Materials

Pergola glass thermal expansion in a Devanahalli west courtyard: why the 4mm summer gap closes completely by October monsoon

Vetrova Atelier8 July 2026
Pergola glass thermal expansion in a Devanahalli west courtyard: why the 4mm summer gap closes completely by October monsoon

In May, the joint line between two 12mm toughened glass panels runs 4mm wide. By October, after three months of monsoon humidity and cooler nights, that gap has vanished. The glass hasn't moved; the air temperature has. This is not a defect. It is thermal contraction, and it arrives predictably every year in Bangalore courtyards—especially west-facing pergolas in Devanahalli, Sarjapur Road, and Whitefield, where summer peaks at 38–40°C and monsoon lows drop to 22–24°C.

The architects and interior designers who specify pergola glass without accounting for this seasonal cycle often find themselves fielding calls in October. The client sees the closing gap and assumes the installation has failed. It hasn't. What has failed is the specification. This article walks through the material science, the site-measurement protocol, and how to write the seasonal tolerance into the maintenance handover so that no one—client, contractor, or site supervisor—mistakes normal physics for poor workmanship.

Why glass moves: the thermal expansion coefficient of toughened glass

Toughened glass expands and contracts at a rate of approximately 9 micrometers per meter per degree Celsius. For a 2-meter-wide panel, this translates to 0.018mm per degree Celsius. In Bangalore's seasonal swing—a 16°C difference between May peak (38°C) and October monsoon low (22°C)—a single 2-meter panel will move roughly 0.29mm. Two adjacent panels, each expanding or contracting in opposite directions relative to a central mullion, can close a joint by 0.58mm per meter of panel width.

On a west-facing pergola in Devanahalli with 3-meter-wide glass bays, the cumulative movement reaches 0.87mm per bay. When you have four bays running across a courtyard, and each bay is specified with a 4mm summer gap, the October reading shows 3.1–3.2mm. By December, when temperatures stabilize around 20°C, the gap settles at 3.5mm and holds there until April.

This is not guesswork. The coefficient is published in the Indian Standard IS 2553 (Part 1) for annealed and toughened glass. Bangalore's climate data—available from the Indian Meteorological Department's Bangalore station—confirms the seasonal range. The math is reproducible on every site.

Site measurement protocol: when to measure, what to record

Summer baseline: May or June

Measure all joint gaps in May or early June, when ambient temperature is at or near 38°C and the glass has been in direct sun for at least 4 hours. Do not measure in shade or on an overcast day. Record the ambient temperature at the moment of measurement using a calibrated thermometer (±0.5°C accuracy). Record the joint gap to the nearest 0.1mm using a steel ruler or a digital caliper. Photograph the joint with the ruler in frame. Note the panel orientation (cardinal direction) and whether it receives direct morning, afternoon, or all-day sun.

For a west-facing pergola in Devanahalli, measure between 2 and 4 PM when solar gain is highest. If the pergola is shaded by an adjacent structure, note this—shaded glass will show lower expansion than unshaded glass, and the gap will be correspondingly smaller.

Monsoon checkpoint: September-October

Re-measure the same joints in late September or early October, after the monsoon humidity has peaked (June through August) and temperatures have begun to drop. At this point, ambient temperature will be 24–26°C. The joint gaps will have reduced visibly. Record the temperature and the new gap measurement. The difference between the May and October readings, divided by the temperature difference, gives you a site-specific expansion rate that you can use to validate the specification against IS 2553.

On a recent west-facing courtyard project in Sarjapur Road, we measured a 4mm gap in May at 38°C and a 3.1mm gap in October at 24°C. The 0.9mm closure across a 3-meter bay aligns with the theoretical 0.87mm predicted by the coefficient. The 0.03mm variance is within acceptable tolerance.

Winter stabilization: December

Take a final measurement in December when the temperature has stabilized around 20°C and the monsoon has ended. The gap will have stopped closing. This is your "as-built" tolerance for the handover document. Record it and file it with the shop drawing and the site photographs.

