ENGINEERING

A general arrangement drawing tells the builder what goes where. A detail drawing tells them how it actually goes together. Most of the questions raised on site, and most of the variations that follow, trace back to a junction that was shown at 1:100 when it needed to be shown at 1:5. This guide covers what construction detail drawings are, what a good one carries, and how they connect to the rest of the drawing set.

What a detail drawing is

A detail drawing is an enlarged, isolated view of one part of a building or structure, drawn at a scale large enough to show every component, its size, its material and how it is fixed to its neighbours. Where a plan or elevation shows a wall as two lines, the detail shows the cladding, the battens, the membrane, the insulation, the framing, the fixings and the flashing that closes the junction against the weather.

It is not a separate design. It is the same design viewed closely enough that nothing is left for the builder to interpret. That is the whole test of a detail: could a competent tradesperson build this junction from the drawing without phoning anyone?

Where detail drawings sit in the drawing set

A construction set works from the general to the particular. Site and floor plans establish layout. Elevations and sections establish heights and vertical relationships. Details then resolve the junctions those views cannot show at their scale. If you are new to reading a full set, our guide to how to read construction drawings walks through the order the sheets are normally used in.

The link between the two levels is the detail callout: a symbol on the plan, section or elevation that identifies the detail by number and names the sheet it is drawn on. A callout reading 4 over A-501, for example, means detail 4 on sheet A-501. The detail, in turn, carries a title that references back to the view it was called from. A reader should be able to move from any callout to its detail and back again without guessing. Broken or circular references are one of the most common defects in issued sets, and they are almost always caused by renumbering sheets late in the project without updating the callouts.

Typical scales for construction details

The right scale is the smallest one at which every component remains legible with its annotation. In practice that gives a fairly settled range:

• 1:20 for wall sections and larger assemblies where the overall build-up matters more than individual fixings.
• 1:10 for most junctions: slab edges, window heads and sills, roof-to-wall connections, balustrade fixings.
• 1:5 for small or layered junctions such as flashings, membrane terminations and joinery.
• 1:2 or full size for very small profiles, custom extrusions or connection plates where hole positions and edge distances must be read directly.

Scales should come from the preferred set used in Australian technical drawing practice, and every detail should state its scale beneath the title. A detail without a stated scale invites someone to scale it off a reduced print, which is exactly what a dimensioned detail is meant to prevent. The broader conventions for line types, lettering and scale are covered in our article on technical drawing standards.

What a good detail drawing carries

Whether the junction is architectural, structural or mechanical, a detail that can be built from without questions usually carries the same things:

• Every layer, named. Each material in the assembly is labelled by product type and, where it matters, thickness or grade. “Membrane” is not enough if two membranes are specified elsewhere on the job.
• Dimensions that locate, not just describe. Set-outs from a gridline, a structural face or a finished floor level, so the junction can be positioned, not merely drawn.
• Fixings and their spacing. Fastener type, size and centres, or a clear reference to the specification clause that governs them.
• Falls and drainage direction on anything exposed to water, with the arrow pointing the way water actually leaves.
• Hatching that distinguishes materials consistently with the rest of the set, so concrete, masonry, insulation and timber read differently at a glance.
• Notes only where geometry cannot carry the information. A note that contradicts the drawn geometry is worse than no note at all.

The junctions that most need detailing

Not every junction earns its own detail. The ones that consistently do are the places where trades, materials or loads change over:

• Weatherproofing transitions such as window and door heads, sills and jambs, roof-to-wall flashings, parapet cappings and penetrations through the roof.
• Slab edges and set-downs at wet areas, balconies and thresholds, where finished levels, waterproofing and step-downs all have to agree.
• Structural connections, such as beam-to-column, column base plates and bracing connections, where the load path has to be explicit. For steelwork these are normally resolved in the shop drawings; see our explainer on structural steel detailing.
• Fire and acoustic separations, where a wall or floor must stay continuous through a junction and the detail is often the evidence that it does.
• Balustrades and handrails, where fixing into the substrate governs whether the barrier will actually perform.

Many of these junctions also carry compliance obligations under the National Construction Code, particularly for weatherproofing, fire separation and barriers. The detail is often where a certifier looks to confirm that what was specified in words is actually achievable in the built form.

Detail drawings, shop drawings and standard details

Three terms get used interchangeably and should not be.

