CFD correlation with wind-tunnel for dry van trailer aerodynamic devices.
The primary purpose of this paper is to correlate the CFD simulations performed using PowerFLOW, a Lattice Boltzmann based method, and wind tunnel tests performed at a wind tunnel facility on 1/8th scaled tractor-trailer models. The correlations include results using an aerodynamic-type tractor paired with several trailer configurations, including a baseline trailer without any aerodynamic devices as well as combinations of trailer side skirts and a tractor-trailer gap flow management device. CFD simulations were performed in a low blockage open road environment at full scale Reynolds number to understand how the different test environments impact total aerodynamic drag values and performance deltas between trailer aerodynamic devices.
There are very limited studies with the Class-8 sleeper tractor and 53ft long trailer comparing wind tunnel test and CFD simulation with and without trailer aerodynamic device. This paper is to fill this gap.
CITATION: Dasarathan, D., Ellis, M., Chinnamani, S., Ayala, R. et al., "CFD Correlation with Wind-Tunnel for Dry Van Trailer Aerodynamic Devices," SAE Int. J. Commer. Veh. 9(2):2016, doi:10.4271/2016-01-8016.
In 2015, the United States Environmental Protection Agency (EPA) and Department of Transportation's National Highway Traffic Safety Administration (NHTSA) released their proposal for the second phase of regulations to reduce greenhouse gas (GHG) emissions from on-road heavy duty vehicles. This second phase of proposed regulations is similar to the first phase of regulations that went in to effect for 2014 model year heavy duty vehicles in that the proposal continues to push for further reductions in GHG emissions. One difference is that this new proposal takes a more systematic approach to reducing emissions by considering the impact that trailers have on the fuel efficiency, and therefore the emissions. The new proposed regulations will continue to allow manufacturers to evaluate aerodynamic performance of designs using alternative methods such as computational fluid dynamics (CFD) simulations and wind tunnel tests. It is important for the industry to have confidence in these alternative methods.
The newly proposed second phase of the GHG emissions standard for heavy-duty vehicles impacts the trailer in addition to just the tractor standard in GHG phase one [1, 2, 3, 4]. The GHG Phase-2 adds the wind yaw conditions in addition to Phase-1. Due to this, aerodynamic performance requirements for the tractor as well as the trailer have become more stringent. Original equipment manufacturers (OEMs) can choose between both experimental tests, such as coast down testing , wind tunnel testing with reduced-scale or full-scale model , and SAE Type II testing , as well as analytical means such as CFD to assess aerodynamic performance. Greenhouse Gas Emissions Model (GEM) is used to determine relative fuel economy savings from aerodynamic drag [8, 9, 10]. GEM uses a standard defined trailer for OEM tractor it takes a drag area ([C.sub.D]A) as aerodynamic input, and for trailer OEM's may use a variety of Class-8 sleeper tractors as the front end for trailer fuel economic evaluations.
This paper investigates both the drag coefficient ([C.sub.D]) and incremental delta [C.sub.D] correlation of a 1/8th scale wind tunnel model and its corresponding CFD simulation for yaw sweep. The model is evaluated with and without trailer aerodynamic devices, specifically a Tractor-Trailer Gap Fairing (TTGF), and Trailer Skirt (TS). It also compared the impact of the scaled model in wind tunnel environment to the open road (OR) condition for following three different configurations.
* Baseline with TTGF
* Baseline with TTGF along with TS
The effects investigated are:
* Influence of the tunnel geometry
* Influence of adding TTGF
* Influence of adding TTGF and TS
* Influence of open road (OR) condition
PREVIOUS RELEVANT WORK
Horrigan K. et all  compared CFD aerodynamic results to 1/5th scale model wind tunnel measurements from the University of Washington Aeronautical Laboratory (UWAL)  using Kenworth T2000 tractor and only smaller trailer was used as the scale model. They found overall very good agreement between simulation and experimentally measured drag. And also compared with the open road condition.
Horrigan K. et all  investigated CFD simulations of a 1:8-scale Generic Conventional Model (GCM), were conducted using a Lattice-Boltzmann based solver. Comparisons were made to experimental measurements from the NASA Ames 12-Foot Pressure Wind Tunnel , including drag coefficients as a function of yaw, static and transient surface pressures, and three-component particle image velocimetry. They found excellent agreement in overall drag prediction as well as local flow results including average static pressure, tractor-trailer gap flow, trailer wake structure, and transient pressure spectra in these regions.
There are very limited studies with the Class-8 sleeper tractor and 53ft long trailer comparing wind tunnel test and CFD simulation with and without trailer aerodynamic device. This paper is to fill this gap.
