Laminated veneer lumber from the species nobody wants
Turkey oak is 11.4% of Hungary’s forest and is mostly burned. Hybrid poplar grows to size in eight years and is mostly pulped. Neither is treated as a structural material. This is an M.Sc. thesis that asked whether a veneer layup could turn them into one — eight boards pressed, tested in three-point bending, and modelled layer by layer in Ansys.
The tool below is that work made usable. Stack veneers in any order and it predicts what the pressed board would do, from a model fitted to the boards that were actually made.
Drag to orbit · click a ply to select it · 1–4 adds a veneer · E explodes the stack · P steps through the ten layouts.
Open full screen ↗Eight boards
Turkey oak veneer at 1.4 mm and poplar I-214 at 2.8 mm, dried at 70 °C for ten hours, spread with MUF or MDI adhesive, pressed at 140 °C — 90 bar for five minutes, then 20 bar for eight, with a slow decompression so trapped vapour cannot blow the board apart.
Three-point bending
Specimens cut to 400 × 19 × 14 mm over a 360 mm span. All-poplar reached 59–77 MPa, all-oak 114.7 MPa. The balanced sandwich — twin oak faces over a poplar core — reached 94.7 MPa on roughly two-thirds poplar by volume.
Modelled, not guessed
A layer-resolved finite element model in Ansys predicted deflection under a 400 N load to within a few per cent of the measured value. Being able to model a layup, rather than press one and hope, is the point of the whole exercise.
Every ply counts by the square of its distance from the middle
Transformed-section beam theory: a veneer’s contribution to stiffness grows with the square of how far it sits from the neutral axis, so the face plies decide almost everything and the core is mostly there to hold them apart. Failure is set by the first ply to reach its own strength.
Ply properties were fitted to the eight pressed boards rather than taken from a table. The fitted poplar modulus came out at 10,303 MPa against the 10,340 MPa literature value used independently in the Ansys model — nothing in the fit was told that number.
Two boards in the set — P1 and P2 — are the same layup pressed twice, and they differ by 21% in stiffness and 30% in strength. That is the noise floor of the experiment. The model sits inside it, and cannot honestly claim to be better than the thing it was fitted to.
What each layup buys for the oak it spends
The last column is the share of the all-oak strength reached per share of oak used. Solid oak scores 1.00 by definition. Three of these layouts were drawn on the original plan sheet and never pressed — the model is the only thing that has an opinion about them.
| Layout | Stack | MOR | of all-oak | oak used | per oak | |
|---|---|---|---|---|---|---|
| 2 oak / 5 poplar | O·P·P·P·P·P·O | 75.8 | 66% | 17% | 3.96× | never pressed |
| L3 — 4 oak / 4 poplar | O·O·P·P·P·P·O·O | 97.8 | 85% | 33% | 2.56× | pressed |
| L1 — 3 oak / 4 poplar | O·P·P·O·P·P·O | 77.5 | 68% | 27% | 2.48× | pressed |
| 4 oak / 4 poplar, inverted | P·O·O·P·P·O·O·P | 62.0 | 54% | 33% | 1.62× | never pressed |
| L2 — 7 oak / 2 poplar | O·O·P·O·O·O·P·O·O | 103.1 | 90% | 64% | 1.41× | pressed |
| 11× Turkey oak | O × 11 | 114.7 | 100% | 100% | 1.00× | pressed |
| 7 oak / 2 poplar, inverted | P·O × 7·P | 62.0 | 54% | 64% | 0.85× | never pressed |
| 6× poplar | P × 6 | 60.1 | 52% | 0% | —× | pressed |
The two inverted layouts land on exactly the same number — 62.0 MPa — although one of them contains nearly twice the oak. Once poplar is on the face, the face fails first and everything underneath it is wasted. That is the whole argument for putting the good species on the outside, stated as a number rather than a preference.
From Development of Laminated Veneer Lumber Using Underused Hungarian Species, M.Sc. Timber Industry Engineering, University of Sopron, 2024 — supervised by Prof. Dr. Tibor Alpár, consulting Prof. László Bejó. The experimental work was carried out in the university’s engineered wood-products laboratory.
Beech and Scots pine appear in the tool as planned layouts that were never pressed; their ply properties come from the literature and are marked as such wherever they appear. Everything else on this page is measured.