3. Dummy Positioning and Belt Routing for ODB MPDB Driver
Why Dummy Positioning and Belt Routing Decide the Injury Numbers
In an ODB MPDB (offset deformable barrier, moving progressive deformable barrier) frontal crash model, the restraint system is only as valid as the initial dummy–belt–seat geometry. A Hybrid III 50th percentile male that sits 15 mm too far forward, or whose shoulder belt rides off the clavicle onto the neck, will load the chest and head through a different path than the physical test, and every injury criterion downstream — HIC36, chest deflection, Nij, femur loads — becomes non-comparable to the reference data. This case establishes a repeatable procedure to place the dummy, route the belt, and prove the setup is correct before the crash run, using pre-simulation checks rather than post-hoc inspection of results.
The task has three verifiable outputs: (1) a dummy H-point and posture that match the seating specification within tolerance, (2) a belt path that passes over the correct anatomical landmarks with defined clearance, and (3) a set of pre-simulation measurements (penetration, belt-to-body gap, initial belt strain, contact status) that pass explicit acceptance thresholds. Everything below is expressed in the model's global coordinate system: +X forward, +Y to the driver's left, +Z up, origin at the vehicle datum. All numeric values are teaching example values for a representative ODB MPDB driver setup and must be replaced with the values from your own test specification.
Step 1 — Establish the Seating Reference Before Touching the Dummy
Positioning is meaningless without a fixed seat state. Lock the seat track to the specification position and record it, because the dummy H-point is measured relative to the seat, not the vehicle.
- Seat track: set to the mid-detent position defined in the test request; record the track travel in mm from full-rear.
- Seat back angle: set the cushion-to-back angle to the specified value (example: 25° from vertical, measured on the back frame reference line).
- Head restraint: set to the specified height and tilt; it must not contact the dummy head at t = 0.
- Confirm the seat cushion is in its undeformed state — no pre-penetration from the dummy before gravity settling.
If the seat is left at an arbitrary detent, the dummy will be positioned correctly relative to the seat but incorrectly relative to the vehicle, and the belt anchor geometry will be wrong. Record the seat state in the setup log so the run is reproducible.
Step 2 — Place the Dummy and Settle It Under Gravity
Import the Hybrid III 50th percentile male and place it with an initial H-point approximately 20–30 mm above the final target, then let it settle. Do not hand-place the final position; gravity settling produces the pelvis rotation and cushion compression that the physical dummy exhibits.
- Set the dummy's initial H-point at the target X and Y, with Z raised by the settle allowance (example: target H-point at X = 1180 mm, Y = −365 mm, Z = 520 mm; initial Z = 545 mm).
- Apply a gravity load case (example: 9.81 m/s² in −Z) for a short settle run of 100–200 ms with the seatbelt unroute and no contacts other than dummy–seat.
- Verify the settled H-point is within ±5 mm of target in X and Z, and that the pelvis pitch is within ±2° of the specification.
- Check that no seat cushion or back element shows penetration greater than 2 mm into the dummy at the end of settling.
If the dummy sinks more than the allowance, the seat foam stiffness or the dummy–seat contact is wrong; if it barely moves, the initial gap was too large or gravity was not applied. Both are setup faults, not crash-physics findings.
Step 3 — Set Limbs and Posture to the Specification
Posture determines where the belt lands. Set the limbs explicitly rather than accepting the import pose:
- Upper arms: 40° from vertical (example), resting against the seat back, not pressed into the torso.
- Forearms and hands: on the steering wheel rim at the specified clock positions (example: 9:15), with a light, non-penetrating contact.
- Thighs: horizontal, knees at the specified spread; feet flat on the toe board or pedals per the test request.
- Head: level, looking forward, neck at neutral pitch.
Verify each joint angle against the specification table and log the measured values. A common failure is a forearm that intersects the wheel or a thigh that lifts off the cushion — both change belt routing and must be corrected before belt routing begins.
Step 4 — Route the Belt Through the Correct Landmarks
Route the lap and shoulder belt as a continuous path over the dummy surface, then let the retractor and buckle define the final geometry. The belt must pass over the pelvis and clavicle, not the abdomen or neck.
- Lap belt: across the anterior superior iliac spine, contacting the pelvis, with the buckle stalk positioned so the belt does not cross the abdomen.
- Shoulder belt: over the mid-clavicle, between neck and shoulder, with the D-ring height set so the belt clears the neck by the specified margin.
- Belt-to-body gap: target 0–3 mm at the chest and pelvis contact zones; a gap larger than ~5 mm means the belt is not engaging the dummy and will produce a delayed, non-physical loading.
- Initial belt strain: confirm the belt elements start at their specified pre-strain (example: 0.5–1.0%) and that the retractor is not pre-tensioned beyond the specification.
After routing, run a short quasi-static belt-tensioning step (or a 50 ms settle with the belt active) and re-measure the gaps. The belt should seat onto the dummy without pulling it out of position.
Step 5 — Pre-Simulation Acceptance Checks
Run the checks below and record pass/fail before launching the crash. These are the gates that prevent a wasted 150 ms run.
| Check | Measurement | Acceptance (example) | Failure symptom |
|---|---|---|---|
| H-point position | ΔX, ΔZ from target | ±5 mm | Dummy too far forward/back; belt geometry shifted |
| Pelvis pitch | Angle vs. spec | ±2° | Lap belt rides onto abdomen |
| Dummy–seat penetration | Max penetration depth | ≤ 2 mm | Initial contact forces bias chest deflection |
| Belt–body gap (chest) | Normal gap at mid-clavicle | 0–3 mm | Delayed belt engagement, low chest loading |
| Belt–body gap (pelvis) | Normal gap at ASIS | 0–3 mm | Submarining risk, lap belt slip |
| Shoulder belt–neck clearance | Minimum distance | ≥ 10 mm | Neck loading, non-physical HIC |
| Initial belt strain | Element strain at t = 0 | Spec value ±0.2% | Pre-loaded or slack belt |
| Contact status | Dummy–seat, dummy–wheel, belt–dummy | Active, no initial penetration | Missing contact, penetration spike at t = 0 |
Reading the Setup at a Glance
The diagram below shows the driver-side layout with the key measurement points and the pass/fail logic that connects them. Use it as a checklist map while performing Steps 2–5.
Typical Failure Symptoms and What They Mean
- Belt gap grows after settling: the belt was routed over a limb that moved during settle, or the retractor pre-tension pulled the dummy. Re-route after settling, not before.
- Dummy drifts forward during the belt-tensioning step: the lap belt is too high on the abdomen and is pulling the pelvis forward. Lower the lap path onto the ASIS.
- Shoulder belt contacts the neck: D-ring too low or too inboard. Adjust the D-ring height and re-check the ≥ 10 mm clearance.
- Penetration spike at t = 0 in the crash run: a contact was defined but the initial geometry already overlaps. This is a positioning fault, not a contact-stiffness fault — fix the geometry, do not soften the contact.
- H-point off by more than 5 mm after settling: seat state changed, cushion stiffness is wrong, or the settle allowance was mis-set. Re-check the seat log before adjusting the dummy.
Once all gates in the table pass, the model is ready for the crash run: the dummy is in the specified posture, the belt engages the correct landmarks with the specified clearance, and no initial penetration or pre-load will bias the injury response. Any deviation found later in the results should first be traced back to one of these gates before the restraint model itself is questioned.