2. Seatbelt geometry and kinematics

Three-point seatbelt performance is determined less by the webbing itself than by where the belt is anchored, how it is routed across the body, and how the geometry changes as the occupant moves. A belt that looks correct in a static showroom check can still allow excessive forward excursion, submarine under the lap belt, or neck contact with the shoulder belt during a crash. This section treats the restraint as a kinematic linkage: the anchors define the fixed points, the D-ring defines the upper turning point, and the occupant's body defines the moving load path. Understanding these relationships is the foundation for every later decision about pretensioners, load limiters, and buckle geometry.

Coordinate Convention and Reference Frame

Before any geometry discussion, fix a reference frame. Use the vehicle coordinate system: X positive forward, Y positive to the occupant's left (driver side), Z positive upward, with the origin at the vehicle datum. All anchor positions in this section are expressed as (X, Y, Z) in millimeters relative to that datum. Occupant excursion is measured along X (forward displacement) and Z (vertical displacement), with positive X meaning forward motion relative to the seat. Keep this convention consistent; mixing seat-relative and vehicle-relative coordinates is one of the most common sources of geometry errors in restraint design.

Anchor Points: The Fixed Geometry

A three-point belt has four anchor points: the lap anchor outboard (nearest the door), the lap anchor inboard (buckle side), the shoulder anchor (upper B-pillar or seat-integrated), and the retractor (usually at the lower B-pillar or seat base). The lap anchors and buckle define the lap belt plane; the shoulder anchor and D-ring define the shoulder belt path.

Regulatory and engineering targets constrain these points. FMVSS 209/210 and ECE R16 define minimum strength and geometry requirements, but the practical design targets are more specific. A typical front-seat lap anchor outboard sits at approximately X = 200–280 mm, Y = ±380–420 mm, Z = 180–260 mm. The inboard buckle anchor is typically at X = 180–260 mm, Y = ±180–220 mm, Z = 160–240 mm. The shoulder anchor is usually at X = −100 to +50 mm (slightly behind or at the occupant's shoulder), Y = ±350–400 mm, Z = 900–1100 mm. These are illustrative ranges, not universal values; they shift with seat position, occupant size target, and vehicle package.

The critical relationship is the lap belt angle. Measured in side view, the lap belt should sit between 45° and 65° to the horizontal when loaded, with the ideal around 55°. Too shallow (below 45°) and the belt can slide over the iliac crest and load the abdomen; too steep (above 65°) and it can dig into the thighs and lose pelvic engagement. The lap belt must engage the anterior superior iliac spine (ASIS) — the bony hip points — not the soft tissue above them.

D-Ring Position and the Shoulder Belt Path

The D-ring (also called the shoulder belt guide or turning loop) is the pivot that redirects the shoulder belt from the retractor to the occupant. Its position controls two things: the shoulder belt angle and the belt-to-neck clearance.

The shoulder belt should cross the occupant's chest from the outboard shoulder toward the inboard hip, passing over the clavicle and sternum. The target angle in front view is 10°–25° from vertical, with the belt crossing the mid-clavicle. If the D-ring is too far inboard, the belt crosses the neck; if too far outboard, it can slip off the shoulder. In side view, the shoulder belt should lie flat against the chest with minimal gap — a gap greater than about 25 mm at the sternum indicates poor routing and increases excursion before the belt engages.

The D-ring height is adjustable in most vehicles, typically over a range of 50–80 mm. This adjustment exists because the same belt must fit a 5th-percentile female and a 95th-percentile male. The design target is usually set so that the mid-range D-ring position gives correct routing for a 50th-percentile male, with the adjustment range covering the extremes. A useful rule: the D-ring should be at or slightly above the occupant's shoulder height, never below it, because a low D-ring allows the shoulder belt to slide off the shoulder during forward excursion.

Occupant Excursion Paths

Excursion is the forward and downward movement of the occupant relative to the seat during a crash, before the belt fully engages. It is the single most important kinematic output of the geometry. Two paths matter:

  • Upper torso excursion: forward rotation of the chest and head about the hip. The shoulder belt limits this by going into tension. Typical target: less than 300 mm forward head excursion in a 50 km/h frontal barrier test, though this varies by regulation and vehicle class.
  • Pelvis excursion and submarining: forward and downward sliding of the pelvis under the lap belt. This is the dangerous failure mode. It occurs when the lap belt angle is too shallow, when the seat cushion is too soft, or when the occupant is out of position. The result is the belt loading the abdomen instead of the pelvis.