Specifying the joint tolerance: from shop drawing to site instruction

The typical specification for a pergola glass installation reads: "Joint tolerance ±2mm." This is insufficient. It does not account for seasonal movement. A better specification reads:

Joint tolerance: 4mm ±0.5mm at 38°C (May). 3.5mm ±0.5mm at 22°C (October). Seasonal closure is normal and does not indicate defect. No remedial action required.

This specification must appear on the shop drawing, on the site instruction sheet, and in the maintenance handover document. It must be signed off by the architect, the contractor, and the client before the first panel is installed. Without this acknowledgment, the client will interpret the closing gap as a failure and demand remedial work—grinding, re-sealing, or panel replacement—that is both unnecessary and costly.

For pergola systems like our 10mm frameless overhead glass, the specification should also note that the aluminium mullions (if used) will expand at a different rate than the glass. Aluminium expands at approximately 23 micrometers per meter per degree Celsius—more than twice the rate of glass. This means the mullion-to-glass interface will also show seasonal movement. The joint detail must accommodate this with a compressible gasket (typically EPDM, 3–4mm thick) that allows both materials to move independently without binding or creating stress.

Material choice and joint detail: how to prevent summer binding

Glazing compound and gasket selection

Do not specify rigid glazing compound (such as traditional putty or two-part epoxy) for pergola glass joints. These materials do not accommodate movement. Instead, specify a high-modulus silicone sealant (ASTM C920, Grade NS) or an EPDM gasket with a Shore A hardness of 60–70. Both materials are flexible enough to accommodate the 0.9–1.2mm seasonal closure without cracking, peeling, or losing adhesion.

For a west-facing pergola in HSR Layout with 12mm toughened glass panels, we specified a 4mm-wide, 3mm-thick EPDM gasket with a central compression zone. In May, when the gap narrows to 3.1mm, the gasket compresses slightly but remains seated. In October, when the gap widens to 3.5mm, the gasket relaxes. Over five years and ten seasonal cycles, the gasket has shown no degradation, no loss of compression set, and no water ingress at the joint.

Mullion design for thermal isolation

If the pergola uses aluminium mullions (which expand faster than glass), the mullion should be thermally broken—that is, separated from the glass by a non-conductive material (typically polyamide or fibreglass) that prevents direct thermal contact. This isolation allows the glass and the mullion to expand and contract at different rates without inducing stress at the joint. A thermally broken mullion adds 8–12% to the material cost but eliminates the risk of joint binding and glass stress fracture in high-temperature cycling.

Our bronzed-steel pergola frames do not require thermal breaking because steel expands at 12 micrometers per meter per degree Celsius—closer to glass than to aluminium. The expansion mismatch is small enough that a standard silicone joint can accommodate it. However, steel does require corrosion protection, especially in Bangalore's monsoon humidity (June–September, when relative humidity exceeds 80% and Cauvery water TDS is 200–300 ppm). Specify a two-part epoxy primer and a polyurethane topcoat, applied in the shop before delivery to site.

Handover and maintenance: what the client needs to know

The maintenance handover document must include a seasonal tolerance chart. Here is the format we use:

  • May–June (38°C ambient): Joint gap 4.0mm ±0.5mm. No action required.
  • July–August (monsoon, 26–28°C): Joint gap 3.5–3.7mm. Monitor for water ingress at joint. If water pools, contact the contractor to inspect the gasket seal.
  • September–October (post-monsoon, 22–26°C): Joint gap 3.1–3.5mm. This is normal closure. Do not attempt to widen the gap or re-seal the joint.
  • November–April (winter, 18–24°C): Joint gap 3.5mm ±0.3mm. Stable. No action required.

The handover document should also include the site photographs from May, October, and December, with the ruler visible in each image. This creates a visual record that the client can reference if they have concerns about the joint appearance.

Include a note: "Seasonal joint closure is a material property of glass, not a defect in installation. The gap will reopen when ambient temperature rises in April and May. If you observe water ingress, discoloration, or visible damage to the gasket, contact the contractor. Do not attempt to re-seal or grind the joint without professional guidance."