Scaled Down Rolling Road Wind Tunnel
Wind tunnel tests were performed on a 1/8th scale long sleeper with a standard 53-foot dry-box trailer in rolling road wind tunnel at the Auto Research Center (ARC) [15-16]. The representative baseline wind tunnel trailer model is shown in Figure 1, for confidentiality reason tractor image is not shown anywhere in this paper. The tractor is a standard Class-8 sleeper cab with fully detailed underbody components, such as engine, transmission, heat exchangers side tank, heat exchangers core (were modelled as perforated vertical flat plates which matches the velocity ratio in the heat exchanger), fan (were replaced by circular anemometer) and fan shroud, steer axle and tandem drive axle with tires, after treatment system, battery box, air tank, fuel tank, exhaust pipe, quarter fender, drive wheel mud flap, deck plate and stair case in the tractor to trailer gap. Tractor main and hood mounted auxiliary mirrors, chassis skirt, side extender and roof extender were also included. Measured wheel base for the scaled model is 28.0 inch, tractor to trailer gap is 5.6 inch, trailer height 2.0 inch taller than the tractor. For the confidentiality reason we could not able to provide the other dimensional details for the tractor, trailer and trailer aero devices.
The approximate boundary layer (BL) thickness at the test section is 1mm, turbulence intensity (TI) is 0.24% and the flow angularity is 0.24[degrees] . Wind off rolling wheel tares were performed at 40% belt speed. The model was tested at a speed of 50 m/s (which corresponds to Reynolds number (Re) of 1.1M based on width of the tractor as characteristic length), for the complete yaw sweep of-9[degrees] to 9[degrees]. Rotating tire and moving floor (floor is not yawed with the vehicle) were modelled, blockage ratio (defined by ratio of vehicle front projected area to tunnel nozzle exit area) of the model were 3.6% and 7.3% at 0[degrees] and [+ or -]9[degrees] yaw condition respectively which is within recommended blockage ratio as per the SAE J1252  for open jet wind tunnel. Forces and moments were measured using an Aerotech 6-axis force measurement balance mounted to the trailer through a mounting sting. The model was tested in three different configurations, and evaluated over the full yaw sweep.
* Baseline with TTGF
* Baseline with TTGF along with TS
CFD Simulation Approach
CFD simulations were performed using the commercially available software, which uses the Boltzmann Equation to solve for the flow field . It can be shown rigorously that the Lattice Boltzmann formulation is equivalent to solving the time dependent compressible Navier-Stokes Equation  and that the formulation imposes fluid boundary conditions on solid surfaces . A review of the fundamentals of the Lattice Boltzmann Method can be found in .
CFD tool uses the Very Large Eddy Simulation (VLES) turbulence model to solve the resolvable scale flow scale and modeling the sub-grid scale using RNG k-epsilon; further details on turbulence modeling are available in . Reference  provides a review of the LBM simulation approach using this turbulence model, and describes capabilities for aerodynamics, thermal, and aero-acoustics applications for production vehicle development. Examples of published validation studies including full-detailed production vehicles using this simulation approach are shown in [23, 24, 25, 26, 27]. Several examples of validations studies using wind-tunnel data from scale-model vehicles are shown in [28, 29, 30, 31].
For the full-scale model the computational grid for the simulations was established using nested regions of refinement of a Cartesian grid, with a finest cell size of 6.0 mm in critical areas, and 12.0 mm cells around the entire tractor according to the recommended best practices procedures for heavy truck Lattice Boltzmann aerodynamics simulations SAE J2966 . For the case of the scaled model the grids are scaled down accordingly. The CFD model was evaluated in three different configurations, the same as it was in the wind tunnel test.
* Baseline with TTGF
* Baseline with TTGF along with TS
There were two CFD simulation environments used in this study.
1. The scaled wind tunnel model for the full yaw sweep of -9[degrees] to +9[degrees], except [+ or -]3[degrees].
2. Open-Road (OR) - The simulation domain was an idealized domain with negligible (0.05%) blockage representing the OR. Tested at 0[degrees], 2[degrees] and 6[degrees] yaw condition
In the scaled wind tunnel model environment [Figure 2], the wind tunnel facility was modelled as per the SAE J2966 . We have replicated WT facility from the measured dimensional data and/or photo images to create the CAD of the facility. Here the wind tunnel setup starting from the settling chamber, contraction section, test section with same length and width of the moving belt, plenum and wind tunnel collector wall followed by diffuser with all the breather were modeled along with the test vehicle. Rolling road boundary condition was applied at the tunnel center belt and rotating tire also modelled. The ground clearance is matched as with the test wheel scrubbing is modelled as like in the WT test. The vehicle is placed at the same location on the belt as in WT test and the rolling road is not yawed with the model. Not modelled were the closed circuit in the tunnel, fan, and distributed suction in front of the rolling belt which is replaced by frictionless wall. Mass flow rate inlet and pressure outlet boundary condition were applied. The mass flow rate at inlet applied based on the test section speed corresponds to 50m/s (which corresponds to Reynolds number (Re) of 1.1M). Simulation condition as low turbulence intensity (0.3%) free stream flow. Cooling flow were modeled as like in the WT test by matching the velocity ratio in the heat exchanger.
