Corrosion

The NZMRM Corrosion Project.
Stuart Hayman & Stuart Thomson

In 2012 the NZMRM Technical Committee launched a most challenging, expensive and exciting piece of research into the causes of corrosion of metal roof and wall cladding systems and how to minimise its effects. The MRM consider that they have a duty of care to prove, by testing, that the solutions published in
the MRM Code of Practice and its recommendations are correct and applicable. This article provides the background to this test project, the methodology adopted, the expected outcomes and details of what has been done to achieve these outcomes.

Metal roof and wall cladding is widely used in New Zealand and Australia for all types of buildings while the rest of the world has tended to restrict the use of metal cladding to farm, industrial and commercial buildings. In New Zealand 70-80% of housing is clad with metal roofs of long run or metal tiles, made, supplied and often installed by members of NZMRM. This is one of the reasons why the NZMRM executive agreed to invest a significant amount of money, time (4-6 years) and effort to investigate the specific causes for the deterioration that can occur when metal claddings are used in corrosive areas (predominantly by the sea) and determine how this can be avoided.

Steel itself, while lightweight and strong, has little corrosion resistance and as dipping of steel sheets in moulten zinc offers good protection from corrosion this became the normal protection of steel sheet for roof and wall cladding for most of its history (150 years). In recent decades the search has been on for ways to improve the metallic coatings which protect the base steel.

Several mixtures of zinc and other metals have been assessed for use to replace zinc alone, and in 1969 the development of 55%/45% aluminium/zinc metallic coating (80%/20% by volume) became pretty much the leader of the zinc alternatives (although there are others on the world market). This product is (perhaps confusingly) called Zincalume® in Australasia, and Galvalume in the US and, more correctly, Aluzinc in Europe. Its manufacture is controlled tightly by the inventors (Bethlehem Steel) and all licensed world manufacturers make to the same standard.

This product was taken up by BHP Steel in Australia, and subsequently in 1993 by New Zealand Steel (then a BHP) subsidiary), and called Zincalume®.

Zincalume® protects the steel by a different mechanism than plain zinc, relying on the inert aluminium oxide film rather than the sacrificial zinc loss of galvanised coating. If undamaged this aluminium/zinc coating will last almost indefinitely in non-extreme environments but unfortunately cutting the edge and
drilling holes and fixing screws breaks the metallic coating which can lead to corrosion in more adverse environments. Aluminium/zinc coated and painted steel (“COLORSTEEL®” or “Colorcote”) performs very well on its own, even in very corrosive places, provided it is rain or manually washed.

Unwashed areas, like walls or under a soffit, do not fare so well if not manually and regularly washed. Aluminium cladding has been used to replace coated steel in more severe environments, at significantly greater expense. While it does not rust, it does pit, so even aluminium has some corrosion issues.

The fixing of the sheets to the building created its own problems. Lead-headed plain nails, and hotdipped galvanised nails and screws have gone out of fashion, or been declared environmentally unfriendly, and subsequently there has been continuing research into a range of coatings to replace the older systems. Currently screws are mechanically zinc plated and then coated with various proprietary metallic paint systems.

One problem is that in severe environments the zinc/paint coated screws used now tend to show rust on the heads well before the cladding and that causes corrosion around the screws. Stainless steel screws, which don’t themselves rust, can damage the cladding more than protected steel screws and although this is not a Building Code failure it ends up as an aesthetic failure of the system.

The generic failure of zinc coated fasteners has been detrimental to the installation of profiled metal roof and wall cladding in extreme environments and a major part of this project is to assess various combinations of metal, screw coatings and isolation techniques to solve these fastening problems.

Unfortunately many of New Zealand cities contain areas that are close enough to the sea to be classified as “marine” and many people choose to build in these desirable but more severe environments.

While several surveys conducted during the early 2000s confirmed these issues, none provided any useful solutions.

