From design to shop drawings
The drawing that proves a facade can actually be made is not the architect's elevation - it's a thousand pages of profiles, gaskets and tolerances the contractor draws after award, and somebody has to review every one.
The architect drew one elegant line. The contractor will draw it again, eight hundred times, at the scale a machine can cut - and that is where a facade is really designed.
A facade arrives on the architect's drawings as a clean elevation - a grid of glass, a rhythm of mullions, a colour. That drawing is _intent_. It cannot be fabricated. Between it and the factory sits the least glamorous and most decisive document set in the whole project: the **shop drawings** - hundreds of sheets where every extrusion section, every gasket lip, every bracket slot and every 2 mm tolerance is finally pinned down. On a serious building in Mumbai or Bengaluru, the specialist facade contractor draws these after award, the facade consultant reviews them against the spec, and the difference between a skin that performs and one that leaks is decided in that review. This lesson follows the line from concept to a cuttable drawing.
Concept to scheme to spec to tender to shop drawing - a deepening, not a redrawing
Concept defines the look, the performance spec defines the numbers, the shop drawing defines the cut
A facade drawing does not appear all at once - it deepens through stages, and each stage answers a different question.
At concept the architect sets the look and the big moves: glazing ratio, grid, materials, the broad ambition. At scheme / developed design the facade consultant resolves the typical section and the key interfaces, and writes the performance specification - the numbers the facade must hit (air, water, wind, thermal, movement). That spec, plus tender drawings showing the design intent, becomes the tender package that specialist facade contractors price.
After award the won contractor produces the contractor's design and then the shop drawings - fabrication-level detailing of every profile, gasket, bracket, panel and fixing, dimensioned to the millimetre a CNC machine can cut. The crucial idea: the architect's drawing says what it should be, the spec says how well it must perform, and the shop drawing says exactly how it is made. Each is necessary; none can substitute for the next.
Concept = the look. Spec = the numbers. Shop drawing = the cut. You can't fabricate intent, and you can't tender a shop drawing.
The design-responsibility split: intent, system, fabrication and the famous grey zone between them
Facade responsibility is deliberately divided, and the division is the source of most facade disputes. A common model on Indian and international projects: the architect owns design intent and aesthetic approval; the facade consultant writes the performance spec, sets design intent details and verifies compliance; the specialist contractor takes on the contractor's design - the detailed engineering and fabrication of a system that meets the spec - and warrants it.
The friction lives at the boundary. A performance specification tells the contractor what to achieve (no water past a defined test pressure, deflection within span/175) and lets them design the system; a prescriptive specification tells them exactly what to build (this profile, this gasket) and keeps responsibility with the consultant. Most modern facade specs are performance-led precisely so the contractor's warranty has teeth - they designed it, they own it. But the consultant must still review the contractor's design, because 'the contractor warrants it' is cold comfort the day it leaks. Knowing which party owns which line on the drawing is half of facade procurement.
Reviewed, not approved - and stamped 'Good For Construction' only when it matches the spec
When the contractor's shop drawings land - often in batches, hundreds of sheets - the facade consultant runs the shop-drawing review: checking every detail against the performance spec, the design intent and the tested mock-up. The reviewer marks each sheet A (no comment), B (proceed with noted corrections) or C (revise and resubmit) - the classic three-status system. Note the careful language: the consultant reviews the drawing, they do not approve the contractor's design - the contractor retains design responsibility, and 'reviewed' wording protects the consultant from inheriting it.
Only when a drawing is clean does it become GFC - Good For Construction: the version the factory is allowed to cut from. Building from a non-GFC drawing is how a thousand wrong panels get made. The review is slow, adversarial and unglamorous, and it is where leaks, thermal bridges and clashes are caught for the price of a comment instead of the price of a recladding.
Your elevation is the start of the conversation, not the end of it. Resist the urge to over-specify the look in a way that fights buildability - a razor reveal or a flush joint may leave nowhere for a gasket or a bracket, and that bill arrives at shop-drawing stage as an RFI. Keep your design-intent drawings clear about what is _aesthetic-critical_ (sightline widths, joint shadows, colour) and what you are happy for the contractor to engineer. And give the shop-drawing review the time it needs: a rushed review is where your intent quietly gets value-engineered away one B-status comment at a time.