Construction detail drawings are part of the design documentation. They show the designer’s intent for a junction and are issued for tender and construction.

Shop drawings are produced by or for the fabricator, after the design is issued. They translate the design detail into what the workshop actually makes: cut lengths, hole positions, weld preparations, part marks. A shop drawing is checked against the construction detail, not the other way round.

Standard details are generic junctions reused across projects, often supplied by manufacturers. They are useful starting points but they are rarely right as issued, because they assume conditions that a specific building may not have. A standard detail dropped into a set without being adapted to the actual adjoining construction is one of the most reliable sources of site RFIs.

Common mistakes on detail sheets

• Details that disagree with the sections. The section shows a 90 mm stud, the detail a 70 mm one. Usually one was revised and the other was not.
• Orphaned details. A detail on the sheet that nothing calls up, or a callout pointing at a detail that was deleted.
• Over-scaled generic content. A manufacturer’s detail pasted at 1:5 that shows a wall build-up this building does not have.
• Missing revision clouds. A detail changed at revision C with nothing to tell the builder which junction moved.
• Annotation that hides the geometry. Leaders crossing the very line they describe, or text sitting over the hatch it is meant to name.

Most of these are coordination failures rather than drafting skill failures, and they are cheapest to catch before issue. A structured check that walks every callout to its detail, and every detail back to the views it serves, removes the majority of them. Our article on structural detailing in design covers the equivalent discipline for structural packages.

Getting detail drawings right

Detail drawings are where a design either becomes buildable or becomes a list of site questions. They deserve the largest scale the sheet allows, a callout system that a reader can follow in both directions, and a final check against the sections they serve. Current Australian building requirements are published by the Australian Building Codes Board, and drawing practice standards by Standards Australia.

If you have a set that needs its junctions resolved or its callouts checked before issue, talk to our drafting team.

Every steel column eventually hands its load to concrete. The steel side of that handover is well covered — AS 4100 governs the base plate itself, the yield line that sets its thickness, and the welds back to the column. What gets far less attention on the drawing board is the other half: the concrete underneath, which is an AS 3600 problem. Bearing is where the two standards meet, and in our experience it is where base plate details most often come unstuck.

This is a detailer’s guide to what the concrete has to do under a base plate, and what the drawing has to show so that it can.

What the concrete under a base plate actually has to do

The base plate spreads a concentrated column load over an area the concrete can tolerate. Steel will happily carry stresses an order of magnitude beyond what concrete will, so the plate exists to convert a small, very high stress into a larger, much lower one. The concrete’s job is to accept that spread load without crushing locally or splitting the pedestal apart.

Two failure modes matter. The first is local crushing directly beneath the plate. The second — less obvious and more dangerous — is splitting: the bearing load pushes outwards inside the concrete, and if there is nothing to restrain that lateral bursting, the pedestal can split before the surface ever crushes. That second mode is why reinforcement detailing under a base plate is not optional dressing.

Why bearing capacity depends on what surrounds the plate

The single most misunderstood point is that concrete bearing strength is not a fixed material property you can look up once and reuse. It depends on confinement — how much concrete surrounds the loaded area.

A plate bearing on a large concrete raft is confined on every side: the surrounding concrete restrains the bursting forces and the permissible bearing stress rises accordingly. The same plate sitting on a pedestal barely larger than itself has almost no confinement, and its permissible bearing stress drops towards the plain compressive case. The governing relationship is between the loaded area and the larger supporting area geometrically similar to it — which is a drawing question as much as a calculation one.

The practical consequence for detailers: a pedestal that gets trimmed on site, or shown undersized on a drawing, can quietly remove the confinement the design assumed. Nothing about the plate changes, and nothing looks wrong, but the capacity the engineer relied on is no longer there.

Edge distance, pedestal size and the detail that gets value-engineered

Because confinement comes from surrounding concrete, edge distance is structural, not cosmetic. A column placed near the face of a blade wall or at the corner of a pedestal has confinement on fewer sides than one placed centrally, and the bearing check has to reflect that reality rather than the idealised case.

This is also the detail most likely to be eroded late in a project. Pedestals get narrowed to clear a service run, a column shifts to suit a setout, or a footing is reduced to save concrete. Each of those is a bearing change, and each of them should go back to the engineer rather than being absorbed silently into the shop drawings.

