I didn't just study engineered wood — I made it. My master thesis, in plain language

What LVL is — and why I spent my thesis making it

Laminated veneer lumber — LVL — is one of the simplest ideas in engineered wood: take thin sheets of real wood veneer, a millimetre or three thick, glue them on top of each other with all the grain running the same direction, and press them under heat into a thick, beam-like material. It looks like plywood's cousin, but there is a key difference: plywood crosses its layers for all-direction stability, while LVL keeps every layer parallel — all the strength pointing one way, like a bundle of fibres. That makes it a beam material: lintels, long shelves, structural frames, anywhere wood has to carry load along its length.

Why does the industry bother? Because a solid beam needs a big, old, defect-free tree — and those are exactly the trees the world is running out of. LVL is built from veneers, so it can be made from small-diameter logs and lower-grade timber that would otherwise be chipped into particleboard. And because any knot or defect ends up spread across many thin layers instead of sitting in one solid piece, the finished material is remarkably uniform and predictable. Kilo for kilo, the literature puts LVL's strength-to-weight ratio ahead of steel's.

For my master thesis at the University of Sopron in Hungary (Wood Engineering and Creative Design, 2024), I didn't just read about this. My question was practical: could two local Hungarian species that the industry mostly overlooks be turned into good LVL? I designed the layups, mixed the glue by the gram, pressed the panels, broke them in a testing machine, and then rebuilt them digitally in a simulation. What follows is that story, with the real numbers.

The trees nobody wanted

My thesis title says "underused Hungarian species", and that is the heart of it. The first species was Turkey oak (Quercus cerris) — in Hungarian, csertölgy. It is the single most common tree in Hungary, covering around 11.4% of the country's forest area, and the solid wood is genuinely strong: bending strength of roughly 94 to 136 MPa and an oven-dry density of 570 to 850 kg/m³. Yet it rarely makes it into high-value products. The second was hybrid poplar I-214 — a plantation tree that races to maturity in 8 to 10 years, with light wood around 300 to 480 kg/m³ and modest strength. One species is dense and strong but overlooked; the other is fast, cheap and abundant but weak.

The industry norm is that LVL gets made from softwoods like spruce and pine. My question was whether these two "leftover" hardwoods — one too unfashionable, one too light — could be combined into a serious structural material. If yes, a country gets to build strong beams out of the forest it actually has, instead of importing prime logs from somewhere else. That thinking — make the most of the resource in front of you — is the same logic I now apply every week in the workshop.

The recipe: glue by the gram, heat, and 90 bar of pressure

Making LVL in a lab is cooking with very expensive discipline. First the veneers — Turkey oak at 1.4 mm thick, poplar at 2.8 mm — went into a drying oven at 70°C for 10 hours, then had their moisture content measured, because trapped moisture inside a hot press is how you ruin a panel. Then the glue: the literature said 70 grams of adhesive per square metre between each layer. Poplar taught me my first lesson in "the material talks back" — it is so absorbent it drank the glue before it could bond, so we raised its dose to 90 g/m². Every layer's portion was weighed on a scale and spread evenly by hand.

We built eight different panel recipes over three production cycles. Controls first: pure poplar panels and a pure Turkey oak panel (eleven oak veneers stacked into one board). Then the interesting ones — mixed-species layouts, inspired by how particleboard puts its best material on the faces: strong Turkey oak veneers on the outside, light poplar in the core, in different ratios. Every layup went into a hydraulic press preheated to 140°C: 90 bar of pressure for 5 minutes, then eased down to 20 bar for another 8. And you release that pressure slowly — let the trapped vapour rush out and it can blow the panel apart from inside.

Thirteen minutes in the press, and out comes a board. Then came the honest part: cutting it into standard test pieces — 400 mm long, 19 mm wide, 14 mm thick — and breaking every single one.

