Sarah Clarity Studios

Chapter 2: Rigging for Movable Parts

Created by Sarah Choi (prompt writer using ChatGPT)

Rigging for Movable Parts (Suspension, Gear, Flaps)

Partnering with Gameplay, Physics, Animation, VFX & Audio

Movable parts are where vehicle concepts meet interaction. Suspension, landing gear, and aerodynamic devices give the player feedback, carry gameplay metrics, and drive VFX and audio exactly where they need to be. A good rig reads the design sentence, respects metrics, and exposes clean controls so animation, physics, and VFX can collaborate without reverse‑engineering intent. This article shows concept‑side and production‑side artists how to plan rigs from the first proportion pass, define the right limits, and hand off jointed systems that are reliable, performant, and easy to automate.

1) Start with the metric sentence and envelopes

Before drawing linkages, declare what the vehicle must physically do. For wheeled platforms, write a short sentence that fixes wheelbase, track, nominal ride height, bump and droop travel, approach/departure/breakover angles, curb hop, and turn circle. For aircraft and VTOL, add rotation clearance, gear retraction volume, disk or jet plume keep‑outs, and minimum landing pad grid. For all vehicles, visualize these as transparent envelopes in side, plan, and front views. These shapes become the non‑negotiable limits your rig will respect; if a linkage violates an envelope, the design changes, not the envelope.

2) Rig philosophy: truth over theatrics

A readable rig privileges believable motion paths and consistent pivots over cinematic tricks. Joints are named by function, axes are declared once and never drift, and limits are given as numbers that match the metric plates. The default pose is neutral and serviceable, not a beauty stance. When the rig is simple and truthful, gameplay can predict clearances, physics can match springs and damping, and VFX can bind emitters to stable sockets. The bonus is that damage and variant automation become trivial because behaviors are portable across the vehicle family.

3) Suspension systems: choices, signals, and constraints

On wheeled vehicles, the suspension rig communicates mass, traction, and intent. A double‑wishbone reads sporty and precise, a MacPherson strut reads compact and utilitarian, a trailing arm reads rugged and tolerant of debris, and swing arms on bikes read agile and exposed. Pick the linkage that supports the silhouette and packaging, then declare its hardpoints on the orthos. In rig, each corner receives a vertical spring travel, a steering angle (front corners by default), and camber/caster/toe behaviors if they matter for the read. When gameplay emphasizes weight shift, add a simple anti‑roll behavior that biases left/right compression under lateral acceleration even if full simulation is not present.

Tracks and walkers obey the same clarity rule. A tracked rig needs road wheels with limited vertical play, an idler and drive sprocket, and a belt deformation proxy that maintains contact with the ground collider. A bipedal or quadrupedal walker needs a stride triangle, hip, knee, and ankle pivots with reachable foot targets that do not penetrate the collision hull. In all cases the foot or tire contact must be explicit so dust, splashes, and skid marks can spawn exactly where the material map says they should.

4) Metrics that drive the suspension rig

The suspension rig lives on a handful of numbers: bump and droop travel per corner, maximum steer angle and rate, scrub radius or kingpin offset for steering feel, and ground clearance at nominal ride. These are written on the metric plates and copied verbatim into the rig spec. The concept artist should show stance at neutral, full bump, and full droop in side and front plates so production can match silhouette consistency at extremes. If a hero fender or skirt collides at full bump, there is still time to cut a relief, re‑contour a lip, or widen the track. The rig must preserve approach and departure angles; if those change during motion, gameplay silhouettes will lie.

5) Steering, Ackermann, and animation readability

Steering reads in the tire and in the fender gap. Even when a full solver is out of scope, build plausible toe‑out on turns by offsetting the steering rack or using a simple driver that gives the inner wheel a few extra degrees. Keep it subtle and consistent with your wheelbase and track; extreme toe change looks broken at gameplay distance. Name steering joints clearly—steer.FL and steer.FR—and parent calipers or hub details appropriately so kitbashed parts don’t pop or slide when the wheels turn.

6) Landing gear: retraction logic and safety interlocks

Landing gear rigs carry sequence and safety. A typical flow is: doors open, gear downlock releases, strut extends, bogie rotates, doors close; or the reverse for retraction. Model the uplock/downlock in silhouette even if the latch is not visible; these define dwell points that animators can hit. Add weight‑on‑wheels and squat proxies so the rig can send events to gameplay, VFX, and audio when the vehicle touches down. Publish clear limits for stroke length, caster shimmy on taxi, and door clearance envelopes so painters can’t tuck complex geometry into impossible volumes.

For VTOL or tiltrotor craft, the gear must clear the disk or jet plume and maintain a three‑point stance under crosswind. The rig exposes a quick “pad fit” control that spreads or narrows stance to match the pad grid metric, and a “kneel” control if gameplay requires ramp loading. Keep pivots real, label axes, and sequence doors and legs conservatively so the default cannot clip the fuselage, even at low frame rates.

