Facade Mockups and Performance Testing

A facade mockup, the full sized stretch of wall that a tower envelope in the development work Michael Stern writes about is sent to a test yard before any production unit is made, is where a design meets water, wind and movement under instruments rather than weather. The yard is worth walking through in order, because the sequence is the point: each station loads the specimen a little harder, and the leakage tests come back after every one to ask whether the seals survived what just happened.

The yard

The specimen hangs on a steel rig standing in for the building. Glass Magazine’s guide to mockup testing insists on a faithful copy of the job at full size, down to the anchors, glass, sealants, gaskets, splices and frame intersections, with the supporting structure simulated as far as practical, and it recommends a specimen two floors high and two bays wide as the normal choice, a corner from near the base as a better one, and a corner from the top, where pressures are greatest, as the best and most expensive.³ A single story one bay wide is not enough.³ Standard systems with a certification history may skip the exercise; custom walls with new, unproven features are the ones where, in the magazine’s words, testing is the only way to verify performance.³ Intertek, which runs chambers across the United States, Canada and China, lists curtain wall, window wall, precast, masonry, rainscreens and metal panel among the assemblies that pass through them.⁹

The chamber

The specimen is sealed to a chamber so a fan can drive a difference in pressure between its two faces. IIBEC’s review of fenestration testing names the methods: ASTM E283 measures the rate of air leakage at a set pressure difference, ASTM E331 sprays water under a steady pressure and looks for any that reaches the inside, and AAMA 501.1 repeats the water test under dynamic pressure, with vertical movement under AAMA 501.7 and drift under AAMA 501.4 drawn from the same family.⁸ Facade Ghana’s testing guide describes the dynamic test as rain driven at the wall by a propeller or an aircraft engine under cycling pressure, closer to a storm than a static spray, and notes that a wall can pass one of the three core tests and fail another, which is why a specification calls for all of them.⁷ WFM Media’s model specification shows the numbers a consultant writes: air at 300 pascals positive and negative, water at 3.4 liters per minute per square meter for 15 minutes under a kilopascal of pressure, with no uncontrolled leak and the source of any leak identified.⁴

Figure

The order of a mockup test

1. 01
   PreloadThe chamber applies a fraction of the inward design pressure and releases it so the specimen seats on its anchors.
2. 02
   AirAir leakage is measured through the whole specimen at a fixed pressure difference.
3. 03
   WaterWater is sprayed under static pressure, then under dynamic pressure from a propeller or engine.
4. 04
   MovementThe rig racks the wall sideways and drops a floor edge to mimic drift and slab deflection.
5. 05
   StructureInward and outward pressure are held at the design value and then at a 50 percent overload.
6. 06
   RepeatAir and water tests run again after every stage to prove the seals survived the movement.

The pressure box

The same chamber then pushes and pulls. ASTM E330 pushes and pulls on the specimen to measure deflection and any permanent damage, which Facade Ghana calls the core structural test for a high rise skin, proving that framing, glass and fixings carry the design wind to the structure.⁷ The design pressures come from a wind study. The B1M’s visit to a wind tunnel laboratory describes scale models set on a turntable with sensors inside recording how the shape responds, and quotes a consultant estimating that testing can save 25 to 30 percent of the embodied carbon in facade and structure by trimming pressures that codes overstate.² Those pressures, and why corners see more of them, are treated under cladding pressures.

Sometimes the box wins. Building Design + Construction’s case study follows a stainless steel clad unitized wall made in Asia for a high rise there and shipped to Florida for its test; it cleared the leakage and racking stages, then its anchorage gave way at 2.5 kilopascals of suction and four units came off the rig’s upper level.¹ Investigators recovered hooks that had slipped off their rods and others bent out of shape, and a metallurgy laboratory supplied the reason: the hooks had been made of a weaker aluminum alloy than the calculations assumed.¹ The hooks were remade, the rods were lengthened and flanged so a hook could not slip, one new hook was sawn through to prove the load would redistribute, and the design went to production.¹

