Performance & Tests

Curtain wall fabricators and installers generally assume that the building is watertight and airtight, that the façade elements are correctly assembled, that gaskets are properly seated, and that the insulation is sound. In practice, the intersections between connections are often left until last, and the performance of the interface components between adjacent systems is rarely checked. Damage over the building's service life tends to occur precisely in these zones, and on large buildings the result can be irreparable failure costing more than rebuilding the envelope itself.

The logical approach, particularly on large projects, is to test a representative mock-up built to the design before tendering. In reality, few firms will bear this cost before winning the work, and few owners will pay for it if the fabricator includes it in the bid. As a result, a sample of a continuously repeating section of the façade is usually prepared and tested only after the client has selected a firm and its system. Two issues are frequently overlooked: what happens if the design fails the tests, and how competent the site operatives and the project's façade consultant actually are.

Where a project is not large enough to justify a full mock-up, an on-site watertightness test can at least verify workmanship during installation. This is done in one of two ways: spraying high-pressure water onto the surface, or creating a pressure difference between the interior and exterior while spraying water.

Even when the right materials are chosen, the critical issues likely to cause problems later must be identified and the building's behaviour established through standard tests. These critical issues — maintenance, structural safety, and water, rain, air infiltration, condensation and water-vapour transmission — are fully interrelated: a poor result in one affects the others. All must therefore be brought under control before the building is occupied and kept under control thereafter.

When a façade mock-up is assessed against the tests described below, defined criteria apply. Depending on the results, the system is either changed or its details are refined; in either case the same tests must be repeated on the revised sample. After testing, the sample is dismantled, any water trapped within the system is recorded, and any deviations from the design are noted. Each test is then evaluated against its own criteria, as follows.

Air permeability

The permissible air-leakage rate (Qo) for fixed panels at the maximum test pressure (po) should not exceed 1.5 m³/h per m². At intermediate pressures (pn), the leakage rate (Qn) is calculated as Qn = Qo (pn / po)^(2/3).

Watertightness

No water should penetrate the interior face during the test, and no water should pool in areas required to remain dry on completion.

Structural performance at design pressure

Deflection of frame members must not exceed 1/200 of the span between the building's fixings or 20 mm, whichever is smaller. Deflection of frame members supporting an insulating glass unit must not exceed 1/175 of the length measured along the glass edge or 15 mm, whichever is smaller. Residual deformation must not exceed 1 mm within the one-hour recovery period allowed for repair.

Structural performance at 1.5 × design pressure

No permanent damage to panels, anchors or frame members is permitted. Frame members must not bend or buckle; panels, glazing and decorative infills must remain as installed, and gaskets must not displace. Residual deformation of frame members must not exceed 1/500 of the span between the points anchored to the building within the one-hour recovery period.

Seismic movement

Apart from designated leakage zones, no permanent deformation is permitted during or after the test.

Temperature differences

Correct profile and glass selection requires the external and internal air temperatures to be taken into account. Typical design values are an external air temperature of −30 °C (winter minimum) to +50 °C (summer maximum), and an internal air temperature of +40 °C (winter maximum) to +10 °C (summer minimum).

Thermal transmittance (U-value, EN 673)

The U-value, also used in the building's HVAC and ventilation calculations, must be determined to allow correct profile and glass selection. The U-value is expressed in W/m²K: Uf for the frame, Ug for the glass, and Uw for the complete window (frame plus glass).

Daylight and solar performance (EN 410)

Daylight transmittance (%) and reflectance (%) must be determined to EN 410. Solar performance must likewise be characterised by the solar heat gain coefficient (SHGC), the shading coefficient, and the solar factor (g-value).

Condensation

Condensation is assessed for summer and winter using the internal and external air temperatures together with relative humidity. Correct profile and glass selection can prevent condensation from forming.

Beyond weathertightness and structural safety, the façade must control sound transmission and contribute to fire safety. Both are governed by harmonised European standards and are increasingly specified as explicit performance targets.

