally at risk, the effect being to greatly reduce the designed allowance for adjustment.

Manufacturing dimensions and construction/erection operations that are likely to cause magnified errors need to be identified and addressed by the arrangement and adjustment provision of members and joints and/or by tighter tolerance limits.
 


Scale/accumulation of errors
In many cases values of induced deviation follow a normal statistical distribution around the mean size or position.  Values may also display a bias reflecting systematic variability where a fixed deviation recurs in batches or groups of components or measurements due to maladjustment of measuring instruments etc.

Although components made in a factory have a higher level of accuracy than those constructed in-situ, manufacturing tolerances are of vital importance in constructions which consist of repeated elements in large buildings.  If all individual units contain the same type of deviation, then the total effect is the sum of them all.  In the case of ribbon glazing or cladding, the problem is compounded if the frame deviates to the maximum permissible conflicting limit.

More stringent tolerance requirements for plumb are essential in very tall buildings to guard against accumulation of deviations - for instance, machined faces may have to be introduced for column splices.

To prevent manufacturing deviations accumulating on a long run of windows, every sixth mullion, say, is given the capability (by a different sort of pointing joint for instance) of absorbing deviations to get back on station.

On the other hand, systematic over- or under-length in five or six windows can be taken up at a column cover.  With steel windows the contractor even incorporates his own adjustment capability with a 6mm flat or a normal 6mm thick coupling bar which allows 8mm adjustment per window to overcome manufacturing deviations.

Fortunately, in most cases it is possible to rely on the inherent improbability of unfavourable extreme deviations (under- or over-length by the permitted deviation) occurring concurrently (SCI, 1994).  Judgement should be exercised on to what extent, if any, adjustment should be provided to accommodate systematic over- or under-length.
 


Design provision for deviations
The building structure will normally dictate the type and positioning of fixings to be used.

The positioning of cast in-situ fixings or pockets for fixings in the structural frame requires close site control.  It is easier to obtain accuracy by positioning the fixings after the structure has been built, but the designer must ensure that with concrete construction, the reinforcement will be sufficiently clear of the fixing position (BRE Digest 235).

As a general principle, it is better to provide a few substantial fixings rather than rely on the correct relative positioning of many small ones (BRE Digest 235).
This can potentially damage fixings and cladding units and undermine the soundness of attachment.
 


Joint design
Sealant joints should be capable of accommodating the accumulated deviations of both frame and cladding, whilst maintaining the joint width within installation and working range of the joint material.

The full-lap joint avoids the need for close dimensional control of the components and the openings to which they are related; the partial lap joint reduces this need.

Joint design should never be based solely on a single consideration such as dimensional variability (BRE Digest 137).  The width of the joint must allow for manufacturing tolerances and for inaccuracies in the setting out of the building and in positioning components of the structural frame and cladding (induced deviations).  Allowance should also be made for the continuing deviations that arise because of the physical properties of materials (inherent deviations).

Depending on the difference between the conditions/deviations assumed when designing joints (e.g. for moisture content or temperature) and the prevailing conditions during erection, buildability will be adversely or beneficially affected.  Proper joint design allows for this fact and is described in detail in Section 01.07.
 


Bracket design
Incorrect positioning, plumbing and lining of the structural frame, and even rolling tolerances for straightness, will give commensurate out-of-plane distortions in the individual cladding panels.  These can be overcome by shims, sliding connections or adjustable bolts coupled with the inclusion of buffer zones.  A disparity between the elevation length and height of the structure and cladding affects the sizes of panel-to-panel joints.  Such discrepancies can be distributed by altering the size of individual (butt) joints or accommodated at a single point with overlapping corner/perimeter flashings.  Obviously, adjustments to individual elevations in either plane need to co-ordinate at corners.