How the two differ in kind
The two beams are not competing versions of the same part; they are two different ways of buying the same stiffness. Steel's modulus of elasticity is somewhere in the region of fifteen to twenty times that of structural timber, so a steel section resists bending with far less material in it. Timber cannot close that gap by being wider — width buys stiffness roughly in proportion, which is to say slowly. It closes it by being deeper, because bending stiffness climbs with the cube of depth: a beam half again as deep is several times stiffer for a modest amount of extra timber. That is why the two solutions to one opening look nothing alike. Steel arrives as a shallow, dense, unhelpfully heavy piece. Engineered timber arrives as a deep, light one. Almost every practical difference below is that single fact working its way outward through the job.
The consequence people meet first is not structural at all, it is logistical. A steel beam sized for a wide domestic opening is routinely past what a crew can carry, and the way it gets to its bearings is a lifting operation — extra bodies, a genie or an engine hoist, sometimes a crane, and in a house with no side access sometimes a window taken out or a section of wall opened up just to admit the beam. Engineered timber has a route steel does not, though it belongs to one form of it. A multi-ply beam — LVL or LSL plies, sometimes built-up lumber — is carried in one ply at a time by hand, stood up on the props, and bolted or nailed together into a single member exactly where it will live. In a mid-terrace with a narrow hall and a turn at the bottom of the stairs, that is not a convenience, it is the difference between a job and a much larger job. A one-piece glulam does not get that trick: it is manufactured whole and has to arrive whole, so the turn at the bottom of the stairs is as much its problem as steel's. What it does get is weight — a glulam sized for the same opening is a good deal lighter than the steel that would replace it, often the difference between a planned lift and a two- or three-person carry. What comes after the beam divides the same way. Timber takes screws and nails, so hangers land straight on the face and the carpenter never stops working. Steel takes bolts, drilled holes and packers, so joists, the ceiling below and the floor above all need something timber bolted to the section before they have anything to fix to.
Where does that leave the answer? Genuinely undecided, for a great many ordinary openings — and it is worth saying so rather than pretending otherwise, because both beams get designed until they work. What is not undecided is which two measurements settle it. The first is the depth budget: from the finished ceiling line you intend to keep, up to the underside of whatever the beam has to carry. If that gap is tight and the span is long, steel is frequently the only thing that fits, and no amount of preferring timber changes it. The second is the route in: door widths, the turn at the stairs, whether there is a hardstanding for plant, whether a window could come out. If nothing heavy can reach the opening, the multi-ply timber beam is the answer and the depth will have to be found somewhere. When neither measurement is binding — the beam can drop below the ceiling and there is a clear route in — then, and only then, the decision moves on to what the beam has to look like, what has to be fixed to it later, and who is standing in the room to build it.
The factors that actually differ
| Structural steel beam | Glulam / engineered timber beam | |
|---|---|---|
| Depth for the same span | The shallowest section available for the load. Where a flush ceiling has to be held across a long opening, this is often the only reason steel is being considered at all. | Wants noticeably more depth for the same stiffness, because depth is how timber gets stiff. On a long span the beam either drops into the room or has to be found room for above. |
| Getting it to its bearings | Heavy enough on a wide opening to be a planned lift: extra hands, hoisting plant, sometimes a crane, and occasionally a temporary opening in the wall so a single long piece can physically make the turn. | Much lighter for the same opening, so the move is usually hands rather than plant. In multi-ply form — LVL or LSL plies bolted together on the props — it need never enter the house as one piece at all, which is sometimes the only way a beam of any size gets in. A one-piece glulam forfeits that: made whole, it must arrive whole, just far lighter than the steel it replaces. |
| Creep under permanent load | None worth counting at room temperature. What it deflects under load on the day it is propped is what it deflects a generation later. | Real, expected, and designed for: sustained load keeps a timber beam moving for years, so the permanent share of the deflection is multiplied in the check. Camber is specified when the beam is ordered, not added on site. |
| Fixing everything else to it | Nothing screws to it. Joists, hangers, ceiling and floor all wait on drilled holes, bolted plates or a timber packer fixed to the web or flange, and every following trade works around that detail. | Screws and nails, anywhere along it. Hangers land on the face, battens go straight on, and no separate operation stands between the beam and the trades behind it. |
| Behaviour in fire | Loses strength as it heats, so where a rating applies it needs protecting — boarded, encased or intumescent-coated to a specified thickness rather than simply painted. | Burns predictably. The outer layer chars, the char insulates the sound timber behind it, and a beam can be sized with a sacrificial thickness so the surviving section still carries the load — which is why exposed timber beams are ordinary and exposed unprotected steel is not. |
| Changing it once it is on site | Length, holes and connection plates are frozen at the fabricator. Altering it afterwards means a magnetic drill or hot work, then making the protective coating good again. | Cut, trimmed and drilled with tools the crew already has, within the notching and boring rules. An opening that turns out different from the drawing is absorbed rather than re-ordered. |
| What happens at the bearing | Concentrates a large reaction on a short length of bearing, so it wants a padstone or spreader onto masonry and a load path underneath that has genuinely been resolved to something solid. | Crushes across the grain long before it fails in bending, so it wants a longer bearing and more timber stacked under each end. Cheap to provide, and one of the commonest things left out. |
| Damp and where it is allowed to live | Rusts wherever it stays wet: primer and paint, and a damp-proof separation where an end bears into masonry that gets wet. | Rots wherever it stays wet, and the adhesive lines carry a service class — an interior-rated glulam does not belong in an exposed or persistently damp position, whatever it looks like it could survive. |
Which one, and when
Choose structural steel beam when…
- The ceiling line is fixed and the span is long — the depth budget has already decided this, and steel is the shallowest beam there is.