Acoustic performance

Airborne sound insulation is characterised by the weighted sound reduction index Rw (in dB), measured in the laboratory to EN ISO 717-1. Typical curtain wall glazing achieves Rw values in the range of 30–45 dB depending on glass build-up, cavity width and the use of acoustic interlayers; laminated glass with a PVB or acoustic interlayer markedly improves performance. The corresponding field-measured value is the standardised level difference DnT,w, which is typically 8–10 dB lower than the laboratory Rw because of flanking transmission around the façade. Common design targets are DnT,w ≥ 40 dB between offices and ≥ 45–50 dB for schools and hospitals.

Fire performance

Reaction-to-fire performance of façade materials is classified under EN 13501-1 using the Euroclass system (A1, A2, B–F) together with smoke (s1–s3) and flaming-droplet (d0–d2) sub-classes; non-combustible cores achieve Class A1. Fire resistance of perimeter barriers — the seal between the floor slab edge and the back of the curtain wall, which prevents fire and smoke spreading between storeys — is tested to EN 1364-4 and EN 1366-4, and rated for integrity (E) and insulation (I) over periods from 30 minutes to several hours. Correct specification of cavity barriers and firestops at every floor line is essential to limiting vertical fire spread within the façade cavity.

Façade testing follows the standards of the CWCT (Centre for Window and Cladding Technology), ASTM (American Society for Testing and Materials), AAMA (American Architectural Manufacturers Association) and TS-EN. Curtain wall systems are generally tested for watertightness, air permeability, and displacement under wind and seismic action. ASTM E331 is used for the static test and AAMA 501.1 for the dynamic water test, and the procedures must align with the CWCT standard test methods for curtain walls.

A typical test sequence for a high-rise façade is listed below. The order may vary from building to building:

1. Air permeability
2. Watertightness (static)
3. Wind resistance (serviceability)
4. Air permeability
5. Watertightness (static)
6. Seismic movement at design displacement
7. Air permeability
8. Watertightness (static)
9. Watertightness (dynamic)
10. Watertightness (hose)
11. Wind resistance (safety)
12. Seismic movement at 1.5 × design pressure
13. Dismantling and reassembly

The tests applied depend on the climatic conditions of the building's location. Beyond those above, ageing and thermal-cycling tests may be carried out to establish material performance over time. In an ageing test, the mock-up is conditioned in a stability chamber under defined temperature and humidity for set periods; in a UV chamber, the ultraviolet component of sunlight is simulated to age the material under controlled temperature and duration. After ageing, a tensile test measures the material's strength. The material tested is typically the silicone used in structural silicone glazing, which is safety-critical: bonded to two surfaces, it fails either by debonding from a surface or by cohesive rupture. The test is performed in a tensile chamber, increasing the force gradually. In the thermal-cycling test, the outer face of the panel is cycled six times between −20 °C and +80 °C while temperature and humidity are held constant on the inner face.

The principal façade tests and their governing standards are summarised in Table 2. The CWCT standard test methods are broadly equivalent to the corresponding EN standards, while ASTM and AAMA designations are common in North American and international specifications.

Test	What it measures	Governing standards
Air permeability	Rate of air leakage through the façade under pressure	CWCT §5 · BS EN 12153 · ASTM E283
Watertightness (static)	Water penetration under uniform static pressure	CWCT §6 · BS EN 12155 · ASTM E331
Watertightness (dynamic)	Water penetration under dynamic (wind-driven) pressure	CWCT §7/8 · BS EN 13050 · AAMA 501.1
Watertightness (hose)	Workmanship check at joints with a handheld nozzle	AAMA 501.2
Wind resistance (structural)	Deflection and strength under design and 1.5 × design load	CWCT §11/12 · BS EN 12179 · ASTM E330
Seismic movement	Performance under inter-storey drift	AAMA 501.4
Thermal transmittance	U-value of frame, glass and whole window	EN 673 · EN ISO 10077
Solar / daylight	Solar factor (g-value), light transmittance and reflectance	EN 410
Acoustic	Weighted sound reduction index Rw	EN ISO 717-1
Reaction to fire	Euroclass material classification; perimeter fire resistance	EN 13501-1 · EN 1364-4 / EN 1366-4

Table 2. Façade tests and their governing standards.