Grout, levelling and the load path you cannot see

Between plate and concrete sits the grout, and it carries the entire load. The assumption behind every bearing calculation is full, continuous contact. Anything that breaks that contact — voids, incomplete packing, a grout that shrinks away from the plate — concentrates load onto whatever is still touching.

Levelling method matters here. Where the plate is set on levelling nuts or shims, the column’s weight and early construction loads can sit on a handful of small contact points until the grout is placed and cured. That temporary condition is a genuine bearing case on a very small area, and it is one the drawing should acknowledge rather than leave to chance.

Non-shrink grout, a grout hole where the plate is large enough to trap air, and a specified bedding thickness are the three things that most reliably convert the designer’s assumption into site reality.

What the drawing has to carry

A base plate detail that respects the concrete side will show:

• Pedestal or footing plan dimensions — not just the plate, so the supporting area is unambiguous
• Edge distances from the plate to each concrete face
• Bedding or grout thickness, with the grout type specified as non-shrink
• The levelling method, and a grout hole where plate size warrants it
• Anchor embedment and the reinforcement that restrains bursting, including ties beneath the plate
• Concrete strength the bearing check assumed, called up on the detail rather than left to the general notes

That last one is worth insisting on. Concrete strength frequently varies between elements on the same project, and a base plate detail that does not state the strength it relies on is an invitation to pour the wrong mix into the one element where it matters most.

Where these details go wrong most often

Four patterns account for most of what we pick up in drawing review:

• The pedestal is not on the steel drawing at all. The plate is detailed perfectly and the concrete it bears on appears only in the concrete package, so nobody checks the two against each other.
• A standard detail is reused across different conditions. The same base plate detail appears at a central column and at a slab edge, where confinement is completely different.
• Bursting reinforcement is omitted or shown generically. Ties under the plate get drawn as typical when the splitting case is what actually governs.
• Grout is left as a note rather than a dimension. “Grout to suit” is not a bedding thickness, and it is not a load path.

None of these are calculation errors. They are drawing errors — which means they are the detailer’s to prevent, and they are cheap to prevent at the right moment.

Reading the two standards together

The cleanest way to approach a base plate is to treat it as two checks that have to agree. The steel side sets plate thickness, weld sizes and anchor design. The concrete side sets what the pedestal has to be, how it is reinforced, and what bearing stress is permissible given the confinement actually available. When the two are developed by different people at different times — which is normal — the drawing is the only place they meet.

Current editions of both standards are published by Standards Australia, and detailing guidance for the steelwork side is available through the Australian Steel Institute. If you are detailing the plate itself, our companion guide walks through the AS 4100 limit state procedure step by step.

If you would like a second set of eyes over base plate and pedestal details before they go for fabrication, talk to our drafting team about how we handle structural steel detailing review.

A welding symbol is the most information-dense mark on a structural steel drawing. It states the weld type, its size, which side of the joint it goes on, how long it runs, how it is finished and whether it is made in the shop or on site — in a notation compact enough to sit on a leader line. Get it right and the fabricator needs no further instruction. Get it wrong and the error is not caught until steel is on the ground.

Two Australian standards govern it. AS 1101.3 defines the graphical symbols themselves. AS/NZS 1554 defines the welding — procedures, categories and acceptance criteria. A drawing has to be right under both: the symbol says what to make, the 1554 reference says to what standard it is made and inspected.

Anatomy of the Symbol

Every welding symbol is built from the same parts, and each position on it carries fixed meaning:

• Reference line — the horizontal line everything attaches to.
• Arrow — points at the joint. The side it touches is the arrow side.
• Weld symbol — the shape (fillet, butt, plug) placed below or above the reference line.
• Dimensions — size to the left of the symbol, length (and pitch, if intermittent) to the right.
• Tail — the fork at the end, carrying the specification, procedure or process reference.
• Supplementary symbols — flag for a site weld, circle for a weld all round, and contour or finish marks.

Below the line, above the line

This is the convention most often misread, and it is worth stating plainly. A weld symbol placed below the reference line means the weld is on the arrow side of the joint. Placed above the line, it is on the other side. Symbols on both sides mean welds on both sides.

