The knowledge of the brake linings coefficient of friction (BLCF) is crucial for
the control of the braking moment in modern vehicles equipped with electric
powertrains. In the case of race vehicles equipped with carbon–carbon brakes,
the coefficient of friction exhibits great variations as a function of the main
influencing factors, namely the pressure, the temperature, and the sliding speed
at the pad–disc interface. In this work, a Le Mans Hypercar instrumented with
more than 150 sensors was adopted to perform the characterization of the BLCF
from racetrack acquisitions. The front and rear left suspensions of the vehicle
were instrumented with strain gauge channels and position transducers to acquire
the reaction loads at the upright and the orientation of the arms. Then, the
geometric matrix method was implemented for calculating the moments at the
upright from which the braking torque was derived without the need to know any
of the wheel inertia, nor the driveshaft torque. Data from multiple acquisitions
across different racetracks, operating temperatures, and ambient conditions were
used to characterize the BLCF of the front and rear carbon brakes equipped on
the vehicle. After implementing pre-processing steps aimed at improving data
homogeneity, two friction maps were characterized for the front and rear
systems, respectively. The friction maps were validated against new experimental
data showing an average 3% error reduction over assuming a constant BLCF.
Accordingly, the characterized friction maps can be integrated in the
brake-by-wire system of the vehicle for accurate caliper pressure control
through real-time estimation of the BLCF from commonly available sensor signals,
such as caliper pressure, wheel speed, and disc temperature. In this context,
the effectiveness of the friction maps was demonstrated by comparing the
predicted brake moments with the torques measured by the instrumented
suspensions, highlighting the advantages over assuming a constant BLCF.