To prepare a curtain wall for testing, a full-scale mock-up incorporating one or more of the building's repeating modules is first built and mounted on one face of the test chamber. The remaining faces of the chamber may be fabricated from steel or plywood. The chamber must have an access door and be fully sealed from the external environment.

The static-pressure chamber must measure its internal pressure and be positioned so that readings are unaffected by air movement inside or outside the chamber. The air-supply system comprises a centrifugal fan with a common duct and control valves capable of generating positive or negative pressure. It must deliver a constant pressure for the duration of the test period and be capable of reaching approximately 600 Pa within one second.

The wind generator, used to create varying positive pressure during the dynamic test, comprises an aircraft-engine-type piston with a 4,000 mm diameter, contra-rotating propeller, mounted close to the outer face of the mock-up. The water-spray system is an evenly spaced grid with spacings no greater than 700 mm, mounted approximately 400 mm clear of the mock-up. The nozzles must cover a square area with a spray spread of no less than 80°, and the water temperature must be between +8 °C and +25 °C, delivering water evenly across the outer face.

For the hose test, brass nozzles delivering a continuous spray with a nominal 30° spread are used. A pressure gauge and a control valve must be fitted between the nozzle and the valve, and the system must be set so that each nozzle delivers 22 ± 2 litres per minute at a pressure of 220 ± 20 kPa.

Figure 1. A typical curtain wall test apparatus.

Air infiltration is the passage of air from outside to inside through the curtain wall. Air leaks through faulty connections between horizontal and vertical profiles or through gaskets, and is the principal cause of energy loss in buildings. Since 2002, UK building regulations have required air leakage in large buildings to remain within set limits, extended in 2006 to buildings of all sizes. On the completed building, the energy calculations require air leakage to be below 10 m³/h per m².

High air-leakage rates increase heat loss and undermine occupant comfort. Because pressure differs between the inside and outside of a building, this test determines whether — and to what extent — air passes through the façade. The limiting test pressures are 600 Pa for air-conditioned spaces and 300 Pa for non-air-conditioned spaces. The test pressures applied under CWCT are given in Table 1.

Wind pressure	Watertightness test pressure	Air permeability test pressure
< 800 Pa	300 Pa	300 / 600 Pa
801 – 1200 Pa	300 Pa	300 / 600 Pa
1201 – 1600 Pa	450 Pa	300 / 600 Pa
1601 – 2000 Pa	600 Pa	300 / 600 Pa
2001 – 2400 Pa	600 Pa	300 / 600 Pa
> 2400 Pa	0.25 × wind pressure	—

Table 1. CWCT pressure levels.

In the air-permeability test, the mock-up is preconditioned with three positive pressure pulses of 660 Pa. Air-leakage readings are then taken at 50, 100, 150, 200, 300, 450, 600, 450, 300, 200, 150, 100 and 50 Pa, holding the sample at each pressure step for at least 10 seconds. The chamber's own leakage is first measured with the sample sealed, then with it unsealed; the difference between the two readings indicates whether air passes through the sample, which is then judged against the limiting values.

Water penetration is the passage of water from outside to inside through the curtain wall. A small, controllable amount of interior water may be acceptable, but the governing principle is to achieve a perfect seal in all curtain walls. The mock-up is tested to establish the façade's resistance to water penetration; this is essential on large projects in particular. In the test sequence above, a static test is repeated after the wind and displacement tests to confirm that watertightness has not been affected by structural movement.

Laboratory tests cannot measure the real building's watertightness exactly, because the scale of the mock-up differs greatly from that of the building and on-site workmanship at full scale is decisive. Test duration and pressure also vary by climate zone. The ASTM E331 protocol is well suited to water testing; the test pressure is set as a proportion of the maximum positive design wind load specified by the standards of the country in which the façade is located.

