Automotive lightweighting: what one kilogram is really worth

The question every automotive engineer is actually asked

Take a kilogram out of a car and you get a chain of consequences. Published work on mass decompounding describes the effect: lighter body structure lets you shrink the brakes, the suspension, the steering, and in an electric car the battery itself — and each of those reductions permits another. Secondary savings can approach the primary saving again, which is why a clean-sheet design beats swapping one panel on an existing platform.

The pay-off is real but modest per kilogram. Published figures put a 10% vehicle mass reduction at roughly 6–8% better fuel economy or range, and the energy reduction value for battery cars in the literature sits around 0.5–1.2 kWh per 100 km per 100 kg. So the engineering question is never simply which material is lightest. It is which material is lightest at the stiffness and strength the part actually needs — and, crucially, in the shape it is.

The mistake: comparing materials without the geometry

Almost every popular comparison quotes specific stiffness, E/ρ, and stops there. That number is correct only for a tie — a bar loaded in pure tension, where stiffness scales with cross-sectional area and nothing else.

Most car structure is not a tie. It is panels and beams loaded in bending, and bending changes the arithmetic completely. For a beam of fixed length and shape, stiffness scales with E·I, and if you are free to change the depth the mass at equal stiffness goes with ρ/E^(1/2). For a flat panel of fixed width and length, free to change thickness, it goes with ρ/E^(1/3).

Those exponents are not decoration. They decide which material wins, and they can reverse a ranking.

The numbers, with a real laminate

Raw fibre properties overstate what a composite part achieves, because a real panel needs fibres pointing in several directions. So instead of quoting E₁ = 181 GPa for unidirectional T300/5208, we ran a genuine quasi-isotropic layup — [0/+45/−45/90]s, eight plies, 1.0 mm — through this site's laminate calculator.

It returns Ex = Ey = 69.7 GPa, Gxy = 26.88 GPa, νxy = 0.296. As a check that the layup really is isotropic in plane, E/2(1+ν) = 69.7/2(1.296) = 26.88 GPa, exactly the computed shear modulus. The laminate behaves like an isotropic sheet in its own plane, which is what quasi-isotropic means.

Now the punchline, and it is the single most useful fact in this article: 69.7 GPa is essentially aluminium's 70 GPa. A realistic carbon panel is not stiffer than aluminium. It is the same stiffness at 1.6 g/cm³ instead of 2.7 — about 59% of the density.

Ranking the three indices (higher is better) gives: for a tie, steel 25.3, aluminium 25.9, carbon 43.6. For a beam, steel 1.79, aluminium 3.10, carbon 5.22. For a panel, steel 0.74, aluminium 1.53, carbon 2.57.

What those numbers mean

Against steel, carbon's advantage grows with the geometry: 1.7× as a tie, 2.9× as a beam, 3.5× as a panel. Put differently, at equal panel bending stiffness a carbon panel weighs about 29% of the steel one, and an aluminium panel about 49%.

Against aluminium, something neater happens. Carbon wins by 1.7× in all three modes — not a coincidence. Because the two moduli are nearly identical, every index ratio collapses to the density ratio, 2.7/1.6 = 1.69, whatever the exponent. The entire carbon-versus-aluminium argument, for a quasi-isotropic panel, is a density argument.

And glass fibre shows why the geometry matters. Quasi-isotropic E-glass/epoxy comes out at just 18.9 GPa. On the tie index it is 9.0, far worse than aluminium's 25.9. On the panel index it is 1.27 — now beating steel's 0.74, though still short of aluminium. Glass earns its place in large, thick, cost-driven panels and loses badly in a tension member. Same material, opposite verdict, decided entirely by shape.

Why cars are still mostly steel

None of the above mentions cost, rate or repair, and those are usually what decide. Steel stamps in seconds, composite parts cure in minutes to hours. Steel body repair is a solved, cheap, distributed process; a damaged structural composite usually means replacement. Steel recycling is mature and profitable.

Published estimates of what manufacturers will pay to remove mass — roughly a few euros per kilogram for a compact car, more for a long-range SUV — set a hard ceiling that carbon fibre generally exceeds for mass-market parts. That is why carbon appears where the value of a kilogram is highest (motorsport, supercars, aerospace) or where a part is low-volume, and why the mainstream lightweighting story is high-strength steel and aluminium rather than composites.

How to use this

Identify the loading mode before you compare anything. Tension, bending of a beam, or bending of a panel each have their own index, and the answer can flip between them.

Use realistic laminate properties, not fibre or UD values. The gap between 181 GPa and 69.7 GPa is the difference between a brochure and a part.

Check whether you are free to change thickness. The beam and panel indices assume you are; if thickness is fixed by packaging or a minimum-gauge rule, you are back to comparing E and ρ directly.

And do the whole-vehicle sum, not the part sum. A kilogram saved on a load path is worth more than a kilogram saved on a bracket, because only the first one lets something else shrink too.

A note on the numbers here

The laminate properties are computed live by this site's reference-tested classical laminate theory engine from ply data cited to Daniel & Ishai. Steel and aluminium are quoted at conventional textbook values (200 GPa / 7.9 g/cm³ and 70 GPa / 2.7 g/cm³) for comparison.

The decompounding, fuel-economy and cost-per-kilogram figures come from published lightweighting literature and vary widely with vehicle class and drive cycle. Treat them as orders of magnitude for reasoning, not as design inputs.