Note that the ISO convention used by AS 1101.3 is the reverse of the American AWS A2.4 arrangement. Imported details, overseas supplier drawings and some software defaults follow AWS. A drawing that mixes the two produces welds on the wrong face of a connection, and it is not obvious on inspection of the drawing alone. If a project takes detail from an overseas source, the convention must be stated on the drawing and checked.

The Symbols You Will Actually Use

A fillet weld is dimensioned by leg length, not throat, unless the drawing states otherwise. Where design throat thickness governs, say so explicitly — the assumption differs between design offices, and a silent drawing invites the fabricator’s assumption rather than the engineer’s.

Intermittent welds

Written as length–pitch to the right of the symbol: 6 50-150 is a 6 mm fillet, 50 mm long, repeating on a 150 mm pitch. Staggered intermittent welds are shown with the symbols offset above and below the line. Intermittent welding saves weld metal and distortion but is not permitted where fatigue or sealing against corrosion governs — a common and expensive substitution on site.

AS/NZS 1554: What the Tail Reference Means

AS/NZS 1554 is a multi-part standard and the part cited changes what the fabricator must do. The parts a structural drawing normally references:

• Part 1 — welding of steel structures. The default for general structural work.
• Part 2 — stud welding.
• Part 3 — welding of reinforcing steel.
• Part 4 — welding of high-strength quenched and tempered steels.
• Part 5 — welding of steel structures subject to high levels of fatigue loading.
• Part 6 — welding of stainless steels for structural purposes.
• Part 7 — welding of sheet steel structures.

Weld category: SP or GP

Part 1 defines two weld categories. GP (general purpose) and SP (structural purpose) differ in permitted imperfections, inspection requirements and the strength that may be relied on in design. SP is the more demanding and the more expensive. Where fatigue governs, Part 5 applies and category FP is used.

The category must appear on the drawing. A drawing that shows weld sizes but never states a category leaves the fabricator to choose, and a design that assumed SP capacity delivered as GP is a structural deficiency that no one on site will see. The usual and correct place for it is a general note — for example, “All welds category SP to AS/NZS 1554.1 unless noted otherwise” — with the exceptions marked individually in the tail.

Six Mistakes That Reach the Shop

1. No weld category anywhere on the drawing set. The single most common defect, and the one with the largest consequence.
2. Arrow-side convention mixed with AWS details. Produces welds on the wrong face.
3. Leg versus throat left ambiguous. A 6 mm throat and a 6 mm leg are not the same weld.
4. Site welds not flagged. The flag drives sequencing, access, scaffold and inspection — omitting it changes the erection plan.
5. “Weld all round” used on a section that is not continuous. The circle means exactly what it says; on an interrupted profile it cannot be executed as drawn.
6. Full-penetration butt welds with no preparation detail. The symbol states the intent; the fabricator still needs the included angle, root gap and root face, or backing where used.

Getting the Notes Right

Most welding information belongs in the general notes rather than repeated on every leader. A workable set states the standard and category, the default fillet size, the consumable class, the treatment of site welds, the inspection regime and the surface treatment after welding. Individual symbols then carry only what departs from the default, which keeps the drawing readable and the exceptions visible.

The current editions of AS 1101.3 and the AS/NZS 1554 parts are published by Standards Australia, and the part and year cited on a drawing should match the edition the project is contracted under rather than whatever a title block template carried over from the last job.

Frequently Asked Questions

Does the symbol go above or below the reference line?

Below the line for the arrow side, above for the other side, under AS 1101.3. The American AWS convention is reversed, so the source of any imported detail matters.

What is the difference between SP and GP welds?

Permitted imperfections, inspection requirements and the design capacity that may be relied on. SP is the higher category. If the design assumed SP, the drawing must say SP.

Is a fillet weld sized by leg or throat?

Leg length by default. If the design is governed by throat thickness, state it on the drawing rather than relying on the reader to infer it.

How is a site weld shown?

A flag at the junction of the arrow and the reference line. It affects erection sequence, access and inspection, so it should never be left off a weld genuinely made on site.

Detailing to the Standard

Consistent welding annotation is a drafting discipline as much as an engineering one — the categories, defaults and exceptions have to be applied the same way across every sheet in a set. Our structural steel detailing work is produced to AS 1101.3 and AS/NZS 1554, and if you have a set you want reviewed before it goes to the fabricator, send it through.