• Temperate zones: test pressure 20% of the maximum positive design wind load; test duration 30 minutes.
• Storm (hurricane) zones: test pressure 80% of the maximum positive design wind load; test duration 60 minutes.

If water penetration occurs during a test, repair is not permitted. Depending on the result, the system or its detailing must be changed and a new mock-up prepared and retested. A successful test is one in which no water penetrates for the full test duration; if penetration is observed, the water test must be stopped. Watertightness is tested in three ways — static, dynamic and hose.

Static (CWCT)

A water grid is placed in front of the façade mock-up and water is sprayed under positive air pressure, representing heavy wind-driven rain. After three preconditioning pulses of 660 Pa, water is sprayed at no less than 3.4 litres/m²/min for 15 minutes at zero pressure. With spraying continuing, the pressure is raised through 50, 100, 150, 200, 300 and 450 Pa for 5 minutes each, then 600 Pa for 15 minutes. The inner face is monitored continuously; if water penetration is observed, the test is stopped.

Dynamic (CWCT)

Water is sprayed onto the mock-up at no less than 2 litres/m²/min. The wind generator, mounted close to the outer face, creates a positive pressure of 600 Pa for 15 minutes. The pressure is determined by increasing the wind speed until the deflection of the frame members on the jet's centre axis reaches the deflection recorded under 600 Pa static pressure. Suction may be applied to the inner face to achieve the required test deflections, but is limited to 25% of the maximum static pressure. The inner face is monitored throughout; if water penetration is observed, the test is stopped.

Hose (CWCT)

Treating the mock-up as a wall section, wetting begins at the lowest horizontal joint and proceeds upward — first the intersecting vertical joints, then the next horizontal joint, and so on. Water is directed onto the perimeter and joints of the outer face from 300 mm away, moving slowly back and forth over each joint and dwelling for 5-minute periods every 1,500 mm. Long and short joints are tested proportionally. The inner face is checked for penetration throughout the test.

Wind is commonly thought to act only as direct pressure on the façade, but where it strikes perpendicularly it creates positive pressure, while at the edges — as it changes direction over the top and sides — it creates suction. Wind reaching a building produces positive pressure at the centre of the windward face and negative pressure (suction) at the corners; slowed at the front, it accelerates over the sides and roof, creating strong pressures there, and produces weaker pressure where it wraps around the rear. Tall buildings are under continuous wind action, and where high structures stand close together, air compression and turbulence create sudden, very high gusts. The durability of a high-rise curtain wall must be designed with these strong wind effects in mind.

The wind test evaluates the structural resistance of the curtain wall, alternately generating positive and negative pressure. Deflection limits must be set for vertical and horizontal loads and must not be governed by the capacity of the profiles alone; the soundest approach is to design to the deflection limit of the glass, since glass will break under excessive deflection. Deflection limits are also important for controlling movement on the inner face, where the building structure may sit very close to the profiles and excessive deflection could cause them to strike the structure. Limits are expressed as the distance between anchor points divided by a fixed number; L/175 is common in curtain wall work. For a 4,320 mm floor height, the permissible deflection is approximately 24.7 mm (4,320 ÷ 175) — meaning the façade may deflect up to roughly 25 mm inward or outward at maximum wind pressure. Profile deflection is controlled in various ways: typically by providing the moment of inertia required to maintain the deflection limits, or by reinforcing the profiles internally with steel, whose modulus of elasticity is three times that of aluminium, allowing the steel to absorb part of the deflection at lower cost.

Serviceability (CWCT)

To precondition the mock-up, the displacement transducers are zeroed and three positive pulses of 500 Pa are applied. The sample is then subjected to five positive pressure strokes (0 to 500 Pa and 500 Pa to 0), each applied as quickly as possible — between 1 and 3 seconds. The procedure is repeated with negative pressure: three pulses of −500 Pa, followed by five negative strokes applied as quickly as possible, held no longer than 3 seconds and no less than 1 second.

