Laying Out a Storefront

Turning a Width Into a Count

The division itself is deliberately blunt: take the surveyed opening width, divide by the limit you settled on, round up to a whole number of lites, then subtract one. That subtraction is the part people query. An opening cut into five lites needs four interior mullions, because its two outermost edges are held by the jambs and the jambs are already sitting in the track order. Counting them twice is the most reliable way to finish a shopfront with a spare stick on the floor and a bay that does not close.

Rounding up is what keeps the answer honest. A twelve metre run at a 1.5 metre limit divides exactly into eight lites. The same run at a 1.6 metre limit divides into 7.5, which becomes eight lites of 1.50 rather than seven of 1.71 — and seven would have broken the limit you spent an afternoon establishing. Round down anywhere in this calculation and the error surfaces either as a rejected batch at the fabricator or, worse, as an accepted batch that sits outside the plant's own envelope and gets made anyway.

Two things the count deliberately does not include. It divides one straight opening, so a corner mullion, a transition into a return elevation or a door jamb inside the run has to be added by hand from the layout. A pair of entrance doors in the middle of a shopfront introduces jambs that are neither interior mullions nor the opening's own jambs, and they are usually a different section entirely. And it divides the OPENING, not the glass. Each lite is smaller than its share of that opening by the frame face on both sides and by the edge clearance the system requires, so what the division produces is a setting-out dimension and not a cutting size.

Keep those two dimensions in separate columns from the first sheet you write. The lite width is what gets marked on the head and sill; the glass size is what the fabricator cuts to; the difference between them comes from the system manufacturer's published glazing details and from the edge cover and clearance conventions set out in the NGA Glazing Manual, never from a judgement made on site with a tape.

The Member You Just Committed To Has to Be Checked

Fewer mullions is an easy thing to sell and every one you delete loads the ones left behind. A mullion carries half the width of the lite on each side of it, so widening the bays widens that tributary strip and raises the line load along the member for the same design wind pressure. The section quoted at 1.2 metre bays is not automatically the section for 1.8 metre bays, and the quotation rarely says which spacing it assumed.

Deflection decides this long before stress does. An aluminium mullion can sit well below its strength limit and still bend far enough to break an edge seal or lose bite on the glass it is supposed to be holding. The convention the industry works to is set out in AAMA TIR-A11 Maximum Allowable Deflection of Framing Systems for Building Cladding Components at Design Wind Loads: within roughly 4.1 metres of span the recommendation is the span over 175, and above that it tightens into a span-over-240 form with an absolute cap. Your project specification may be stricter, and where it is, the specification is what you build to.

The other half of that check is what the structure above intends to do. A storefront installed tight under a slab edge or a transfer beam receives that structure's live-load deflection straight into the head, and there is no glass make-up that survives being used as a prop. A head receptor — a deeper channel that lets the head track move vertically inside it — is the standard answer, and it belongs in the layout rather than in a remedial visit. Get the anticipated movement from the structural engineer before the head is ordered, because a receptor changes the head profile, which changes the daylight opening, which changes every glass height in the run.

What One Lite Weighs Coming Off the Van

Once the division is fixed each lite becomes a mass, and that number is worth producing while the order is still open rather than when the vehicle arrives. Soda-lime float runs at about 2,500 kilograms per cubic metre, so it is area multiplied by the total glass thickness multiplied by that density — total glass thickness meaning the plies only, with an insulating unit's cavity and a laminate's interlayer excluded from the glass volume even though the nominal figure on the schedule includes them.

The number moves quickly with size, which is the whole point of working it out at the layout stage rather than on delivery day. Going from a 1.2 metre lite to a 1.8 metre lite at the same height and the same make-up is half as much glass again in one piece, and something that was a two-person lift becomes a lifter, a frame, a spotter and a different day on the programme. That is a layout decision wearing a logistics costume, and it is reversible right up until the glass is cut.

