Characterisation of Aluminium Composite Cladding

By T. Byrnes, L. Angel, K. Hunt and T. Flynn

The accurate and rapid identification of cladding material and insulation on existing building facades is of interest to fire safety engineers, building consultants and building owners in light of the numerous high-rise fires over the last decade and the mounting legislative pressure to rectify any such occurrences of flammable cladding.

In this article, we discuss an effective process for rapid characterisation of aluminium composite panels (ACP) – the most common ‘combustible’ cladding type – and discuss the limitations of all available methods.

ACPs are typically composed of three layers – namely two sheets of ~0.5mm aluminium glued either side of a 2-3 mm thick central polymeric core. The core can be pure polyethylene (PE), ethylene vinyl acetate (EVA) or a copolymer thereof, cellulose reinforced phenolic (Bakelite) or other. Fire retardant cores have at least 70% by weight of the mineral filler ATH (known also as alumina trihydrate or aluminium hydroxide). The endothermic decomposition reaction of ATH at temperature is described in Equation (1)

Other fire retardants such as MDH (magnesium hydroxide) are occasionally employed. The properties of ATH and MDH are summarised in Table 1.

To completely characterise the panel core type, it is necessary to identify the polymer matrix, the fire-retardant mineral filler and the concentration of both by one or more methods.

Visual

Firstly, ACP can be easily identified by its three-layer structure if an exposed edge is available. Formed edges will be of small radius compared to aluminium sheet where the radius is quite large and accompanied by slight cracking parallel to the bend. The core colour is a good indication of the panel type. Black cores are very often carbon black filled polyethylene. (Pure polyethylene is clear.) White to dark grey cores are often fire-retardant ACP. However, this relationship is not dependable enough to rely upon it as a sole means of identification.

Density

Low density polyethylene (LDPE) has a specific gravity of 0.91 - 0.94 g/cm³. A PE panel core containing 70% ATH will have a density of 1.64 g/cm³. Refer to Figure 1 for the relationship between filler content and density. Core density may be measured through volumetric or hydrostatic means or using a pycnometer. With the density established, it is possible to calculate the w/w% filler concentration using Equation (2), provided the FR agent and polymer is 100% pure and the polymer and filler type and densities are known.

Since water has a density of 1 g/cm³, a pure PE core will float, while any significant level of filler will result in the sample sinking, making it a useful screening test for unretarded polyethylene.

Flame

Polyethylene ignites readily under application of a gas flame and melts between ~120 to 160°C. It will also continue to burn with flaming droplets after removal of the ignition source.

Fire retarded (FR) polyethylene will foam due to the steam generated during the decomposition of the fire retardant within the molten polymer. It should ignite with difficulty and self-extinguish after removal of a flame, at least until the fire retardant has been 100% consumed. Flame testing is a useful screening tool which is highly sensitive to combustible material. Results that are ambiguous are forwarded on for more expensive laboratory testing.

Laboratory

1. FTIR

The chief method for identification of both polymer and fire retardant is infra-red spectroscopy (FTIR). Each constituent will generate a characteristic peak in the infra-red spectra (Figure 2). Table 2 is a list characteristic peaks (cm^-1) for common ACP constituents.

2. TGA/DSC

TGA (thermal gravimetric analysis) measures the mass change as a sample is heated in a controlled environment up to 850°C. Air is used for testing of ACP cores. The organic (combustible) portion is burnt off, leaving behind the inorganic (non-combustible) portion as an ash residue. The weight is continually monitored throughout. See Figure 3.

Some practitioners rely upon TGA as a primary means to identify the polymer and fire retardant as well as the concentration of each. While all materials have characteristic decomposition temperatures, these can occur over a broad range and it is not always possible to effectively differentiate the contribution of each component in complex mixtures.

Also, TGAs unable to quantify the impurity levels in the mineral filler leading to an overestimate of the fire-retardant content.

