Guides / Design

Landing gear: designing the part you want to fail first

August 9, 2026

Landing gear is the only structure on the aircraft designed around an event that should not happen. Normal landings are gentle. The design case is the arrival that was not.

That changes how the parts should be drawn.

Work out the load path first

Trace the energy from the ground to the airframe and write down every part it goes through. On a typical commercial multirotor it looks like: foot, leg, leg clamp, arm or fuselage boss, structure.

Now decide which of those you are willing to replace. That part should be the weakest one in the chain, by design, and the rest should have margin over it. Without that decision, the weakest link is whichever part happened to be thin, which is usually the expensive one.

The parts, and what each needs

Feet and skid blocks. They take abrasion, grit and impact. PA66 is tough and cheap, POM machines to better tolerance and slides, and a rubber overshoe adds grip and damping. Replaceable, held with two screws, and it should be obvious to a technician which way round it goes.

Legs. Usually carbon tube or aluminium tube. Not a machined part, but the machined parts at both ends decide whether the tube is loaded in line or in bending.

Leg clamps. 7075-T6, split clamps with real ears. The rules are the same as an arm clamp, and the same crushing and slipping failures apply. See boom clamp fit on a carbon tube.

Hinges and pivots on folding gear. A pin, two lugs and a bush. Get the bush material right, POM or a bronze bush rather than metal on metal, or the hole becomes oval after a season.

Latches. The part that fails quietly. A latch that loosens lets the leg fold in flight, so design it so that wear increases friction rather than releasing it.

Sizing for impact rather than for weight

A static weight calculation will under size landing gear badly. A descent arrested in a short distance produces a peak force many times the aircraft weight, and the energy has to go somewhere.

Three practical approaches, in order of how much they are worth:

1. Make something compliant. A leg that flexes spreads the arrival over more time and reduces the peak force. This is why sprung or angled legs beat vertical struts.
2. Make something sacrificial. A shear feature, a plastic foot, a frangible link. Choose it and price it into the maintenance plan.
3. Make the rest strong. Once the first two are decided, everything upstream needs margin over the sacrificial part, not over the nominal landing.

Machining details that decide the failure mode

• Radius every internal corner in a loaded part. A sharp corner in a 7075 clamp ear is where the crack starts. Larger radii are also cheaper to cut, which is the rare case where strength and cost agree.
• Keep material around the fastener holes. A lug with a hole close to the edge tears out through the edge. Two diameters of material from hole centre to edge is a reasonable starting point.
• Avoid machining across the grain in plate. For highly loaded 7075 parts, orientation in the plate matters. Say so on the drawing if the part is critical.
• No cosmetic weight saving in the load path. Pocketing a clamp ear to look aggressive is how a clamp becomes a hinge.
• Deburr properly. Sharp edges concentrate stress and cut technicians who are catching a falling aircraft.

Finish, because landing gear lives in mud

Hard anodise on the aluminium parts. It resists abrasion from grit, it protects 7075 which needs it, and it survives washing. Feet in plastic do not need a finish, and a bright colour on a replaceable foot is a free way to tell a technician which part is meant to be replaced.

For agricultural aircraft, add the drainage and sealing notes in stainless or aluminium for agricultural parts. Landing gear sits in whatever is on the ground.

Testing, cheaply

You do not need a drop tower. A repeatable drop from a known height onto a known surface, with the aircraft mass simulated, tells you where the part breaks and whether it broke where you wanted.

Do it before the first production lot, not after. Changing a foot design after 500 pieces are anodised is an expensive way to learn which part is the fuse.

What to send

The aircraft mass, the intended landing surfaces, the descent rate you want it to survive and which part you have chosen to be sacrificial. With that we can look at the parts and say which ones we think are under built before we quote them.

Reply comes back within 24 business hours through the RFQ page, and 7075 is on the material list as standard stock.

Frequently asked questions

Keep reading

• UAV component machining: what to put on the drawing
• Taking weight out of a machined drone part without losing the part