You write the performance spec and you run the shop-drawing review - the two documents that bracket the contractor's design. Make the spec testable (every clause a number a mock-up can prove) and make the responsibility boundary explicit in writing: state plainly that the contractor's design is theirs to warrant and yours to review, not approve. In review, check continuity of the four control layers across every detail and interface, flag thermal bridges at every bracket, and never let a C-status drawing slip to GFC under programme pressure. The grey zone between 'design intent' and 'contractor's design' is exactly where you earn your fee and avoid a lawsuit.
On site the only drawing that matters is the **GFC** one - Good For Construction. Before you let a panel be installed, check the sheet is the latest GFC revision, not a superseded tender drawing or an un-reviewed shop drawing. The revision cloud and the status stamp (A / B / C) on the title block are your safety check. Learn to read a shop drawing as the real thing: it shows the actual extrusion section, the gasket, the bracket and the slot tolerances you'll be setting out to. The gap between the architect's pretty elevation and this dense, dimensioned sheet is the gap between design and build.
CWCT Standard (UK)
Performance specification framework
The Centre for Window & Cladding Technology standard underpins most performance-led facade specs globally and on Indian premium projects - but it is a benchmark you cite, not a substitute for project-specific clauses; a generic CWCT reference without project numbers is not a spec.
NBC 2016 (India)
The umbrella building code
National Building Code Parts 6 (structure) and 11 (climate) frame how facade design intent is set in India and what state bye-laws reference - but it does not give you shop-drawing-level detail; that is the contractor's design to produce.
IS 16231 / fenestration standards
Indian fenestration performance
IS 16231 (and related window/door standards) set Indian performance baselines a spec can call up - useful for opening-light performance, but most curtain-wall specs still lean on CWCT/ASTM because the Indian facade-system standard base is thinner as of 2026.
“The architect designs the facade fully, and shop drawings are just the contractor tracing those drawings at a bigger scale.”
Shop drawings are not a tracing - they are where a huge amount of the real engineering happens. The architect's drawings carry intent and aesthetics; the performance spec carries the numbers; but the contractor's shop drawings resolve every profile, gasket, bracket, tolerance and fabrication sequence that makes the facade actually buildable and watertight. On most projects the specialist contractor carries genuine 'contractor's design' responsibility for that detailing - it is co-authored, not copied, which is exactly why it must be reviewed against the spec before it goes GFC.
Worked example - mark up a shop drawing against the spec
The defining facade-procurement skill is reviewing a shop drawing against a performance spec and writing a defensible comment. Let's run one review on a single unitized curtain-wall transom detail.
The performance-brief skeleton from Lesson 0.3, a printed or on-screen facade detail (any manufacturer's curtain-wall transom section works), a red pen or markup tool, and the A/B/C status convention.
You are reviewing one shop-drawing sheet. The spec says: WATER : no penetration to 600 Pa (static) AIR : <= 1.5 m3/hr.m2 at 300 Pa DEFLECT. : <= span/175 THERMAL : continuous insulation, no un-broken metal bridge STATUS : assign A (no comment) / B (proceed w/ corrections) / C (revise + resubmit) The transom detail shows: an outer pressure plate, a gasket, a drained cavity with one weep slot, and an aluminium bracket bolted straight through the insulation line.
- 1Trace the water line. Follow the drainage path: rain past the outer gasket must reach the weep slot and exit. Ask - is the cavity drained AND pressure-equalised, or is it a sealed trap? One weep slot per bay may be too few. Write the comment: 'Confirm weep slot size/spacing achieves pressure equalisation per spec 600 Pa - provide drainage calc.'
- 2Check the air line. The air barrier (here the inner gasket / pressure plate seal) must be continuous and meet the next panel's seal. Look for the splice. If the detail stops the air seal at the panel edge without showing the inter-panel joint, that is a real gap. Comment: 'Show air-seal continuity at panel-to-panel joint.'
- 3Find the thermal bridge. The aluminium bracket bolted straight through the insulation is an un-broken metal path from outside to inside - a thermal bridge that will cool the inner face and risk surface condensation in an AC building. Comment: 'Bracket penetrates thermal line uninterrupted - provide thermal break or isolator; confirm no condensation risk at design dewpoint.'