Gravimetry

Gravimetry is the same process as TGA but is used solely to determine the respective proportions of the polymer and filler. The drop in weight correlates to the organic (combustible) content of the core which is usually the polymer fraction. The weight of the inorganic fraction allows one to calculate the original fire-retardant concentration according to Equation (3)

Where M is the non-combustible portion of the core (w/w%), A relates to the purity of the non-combustible portion (in terms of percent alumina or percent magnesia as determined by XRF) and W is the theoretical hydrate loss for that oxide (w/w%) obtained from Table 1.

W is theoretically 100 – 100(MW_AO/MW_A) where MW_AO and MW_A are the molecular weights in g/mol of the decomposed and undecomposed forms of the FR agent. If there is no transformation, W = 0. The organic component of the core (in w/w%) can be described by Equation (4);

Where Equation (3) and Equation (4) do not add to 100%, any outstanding w/w% is due to an inert oxide or impurity in the FR agent.

4. XRF

As most FR agents have some level of impurity, X-ray fluorescence (XRF) is used to differentiate each constituent (w/w%) of the non-combustible ash after gravimetry. Failure to employ XRF will likely result in an overestimate of the fire-retardant level. As XRF only reports the percent oxide, compounds like carbonates that may not decompose at the temperature employed in TGA analysis will skew the result.

5. XRD

Quantitative X-ray Diffraction (XRD) is useful in that it can determine the structure and relative proportion of minerals within the inorganic fraction. It cannot detect amorphous materials like plastics or amorphous silica and therefore still requires gravimetry to determine the polymer content of the core. The mineral mass in the as-received core (mc) in grams can be estimated through Equation (5).

where m_ash = weight of the ashed residue after gravimetry (g) MWN = the molecular weight of the compound formed from element ‘N’ before decomposition (mols/g) MWNO = the molecular weight of the compound formed from element ‘N’ after decomposition (mols/g) %Nc = the proportion of the mineral part of the core that is a compound formed from element ‘N’ as determined by XRD (w/w%)

Each unique element is assigned a label A, B, C… to N in the equation.

Note this equation is only valid if all the mineral components are identified (adding to 100%), including those that undergo no decomposition (ie; MW_N = MW_NO). If any amorphous material like silica is present, it will not be detected, and the mineral mass will be underestimated. Note that if mineral N is present alone, then it obviously makes up 100% of the mineral contribution such that Equation (5) simplifies to Equation (6);

If MW_N = MW_NO (i.e., the mineral undergoes no change during pyrolysis), then equation (6) simplifies further to m_c = m_ash as one would anticipate.

6. SEM

Scanning Electron Microscopy (SEM) coupled with Energy Dispersive X-Ray Analyser (EDX) can be used to identify the filler type, purity and concentration. Phase image analysis will yield the v/v% if the filler particle size is uniform and equiaxial. From this the w/w% filler can be calculated using Equation (7).

where v/v% FR is the volume fraction occupied by FR particles (in percent), v/v% PE = 100 - v/v% FR, ρFR = density of the FR mineral and ρPE = the density of polyethylene (both in g/cm3).

Drawbacks are the large number of assumptions in the v/v% to w/w% conversion and the insensitivity of EDS to the mineral form of the element detected. For example, both Al₂O₃ and Al(OH)₃ in the as-received core will generate Al and O peaks.

7. Calorific Method

Calorimetry simply measures the calorific content of a core sample by measuring the energy released when a sample is burnt in a calorimeter. For example, pure polyethylene contains ~45 MJ/kg, while a typical FR core should not exceed 15 MJ/kg. The relationship between filler content and calorific content can be seen in Figure 1. The test suffers from the shortfall that it concentrates on the polyethylene content of the core rather than the mineral content. That is, it is unable to distinguish between a core filled with active fire retardant from one filled with an inert material (eg; chalk) which has no ability to retard ignition.

Conclusion

Our recommendation for core testing is a combination of a flame screening test for polyethylene, coupled with FTIR + Gravimetry + XRF. It has been our experience that this offers the highest sensitivity and specificity for aluminium composite panel core characterisation.

T. Byrnes, L. Angel, K. Hunt are with the Facades Diagnostic Team, Arcadis, Sydney, Australia

T. Flynn is with Mark Wainwright Analytical Centre, University of NSW, Sydney Australia