Since 2009 NZMRM has been working on a programme to investigate all of these factors with the aim of providing the optimum cladding system designs and installation methods for designers, builders and roofers and to provide the best possible solutions for use of metal roofing products. While metallic coated steel and aluminium manufacturers, fastener manufacturers and paint manufacturers have all been carrying out research into their own products, there has been no coordinated attempt to look overall at the factors involved and their interactions. Meanwhile new substrates and fasteners have been developed so that while the delays in launching the project have been frustrating, the upgrade to involve new materials would not have occurred had the project started in 2009.

The programme’s outcome has thus changed from being the best method of using existing materials to that of finding the best possible methods under various conditions of all materials likely to be available in the foreseeable future.

Issues for the project to look at now included:-

• Need to have fastener/substrate system in which neither corrodes
• Understanding the factors leading to aluminium roof corrosion
• Effect of different paint systems on metallic substrates
• Effect of underlay and spacing/ventilation on underside corrosion
• Benefits of new substrates and new fastener

The expected outcomes:

• The ability to use stainless steel fasteners with metallic coated steel either by isolation or with new fastener coatings
• How to best use new metallic coating materials to improve performance.
• Acceptance of new class 5 coated steel fasteners for use in very severe environments
• Determining the necessity for the use of underlayment and scribed metal flashings to avoid corrosion at the gutter line and determine the effect of synthetic underlays
• Methods to prevent aluminium pitting corrosion
• Determining the effect of paint as an inhibitor to the natural oxidisation of aluminium or metallic coatings in extreme environments

The first decision to be made was what materials, combinations and systems to expose (which created an increasing number of possibilities).

Laboratory testing has been undertaken in the last couple of years to eliminate what clearly doesn’t work in severe environments.

It is important to note that this project is not intended to confirm what doesn’t work, as we know that already from earlier assessments, but to determine what does work best in different conditions.

The next decision was where to locate test panels and how many sites we could manage. In selecting sites there was a fine line between being too extreme and not extreme enough as all exposure testing is a balance between these two. When looking at building products with a consumer expectation of decades
of life a “normal” life span is too long to assess different materials. However, if the exposure site is too extreme it becomes unrealistic. The third decision then was to sort out a balance between these extremes of too mild or too severe, and this is assessed by the characterisation of each of the sites to determine just how corrosive they actually are. This has been done using two different internationally recognised methods described below.

The design of the test building and panels, what was to go onto them, and the method of exposing these in a realistic manner to the weather at each site was necessarily complicated. The number of permutations presented a logistical puzzle that was solved only by compromising the design so the test shed sheeting was never intended to be an ‘as built’ design. The test shed size was restricted to 10m2 as being the largest ‘shed’ that can be built without going through a building consent process.

The Project.

Members of the project team were:
Stuart Hayman Manager
Stuart Thomson Designer
Alistair Fleming Planner
Rod Newbold Procurement
Ross Simpson Installation

The $100,000 plus project was financed by the NZMRM with contributions from each of the partners.

Partners with MRM in this project include both the coil suppliers – NZ Steel and Pacific Coil Coaters – and the major fastener manufacturers – Hylton Parker, Ramset and Bremick, and other fastener and underlay suppliersl. The intention is that this is a roofing industry project aimed at producing the best outcomes for the NZ building industry, not just promoting specific products.

The Sites.

Six exposed coastal sites around New Zealand were chosen based on availability, assessed severity of environment and where security and access could be assured. These were:

1. House at Bell Block NP
2. Waipu Golf Club
3. Invercargill Gun Club
4. Nelson Boulder Bank (Cawthron Institute)
5 Taharoa NZ Steel mine site
6. Muriwai Golf Club (AKZO/PCC Site).

The Shed.

The sheds were designed by Stuart Thomson and built at the factory of Roll Forming Services in East Tamaki who manufactured the galvanised steel framing from Z450 g/m2 and also provided a lot of logistical assistance. After an initial test assembly, these were made up as kitsets, transported to the sites and erected by Ross Simpson on a previously poured concrete slab. The intention was that the land owners would be able to use the sheds for their own purposes during and after the trials so that each of the sheds had a slightly different design.