Here we have used similar tractor and trailer model in CFD and matched the WT test vehicle dimensions from the measurement. Boundary layer (BL) on the floor in the CFD WT simulations were matched as per the WT test (1mm). Boundary layer on the vehicle is unknown for the WT test and turbulent boundary layer for the CFD simulation.
For the OR environment, the same vehicle geometry used for the CFD wind tunnel simulation [Figure 3, 4, 5] was scaled up to full scale size, and the corresponding on-road simulation boundary conditions such as vehicle speed (65mph, full scale Re of 5M), moving ground, and rotating tires were applied. Similar turbulence intensity as in scale model was used, but with a 0.05% blockage environment representative of an ideal open road environment as per SAE J2966 , so that the extents of the flow domain do not impact the vehicle. The vehicle and moving floor is yawed about the freestream direction, simulation done at three different yaw angles 0[degrees], 2[degrees] and 6[degrees].
Figure 3 shows the CFD baseline simulation trailer, it uses the same exact CAD from the WT test. Figure 4 shows the baseline with TTGF configuration and Figure 5 shows the baseline with TTGF and TS.
All simulations were performed from an initial flow field obtained from a coarse-grid simulation in order to reduce initial transient oscillations. All simulations were run transient and time-accurate, until forces were stabilized and averaged forces were converged. Analysis of the transient forces was used to determine the end of the initial transient and the amount of time required for averaging. The results are reported for the time-averaged flow field, averaged over 10 seconds of physical time.
RESULTS & DISCUSSION
The wind tunnel testing was performed at yaw sweep from -9[degrees] to +9[degrees] yaw. Table 1 shows the wind tunnel test results. Here, the drag coefficient ([C.sub.D]) values are normalized ([C.sub.Dn]) based on the 0[degrees] yaw baseline wind tunnel test value. We assumed that the dynamic pressure correction was applied in the wind tunnel test data but not have information about the dynamic pressure calculation method, here the [C.sub.Dn] (normalized drag coefficient) are presented as it is received from the wind tunnel test, i.e. no correction was applied after that. We also don't have any information about the uncertainty of the wind tunnel test.
For the baseline, the [C.sub.Dn] varies from 1.000 to 1.304 for 0[degrees] to [+ or -]9[degrees] yaw conditions. [C.sub.DWAn] (normalized wind average drag as per SAE J1252  based on 65mph vehicle speed with 7mph wind speed) is 1.091. For the baseline w/ TTGF, the [C.sub.Dn] varies from 0.994 to 1.246 for 0[degrees] to [+ or -]9[degrees] yaw conditions, [C.sub.DWAn] is 1.073. For the baseline w/ TTGF and TS, the [C.sub.Dn] varies from 0.873 to 1.055 for 0[degrees] to [+ or -]9[degrees] yaw conditions and [C.sub.DWAn] is 0.947.
The data clearly shows that adding the TTGF improved the wind averaged drag coefficient by 1.6% and adding the TS in addition to the TTGF improves the wind averaged drag coefficient by 13.1% over the baseline.
Figure 6 shows that the drag coefficient polar are slightly asymmetric between the positive and negative yaw angles for all the test configurations. This may be due to the difference in blockage to the sides of the model in the plenum (one side is 25% more room than the other side) and/or some extent of the test vehicle upper body and underhood components. A more parabolic drag profile (more steep increase of drag wrt yaw) is seen for the baseline configurations as well as the baseline with TTGF only. As the trailer becomes more aerodynamic by adding the TTGF with the TS, the drag polar increases more gradually (less slope).
CFD SIMULATION RESULTS
Scaled CFD Model
Table 2, 3, 4 shows the CFD drag coefficient data for the scaled model in the wind tunnel environment with the drag values normalized based on the 0[degrees] yaw baseline wind tunnel test data, since there is not much change in the drag coefficient between WT test and CFD WT at 0[degrees] yaw condition.
