4. Human surrogate selection and limitations

Impact biomechanics experiments require a physical body to absorb, transmit, and respond to mechanical loading. Since ethical and practical constraints prevent subjecting living humans to injurious impacts, researchers rely on three surrogate classes: Post-Mortem Human Subjects (PMHS), Anthropomorphic Test Devices (ATDs), and human volunteers. Each occupies a distinct region of the fidelity–ethics–cost–repeatability trade space. Selecting the wrong surrogate invalidates the experiment before data collection begins; understanding each surrogate's limitations is therefore a core engineering competency, not a background detail.

4.1 The Three Surrogate Classes at a Glance

The fundamental distinction lies in what each surrogate can and cannot represent. PMHS provide real biological tissue — bone, ligament, muscle, and viscera — with geometry and material properties that no man-made device fully replicates. ATDs are precision instruments: repeatable, instrumented, and durable, but constructed from steel, rubber, and foam whose mechanical behavior only approximates human tissue. Human volunteers provide the only living, physiologically responsive system, but can only be exposed to non-injurious loading levels.

4.2 PMHS: The Reference Standard with Real Constraints

PMHS testing remains the gold standard for establishing injury tolerance curves and validating ATD biofidelity. A typical PMHS experiment for thoracic injury tolerance might proceed as follows. A fresh-frozen specimen is thawed, screened for communicable disease, and instrumented with accelerometers on the sternum and spine, plus a chest band for deflection measurement. The specimen is positioned supine on a sled or seated in a test buck, and a pneumatic impactor delivers a controlled sternal displacement at a defined rate — for example, 3.0 m/s to a peak deflection of 60 mm. Force and deflection are recorded at 10 kHz, and the resulting force–deflection curve is analyzed for peak force, energy absorbed, and the deflection at which rib fracture occurs.

The limitations are substantial. Specimen age typically ranges from 60 to 90 years, meaning bone quality is osteoporotic and not representative of the 20–40 year old population most at risk in motor vehicle crashes. The absence of active muscle tone means the PMHS cannot brace or respond to impending impact, altering kinematics. Post-mortem tissue degradation, even with careful freezing, changes material properties. And the sample size is inherently small — a study with 12 valid specimens is considered large. These factors mean PMHS data carry wide confidence intervals and must be interpreted with age and condition adjustments.

4.3 ATDs: Repeatable but Approximate

ATDs — the Hybrid III family, THOR, WorldSID, and others — are engineered to produce repeatable, instrumented responses under standardized test conditions. Their value lies in regulatory compliance testing: every frontal crash test conducted for FMVSS 208 uses a Hybrid III 50th percentile male ATD, and the resulting injury values (HIC, chest deflection, femur force) are compared against pass/fail thresholds. This repeatability is essential for comparing restraint system designs across hundreds of tests.

The limitations are equally clear. An ATD's biomechanical response is tuned to match a specific set of PMHS corridors, typically for a single loading direction and severity. The Hybrid III, for example, was designed for frontal impact and performs poorly in lateral or rearward loading. Its neck is a segmented rubber and metal structure that does not replicate the cervical spine's complex facet joint behavior. Its chest is a steel rib cage with a damping material that approximates, but does not match, human thoracic force–deflection characteristics. The ATD also lacks active musculature, so it cannot simulate bracing or the post-impact muscle response that influences occupant kinematics.

When selecting an ATD, the engineer must match the device to the loading scenario. A frontal sled test calls for a Hybrid III; a side impact test calls for a WorldSID or ES-2re; a rear impact test may use a BioRID or RID2. Using a frontal ATD in a lateral test produces kinematics that are not just inaccurate but potentially misleading — the shoulder and pelvis structures are not designed to load in that direction.

4.4 Human Volunteers: The Living Reference

Human volunteer testing is restricted to non-injurious loading — typically low-speed sled tests (delta-v below 8 km/h), low-amplitude pendulum impacts, and volunteer-seated vibration studies. These experiments provide data that no other surrogate can: active muscle response, real-time physiological reaction, and subjective feedback on discomfort or perceived injury risk. A typical volunteer rear-impact study might use a sled acceleration of 4–6 g at the seat track, with surface EMG on the sternocleidomastoid and cervical paraspinal muscles, and head accelerometers to capture head–neck kinematics.

The limitations are ethical and practical. Volunteers must be fully informed, can withdraw at any time, and cannot be exposed to loading that carries any meaningful injury risk. This caps the severity at a level well below injury thresholds, meaning volunteer data cannot directly establish injury criteria — only the sub-injurious response that informs the lower end of the injury risk curve. Volunteer studies also require extensive medical screening, and the sample is typically young, healthy, and physically fit, which is not representative of the general occupant population.

4.5 Selection Framework and Practical Guidance

Choosing a surrogate is a structured decision. The following decision logic applies:

1. Define the loading severity and direction. If the test must produce injury or exceed volunteer-safe limits, human volunteers are excluded. If the loading is lateral or rearward, a frontal ATD is excluded.
2. Determine the required output. If the goal is to establish or refine an injury tolerance curve, PMHS is the reference. If the goal is regulatory compliance or design comparison, an ATD is appropriate. If the goal is to understand low-severity kinematics or muscle response, volunteers are the only option.
3. Assess the acceptable variability. If the experiment requires tight repeatability across many runs, ATDs are preferred. If biological realism is paramount and variability can be managed statistically, PMHS or volunteers are used.
4. Check ethical and regulatory feasibility. PMHS requires donor consent and institutional review; volunteer testing requires human subjects approval and medical oversight. ATD testing requires neither.

4.6 Cross-Surrogate Validation and Scaling

Because no surrogate is perfect, researchers often combine them. A common workflow: PMHS tests establish the thoracic force–deflection corridor; the ATD is then tuned so its chest response falls within that corridor; volunteer tests confirm that the ATD's low-severity response matches living human kinematics. Discrepancies between surrogates are not failures — they are data. If the ATD's chest deflection is 15% lower than the PMHS corridor at the same impact energy, that difference quantifies the ATD's biofidelity gap and informs how much uncertainty to assign to ATD-based injury predictions.

Scaling is another critical tool. Because PMHS specimens are often elderly and small, and ATDs are available in only a few sizes (5th percentile female, 50th percentile male, 95th percentile male), researchers use mass and stiffness scaling to translate results between populations. The classic approach uses the ratio of effective mass and a characteristic length to scale force, deflection, and time. For example, a force measured on a 50th percentile ATD can be scaled to a 5th percentile female using a mass ratio of approximately 0.6 and a length ratio of approximately 0.9, yielding a scaled force that is roughly 0.6 × 0.9² ≈ 0.49 of the original. These scaling factors are approximations and introduce additional uncertainty, but they are essential for applying experimental data across the occupant population.

4.7 Key Takeaways for Experimental Design

Surrogate selection is not a matter of preference but of matching the surrogate's capabilities to the experimental question. PMHS provide the biological reference but carry age, sample size, and ethical constraints. ATDs provide repeatability and instrumentation but approximate human response within a narrow loading envelope. Human volunteers provide living tissue response but only at sub-injurious levels. The most robust experimental programs use all three in a coordinated validation chain, explicitly quantifying the uncertainty introduced at each transfer step. When reporting results, always state the surrogate used, its known limitations for the loading condition, and the scaling or correction factors applied — this transparency is what allows the broader biomechanics community to interpret and build upon the findings.