A steel column is only as reliable as the connection that transfers its load into the concrete below. Base plate design sits at the boundary between two Australian Standards — AS 4100 for the steelwork and AS 3600 for the concrete it bears on — and most of the errors we see in shop drawings come from treating it as one problem rather than two. This is the limit state procedure we follow when detailing column base plates for Australian projects.

What the limit state check actually covers

Limit state design asks a single question at each interface: is the design capacity greater than the design action? For a column base that means four separate checks, and passing three of them is not passing.

• Concrete bearing — the plate must spread the column load over enough concrete that the bearing pressure stays within the design capacity of the pedestal or footing.
• Plate bending — the cantilevering plate outstand behaves as a yield line; plate thickness is governed by the bending moment that outstand attracts, not by rule of thumb.
• Anchorage — holding-down bolts must transfer uplift and shear, and the failure mode is usually in the concrete (cone breakout, edge splitting, pull-out) rather than in the bolt steel.
• Weld capacity — the column-to-plate weld has to carry the same actions, including any moment the connection is assumed to resist.

Step 1 — establish the design actions and the connection model

Before any geometry, settle what the base is assumed to do in the analysis model. A pinned base carries axial force and shear. A fixed base also carries moment, which changes the bolt layout, the plate thickness and often the footing itself. The single most expensive mistake in base plate detailing is a drawing that shows a nominal four-bolt pinned detail while the frame analysis assumed a moment-resisting base.

Take the governing load combination for axial compression, the combination that produces maximum uplift, and the combination that produces maximum shear. These are frequently three different combinations, and the base plate has to satisfy all of them.

Step 2 — size the plate for concrete bearing

The plate area follows from the design bearing strength of the concrete beneath it. Where the plate is small relative to the supporting concrete, the confinement of the surrounding material permits a higher bearing strength than the unconfined value — this is the dispersion allowance, and it is capped. Two practical consequences:

• Grout type and thickness matter. A non-shrink cementitious grout of the specified strength is part of the load path, not a levelling convenience.
• A plate sitting on a small pedestal gets little or no confinement benefit. Check the pedestal dimensions before claiming the enhancement.

Step 3 — determine plate thickness from the yield line

With the plan area fixed, the plate outstand beyond the column footprint acts as a cantilever carrying the bearing pressure. The design moment per unit width comes from that pressure acting over the outstand, and the required thickness follows from the plate’s section modulus and yield stress. For grade 250 and grade 350 plate the difference in required thickness is significant, so specify the grade on the drawing rather than leaving it to the fabricator.

Where the base resists moment, the pressure distribution is no longer uniform. Part of the plate lifts, the bolts on the tension side pick up the uplift, and the compression side sees a higher peak pressure over a smaller area. Designing that plate on an average pressure will under-thickness it.

Step 4 — check the anchorage, in the concrete

Holding-down bolts rarely fail as steel. The governing modes are concrete cone breakout under tension, edge breakage and pry-out under shear, and pull-out where the embedded head or plate washer is undersized. Edge distance and embedment depth do more for capacity than bolt diameter does, which is why moving a column 50 mm off a pedestal edge can be worth more than upsizing every bolt.

Detail the bolt holes generously — oversized holes with a plate washer are standard for setting-out tolerance — and then make sure the shear path is honest. If the design relies on shear transfer through the bolts, the oversized holes have to be accounted for, or a shear key or recessed plate provided instead.

Step 5 — the weld, and what the drawing must say

The column-to-plate weld carries whatever the connection is assumed to carry. A nominal fillet all round is adequate for a genuinely pinned base with modest shear; it is not adequate for a moment base, where the flange welds are doing the work. Show the weld category and size explicitly.

A base plate detail is complete when a fabricator can build it and an inspector can check it without ringing the engineer. That means plate size, thickness and grade; bolt size, grade, embedment and edge distance; hole sizes and washer requirement; grout type and thickness; weld size and category; and the setting-out datum. Our fabrication shop drawing checklist covers the wider drawing set, and the AS 4100 structural steel design manual gives the broader clause-by-clause context this connection sits inside.

Where base plate details go wrong most often

• Plate thickness carried over from a previous job with a different column, load or plate grade.
• Moment assumed at the base in analysis, pinned detail drawn on the shop drawings.
• Anchor capacity checked as steel only, with no concrete breakout check and no edge distance stated.
• Grout ignored, so the plate is drawn hard down on concrete that was never finished to that level.
• Bolt setting-out template not issued, so the cast-in bolts do not line up with the fabricated plate.

