b'Engineer Innovation | Automotive60 The methodology has been validated Vehicle Speed [km/h] on different conventional driving 50 Reference Vehicle Speed [km/h] cycles, like WLTC. The number of calls 40 generated can be easily observed as in Figure 3 when a 3D vehicle model has 30 been called, as each vehicle speed step 20 of 5 meters/s, except at low speed, generates a call. The number of calls 10 can be modified by changing the vehicle speed step: the lower it is, the 0 050100150200250300350400450 higher the number of calls is. A minimal X: Time [s] reference speed is not set at 0 km/h Fan Activation [null] when vehicle is standstill, and therefore 1.00.8 the 3D model runs with a minimal 0.6 velocity. The 3D simulation with free 0.4 convection will be investigated later. 0.2 Real driving cycles have been 0 performed at low speed (urban 050100150200250300350400450 conditions) and at higher speed (extra X: Time [s] urban conditions), and we can observe Blower Position [null] the same coupling mechanism, but [null][null]Controlled Cabin Temperature [degc] with lower vehicle speed step (2 m/s 1.6 36 instead of 5 m/s). The fan is activated 1.4 34 because the condenser inlet pressure 1.2 32 increases up to 25 barA but is 1.0 30 deactivated when vehicle speed reaches 50 km/h or higher. The 0.8 28 alternating current compressor electric 0.6 26 consumption is not negligible in 0.4 24 comparison to other electric devices. 0.2 22 The powertrain cooling circuit is 0 20 activated when the coolant 050100150200250300350400450 temperature achieves 45 degrees X: Time [s] Celcius. During the extra urban cycle, Figure 3: CFD vehicle and cabin calling the blower position switches from high level to lower level in relation to the controlled cabin temperature. The HPC for predictive simulations toinfluence of the blower air mass flow optimize energy-efficiency rate can be observed, especially The objective of high performancebetween 365 seconds and 370 seconds, computing is to provide thewith an increase of average cabin computational power for thetemperature due to low blower air mass simulations needed for virtual designflow rate. and validation phases. The main application that requires the power ofBy simulating the flow field around a HPC is the 3D simulation as it relies onvehicle, under the hood and inside the resource-consuming data obtained fromcabin, CFD can provide all the CFD analysis. 3D models are runinformation that is needed. Following separately on a HPC server throughthis process, the great benefit is that specific scripts based on call strategies,boundary conditions are no longer which are different for both models.estimated but are calculated through a The entire simulation with all driving1D Simcenter Amesim model. Therefore, cycle data that need to be consideredpotentially every relevant instant or Worldwide harmonized Light vehiclesphase in a driving cycle can be studied in Test Cycles (WLTC) and Real Drivingdetail for each type of application. In the Cycle for urban and extra-urban)Figure 4, some 3D visualizations are becomes affordable in terms of time:shown, coming from WLTC simulation. In using ~300 CPUs, the whole simulationparticular they focus on measures that (1D model calling several 3D models)Simcenter STAR-CCM+ calculates in order runs in a few hours, depending on theto use in 1D model, for both vehicle and driving cycle duration. cabin models.38'