The excursion path is not a straight line. As the torso rotates forward, the shoulder belt pays out from the retractor, the D-ring becomes the pivot, and the lap belt tightens against the pelvis. The sequence matters: if the lap belt engages first, the pelvis is locked and the torso rotates about a stable base. If the shoulder belt engages first, the torso can rotate and pull the lap belt upward, creating a path to submarining.

Design principle: The lap belt must engage before or simultaneously with the shoulder belt. This is achieved by keeping the lap belt path short and direct, and by ensuring the retractor locking mechanism and pretensioner (if fitted) act on the shoulder belt without pulling the lap belt upward.

Worked Example: Checking a Front-Seat Geometry

Consider a mid-size sedan front seat. The occupant is a 50th-percentile male, H-point at X = 0, Z = 0 (seat-relative). The lap anchor outboard is at (240, 400, 220) mm, the inboard buckle at (220, 200, 200) mm, and the D-ring at (−20, 370, 980) mm. The shoulder anchor (retractor) is at (−150, 390, 300) mm.

Step 1 — Lap belt angle. The lap belt runs from (240, 400, 220) to (220, 200, 200). The vertical drop is 20 mm over a horizontal run of 200 mm in Y and 20 mm in X. The angle to horizontal is approximately arctan(20 / √(200² + 20²)) ≈ 5.7°. This is far too shallow — well below the 45° minimum. The belt would sit on the abdomen, not the pelvis. Correction: raise the outboard anchor to Z = 260 mm and lower the inboard to Z = 180 mm, giving a drop of 80 mm over 200 mm, or about 21.8°. Still too shallow. The real fix is to move the anchors rearward and downward relative to the occupant, or to use a seat-integrated buckle that positions the lap belt across the ASIS. In practice, the lap belt angle is set by the seat structure and the buckle stalk length, not by the anchor alone.

Step 2 — Shoulder belt angle. From D-ring (−20, 370, 980) to the occupant's shoulder at approximately (0, 200, 900), the belt runs 170 mm inboard and 80 mm down. The front-view angle from vertical is arctan(170 / 80) ≈ 64.8°, which is far too horizontal. The belt would cross the neck. Correction: move the D-ring outboard to Y = 300 mm and up to Z = 1020 mm. New angle: arctan(100 / 120) ≈ 39.8°. Still above the 25° target. The D-ring must be closer to the occupant's shoulder line — typically Y = 250–280 mm for this seat position.

Step 3 — Excursion estimate. With the corrected geometry, the shoulder belt engages after approximately 80–120 mm of forward torso rotation, depending on belt slack and retractor lock time. The lap belt engages almost immediately because the path is short. The expected head excursion is in the 250–320 mm range for a 50 km/h sled test, which is within typical targets. Verification would be by sled test with a Hybrid III 50th-percentile male dummy, measuring head CG forward displacement and checking for submarining via lap belt force and pelvis acceleration traces.

Common failure symptom: If the lap belt force trace shows a late rise relative to the shoulder belt force, or if the pelvis acceleration shows a downward component before the lap belt peaks, submarining is likely. The fix is almost always geometric: increase lap belt angle, move the buckle rearward, or stiffen the seat cushion.

Practical Geometry Check Sheet

ParameterTargetCheck Method
Lap belt angle (side view)45°–65°, ideal 55°CAD measurement at loaded position
Shoulder belt angle (front view)10°–25° from verticalCAD measurement at mid-clavicle
D-ring height vs. shoulderAt or slightly aboveSeat position sweep, 5th–95th percentile
Belt-to-sternum gap< 25 mmPhysical check with occupant
Lap belt engagement pointASIS, not abdomenPalpation or pressure film
Head excursion (50 km/h)< 300 mm typical targetSled test, Hybrid III

These checks are not independent. Changing the D-ring height alters the shoulder belt angle and the excursion path. Moving the lap anchor changes the lap belt angle and the buckle position. The geometry must be evaluated as a system, with the occupant's H-point and seated posture as the reference. The next section builds on this by examining how pretensioners and load limiters modify the force path once the geometry is fixed.

Three-Point Belt Geometry: Side View with Excursion Path Seat back Seat cushion Head Torso Pelvis Lap anchor outboard (240, 400, 220) mm Buckle anchor (220, 200, 200) mm D-ring (−20, 370, 980) mm Retractor (−150, 390, 300) mm Lap belt Shoulder belt Head excursion ~250–320 mm forward Pelvis path θ_lap θ_shoulder Legend Lap belt Shoulder belt Excursion path Anchor point D-ring / pivot Coordinates in vehicle datum (X,Y,Z) mm X forward, Y left, Z up