All articles

Honeycomb sandwich panels: stiff, light, and how they fail

Two thin skins, a light honeycomb or foam core between them, and a panel many times stiffer than the same material solid. How sandwich beams bend, why the core's shear adds deflection, and the four ways a sandwich fails: facing strength, core shear, wrinkling and intracell buckling.

The I-beam idea, spread into a panel

Bend a plank and its top surface squashes while its bottom stretches; the middle barely works at all. Material far from the middle — the neutral axis — does nearly all the resisting, and its contribution grows with the square of its distance from that axis. An I-beam exploits this by putting its steel into two flanges held apart by a thin web.

A sandwich panel is the same idea spread into a sheet: two thin, stiff skins, called facings, bonded to a thick and very light core. The skins are the flanges and the core is the web. Push the skins further apart with a thicker core and the bending stiffness climbs steeply, while the weight hardly moves, because the core is mostly air.

Hexcel's classic comparison shows how steeply. Split a solid sheet into two skins and bond them either side of a honeycomb core as thick as the sheet was: the panel is about 7 times stiffer and 3.5 times stronger in bending, for roughly 3% more weight. Make the core three times that thickness and it is about 37 times stiffer and 9 times stronger, for roughly 6% more weight.

The skins carry bending, the core carries shear

Thin-face sandwich theory splits the work cleanly. The facings carry the bending moment as tension in one skin and compression in the other, at a stress of M/(h·t_f·b): the moment divided by the skin's cross-section and by h, the distance between the centres of the two skins. The core carries essentially all the transverse shear, at a stress of about F/(h·b).

That gives the core three jobs. It must hold the skins at a fixed distance without crushing, carry shear from one skin to the other, and stay bonded to both. Honeycomb made from expanded aluminium foil or aramid (Nomex) paper does all three at densities of roughly 30 to 150 kg/m³, a few percent of solid aluminium. Foams and end-grain balsa are the other common cores.

The deflection a solid-beam formula misses

Because the core is soft in shear, a sandwich beam deflects in two ways at once. The skins bend the beam as usual, adding a deflection proportional to P·L³/D, where D is the flexural stiffness. The core also shears, sliding one skin along the other, and adds a second deflection proportional to P·L/S, where S = b·h·G_c is the shear stiffness. A textbook beam formula has only the first term.

Which term matters depends on the shape. Bending deflection grows with the cube of the span, but shear deflection only in proportion to it, so a long, slender panel deflects almost purely in bending, while a short, deep panel on a soft core can take a large share of its deflection in core shear. The calculator reports both parts and the shear share, so you can see which regime a design sits in.

Four ways a sandwich fails

Facing failure: the skins yield or fracture when the bending stress reaches their strength. Core shear failure: the core tears in shear where the shear force is highest, near the supports under a uniform load.

Facing wrinkling: the compressed skin buckles locally in short waves, pushing into the core or pulling away from it. How stiffly the core supports the skin sets the critical stress — 0.5·(G_c·E_c·E_f)^⅓ in Hexcel's form — so a stiffer core resists it. Intracell buckling, or dimpling: over a honeycomb, the skin can buckle into the empty cells like a drum skin between its supports, at 2·E_f·(t_f/s)² for cell size s. Thin skins over large cells are the risk.

Whichever check has the lowest safety factor governs the design. Real panels can fail in other ways too: the core crushing under a local load such as a fastener or a foot, the skin-to-core bond letting go, or the whole panel buckling in compression. Each needs its own check, and the calculator says which four it covers.

Honeycomb has a grain

Honeycomb is made by bonding stacked sheets along offset stripes and pulling the stack open, so it has two directions: the ribbon direction (L), along the continuous sheets, and the transverse direction (W), across them. In shear, L is the stiffer and stronger one — for common Nomex and aluminium cores, roughly 1.7 to 2 times W.

So the direction a beam spans matters. Use the L shear modulus and strength when the span runs along the ribbon, and the W values when it runs across. Core datasheets list both, and the calculator's core presets come in L and W pairs for exactly this reason.

Where you meet sandwich panels

Aircraft floors, flaps and fairings; helicopter rotor blades; the monocoque of a racing car; satellite panels; boat hulls and decks on foam or balsa cores; the shells of wind turbine blades. Wherever stiffness per kilogram decides the design, a sandwich is usually the answer.

You have handled them too. Corrugated cardboard is a sandwich of two paper skins on a fluted core, and a hollow-core door is two thin skins held apart by a paper honeycomb. Both are light for exactly the reason an aircraft floor is: the skins sit far apart, and the core between them weighs almost nothing.

solid sheetstiffness ×1sandwichstiffness ×7same material, same load: stiffness = (c + 1)³ − c³

Split a solid sheet into two skins and hold them apart with a light core. For a core c sheet-thicknesses deep, bending stiffness grows by (c + 1)³ − c³, so a core as thick as the sheet makes the panel about 7 times stiffer for almost no extra weight. The picture shows bending only: a soft core adds its own shear deflection, which the sandwich calculator includes.

1. Why is a sandwich panel so much stiffer than a solid sheet of nearly the same weight?

2. A sandwich beam deflects more than a standard solid-beam formula predicts. Why?

3. What is intracell buckling?

Sandwich panel calculator →