62 RCI 0714

RCI JULY 2014

SPRA MATTERS Wind loading Part Two Key factors at installation Part One in the January issue of RCI dealt with the process of deriving a set of wind load values which the attachment of the roof components must meet with a suitable factor of safety. The key points were: • Wind load calculation must be to the current European Standard. • The calculation parameters must be derived from the ‘to be built’ dimensions of the building. • Changes to perimeter design (e.g. parapet cladding and capping) can affect results. • Because wind load acts on all air impermeable layers in the system, all must be attached properly. Part Two looks in more detail at how this final point, the process of installation to ensure the resistance to wind load, is achieved in practice. Coincidentally, just after the Part One copy was sent to RCI, serious gales affected Scotland, the south coast, South Wales and Essex, the latter in an intense short sharp gale. It looks highly unlikely that any of these events exceeded design ten minute mean wind speed values, as set out in the UK wind map of the national annexe to the BSEN. However, as is normal procedure, SPRA Technical Committee called for feedback from its members on the effects of these gales. This suggests that for the most part, roofs resisted the winds, but there were a few cases of very serious damage. So are we to conclude that the 062 JULY 2014 RCIMAG.COM By Jim Hooker, SPRA technical director calculation, safety factor or methods of attachment are becoming unsuitable, or was there another explanation? This is an important question, given the predicted trend towards more severe weather – high winds, high intensity rainfall etc. – as climate change starts to bite. There is another trend which should also be recognised; the increased popularity of adhered specifications as against mechanically fixed. Does this pose an increased risk of wind-related damage? Mechanically fixed systems In a mechanically fixed (warm roof) system, unless the deck is sealed, there will be a vapour control layer and it will be loose-laid. This allows cost-effective Polyethylene sheet to be used because it can be restrained by the insulation placed above it. Decades of experience has demonstrated that the multiple puncturing of a PE vapour control layer by the fasteners does not affect air permeability and condensation control in the majority of design conditions – swimming pools and other high humidity environments excepted. But often overlooked is the fact that wind load acting on this continuous layer will be transferred to the insulation fasteners. This requires great care in the attachment of the insulation. Because of its weight and permeable composition, the number of fasteners required per mineral wool board is generally one (subject to checking with manufacturer). However the oft-quoted minimum number of fasteners for Polyisocyanurate PIR insulation (see table below) is a minimum, it is not a rule of thumb. Indeed, as many as seventeen (5.9/m²) or more may be required in very high wind load environments (see BRUFMA Guidance and manufacturers’ literature). Above: PIR Insulation – Min no. of fixings per board size Above: PIR delaminated by traffic The critical installation factors for resistance to wind load are: • Has an accurate pull-out resistance in the deck been established? (by test for all but new steel decks, for which values are known) • Are the correct fasteners being used to suit the deck? • Is the setting out of boards correct? (broken bond, unless they are tapered) • Is fastening to the correct density and pattern? • Is the screw gun drive torque correct? Excess torque will reduce, not increase, resistance to wind load. Size (mm) No. of fasteners Fasteners/m² 1200 x 600 4 5.55 1200 x 1200 4 2.77 1200 x 2400 6 2.08 In a mechanically fixed system, insulation fasteners must resist wind load acting on the boards and the vapour control layer In an adhered system, wind load on the single ply membrane will be transferred to the deck via the insulation and vapour control layer


RCI JULY 2014
To see the actual publication please follow the link above