Question 4 of 7: Precast Concrete Spandrel Panel – Thermal Movement and Sealant Joint Design

Question 4: Precast Concrete Spandrel Panel – Thermal Movement and Sealant Joint Design (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. Panel width $=5$ m, anchored at its own mid-point (so each free edge is $L=2.5$ m from the anchor). $\alpha=11.7\times10^{-6}$/°C. Design winter $T_{min}=-25^\circ$C, design summer $T_{max}=35^\circ$C (Montreal). Sealant installed at the annual mean temperature. Sealant movement capacities of ±25% and ±50% are to be checked.

Find. (1) The panel's maximum thermal movement; (2)–(3) the minimum vertical joint width for each sealant; (4) a labelled joint sketch; (5) failure modes from a badly-proportioned joint; (6) single-stage vs. two-stage joint design and the advantage of the latter.

Approach. Compute the free-edge movement of one panel from its own mid-point anchor over the full design temperature range (ASTM C1472 method); since the sealant is installed at the annual mean, the joint's total range splits symmetrically into extension (winter) and compression (summer) about the installed width, and the minimum joint width follows directly from dividing that one-directional movement by the sealant's rated movement capacity.

1. Part (1) — Maximum panel movement. Each panel is free to move relative to its own mid-point anchor over $L=2.5$ m, across the full annual range $\Delta T=35-(-25)=60^\circ$C: $$\Delta L_{edge}=\alpha\,\Delta T\,L=11.7\times10^{-6}\times60\times2.5=1.755\times10^{-3}\ \text{m}$$ $$\boxed{\Delta L_{edge}=1.76\ \text{mm (per panel edge, full annual range)}}$$ At a vertical joint between two identical panels, both adjacent edges move symmetrically toward or away from each other, so the joint itself sees a total one-directional movement (extension or compression, referenced to the mean-temperature installed width) numerically equal to this same 1.76 mm — confirmed independently in Step 2 by working from the mean temperature.
2. Parts (2)/(3) — Joint movement referenced to the mean-installed width. Sealant is applied at the annual mean temperature $T_{mean}=(35-25)/2=5^\circ$C, so each panel edge moves $\pm\alpha(T_{max}-T_{mean})L=\pm11.7\times10^{-6}\times30\times2.5=\pm0.878$ mm from its installed position. The joint sees both adjacent edges move together, so the total one-directional joint movement is: $$\Delta L_{joint}=2\times0.878=1.755\ \text{mm}\quad(\text{matches Step 1, as expected})$$
3. Part (2) — Minimum width, ±25% sealant. ASTM C1472: the design movement must not exceed the sealant's rated fraction of the installed width, $\Delta L_{joint}\le MC\times W$: $$W_{min}=\frac{\Delta L_{joint}}{MC}=\frac{1.755}{0.25}=7.02\ \text{mm}\ \Rightarrow\ \boxed{W_{min}\approx7\ \text{mm (25\% sealant)}}$$
4. Part (3) — Minimum width, ±50% sealant. $$W_{min}=\frac{1.755}{0.50}=3.51\ \text{mm}\ \Rightarrow\ \boxed{W_{min}\approx3.5\ \text{mm (50\% sealant)}}$$ A higher-movement-capacity sealant permits a proportionally narrower joint for the same design movement — in practice a designer would still round up to a constructible minimum (commonly ≥10 mm) to allow for tooling tolerance and to keep the sealant's width:depth aspect ratio in its effective range, which the bare thermal-movement calculation does not by itself enforce.

Part (4) — Vertical joint components. Figure 2 shows the standard anatomy: the panel edges are prepared with a clean, primed bond face; a closed-cell foam backer rod, sized slightly larger than the joint width, is compressed into the joint to control the sealant's bond depth and establish the correct hourglass profile; a bond-breaker tape is used instead where the joint is too shallow for a backer rod, to stop three-sided adhesion; the sealant is then tooled over the backer rod to a width:depth ratio of roughly 2:1 (never depth > width) so it can elongate and compress in a simple shear/tension pattern without over-straining at the bond line. The joint width $W$ must meet or exceed the Part (2)/(3) minimum for the specified sealant, and depth $D$ is set by the sealant manufacturer's width:depth table (typically $D\approx W/2$ for $W$ up to about 13 mm, tapering to a fixed minimum depth for wider joints).

Part (5) — Failure from a badly-proportioned joint. A joint that is too narrow for the actual thermal movement forces the sealant to stretch beyond its rated elastic movement capacity every cycle; the sealant either tears cohesively (splits through its own body) or fails adhesively (pulls away from the concrete bond face), both of which open a direct leak path. A joint that is too deep relative to its width (poor aspect ratio, sealant depth exceeding roughly half the width) develops a highly non-uniform, concentrated strain at the bond line when the joint opens — instead of the sealant's mid-section stretching uniformly (the intended hourglass shape), the corners at the substrate interface see the highest strain and tear first, again causing adhesive failure. This is exactly why a backer rod (or bond-breaker tape) is specified: it fixes the depth independently of the width and prevents three-sided adhesion, which would otherwise restrain the sealant on three faces and guarantee premature tearing under the very first thermal cycle.

Part (6) — Single-stage vs. two-stage joint. A single-stage joint relies on one exterior sealant bead (over backer rod) as the sole barrier to both bulk water and air/vapour leakage; it is fully exposed to UV, thermal cycling and weathering, so any degradation or loss of adhesion immediately becomes a full-depth leak with no redundancy. A two-stage joint splits the two jobs: an outer, vented rain seal (sometimes just a baffle/loose gasket, not required to be perfectly airtight) sheds the bulk of wind-driven rain, backed by a pressure-equalized, drained/vented air chamber, with the true, continuous air/vapour seal placed at the inner face, protected from UV and weather. Advantages of the two-stage design: (i) redundancy — if the outer rain seal degrades, the inner air seal still keeps the building airtight and largely dry; (ii) the critical inner seal is shielded from UV/thermal cycling and lasts far longer; (iii) pressure equalization in the vented cavity reduces the air-pressure differential across the outer seal, which is the main driving force for wind-driven rain penetration in the first place (the rainscreen principle); and (iv) the outer seal can be inspected and replaced periodically without disturbing the building's primary air barrier.