PSI - Issue 24

Dario Vangi et al. / Procedia Structural Integrity 24 (2019) 423–436

427

D. Vangi et al. / Structural Integrity Procedia 00 (2019) 000–000

5

Fig. 3. SiL functioning scheme of an adaptive ADAS.

of braking and steering system is not instantaneous, and setting time to the desired value should be considered. The iterative cycle starts once more, with the identification by sensors of new position and velocity of the opponent. For the calculation of outcomes associated with each manoeuvre, starting from a specific scenario, it is necessary to simulate each possible intervention on braking and steering. The simulation of kinematics (free motion of vehicles) is primarily significant: if no impact between vehicles occurs, the system must identify the resulting clearance. If the manoeuvre results in an impact, IR derivation is obtained simulating also the impact phase: as expressed by Eq. 1 and highlighted in Figure 4 (modified from Jurewicz et al. (2016)), IR is a function of the impact type (which also depends on occupant’s position) and of ∆ V , which is a typical post-impact parameter: a 0 , a 1 , a 2 in Eq. (1) are logistic regression coe ffi cients: the corresponding reference curves in Figure 4 are referred to values of maximum abbreviated injury scale (MAIS, as defined in Gennarelli and Wodzin (2008)) equal to or higher than 3 (MAIS 3 + ). As depicted in Figure 4, if the intrusion involves the compartment on the same side of the occupant (near side collision), the crash is characterized by maximum potential severity ( ∆ V being the same). Even if approximated methods exist for the calculation of impact severity starting from sole pre-impact parameters (as V r in rear-end collisions considered by Da´vid et al. (2019)), simulation of the impact phase remains the most accurate solution for a proper estimate of ∆ V . The impact phase between vehicles can be simulated through di ff erent calculation methodologies, i.e., through analytical models, finite element method (FEM), multi-body systems (MBS) and reduced order dynamic models (RODM). Analytical methods are typically based on impulse-momentum models and allow solving the problem ana lytically at the cost of rough approximations, deriving from the choice of a plane and a centre of impact as explained by Kolk et al. (2016). FEM and MBS are employed in the accurate reconstruction of impacts, for instance in vehicle crashworthiness analysis: calculation time for the single impact configuration between vehicles is in this case high, typically variable between an hour and a day with modern calculators. In such context, the use of special-purpose RODM is justified, featuring intermediate characteristics of calculation time and accuracy in respect to analytical methods and FEM / MBS - Le Guennec et al. (2018). Once the simulation results of all possible manoeuvres are available for the specific critical scenario, the ADAS is capable of associating a specific value of clearance (collision avoided) or IR (collision not avoided) with each braking and steering intervention. This capability is summarized in an intelligible way by the map of outcomes reported in Figure 5, whose analysis allows the ADAS to select the best intervention in the specific critical scenario. In case the IR = 1 1 + e a 0 + a 1 · ∆ V + a 2 · ImpactType (1)