7) Flaps, slats, spoilers, and thrust vectoring

Aero surfaces sell speed, control, and damage. Flaps read as lift/drag devices during takeoff and landing, slats read as high‑lift aids at the leading edge, spoilers read as braking plates, and elevons read as agile control on deltas. Give each device a clean hinge axis, a default neutral angle, and a realistic max deflection. If the design uses flaperons or elevons, couple roll and pitch behaviors with simple drivers so the read is believable. Spoilers should cast shadows onto the wing and interrupt the smooth crown so their presence is obvious even in clay. For rockets or VTOL, thrust vector nozzles need yaw and pitch axes and limits that align with the plume keep‑outs drawn on the metric plate; the rig should not allow vectors that ingest hot gases or impinge on the hull.

8) VFX and audio hooks: where the senses attach

Movable parts generate sound and effects. Place sockets at wheel contact patches, brake calipers, suspension top mounts, and bump stops so dust, sparks, and clunks can trigger from motion. For gear, add sockets at door latches and strut pistons for hydraulic hiss, at tire contact for screech and rubber thump, and at uplock catch for a metallic click. For flaps and spoilers, give a “separation onset” event near 70–80% deflection so aero noise layers can ramp. Mark exhaust, intake, and downwash cones with named volumes so particles and audio falloffs are consistent across variants. All of these hooks should be named consistently within the vehicle family to enable automation.

9) Physics handshake: simple bodies that behave

The rig must come with matching collision and mass proxies. On cars, this means a low‑poly chassis hull, four wheel colliders, and optional axle bodies for quick suspension tests. On tracked vehicles, use a single belly hull and convex hulls for skirts and turret to avoid snagging, then place a rectangular footprint for track friction. On aircraft, define fuselage and gear hulls and simple capsules for wings and tails to get contact and occlusion working. The mass distribution should mirror the kitbashed internals: engine or battery block, crew, and payload. If physics runs ahead of art, these proxies still let the team drive, land, and bounce the vehicle believably.

10) Control surfaces and UI feedback

Players and animators benefit from readable controls. Expose a small set of rig attributes—rideHeight, steer, brake, flap, spoiler, gearDeploy, kneel, tilt, vectorYaw, vectorPitch—and keep their ranges aligned with the metric plates. Animate indicator geometry or emissive accents for critical states like gear unsafe, flaps full, or suspension max bump so a quick silhouette check in engine tells the truth without a debug overlay. These reads also help QA catch regressions when ranges drift between versions.

11) Handoff: the rig spec that prevents drift

A proper handoff includes a clean hierarchy tree, a plate showing joint locations and axes, a limit table with numbers, and a short state matrix listing named clips such as Deploy Gear, Retract Gear, Brake Hard, Land, Taxi, Takeoff, Kneel, and Bank. Add matching collision proxies, a socket list for VFX/audio, and a unit/axis declaration. Provide clay renders at neutral and extremes so anyone can eyeball whether later changes broke clearances or signatures. Version the rig spec and annotate what changed when limits move; VFX and audio rely on names and ranges staying constant.

12) Common pitfalls and their fixes

Rigs often fail in predictable ways. Negative scales used to mirror corners flip normals and break animation; fix by instance‑mirroring or duplicating with positive transforms. Pivots placed for modeling convenience drift from real hinge lines; fix by snapping pivots to geometry and recording coordinates in the spec. Over‑constrained rigs that hide collisions with fake offsets backfire under latency; fix by simplifying constraints and honoring volumes. Unnamed or inconsistent sockets break automation; fix by adopting a family naming convention and validating it on export. Over‑aggressive spring and damper settings create floaty arcade reads; fix by matching visual travel to metric bump/droop and tuning rates until cameras and VFX feel grounded.

13) Case study A: light recon buggy

The brief demands alley fit, curb hop, and agile handling. The team locks ride height at 280 mm, bump/droop ±80 mm, approach 27°, departure 25°, and a curb‑to‑curb turn of 9.5 m. The rig exposes rideHeight and steer, with a simple Ackermann driver giving the inner wheel three extra degrees. VFX sockets sit on the four contact patches and the two brake calipers; audio hooks live at the shocks, diff, and exhaust. In graybox the buggy climbs a 30% ramp, drops 0.6 m without belly contact, and shows dust precisely at tire contact. Painters keep the fenders clear of full‑bump envelopes and the silhouette remains consistent at extremes.

14) Case study B: tiltrotor troop carrier

The design must fold for storage, land on urban pads, and taxi with confidence. The rig sequences doors → legs → doors with an uplock/downlock pause for readability. Stroke is 220 mm with a subtle squat‑to‑idle under weight‑on‑wheels. Flaperons couple to roll and pitch, spoilers deploy on braking, and the gear has a kneel control for ramp loading. VFX hooks include downwash cones that expand with throttle and tire dust that blends to skid screech on hard yaw. Audio taps hydraulic hiss on door open/close and a gear thump on uplock capture. The pack ships with neutral and extreme clay renders and a limit table; animation and VFX plug in without renaming anything.

15) Closing thoughts

Rigging movable parts is about turning constraints into readable motion. When concept publishes envelopes, hardpoints, and signature reads, and production binds joints, limits, and sockets to those truths, the whole studio wins. Gameplay gets honest traversal and clear states, physics gets stable contacts and masses, animation gets controls that perform, VFX gets anchors that never drift, and audio gets events that tell the right story. Do the simple things—real pivots, fixed axes, named sockets—exceptionally well, and the rest of the pipeline becomes faster, cleaner, and more fun.