Walk through any furniture showroom in Sydney or Melbourne right now and you’ll see curves everywhere. Curved sofa backs, arched dining chairs, half-moon cabinets, rounded console tables. The aesthetic has been gaining momentum for three years and isn’t slowing down. Clients are asking for it. Designers are specifying it. And the question that keeps coming up in our workshop is which technique we should use to build the curves: traditional steam bending, or CNC-cut laminated curves.

Both work. They produce different results. They cost different amounts. And the right choice depends on factors clients don’t always think about. Let me walk through how we approach it.

What steam bending actually is

Steam bending has been around for centuries. The basic idea is straightforward. You take a piece of solid timber, expose it to steam at around 100 degrees Celsius for an hour per inch of thickness, then bend it around a former while it’s hot and plastic. As it cools, the timber locks into the new shape.

The best timber species for steam bending are ones with long, flexible fibres: white oak, ash, elm, beech. Australian native timbers are mixed for this. Some species bend beautifully (Tasmanian myrtle, certain eucalypts), some don’t (Jarrah is notoriously stubborn). The species choice constrains the design.

Steam bending produces a curve in solid timber. That’s the key advantage. The grain runs continuously through the curve, and visually you can see it’s a single piece of wood that’s been persuaded into shape. There’s a subtle but real beauty to that which I think most clients can feel even when they can’t articulate why.

What CNC laminated curves actually are

CNC laminated curves are built differently. You take thin veneers of timber, typically 1-2mm thick, apply glue between them, and clamp them around a former until the glue cures. The CNC piece comes in shaping the former, and sometimes machining the final piece after lamination. The result is a curve that looks solid but is actually a stack of laminated layers.

The advantages are substantial. Almost any timber species can be used. The curve geometry isn’t constrained by the timber’s natural bending behaviour. You can do tighter radii than steam bending allows. And the result is dimensionally stable, because the laminations resist movement in ways solid timber doesn’t.

The visible disadvantage is the lamination lines. From a couple of metres away, you can’t see them. Up close, particularly on a horizontal surface with raking light, they’re sometimes visible. Some clients love this aesthetic. Some don’t.

What we actually use, and when

Here’s how we currently make the call.

Sofa frames and curved bench seating. Almost always laminated. The structural demands are high, the geometry is often complex, and the lamination process produces a frame that won’t move with humidity. Steam-bent sofa frames look beautiful but I’ve seen too many of them shift over five years in Australian conditions to recommend them as a primary structural element.

Dining chair backs and arms. Steam bending wins when the design allows it. The visual quality of a solid timber chair back with continuous grain is one of those subtle markers of a piece that’s been made properly. The cost premium is real, but if you’re paying for a chair to live in your house for 30 years, the difference is worth it for most clients.

Curved cabinet doors. Almost always laminated. Steam bending a panel large enough to be a cabinet door is mechanically difficult, the failure rate during bending is high, and the result moves more with humidity than the rest of the cabinet body. Lamination produces a stable door that fits its frame consistently.

Console tables and curved tops. Mixed. If the table is small and the curve is gentle, steam bending works beautifully. If the table is large or the curve is tight, lamination wins on both feasibility and stability.

Arched headboards. This is where we’ll often have the longest conversation with clients. Steam-bent arched headboards in Tasmanian oak or American white oak look magnificent. Laminated arched headboards in any timber the client wants are more flexible and cheaper. The right answer depends on the client’s priorities, the room’s lighting, and the specific design language.

Where 3D design tools have helped

The shift in our workflow over the last few years has been driven by 3D design tools that let us model the final piece in either technique before we commit to one. We can show a client what a steam-bent piece will look like with continuous grain, then toggle the model to show the laminated version with visible layer lines, then toggle again to show a hybrid approach with steam-bent visible elements and laminated structural components.

That visualisation step has reduced the number of “I didn’t expect it to look like that” conversations to almost zero. Clients can see the difference before we commit, and they can make the decision with their eyes rather than their imaginations.

The cost question

Steam bending costs more per linear metre of curve, almost without exception. The reasons are structural: the fail-rate during bending is non-zero, the species choice is more restricted, the timber stock has to be straight-grained and unblemished, and the setup time per curve is meaningful.

For a single curved chair back at 30cm of curve, the cost difference might be $80-$150 between techniques. For a sofa frame with 4 metres of total curved material, the difference could be over a thousand dollars. The cost premium scales roughly linearly with the amount of curve in the piece.

The lifespan argument is real but easy to overplay. A well-laminated curve will last as long as the rest of the piece, easily 50+ years. The difference isn’t durability. It’s aesthetic intimacy with the material.

What I’m watching in the technique evolution

A few things on the horizon.

Better glues for lamination. The current generation of polyurethane-based glues for cold lamination are excellent. The next generation appears to be addressing the springback question, where laminated pieces try to flex back toward flat over time. If that gets solved cleanly, the case for lamination strengthens further.

Steam-bending automation. A few workshops in Europe are experimenting with computer-controlled steam-bending rigs that apply pressure more consistently than a human can. Early results look interesting. The cost of the equipment puts it out of reach for most small workshops, but it might filter down.

CNC compound curves. The biggest current limitation of CNC laminated work is that compound curves (curves in two axes simultaneously) are hard to laminate cleanly. New machining techniques are starting to address this, opening up design possibilities that weren’t practical even three years ago.

What to ask your maker

If you’re commissioning a piece with curves, three questions worth asking.

Which technique are you using, and why? A confident maker will explain the choice. A maker who shrugs or gives a vague answer hasn’t thought it through.

What timber species are you using, and how does it move? Curved pieces in unstable species will shift over time. Australian conditions are not gentle on timber that wants to move.

Show me a previous piece you’ve made using this technique. A photograph isn’t enough. Look at the actual piece in person if you can, in raking light, at the curve’s tightest point. That’s where any compromises show.

The curved furniture trend will continue running for at least another two or three years. The pieces being made well will outlast the trend. The pieces being made fast and cheap, on either technique, won’t.

Choose your maker before you choose your technique. Then trust the maker to make the right call.