Major Poisson's Ratio of a Lamina

Pull a lamina along its fibres and it narrows across them. The ratio of that contraction to the extension causing it is the major Poisson's ratio $\nu_{12}$, and it is taken as the same volume-weighted average that gives the longitudinal modulus: $\nu_{12} = \nu_f V_f + \nu_m (1 - V_f)$. The justification is the same iso-strain argument, and it works for the same reason.

The subscripts carry more information than the number does, and this is where an isotropic habit does real damage. In $\nu_{12}$, the first index names the direction of the applied stress and the second the direction of the strain being measured — so this is: load along the fibres, measure across. The minor ratio $\nu_{21}$ — load across, measure along — is a completely different and much smaller number. They are tied by the reciprocal relation $\nu_{21} = \nu_{12} E_2 / E_1$, and with a longitudinal modulus an order of magnitude above the transverse, $\nu_{21}$ typically lands near 0.02 to 0.03 rather than near 0.3.

Both are correct; they answer different questions; and putting the wrong one into a lamination-theory compliance matrix produces a matrix that is not symmetric. That asymmetry is a good self-check, because the reciprocal relation is not a modelling choice — it follows from the existence of a strain energy density, and any set of lamina constants that violates it is wrong somewhere.

Now the reassuring part. Of all the micromechanics relations, this weighted average is the crudest, and it is also the least consequential. $\nu_{12}$ enters the ply stiffness matrix mainly through the term $1 - \nu_{12}\nu_{21}$, and since $\nu_{21}$ is around 0.03, that term is within a percent of 1 for any real lamina. A rough $\nu_{12}$ is usually good enough, which is fortunate, because measuring a Poisson's ratio well is harder than measuring a modulus well and the scatter is correspondingly wider. That is also why solving this equation backwards for a volume fraction, though arithmetically valid, is the worst method on this site: the two constituent ratios differ by perhaps 0.15, so the entire range of $V_f$ from 0 to 1 moves $\nu_{12}$ across a span narrower than ordinary strain-gauge scatter.

One genuine surprise is worth preparing for. For a LAMINATE rather than a single ply, the in-plane Poisson's ratio can exceed 0.5, and an angle-ply lay-up near ±25° can show ratios above 1. Every engineer who meets this for the first time assumes an error, because 0.5 is drilled in as a hard ceiling. It is a hard ceiling for isotropic solids, where it corresponds to incompressibility and where exceeding it would mean a material that gains volume as you squeeze it. An angle-ply laminate is not isotropic and the mechanism is different: the contraction comes from the plies scissoring against one another, a geometric effect of the lay-up rather than a material one, and no thermodynamic limit is being violated. The same freedom is what makes auxetic laminates with genuinely negative Poisson's ratios possible.

Two last cautions. The constituents themselves are not isotropic — a carbon filament has different ratios along and across its axis, and the value that belongs in this average is the one governing contraction transverse to the fibre under axial load. And this equation gives an elastic constant, not a strength. It says nothing about when the ply fails, which is a separate question with a separate criterion.