How the resulting mass gets handled is not a number you can look up. In the United Kingdom the assessment behind it is required by the Manual Handling Operations Regulations 1992, which publish no threshold that makes a lift acceptable — any figure quoted as though they do has come from somewhere else entirely, and the assessment still has to be done for the lite, the route and the crew in front of you.

Weight also settles two things people forget to ask. It is what the setting blocks and the sill are being asked to carry for the life of the shopfront, and it is what decides whether a lite can stand against the hoarding overnight or needs a rack and a strap. Work it out per lite type and put it on the same sheet as the sizes, so nobody has to reconstruct it from the schedule at seven in the morning.

Head, Sill and Two Jambs, and Whatever the Corner Adds

Track is the simplest quantity on the job and the one most often short, because it gets estimated off the elevation instead of measured round the perimeter. The framework of a single opening is four runs — head, sill and a jamb at each end. Add those, add every corner and transition condition, add the door frames where they are part of the same system, and what you have is a perimeter run before any waste is applied.

Interior mullions are not inside that perimeter figure. They are a separate line, priced as the count you established multiplied by their cut length, and they are very often not the same extrusion as the jamb even though both look like a vertical stick from the pavement. Ordering them as one item is how a run arrives with the right total length in the wrong profile.

Waste on aluminium framework is cut waste, not breakage, and it scales with the number of ends rather than the length of the run. Four square openings in a straight wall lose very little. A run with two internal corners, a return, a stepped head and a pair of door jambs loses a great deal more, because every mitre consumes the offcut behind it. Five per cent is what a plain layout carries; a complicated one carries more, and the honest way to pick the figure is to look at your own cutting list rather than at a default in a box.

Mill length is the other half of the same question. Extrusion arrives in whatever stock length the system supplier holds, and that length decides where the splices land in a long head run. A twelve metre head is not one piece of aluminium. Find out the stock length before setting out, because a splice sitting a hundred millimetres away from a mullion is a detail nobody wants to glaze around, and moving it afterwards means recutting the run.

Perimeter anchorage rides along with the track quantity because it is counted off the same run. Fixing centres come from the system manufacturer's published installation instructions and from what the substrate actually is — the schedule for a concrete upstand is not the schedule for blockwork or for a timber sub-frame — and AAMA CW-DG-1 Aluminum Curtain Wall Design Guide Manual is the reference behind most of the framing conventions a storefront borrows from its taller relatives.

When the framework lands, check the delivered profile against the one the deflection check was run on. Systems are sold in families where the face width is identical and the depth is not, and the shallow member in a family is the cheap one that quietly gets substituted whenever the deep one is on lead time.

The Sill Is a Gutter That Happens to Hold Glass

Every storefront sill is wet at some point in its life. Water gets past a gasket, condensation runs down the inboard face of the glass, and wind drives rain into the joint between two sticks. The system is designed on the assumption that all of this happens, which is why the sill is a drained chamber with weeps to the outside rather than a sealed box, and why treating it as a sealed box is the fastest way to make a shopfront leak.

Two things ruin it and both happen during installation. The first is a bead of sealant run along the outside of the sill to tidy the joint, which closes the weeps and converts a drained chamber into a reservoir. The second is an end dam that was never formed, so everything the sill collects runs sideways out of the open end of the extrusion into the wall construction — a failure that presents as damp inside a stud wall, or as an efflorescence stain on brickwork a metre away from the shopfront, which is precisely why it gets blamed on something else for two seasons.

The sub-sill flashing under the framework catches what the sill hands on. It runs the full length of the framework, laps in the shedding direction with whatever weather barrier the wall carries, turns up at both ends to form dams, and is sealed to the substrate before any track is set down on it. The lapping and sequencing follow the installation practice in ASTM E2112. It is measured as a length with an allowance for laps and end formation rather than as a sheet, and it is simultaneously the cheapest item on this list to buy and the most expensive one to add later.

Where the finished job has to be proved, water penetration is measured to ASTM E1105 in the field or ASTM E331 in the laboratory, both under a static pressure difference across the assembly. Neither test has ever been kind to a sill whose weeps were sealed by somebody tidying up.