AS 4100 and AS 3600 are both published by Standards Australia, and the referenced editions are called up through the National Construction Code. Always design and detail to the edition your project’s building approval cites, not the newest one on the shelf.

If you need base plate and connection details drawn to Australian Standards, our structural steel detailing team produces the fabrication and erection drawings, and we can work from your engineer’s design or mark up the connection detail for their review.

If you exchange CAD files with manufacturers, you have almost certainly hit the moment where a supplier asks for a STEP file and all you have is an IGES — or the reverse. The IGES vs STEP question comes up constantly, because both are neutral formats designed to move geometry between different CAD systems — yet they are not interchangeable, and choosing the wrong one (or converting carelessly between them) can quietly cost you topology, assembly structure, and hours of rework.

This IGES vs STEP guide explains what each format actually is, when to use which, and how to convert between them without losing the data that matters. It is written from the perspective of a drafting team that converts these files every week, including a recent job where an IGES-to-STEP conversion went wrong in an instructive way.

What IGES is

IGES (Initial Graphics Exchange Specification) is one of the oldest neutral CAD formats still in active use. The last formal revision, IGES 5.3, dates to 1996, and that age tells you most of what you need to know about its strengths and limits. IGES was built primarily to exchange surfaces and wireframe geometry. It represents a model as a collection of entities — curves, surfaces, points — rather than as a single watertight solid. That makes it extremely flexible for moving surface data between systems, but it also means an IGES file can arrive as a loose collection of surfaces that look like a solid on screen but are not stitched into one.

The .igs and .iges extensions refer to the same format; .igs is simply the older 8.3-style short extension. There is no functional difference between them.

What STEP is

STEP (Standard for the Exchange of Product model data) is the modern successor, governed by the ISO 10303 standard. The application protocols you will see most often are AP203, AP214, and the newer AP242. Where IGES thinks in surfaces, STEP is designed to carry solid models — a proper boundary representation (B-rep) where faces, edges, and vertices are topologically connected into a closed volume. AP242 goes further again, carrying product manufacturing information (PMI) such as tolerances and annotations, plus assembly structure.

In practice this is the headline difference in the IGES vs STEP comparison: a STEP file usually arrives as a clean solid you can immediately measure, modify, and put into CAM. An IGES file often needs healing first.

IGES vs STEP: when to use each

The IGES vs STEP decision usually comes down to who is on the other end and what they are doing with the model:

• Choose STEP for manufacturing and machining. If the file is going to a CNC shop, a fabricator, or into CAM software, STEP (AP214 or AP242) is almost always the right answer because it carries a true solid and, in AP242, the tolerancing.
• Choose STEP for assemblies. STEP preserves assembly hierarchy and part relationships. IGES flattens everything into geometry and loses the structure.
• IGES still has a place for pure surface data. Some legacy systems, older CAM packages, and certain industrial-design surfacing workflows still expect IGES, and for transferring complex freeform surfaces it remains perfectly serviceable.
• When a client specifies a format, give them what they asked for. Their downstream toolchain dictates the requirement; do not “upgrade” an IGES request to STEP without checking, because their importer may be tuned for one or the other.

A real conversion that went wrong — and what it taught us

On a recent mechanical assembly job for a Brisbane manufacturer, we received the parts as IGES files. The client’s own SolidWorks installation did not have a working IGES import path that produced usable solids, so they needed the parts as STEP before they could do anything with them. On the surface a trivial conversion — open the IGES, export STEP. In reality, several of the parts came across as disconnected surface bodies rather than solids, because the original IGES export had never stitched the surfaces into a closed volume.

The fix was not to re-export and hope. We ran each part through a knit/heal step: importing the IGES surfaces, identifying the gaps where adjacent surfaces failed to meet within tolerance, knitting them into a closed boundary, and only then exporting to STEP as a true solid. Two parts had genuine geometry gaps that no automatic heal could close, and those had to be rebuilt locally before the solid would form. The lesson we took from it: always verify that an IGES import produces a single solid body before converting downstream. A surface model that looks complete on screen can still be open, and the failure only surfaces when CAM or a Boolean operation rejects it.