Safety (CWCT)

The mock-up is subjected to the pressures below, each applied as quickly as possible and held between 1 and 3 seconds. After preconditioning at 500 Pa with three positive pulses, a single pulse of 2,340 Pa (1.5 × design pressure) is applied. The procedure is then repeated in the negative direction: three pulses of −500 Pa followed by a single −2,340 Pa pulse (1.5 × design pressure). The displacements measured at maximum pressure are recorded.

Seismic loads must be established in advance for the design of the anchors and the curtain wall system. Because adequate clearance is normally provided between profile and glass, the façade naturally resists seismic and wind loads, and standard curtain wall systems can accommodate 75 mm of floor movement without water leakage or glass breakage. Where floor-to-floor spans are large, however, the anchors carry greater loads and their design must be reviewed.

Beyond seismic action, accidental explosions and terrorist attacks are increasingly common. As the building envelope, the curtain wall is expected to provide the first line of defence against these large, short-duration blast loads; assessing the façade's response requires a dynamic load analysis on a mock-up of a defined size, and laminated glass is generally specified in such buildings.

A curtain wall system comprises cladding panels, the frames supporting them, and a carrier connecting the frames. Thermal and wind loads act directly on the façade and can readily be resisted with adequate safety margins — for example, the system can be designed for the wind load arising from the maximum wind speed recorded over the previous 50 years in the local climate data. The same cannot be done for seismic action, because its timing, magnitude and frequency cannot be predicted.

Seismic events damage the surroundings of a building through horizontal and vertical displacements. Because the building frame and floor slabs are very large relative to the mass of the curtain wall, the system cannot resist the internal inertia generated by horizontal and vertical movement. The curtain wall must therefore be designed not only for the seismic-zone estimate but also for the building's robustness and for the displacements that seismic action imposes on the building structure.

Design displacement (AAMA 501.4-00)

The intermediate steel support beam is moved ±xx mm from its starting point and returned, with the cycle repeated three times. Movement is recorded during the test and is measured at the support beam, not at the mock-up.

1.5 × design displacement (AAMA 501.4-00)

Following the same procedure, the intermediate steel support beam is moved ±xx mm from its starting point and returned, with the cycle repeated three times. Movement is recorded during the test and is measured at the support beam, not at the mock-up.

Laboratory mock-up testing confirms that the design and the chosen system perform as intended, but it cannot reproduce the workmanship of full-scale installation on the building itself. Field testing therefore complements the mock-up programme by verifying the as-built quality of the installed façade, either on a sample of completed bays or in response to reported leaks. Field tests use lower pressures and simpler equipment than laboratory tests, and are diagnostic rather than pass/fail certifications of the design.

Water leakage field check

The most common field test is the diagnostic water-leakage check to AAMA 501.2, in which a calibrated handheld nozzle is played slowly along the perimeter joints, gaskets and frame intersections of an installed bay while the interior is observed for penetration. It is quick and inexpensive and is intended to confirm the general quality of installation rather than to measure resistance to a specified pressure. Where compliance with a defined air- and water-resistance level must be demonstrated on the completed building, the more rigorous chamber-based field procedures of AAMA 503, ASTM E1105 (water) and ASTM E783 (air) are used, applying a sealed chamber to the installed bay.

When to test in the field

Field testing is appropriate at the start of installation, to validate workmanship before the trade is fully mobilised; periodically during installation, as a quality-assurance sample; and reactively, to locate the source of a reported leak. Early field testing is the most cost-effective, because correcting a recurring detailing or workmanship fault before the whole façade is installed avoids remedial work across the entire building.

The following organisations publish the test standards and guidance referenced in this document:

• CWCT — Centre for Window and Cladding Technology; standard test methods and guidance for building envelopes.
• ASTM International — test standards for air leakage, water penetration and structural performance (E283, E331, E330, E1105).
• AAMA / FGIA — Fenestration and Glazing Industry Alliance; window-wall and curtain-wall test methods (501 series).
• CEN / ISO — European and international standards (EN 12153, EN 12155, EN 13050, EN 410, EN 673, EN ISO 717-1, EN 13501-1).
• TSE — Turkish Standards Institution; national adoption of EN standards (TS EN series).