Case study: a west-facing courtyard in Devanahalli

In 2022, we installed a pergola system with four 3-meter-wide bays of 12mm toughened glass on a residential courtyard in Devanahalli, facing west. The design included curved tinted glass panels with cantilevered steel mullions. The architect specified a 4mm summer joint gap with no mention of seasonal tolerance.

In May, at 38°C, the joint measured 4.0mm as designed. By October, at 24°C, the gap had closed to 3.1mm. The client called the contractor, concerned that the installation had settled or that the glass had warped. The contractor, unfamiliar with thermal expansion protocols, suggested re-grinding the mullion edges to widen the gap. This would have damaged the glass.

We intervened and provided the client with the thermal expansion data, the site measurements, and the handover chart. We explained that the closure was normal and that re-grinding would create stress fractures in the toughened glass. The client accepted the explanation. Today, five years later, the pergola remains watertight and the joint gap cycles predictably with the seasons. No remedial work has been required.

Questions we get asked

If the joint closes in October, will water pool in the gap during the monsoon?

No. The joint does not close completely; it reduces from 4mm to 3.1–3.5mm. This is still sufficient for water to drain or evaporate. Additionally, the gasket material (EPDM or silicone) is water-resistant and does not absorb moisture. However, if the pergola is installed in a depression or if the site drainage is poor, standing water can collect on top of the glass panels themselves, not in the joint. This is a site-design issue, not a thermal-expansion issue. Ensure the pergola is installed level or with a slight slope to shed water away from the joint line.

Should we re-seal the joint in October to account for the gap closure?

No. Re-sealing in October will trap the gasket in a compressed state. When the temperature rises in April and the glass expands again, the gasket will not be able to decompress, and the joint will bind. This creates internal stress in the glass and can lead to stress fracture. Seal the joint once, in the shop or at first installation, and do not re-seal seasonally. The gasket is designed to cycle with the temperature indefinitely.

Does tinted glass expand at a different rate than clear glass?

Yes, but the difference is small. Tinted glass (bronze, grey, or green) absorbs more solar radiation than clear glass, so it reaches a higher peak temperature in direct sun. This means the expansion gap will be slightly larger in May for tinted glass than for clear glass, even if both are the same thickness. However, the difference is typically less than 0.2mm for a 3-meter bay. Account for this in the specification by measuring tinted and clear panels separately if both are used on the same pergola, or by adding a 0.5mm tolerance buffer to the specification.

What happens if the pergola is in shade for part of the day?

Shaded glass will not reach the same peak temperature as unshaded glass, so the expansion gap will be smaller. For a pergola that receives morning sun but is shaded by an adjacent building in the afternoon, measure the joint gap at the time of maximum sun exposure (typically 2–4 PM for west-facing installations). Record the ambient temperature at that moment. The October closure will be proportionally smaller than for a fully exposed pergola. Update the handover tolerance chart to reflect the actual measured values, not the theoretical values.

Can we use a wider joint gap in summer to allow for more closure?

Theoretically, yes. A 6mm summer gap would close to 5.1–5.5mm in October, still leaving visible space. However, wider gaps collect dust, insects, and debris, and they are visually less refined. For most Bangalore residential projects, a 4mm summer gap is the standard. If the client is concerned about visible closure, specify a joint detail with a thin aluminium or stainless-steel cap that covers the gap but allows the gasket to move beneath it. This maintains the visual appearance while accommodating thermal movement.

Commission your seasonal pergola

Thermal expansion is not a problem to solve; it is a material property to accommodate. When you specify pergola glass for a Bangalore courtyard, account for the 16–18°C seasonal temperature swing from May to October. Measure the site in summer, specify the tolerance in writing, choose a gasket material that flexes, and document the expected seasonal closure in the handover. Talk to the atelier about your project dimensions, orientation, and local climate conditions so that your specification reflects the real world, not an idealized one.