Grids, Spans & the Timber Logic

The first thing to internalise is that mass-timber elements have a comfortable span range, and staying inside it is what keeps a building efficient. Floor panels - typically cross-laminated timber (CLT) or similar - span one way between supports, and like any spanning element they get deeper (and heavier, and dearer) as the span grows. There is a sweet spot where a standard, sensibly thick panel spans economically; push much beyond it and the panel must get thick and heavy, or you must add a structural topping, or you switch to a ribbed or cassette system - all of which cost more. Beams - usually glulam or LVL - have their own comfortable range, longer than the panels because a beam is a concentrated, deep element, but still bounded before the beam becomes uneconomically deep. Columns are rarely the limit; the spans between them are.

So a timber floor is really a conversation between two spans: the panel span (support to support, the shorter, closely-spaced direction) and the beam span (the longer bay the beams cross, carrying the panels). A common and efficient arrangement is one-way CLT panels spanning a modest distance onto glulam beams, which span a longer distance onto columns - a clear, repeating hierarchy. Get those two spans into their comfortable ranges and the whole floor is light and cheap; let either run long and the members bloat.

Here is the honest boundary: the actual economical spans depend on the product, its grade and thickness, the loads, the fire requirement and deflection and vibration limits - all of which the engineer determines against the code. A panel that spans happily in an office may be marginal under a heavy assembly load or a long-term deflection limit; vibration, not strength, often governs timber floors. Treat any span figure you see - including rules of thumb in manufacturers' literature - as illustrative of the principle, not a design value. Your job as the designer is to know that timber has a comfortable range and to draw a grid that respects it, then let the engineer confirm the numbers. Design within the range and timber rewards you; design as if timber were concrete with no span limits and it will quietly punish the budget.

If span sets the size of a timber building's rhythm, repetition sets its economy. Mass timber is a prefabricated, manufactured product: every panel and beam is cut, drilled and finished in a factory to a precise digital model, then trucked and craned into place. That production model rewards sameness. A grid of identical, repeating bays means identical, repeating elements - the same panel size made again and again, the same beam, the same connection detail - which is cheaper to fabricate, faster to erect, easier to coordinate and far less wasteful than a grid of many different bespoke pieces. The discipline is close to industrial design: you are designing a kit of parts, and fewer, more-repeated parts is a better kit.

This changes how you think about the plan. In a concrete building an irregular column grid costs relatively little, because concrete is poured to fit. In a timber building, irregularity is expensive, because every odd bay is a bespoke element, a bespoke connection and a bespoke coordination problem. So the timber-literate designer tunes the plan toward a regular, rationalised grid - lining columns up, keeping bays consistent, letting the same element repeat down a floor and up the building - and treats each departure from that grid as a deliberate, costed decision rather than an accident. Where the brief genuinely needs a big clear span or an irregular space, you concentrate the special elements there and keep everything else regular, so the exceptions read as exceptions.

Repetition also compounds vertically. If the grid stacks - columns landing on columns, walls landing on walls, all the way down to the foundation - loads travel cleanly and the same details repeat floor to floor. Transfer structures, where load has to jump sideways because something below moved, are costly and heavy in any material and especially so in timber. So the timber logic is: line things up in plan, stack them in section, and let the same few elements repeat as many times as possible. None of this is a constraint on architecture so much as a different craft of it - the elegance of a well-tuned timber grid, legible and rhythmic, is one of the quiet pleasures of the material, and it is designed in at concept, not rescued later.

Putting span and repetition together gives the central move of designing with mass timber: let the material shape the plan. In practice this means drawing the structural grid early, in dialogue with the timber's comfortable spans and standard sizes, and letting that grid inform the architecture - rather than drawing a free plan and hoping timber can deliver it. It is a reversal of a habit many designers have picked up from concrete, where structure is often treated as infinitely accommodating and resolved late. In timber, structure is the architecture, and it is resolved early.

Concretely, at concept stage you sketch a grid whose bays sit in timber's efficient range, whose columns line up and stack, and whose panels run one clear way; you locate the spaces that need to be column-free (the hall, the large room) and decide how to give them a longer span deliberately - a deeper glulam, a truss, a hybrid element - rather than letting long spans creep in everywhere. You let corridors, service zones and repetitive rooms fall naturally onto the regular grid, and you save the structural drama for where it earns its place. Done well, the plan and the structure are the same idea, and the building is both efficient and honestly expressive of how it stands up - which, with timber usually exposed, the occupants will actually see.

This is also where the architect and engineer collaborate earliest and most productively. You bring the spatial and architectural intent and a first, timber-literate grid; the engineer brings the real spans, sizes, load paths and code limits, and together you converge on a grid that is both good architecture and good structure. Because so much of a timber project's cost, carbon and programme is locked in by the grid, this early conversation is worth more than almost any later optimisation. Get the grid right - drawn around the material - and the rest of the design flows; get it wrong, and you spend the project fighting the consequences. The judgement is architectural and yours; the confirming numbers are the engineer's; the grid is where the two meet, and it is the heart of designing with mass timber.

A grid does not exist in the abstract - it meets the market it is built in, and in an emerging timber market like India that matters even more. Where a mature market has local CLT and glulam factories making a range of standard sizes, a designer can tune a grid to those catalogues confidently. Where the supply chain is nascent and much material is imported - India's situation today for larger mass-timber elements - the available panel and beam sizes may be narrower, the lead times longer, and the cost premium real, all of which push even harder toward a simple, regular, highly repetitive grid that uses standard sizes and avoids anything bespoke or oversized. The timber logic is universal; in a constrained market it is not optional.

There are other real-world pulls on the grid. Transport limits panel length (the truck and the road; Lesson 4.4 covers this), so a grid that needs very long single panels may be undeliverable and must be split, which affects joints and the plan. Craneage and site access reward a grid that can be erected in a sensible sequence with elements a crane can place. Services - the ducts, pipes and cable trays a building needs - must be planned into or under the timber floor zone from the start, because you cannot freely drill a structural CLT panel later; the grid and the services strategy are designed together. And fire and acoustic requirements may set minimum panel thicknesses or added layers that change the efficient span, which again the engineer and fire engineer confirm.

The honest summary is that a good timber grid is a negotiated thing: architecturally intentional, tuned to the material's spans and standard sizes, disciplined toward repetition and stacking, and shaped by transport, craneage, services, fire, acoustics and - especially in India - by what the supply chain can actually deliver at a sensible cost. None of these are reasons to avoid timber; they are the design parameters of doing it well. Bring them in early, draw the grid around them with your engineer, and you get the efficient, elegant, buildable timber structure the material is capable of - rather than a concrete plan translated into expensive wood.