The sheds were all oriented to the prevailing wind with the lee side having the openings and a canopy overhang to provide an ’unwashed” area for exposing fasteners on the horizontal wall cladding. This 5-rib wall cladding doubled as the wind bracing for the buildings.

The 30° pitch roof of each test shed consisted of nine prefabricated panels (1.2m x 1.2m x 18mm of untreated plywood) with four different metallic and paint coated corrugated profile sheets on each panel.

The roof sheets

Each of the sheets was water cut to provide four exposed corrugations and included aluminium, ZAM, Zincalume®, a new metallic coated product (currently called Zincalume® Activate, Type AM in AS 1397:2011) and galvanised steel. Combinations of painted and unpainted sheets and some water borne paint coatings provided 12 different options.

The fasteners

The screws were sourced from three main suppliers;
Bremick, Buildex, Hylton Parker, (plus some others), The screws and load spreading washers included:

• Class 4 metal plated, painted and unpainted
• Class 5 (B8) metal plated painted and unpainted
• Stainless steel painted and unpainted
• Aluminium painted and unpainted

The test consisted of 10 different fasteners with each of these driven through the sheet and also through a 9mm oversized hole and fitted with a load spreading washer. This provides a test of the fastener/substrate combination in contact and also tests the shank of the screw. The screws are staggered across the roofs so that all combinations of substrate and screw are exposed. Although this cuts down the number of replicates there are still 1440 fasteners

exposed on each shed. The aim was to have 100 samples of each fastener as the AS/NZS 3566 standard requires a 95% pass rate from 100 exposed screws.

The Gutter Details

As many variants as possible were used including kraft and synthetic underlays as well as underlayment, different tapes, aprons and scribed metal flashings all with the intention of minimising corrosion at the gutter line. Some plastic fillers were used as well as safety mesh (as distinct from wire netting) and different underlay fixings such as “little grippers” and staples.

Corrosivity assessment

There are two methods for this, different in means and slightly different in objective.

Zinc coupons

Method 1 – is to expose zinc coupons on the shed and measure the rate of loss of zinc (or actually growth of zinc corrosion products). These are measured at 6 month or annual intervals and after a certain loss of zinc has occurred the fasteners and substrates are examined. This method is taken from the Australian standard AS 3566.2 for the durability of screws. Both Stuart Thomson and Alistair Fleming are on the 3566.2 committee looking at the revision of this standard and were able to provide expert advice on what to use. The composition of the zinc is critical, and a zinc alloy known as Zintane 83 is being used as the method of determining the corrosivity of the site. MRM was able to import this material from the manufacturers in France.

The coupons have been coded to each site and are numbered to provide traceability over the four plus year test period, and they are periodically exchanged and weighed to provide the zinc loss data. This method is to assess the performance of the exposed items after a certain loss of zinc. It is basically to calibrate the exposure of the samples, and the speed of corrosion is also a measure of the site.

Salt Candle

Method 2 – Salinity and time of wetness. Time of Wetness (TOW) is described as the period during which a metallic surface is covered by adsorptive or liquid film capable of causing corrosion. This method complies with ISO 9223 and is Kitset ready to go to site First shed Under construction First shed finished
Zinc coupons new and after 6 months the method used internationally to assess the corrosivity of exposure sites. It has several parts:

a) The salinity is measured by the “salt candle” method. A bottle of water with an exposed wick supported on a “candle” is exposed at the site on a specifically designed stand for 30 days. The liquid in the bottle is measured for its chloride content and this is expressed as mgms chloride per m² per day. This is repeated monthly on a continuing basis during the trial.

b) The humidity and temperature is measured continuously by a data logger mounted near to the candle this provides a measure of how much time the site would actually be damp for on a daily basis (TOW).

c) The local weather data (from a nearby NIWA site) is checked to see that it agrees with the TOW measurements.

d) The time of wetness and the salinity together provide a measure of the corrosive atmosphere at the site on a time basis – i.e. for how long it and how corrosive it is.