Common conversion pitfalls

• Surfaces vs solids. The single biggest issue. An IGES import can give you surfaces that need knitting before they behave as a solid. Always check the body type after import.
• Lost assembly hierarchy. Converting an assembly to IGES collapses it into geometry. If structure matters, use STEP from the start.
• Tolerance and gap problems. Surfaces that do not meet within the stitching tolerance leave gaps; tightening or loosening the knit tolerance is often the difference between a clean solid and a failed heal.
• Units drift. Confirm the unit system on import — a model that imports at the wrong scale is easy to miss until a dimension looks wrong.
• PMI loss. Tolerances and annotations do not survive a trip through IGES. If you need them, stay in STEP AP242.

Four practical ways to convert

• SolidWorks / Inventor. Open the source file and use Save As to the target format. Both let you set import options for whether to knit surfaces into solids automatically — turn that on and verify the result.
• AutoCAD. Capable of importing and exporting both formats, best suited where you are working with 2D-derived or simpler geometry rather than complex surfacing.
• FreeCAD. A no-cost option that handles IGES and STEP via its OpenCASCADE core. Useful for one-off conversions when you do not have a seat of a commercial package free.
• Online converters. Fine for a quick, non-confidential single part. Avoid them for anything under NDA or for assemblies where structure must be preserved.

What Australian shops tend to expect

In our experience across Australian manufacturing and fabrication, STEP has become the default request — in the IGES vs STEP choice most CNC shops, laser cutters, and fabricators will ask for a STEP file first because their CAM software ingests it cleanly as a solid. AP214 is the version you will be asked for most often; AP242 is appearing more where tolerancing needs to travel with the model, particularly in defence and aerospace-adjacent work. IGES requests now tend to come from two places: older equipment and tooling suppliers still running legacy systems, and industrial-design or surfacing workflows where freeform surface data is the point. If a supplier does not specify, sending STEP and keeping the native file on hand is the safe default.

One practical habit worth forming: when you send a neutral file, send it alongside a short note of the units and the source CAD system. A surprising number of conversion headaches are not really format problems at all — they are a part that imported at the wrong scale, or a recipient who did not know whether to expect surfaces or a solid. A one-line note prevents both.

A simple decision framework

When a file lands on your desk and the IGES vs STEP choice is not obvious, ask three questions. Is it going to manufacturing or CAM? Send STEP. Does it contain an assembly whose structure matters? Send STEP. Is the recipient on a legacy surfacing system that specifically asks for IGES? Send IGES, and verify the surfaces are clean before you do. When in doubt, ask the client which their downstream tool expects — five minutes of asking saves an afternoon of re-converting.

If you would rather hand the whole problem to a team that converts these formats daily, that is exactly the kind of work we do. See our CAD conversion services, our deeper explainer on IGES and STEP files, or our guide to converting PDF drawings to DWG. Send us a sample file and we will tell you exactly what is in it and what it will take to get it into the format you need.

AS 1657 is the Australian Standard that governs fixed platforms, walkways, stairways, and ladders — the access structures that let people work safely at height around plant and equipment. It is one of the most frequently referenced standards in Australian industrial design, and also one of the most frequently got wrong. Over the last 18 months we have reviewed more than 30 platform and walkway designs for civil, structural, and mechanical clients, and the same compliance failures keep reappearing.

This article walks through the seven we see most often, what the standard actually requires in each case, and — crucially — how to fix each one without tearing up the whole design. None of these are exotic. They are the ordinary mistakes that come from designing to memory rather than to the clause.

1. Guardrail height set to the wrong dimension

The most common failure of all: a top rail set below the minimum height the standard requires for the platform’s height above ground. Designers often carry a single “handrail height” number in their head and apply it everywhere, but the requirement is not a single universal figure. The fix is usually straightforward — raising the top rail and adjusting post lengths — and rarely forces a structural redesign, but it must be checked against the actual platform height rather than assumed.

2. Missing or undersized toeboards (kickplates)

Toeboards stop tools and debris falling onto people below, and they are mandatory on platforms above the relevant height. We regularly see them omitted entirely, or specified too short. Adding a compliant toeboard is one of the cheapest fixes on this list because it bolts to the existing edge structure — but it is also one of the most commonly flagged in audit, precisely because it is easy to forget at the modelling stage.

