24-Bld-A7 Building Envelope Design · May 2016

Question 3 of 6

Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)

Question 3 (20 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

Given.

Find. (1) The panel's total thermal movement. (2)–(3) The minimum vertical joint width for two different sealant-installation temperatures.

Approach. Figure 1 shows the panel fixed only at its mid-point, so each half of the panel (from the anchor out to its own vertical joint) moves independently with an effective length of $L_{eff}=W/2$; per ASTM C1472, a joint between two identical panels sees the sum of both panels' edge movements, and the sealant's ±25% capacity is applied against the width as actually installed, which is why the installation temperature matters.

1. (1) Total movement of one panel edge. Over the full service temperature swing, $\Delta T = T_{max}-T_{min}=65-(-25)=90^\circ\text{C}$, so $$\Delta L=\alpha\,\Delta T\,L_{eff}=11.7\times10^{-6}\times90\times2500=\boxed{2.63\ \text{mm}}$$ — the maximum total (peak-to-peak) movement one panel edge can undergo relative to its own mid-point anchor.
2. (2) Joint installed at the annual mean temperature. Taking the mean as the midpoint of the design extremes, $T_{mean}=(65-25)/2=20^\circ\text{C}$: $$e_{expand}=\alpha(T_{max}-T_{mean})L_{eff}=11.7\times10^{-6}(45)(2500)=1.316\ \text{mm per panel edge}$$ $$e_{contract}=\alpha(T_{mean}-T_{min})L_{eff}=11.7\times10^{-6}(45)(2500)=1.316\ \text{mm per panel edge}$$ Since two identical panels flank the joint, the joint itself must absorb double each edge value: $2(1.316)=2.63\ \text{mm}$ in both directions (symmetric here because the mean sits exactly halfway between the extremes). With a ±25% sealant, $$W_{min}=\frac{2.63}{0.25}=\boxed{10.5\ \text{mm}}$$
3. (3) Joint installed at 5°C (off the mean). $$e_{expand}=\alpha(65-5)(2500)=11.7\times10^{-6}(60)(2500)=1.755\ \text{mm per edge}\Rightarrow 2(1.755)=3.51\ \text{mm at the joint}$$ $$e_{contract}=\alpha(5-(-25))(2500)=11.7\times10^{-6}(30)(2500)=0.878\ \text{mm per edge}\Rightarrow 2(0.878)=1.76\ \text{mm at the joint}$$ The governing (larger) demand sets the minimum width: $$W_{min}=\frac{3.51}{0.25}=\boxed{14.0\ \text{mm}}$$

Installing at 5°C — well below the 20°C annual mean — consumes most of the sealant's stretch capacity closing the joint on the hot side, so the required joint is 33% wider than the 10.5 mm needed when the same sealant is installed at the true mean temperature. This is the practical reason specifications call for sealant installation "at or near mean temperature": it minimises the joint width (and hence the sealant volume and cost) needed to survive the full annual cycle.

(4) Vertical joint sketch and relative dimensions. The joint consists of a compressible closed-cell backer rod set to the correct depth, over which the sealant is tooled to a concave profile. Sealant manufacturers size joint depth relative to joint width (the "shape factor") so that the sealant strains evenly rather than concentrating stress at the substrate bond line: for joints up to about 13 mm wide, depth is set equal to width (1:1, minimum 6 mm); for wider joints, depth is capped near 10–13 mm regardless of width (roughly 1:2 or leaner), because a sealant that is too deep for its width restrains its own mid-plane and cannot elongate uniformly. The backer rod is sized about 25% larger than the joint width so it wedges in place without bonding to the sealant (a "three-sided adhesion" bond-breaker), which is what allows the sealant to work in pure extension/compression against only the two panel faces.

(5) Failures from a joint that is too wide or too deep. A joint that is too wide for the sealant's rated movement capacity forces greater elongation per cycle than the sealant can sustain, producing adhesive failure (the bead pulls away from the panel face) or cohesive failure (the bead tears internally); a joint that is too deep relative to its width prevents the sealant from achieving the intended hourglass strain profile — the sealant restrains itself at mid-depth, concentrates strain at the bond line, and fails there prematurely, and a too-deep joint with a poorly-set backer rod can also let the sealant three-side-bond to the rod, tearing it apart from within on the very next movement cycle.

(6) Single-stage vs. two-stage joints. A single-stage joint (left panel of the sketch above) uses one exterior sealant bead to resist both rainwater and air/vapour leakage simultaneously; it is fully exposed to UV, thermal cycling and standing water, so it carries the entire performance burden on one line of defence. A two-stage joint (right panel) splits the two functions: an outer, vented and drained rain baffle sheds the bulk of the water (pressure-equalised, so it never has to be perfectly airtight), while a separate, protected inner sealant bead — set back from weather and UV — provides the actual air/vapour seal. Advantages of the two-stage joint: (i) redundancy — a flaw in the outer baffle does not by itself cause an air/water leak; (ii) the inner seal, being shielded from UV and bulk water, lasts far longer than an exposed single-stage bead; (iii) pressure equalisation of the outer cavity greatly reduces the driving force for water to reach the inner seal in the first place, which is the same rainscreen principle used at the rest of the building envelope.

Part B — rainwater management at the window/wall connection. The exam figure's numbered legend describes a sill detail built on redundant, gravity-drained lines of defence rather than a single perfect seal. Bulk rain first meets the primary seal at the window-to-frame interface (items 5 and 12, sealant and backer rod) and the perimeter sealant continuing along the jamb (item 2, "sealant beyond"); any water that gets past this primary seal lands on sloped blocking under the window (item 11), which is pitched to drain outward rather than let water pond at the sill. That water is intercepted by the pre-finished metal flashing with end dams (item 6), which is lapped over the vapour-permeable sheathing membrane (item 8) in shingle fashion so gravity carries any water down and out over the wall's water-resistive barrier rather than into the stud cavity; the foil-faced membrane at the sill (item 7) is a secondary, fully-adhered backup layer directly under the window nailing flange, catching anything that bypasses both the primary seal and the sloped blocking before it can reach the wall assembly. The intermittent shims (item 4) support the window without blocking this drainage path, and the insect screen (item 13) protects the drainage/pressure-equalisation opening beneath the sill without impeding it. The overall strategy is the standard "seal at the window, drain at the sill": accept that some water will get past the face seal, and give it a continuous, sloped, shingle-lapped path back to the exterior before it can reach the wood framing.