What the bending machine said

Every sample went into a three-point bending test: supported at two points 360 mm apart, pushed down in the middle until it broke, while the machine recorded everything. The controls set the two ends of the scale. Pure poplar LVL averaged about 59 MPa in flexural strength with a stiffness (modulus of elasticity) around 9,200 MPa. Pure Turkey oak LVL averaged about 115 MPa and roughly 12,700 MPa — nearly double the strength, exactly what its density promised. The poplar samples also scattered more in their results, while the oak was consistent. Across every panel we made, one relationship held firm: as density went up, strength went up with it.

The mixed panels landed in between, and the more oak veneers a layup carried, the stronger it got. But the layout that excited me most was the balanced sandwich we called L3: a double layer of Turkey oak veneer on each face, poplar filling the core — about two-thirds poplar by volume. It scored around 95 MPa in flexural strength, close behind the oak-heavy layup and in some measurements even passing it. Two-thirds cheap, fast-growing wood, delivering most of the performance of the dense, strong one.

That is the finding I still carry with me: in a bending beam, the stress lives at the faces, not the middle. Put your strong, expensive material exactly where the stress is, fill the rest with light and affordable wood, and you lose very little. That is not a compromise — that is engineering.

Glue matters — and so does the computer

We also tested two industrial adhesives against each other. MUF (melamine-urea formaldehyde) is the classic — it needs a catalyst mixed in to start curing, like yeast in dough. MDI (methylene diphenyl diisocyanate, from the polyurethane family) cures on its own under heat. The panels glued with MDI came out measurably denser, and on the oak-lean layout the MDI version beat its MUF twin in both the deflection it could take and the load it took to break it. On the balanced layout the two glues finished nearly level. So MDI looked promising — but I will say it the way the thesis says it: the evidence pointed that way, and more testing is needed before calling a winner. An honest result is worth more than a confident one.

The last chapter left the lab entirely. Using finite element analysis (FEA) in Ansys, we rebuilt the panels digitally — every veneer layer modelled in its place — loaded the virtual beams with 400 N, and compared the predicted bending against what the real samples did. The simulation landed within 3% to 9.2% of the measured deflections. That gap matters less than what it proves: once your model is calibrated against real broken samples, you can try a new layup on screen before ever heating the press. Test in the computer, confirm in the wood.

Why this matters for your furniture in Amman

You will probably never order a panel of Turkey oak LVL from me. But the thinking behind that thesis is in every quote I write. When a client asks for a 2.4-metre bookshelf, the question is never "what is the best wood?" — it is "where does the load live, and what does each part of this piece actually need?" A shelf that must not sag wants stiffness along its length; a tabletop wants a stable core with a beautiful face; a drawer bottom needs almost nothing. Paying for premium material in places that carry no stress is not luxury — it is waste.

It is also why I refuse the phrase "fake wood" for engineered boards. I have stood at the press that makes this material. Plywood, blockboard, LVL — these are real wood, re-arranged by engineering to be stronger, more stable and more resource-efficient than the tree could manage on its own. A veneered engineered core in the right place is often the more honest choice, not the cheaper trick. And the reverse holds too: where solid wood is genuinely the right answer — a tabletop edge that will be sanded for decades, a chair joint under racking stress — I will tell you that, with the numbers behind it.

Resource honesty is the deepest lesson. Hungary looked at its most common, least glamorous tree and asked how to make it structural. In Jordan we import nearly all our timber, which makes the same question sharper: every board that arrives here has travelled far and cost real money, so it deserves to be used where it works hardest. That is what I learned pressing panels in Sopron, and it is how I build in Amman.

From the lab bench to the workshop bench

People sometimes ask why a furniture maker needs a wood engineering degree. The thesis is my answer. Breaking sample after sample teaches you a respect for wood that no catalogue can: you learn that density is destiny, that glue is a design decision, that poplar will drink your adhesive budget if you let it, and that a confident claim without a test behind it is just a guess wearing a suit.

Nothing in my workshop gets pressed at 90 bar. But every piece that leaves it is built on the same three questions I asked in that lab: what is this material actually good at, where does the load really go, and can I prove it — not just feel it? When I tell you a shelf will not sag, or that an engineered core will outlive a solid panel in your climate, that confidence was earned in Sopron, one broken sample at a time. That is what I brought home to Amman.