3. Stair pitch and going outside the allowed range

AS 1657 distinguishes between stairways, step-type ladders, and ladders by their angle (pitch), and each has a permitted range. A “stair” drawn at an angle that actually falls into the step-ladder band changes the going, riser, and handrail requirements entirely. We see designs that sit right on a boundary and unknowingly inherit the wrong set of rules. The fix is to confirm which access type the pitch puts you in before detailing the treads, not after — once the geometry is locked, correcting the pitch can cascade into the supporting structure.

4. Inadequate clearances and headroom

Walkways routed under pipework, cable trays, or structural members frequently lose the required headroom, and platforms tucked against equipment lose the required width or side clearance. These are layout failures rather than detailing failures, which is why they are expensive to fix late — the cure is often rerouting the walkway. Catching them early, during the general arrangement, is far cheaper than discovering them at the model-review stage.

5. Gaps and infill that exceed the allowable opening

The space between the top rail and mid rail, and between the mid rail and platform, is limited so that a person cannot fall through. Designs with a single mid rail on a tall guardrail often leave an opening larger than permitted. Adding a second intermediate rail or infill mesh closes the gap and is a minor addition to the rail assembly — but it has to be designed in, because retrofitting mesh to an installed rail is awkward and unsightly.

6. Ladders without the required safety cage or landing provisions

Fixed ladders above a certain height have requirements around fall protection and rest/landing provisions. We see tall single-flight ladders drawn without addressing these, usually because the height crept up during design without anyone rechecking the clause. The remedy depends on the height — sometimes a landing, sometimes an alternative fall-arrest provision — and it is much easier to allow space for it in the layout than to bolt it on afterwards.

7. Detailing to an outdated edition or a mix of editions

Finally, the quiet one: designing against an old copy of the standard, or mixing clauses from different editions because that is what was in the office template. Standards get amended, and a detail that complied a few years ago may not today. The fix is process rather than geometry — work from the current edition, and keep templates under version control so an outdated note does not propagate across every drawing in the set.

The pattern behind all seven AS 1657 mistakes

Almost every one of these comes from the same root cause: applying a remembered rule of thumb instead of checking the clause against the specific geometry in front of you. The platform height, the stair pitch, the ladder height — these are the inputs that decide which requirements apply, and they vary job to job. A design that was perfectly compliant on the last project can fail on this one simply because the height changed.

The good news is that most of these are inexpensive to correct when caught at the general-arrangement or model-review stage, and ruinous only when they reach fabrication. That is the argument for a compliance review before the drawings are issued for construction.

How an AS 1657 compliance review actually works

When a design lands with us for review, we work through it in a fixed order, because the inputs cascade. First we establish the governing dimensions: the height of each platform above the floor or grade below, the pitch of every stair and ladder, and the height of each ladder flight. Those three numbers decide which clauses apply, so they are settled before anything else is checked. A great many “failures” are really just a design built against the wrong assumption about one of these inputs.

From there we check the guarding: top rail height against platform height, mid rail and infill against the maximum opening, and toeboard presence and size on every exposed edge. Then access: stair going and riser consistency, ladder rung spacing, cage or fall-arrest provisions where the height triggers them, and landings on long flights. Finally clearances: headroom along every walkway and width past fixed obstructions. Each item is recorded as compliant, non-compliant, or needs-information, and the non-compliant items are sorted by how much they cost to fix — rail and toeboard changes are cheap, layout-driven clearance failures are not.

The output is a marked-up drawing and a short schedule of findings, so the engineer can see exactly which clause each item relates to and what the lightest-touch remedy is. The aim is never to redesign the structure — it is to get the existing design compliant with the smallest possible change.

Why these slip through in the first place

It is worth being honest about why competent engineers produce non-compliant access designs. It is rarely incompetence. It is that access structures are usually the last thing added to a model — bolted on around plant and structure that are already fixed — so they inherit whatever space is left rather than being designed to the standard from the start. By the time the platform goes in, the headroom under that pipe rack is whatever it is. Designing the access route early, while the surrounding geometry can still move, prevents most of the clearance and layout failures on this list before they happen.

If you have a platform, walkway, or access design you want checked against the standard before it goes out, that is work we do regularly. See our page on AS 1657 fixed platforms and walkways, and send us your general arrangement — we will tell you which of these seven (if any) apply and what the lightest-touch fix is for each.

Note: this article is general guidance, not a substitute for the current published text of AS 1657. Always design and verify against the latest edition of the standard.