Figure

One mockup that fell down

specimen width, including a straight wall, an arc, column cladding and an outside corner

22.7 m

height of the specimen, two full stories

12 m

suction that defeated the anchor hooks and sent four units off the upper rig level

2.5 kPa

the cause: hooks made of a weaker aluminum alloy than the calculations assumed

1 alloy

The racking frame

Next the rig moves the building. In the inter story drift test the floor edge above is pushed sideways relative to the one below, three cycles each way, then the exercise is repeated at a greater displacement after the overload test; thermal cycling follows, with the whole specimen taken from a cold day to a hot one and back three times while design pressure is applied at each extreme.¹ UL Solutions groups its chamber work the same way, air, water, inter story drift, structural, thermal and, for hurricane regions, impact and cycling to the Florida protocols.¹⁰ WFM Media’s specification adds a tie back test for the window washing equipment, loaded in every direction while a static pressure sits on the wall, with no permanent distortion allowed.⁴ What those movements are, and how joints are sized for them, has its own entry under movement and tolerances.

The visual mockup across the yard

A second, quieter specimen often stands nearby. INSOL, a facade maker in New Zealand, separates the visual mockup, placed on the real elevation at the real orientation so the architect and client see the same element under changing daylight, from the performance mockup that is tested; the visual one catches a color that reads wrong at noon or a shadow nobody intended, while the performance one becomes the standard for installation method, hardware fit, trade interfaces and maintenance access.⁶ The firm’s bluntest argument is for the unknown: some characteristics of a facade element only show up once it exists at full size.⁶

Figure

Two mockups, two jobs

	Visual mockup	Performance mockup
Purpose	Let everyone see and touch the same facade element	Prove air, water, structural and movement performance
Where it stands	On the building elevation it is designed for, in real daylight	On a steel rig in a yard or a laboratory chamber
What it catches	Color, texture, perforation pattern, an unintended shadow	A leaking joint, a weak anchor, a seal that opens under drift
Afterlife	A reference the architect approves against	A control sample for installation method and tolerance
Side benefits	Stakeholder agreement before fabrication	Maintenance access, hardware fit and trade interfaces checked early

Who signs the report

Accreditation sits behind every station. IIBEC explains that in the United States and Canada laboratory and field tests must be carried out by an independent, accredited laboratory before the codes will accept the result, and it names AAMA 501 as the principal guide specification for mockup performance of wall systems.⁸ Glass Magazine adds that AAMA protocols require an accredited testing agency operating under a documented quality system, and points specifiers to ASTM E2099, as the document that governs how a specimen is specified, built and evaluated.³

Back on the building

The last station is the building itself. Building Envelope Allies lays out the AAMA 501.2 field check on installed curtain wall meant to remain permanently sealed: a calibrated nozzle held a foot from the glass, moved across at least 100 square feet, working upward, for five minutes while a technician inside watches for water, with more than half an ounce in that time counting as a failure.⁵ The firm recommends testing before interior walls are finished, when a leak is easier to find and cheaper to fix.⁵ WFM Media’s practice is to sample random locations chosen by the consultant and to record the whole thing on video, with any failure sending the work completed so far back for review.⁴

What the test was worth

The argument for the yard is the cost of its absence. Building Design + Construction’s investigators conclude that without the mockup the Asia tower would have gone up with anchors that modest wind suction could defeat, and that the flaw would have shown itself only as a failure after occupancy.¹ Facade Ghana calls the mockup the point in a project where a problem costs least to fix, a drawing change rather than a recladding, and WFM Media observes that once a facade is up on a tower, replacing an underperforming component can be close to impossible for want of access, so the residents live with it.⁴ The full size specimen is, in that sense, the cheapest version of the building that will ever be built.⁷

Figure

The hose test on the finished wall

minimum test area, representing typical construction with frame intersections and splices

100 sq ft

distance the calibrated nozzle is held from the glass

1 foot

spray time across the area, bottom to top, while a second technician watches inside

5 minutes

water collected in that time that counts as a failure

½ ounce