02 06 2015 Annex 4 TF WLTP ANNEX

02. 06. 2015 Annex 4 TF WLTP ANNEX 4 AND 7, DELTA CD*A ALTERNATIVE DETERMINATION METHODS

WLTP ANNEX 4 AND 7, ALTERNATIVE DELTA CD*AF PROPOSAL FOR ALTERNATIVE CD*AF DETERMINATION Proposal: Ø Allow for other determination methods (e. g. simulation), in a very restricted an guided way. Justification: Ø The difference between very similar options (e. g. same rim on different bodystyles) may be in the order of the measurement tolerance of cd*Af. In this case we could reduce unnecessary effort (wind tunnel measurements are very costly and time consuming), while at the same time providing more accurate data on the influence of options or body shapes. Ø Measurement will remain the basic principle anyway! History: Ø IWG in Stockholm meeting agreed on approach. Requested improvements have been included. WLTP Annex 4 and 7, alternative cd*A determination, BMW, 02. 06. 2015 Page 2

WLTP ANNEX 4 AND 7, ALTERNATIVE DELTA CD*AF DRAFTING TEXT At the request of the manufacturer and with the agreement of the responsible authority, an alternative method (e. g. simulation, wind tunnel not fulfilling the criterion in Annex 4) may be used for the determination of delta (CD×Af) if the following criteria are fulfilled (a) The alternative determination method shall fulfill an accuracy of ± 0. 015 m² delta (C D×Af) and additionally in the case simulation is used, the Computational Fluid Dynamics method should be validated in detail, such that the actual air flow patterns around the body, including magnitudes of flow velocities, forces, or pressures, are shown to match the validation test results; (b) The alternative method shall be used only for those kind of aerodynamic-influencing parts (e. g. wheels, body shapes, cooling system), for which equivalency was demonstrated; (c) Evidence of equivalency shall be shown in advance to the responsible authority for each road load family in the case of a mathematical method or every four years in the case of a measurement method and in any case shall be based on wind tunnel measurements fulfilling the criteria of this gtr. ; (d) If the delta (CD×Af) of an option is more than 100% larger than the option with which the evidence was given, aerodynamic drag shall not be determined with the alternative method; and (e) In the case of a simulation model change, a revalidation shall be necessary. WLTP Annex 4 and 7, alternative cd*A determination, BMW, 02. 06. 2015 Page 3

WLTP ANNEX 4 AND 7, ALTERNATIVE DELTA CD*AF BACKUP Backup WLTP Annex 4 and 7, alternative cd*A determination, BMW, 11. 05. 2015 Page 4

WLTP ANNEX 4 AND 7, ALTERNATIVE DELTA CD*AF TOLERANCE New proposed text within Annex 7: 3. 2. 2. 3. Aerodynamic drag of an individual vehicle The aerodynamic drag shall be measured for the drag influencing options and body shapes at a wind tunnel fulfilling the requirements of paragraph 3. 2. of Annex 4 verified by the responsible authority. [ put here "Alternative cd*Af"] Δ(cd*Af)LH (m²) is the difference of the aerodynamic drag of the test vehicle H compared to the test vehicle L and shall be recorded. Δ(cd*Af)ind (m²) is the difference in aerodynamic drag between an individual vehicle and the test vehicle L, due to options and body shapes on the vehicle that differ from those of the test vehicle L. These aerodynamic drag differences shall be determined with an accuracy of 0. 015 m² (c d*Af). Δ(cd*Af) may be calculated according to the following equation maintaining the accuracy of 0. 015 m² also for the sum of options and body shapes: Ø The accuracy for aerodynamic options and body shapes will be anyway defined more clearly by that proposal within the drafting taskforce. Ø OEM is responsible for correct Delta cd*A, however it is determined. WLTP Annex 4 and 7, alternative cd*A determination, BMW, 31. 03. 2015 Page 5

WLTP ANNEX 4 AND 7, ALTERNATIVE DELTA CD*AF SIMULATION CRITERIA Ø Out of the IWG meeting in Pune there was a request on showing possible simulation criteria. Ø IWG agreed, that there should be no criteria in the GTR. Simulation criteria for aerodynamics (CFD simulation): General simulation properties • Three dimensional simulation involving friction. • Direct simulation or modeling of turbulence via at least one-equation model (e. g. Spalart Almaras). • Navier-Stokes or Lattice-Boltzmann based methods are allowed. • Numerical solution needs to be evaluated over an averaging interval in which cd*Af may not exceed [5/1000]. Model accuracy • Ground movement shall be modeled (either as an infinite belt or a five-belt or one belt). • Wheel rotation shall be modeled. • The cordial deviation of surface elements may not exceed [0. 1] mm. • The volume deficit between full CAD geometry and the discretized simulation model may not exceed [tbd] %. List of software vendors / products: EXA: Power. FLOW; Open. CFD: Open. Foam; Adapco: CCM+; Ansys: Fluent; Ansys: CFX WLTP Annex 4 and 7, alternative cd*A determination, BMW, 31. 03. 2015 Page 6