- The load is heavy or arrives as points: a wall above, a roof, another beam framing into it.
- There is a route and room for the lift, or the wall is coming out anyway and the opening exists while the beam goes in.
- You want the deflection you designed for to be the deflection you still have in twenty years, with no creep allowance sitting in the calculation.
Choose glulam / engineered timber beam when…
- Nothing heavy can reach the opening and there is nowhere to lift from — a multi-ply beam, carried in one ply at a time and fastened together on the props, is the only way this beam arrives at all. That is the multi-ply route specifically; a one-piece glulam still has to come in whole.
- The depth is available: the beam can drop below the ceiling, or hide within the joist zone, or become the thing everyone looks at.
- The beam will be seen, and you would rather it were timber than a boxed-in steel section pretending to be a bulkhead.
- Everything that follows the beam is carpentry, and you want the crew already in the room to keep going without a fabricator, a lift and a hot-work day in the middle of it.
Now run your own numbers
This page holds no prices on purpose — a national average is wrong for almost every real project. Quantify both options with your dimensions and your local quotes.
Frequently asked questions
- Which one is cheaper?
- They are cheaper in different shapes, which is why no figure belongs here. Steel's cost is material plus fabrication plus a lifting operation, and that last part is largely fixed — much the same crew and much the same plant whether the beam is modest or long — so a small opening amortises it badly and a difficult access route inflates it sharply. Engineered timber's cost sits much more in the material and tracks the beam more nearly in proportion, with no plant line at all. Then there is the cost that appears in neither quote: if steel's shallower section is what saves you from re-working a ceiling or losing head height, it has already paid for itself, and if timber's extra depth costs nothing because the beam was dropping into the room anyway, so has it. Price both against your own opening and your own access, because that is where the difference actually lives.
- Why does the timber beam have to be so much deeper?
- Because what is being bought is stiffness, and timber buys stiffness with depth. Structural timber's modulus of elasticity is a small fraction of steel's — broadly a fifteenth to a twentieth — so a timber section of the same size bends far more under the same load. Widening it recovers that roughly in proportion, which is a poor deal. Deepening it recovers it far faster, because bending stiffness rises with the cube of the depth. That is why the timber answer to a longer span is always a deeper beam rather than a fatter one, and why the depth you can spare is the first thing to measure and the last thing to compromise.
- Is LVL the same thing as glulam?
- Both are engineered timber and both sit on the same side of this comparison, but they behave differently to buy and to build with. Glulam is dimension lumber glued up in laminations into a large solid member, commonly made to order, available in big sections and long lengths, cambered on request, and it is the one you leave exposed because it looks like something. LVL is thin veneers bonded into a stock product carried at the yard in standard depths, and it is the one you buy as separate plies and fasten together on site into whatever width the design calls for. For a retrofit where every piece has to be carried in by hand, the multi-ply route is usually the practical one. For a beam that is meant to be seen, glulam. The deflection check is the same exercise either way — only the modulus and the section change.
- What about a flitch beam — is that a real compromise?
- It is a real option and sometimes the right one. A flitch beam is a steel plate sandwiched between timber members and bolted through, so it arrives as separate pieces light enough to handle, takes screws and nails on its timber faces like any joist, and is stiffer and shallower than the timber alone would be. The trade is that it is a designed assembly rather than a product: the plate has to be drilled to a pattern, the bolt size and spacing are part of the calculation rather than a site decision, and the finished beam is still not as shallow as an equivalent rolled steel section. Where the depth budget is tight and the access is impossible, it is often the only thing that satisfies both.
- Does the beam need fire protection?
- Both might, for genuinely different reasons, and the answers do not look alike. Steel weakens as it heats, so wherever a rating applies it has to be boarded, encased or coated — and an intumescent coating is a specified dry film thickness, not a paint you choose by colour. Timber burns, but at a rate that can be relied on: the surface chars, the char shields the sound timber behind it, and the beam is sized with a sacrificial layer so that what is left after the required period still carries the load. Which of these your building actually needs is a local code question tied to the storey, the use and what is above the opening. Ask it before the beam is ordered rather than after it is propped in place, because it can change the section either way.
- Can these calculators size the beam for me?
- No, and it matters that you know exactly what they do. The steel tool returns weight per unit length from the section's dimensions — useful for the lift, the bearing reaction and the tonnage, not for choosing a section. The glulam tool checks deflection against the code ratio for a uniformly loaded simple span and says plainly that passing it is serviceability only, with bending stress and shear still to be verified. Neither knows your loads, and on a wall removal the load is the hard part: what sits above, how it arrives, where it goes once the wall has gone, and whether the existing structure will take the reaction at each end. That is an engineer's calculation, building control will want it on paper, and the temporary propping while the wall comes out is a separate design again. Use these to compare options and to understand what a specification is doing — not to write one.