- 4Check deflection. The spec says transom deflection <= span/175. The shop drawing should cite the section's moment of inertia and the resulting deflection under design wind. If the drawing shows the profile but no deflection note, request it. Comment: 'Provide transom deflection check against span/175 at design wind pressure.'
- 5Assign the status. With four substantive comments - one of them (the thermal bridge) a performance failure, not a clarification - this sheet cannot go GFC. Mark it C - revise and resubmit. A drawing with only minor clarifications would be B; only a clean sheet is A.
- 6Log it. Record each comment in the review register with sheet number, revision and date, so the resubmission can be checked against your comments - the audit trail that protects everyone when a dispute arises.
You’ll walk away with
A marked-up shop-drawing detail with four spec-anchored comments and a defensible C-status decision - the exact deliverable a facade consultant produces dozens of times on a real project, and the moment design responsibility is verified before fabrication scales up.
Two quick ways to feel the design-to-fabrication gap.
- 01Find any architect's facade elevation and any manufacturer's curtain-wall shop-drawing detail online. Put them side by side and count the information that exists on the shop drawing but nowhere on the elevation - gaskets, tolerances, weeps, brackets. That delta is the contractor's design.
- 02Take one facade detail and try to mark whether each line is 'aesthetic-critical' (the architect cares) or 'performance-critical' (the engineer cares). The overlap and the gap are where coordination lives.
A facade is designed in deepening layers - the architect's intent, the consultant's performance spec, the contractor's fabrication-level shop drawings - and reviewed, not approved, at each handover. Responsibility is split by design: intent, system and fabrication are owned by different parties, and the shop-drawing review against the spec is where that split is verified before a drawing goes Good For Construction. Knowing who owns which line is half of facade procurement.
Concept (look) -> scheme/dev design + performance spec (numbers) -> tender (priced) -> contractor's design + shop drawings (the cut). Performance specs let the contractor design and warrant the system; prescriptive specs keep responsibility with the consultant. The shop-drawing review marks each sheet A/B/C; only a clean sheet becomes GFC. The consultant reviews, never approves - the contractor keeps design responsibility.
What are facade shop drawings?
Shop drawings are the fabrication-level drawings the specialist facade contractor produces after winning the job - hundreds of sheets detailing every extrusion section, gasket, bracket, panel and fixing, dimensioned to the millimetre the factory cuts to. They translate the architect's design intent and the consultant's performance specification into something a CNC machine and a fabrication line can actually make. They are reviewed against the performance spec before any panel is fabricated.
What does GFC (Good For Construction) mean on a facade drawing?
GFC means the drawing has passed the consultant's shop-drawing review against the performance spec and is the version the factory is allowed to fabricate from. Before GFC, drawings cycle through review statuses - A (no comment), B (proceed with noted corrections), C (revise and resubmit). Building from a non-GFC drawing is how large batches of wrong panels get made, so checking the GFC revision is a basic site control.
What is the difference between design intent and the contractor's design?
Design intent (owned by the architect and facade consultant) sets what the facade should look like and how well it must perform - the aesthetics and the performance specification. The contractor's design (owned by the specialist facade contractor) is the detailed engineering and shop-drawing detailing of a system that achieves that intent, which the contractor warrants. Performance-led specs deliberately push system design to the contractor so their warranty has teeth; the consultant then reviews, but does not approve, that design.
Peer-reviewed journals & authoritative standards
- 01Material Selection and Characterization for a Novel Frame-Integrated Curtain Wall. (PMC8069006). — Materials / NCBI-PMC, 2021.
- 02Su, Z. et al. Multi-Disciplinary Characteristics of Double-Skin Facades for Computational Modeling Perspective and Practical Design Considerations. Buildings, 12(10):1576. — Buildings (MDPI), 2022.
- 03Ventilated facade system: A review (system families, fabrication and practical design considerations). — ScienceDirect (Elsevier), 2025.
A reviewed shop drawing proves the facade can be made - but not that it performs. Before thousands of panels are fabricated, the design has to be physically proven on a test rig. That ritual, the mock-up, is next.