Present we don't have any uncertainty data for the CFD as well as WT test. Reported [C.sub.Dn] will have slight influence after adding the uncertainty values.
Not knowing the exact dynamic pressure calculation method applied on the WT test, we have tried following three different dynamic pressure correction methods
1. Nozzle Method 
* Dynamic pressure is calculated by taking the pressure difference between the settling chamber (p.sub.sc) and inside the nozzle ([p.sub.N]), shown in figure 7
* Dynamic pressure [[DELTA]p.sub.N] used in the [C.sub.D] calculation.
2. Plenum Method 
* Dynamic pressure is calculated by taking the pressure difference between the settling chamber ([p.sub.sc]) and plenum ([p.sub.p]), shown in figure 7
Dynamic pressure [[DELTA]p.sub.p] used in the [C.sub.D] calculation.
3. Modified Nozzle Method
* Nozzle method dynamic pressure corrected for continuity blockage
* Dynamic pressure = [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
Where [BR.sub.yaw] - Blockage Ratio (BR), vehicle area to nozzle exit area at different yaw
The results [Table 2, 3, 4] shows that the CFD, predicts the performance delta between the baselines and trailer aero device within 0.6% [C.sub.SWAn]' , for both the TTGF and TTGF with TS configurations. Based on the various dynamic pressure correction methods studied for CFD WT simulation modified nozzle method showed [Figure 8, 9, 10] good agreements with the WT test. Nozzle and plenum methods [figure 8, 9, 10] are offset of 3% and 16% lower normalized drag coefficient than the WT test. There is an increasing deviation observed when the trailer becomes more aerodynamic by adding TTGF and/or TS.
Figure 6 shows the drag polar for both the wind tunnel test and the CFD simulations (as per modified nozzle method). Here, the overall trend is very similar to the wind tunnel test. Similar asymmetries in the drag coefficient values can be seen between the positive and negative yaw angles for all the test configurations.
There is an inflection at 0[degrees] yaw condition seen for the baseline and TTGF configuration [Figure 6] in WT test as well as WT CFD simulation. This is due to the slightly higher trailer base pressure at [+ or -]1[degrees] yaw [Figure 11], but the trend is not noticeable for the TS configuration.
One interesting observation is the increasing error as the model is yawed towards the maximum of 9[degrees]. After reviewing the WT test, it was hypothesized that the compliance in the 10 independent wheel/suspension assemblies could introduce a tare error. The thrust load on the axle assemblies would change at full-belt speed since the wind off rolling tares were performed at 40% belt speed.
This result clearly shows that the CFD simulations predict total drag coefficient very close to wind tunnel test, not only for the baseline, but also for the configurations with added trailer aerodynamic devices and their delta drag coefficient.
CFD Scaled to Open Road
Table 5 summarizes the drag data for the CFD OR simulations, WT test and CFD WT (modified nozzle method) simulation. When compared to CFD OR result, CFD WT simulation show similar normalized drag coefficient reduction for the TTGF configuration and 0.6% less reduction for the TTGF with TS variant.
Total [C.sub.Dn] results highlight the impact of the tunnel boundary conditions on the aerodynamic forces of the test model. There is significant shift in the total value of CDWAn between the tunnel simulation and the open road [Table 5], and also the shape of the yaw curve. The WT reduces the model's sensitivity to yaw angle.
Figures 12 and 13 show the trailer back pressure for CFD WT and OR environment respectively at 0[degrees] yaw. CFD WT model shows overall increase in base pressure which is assumed due to close proximity to the tunnel collector wall. This effect increases with adding TS. There may be some other factor like the rolling road end distance relative to the trailer back surface (we tried to keep the similar distance) which is also plays some role in disturbing the trailer back pressure.
Even though higher back pressure on the trailer back face which lead to lower trailer back face drag coefficient, overall higher total drag coefficient in the WT environment is due to a slightly wider shear layer [Figure 14, 15, 16] which increases the pressure around the vehicle [Figure 17, 18, 19], because of the open jet type flow in the wind tunnel result in a smaller trailer underbody wake [Figure 20, 21, 22], more exposed trailer bogie which leads to increased surface pressure on the bogie than in an OR environment. In addition to this, WT environment were run at lower Re [1.1M] compared with the OR [5M] which will increase the skin friction in the WT environment and also affect the underbody flow.
From this detailed study it is shown that CFD correlates well with the wind tunnel test for the normalized drag coefficient performance, within 0.6% [C.sub.DWAn]. CFD predicted total drag values very close to wind tunnel test based on modified nozzle method, not only for the baseline, but also for the configurations with added trailer aero devices. Similar trends and asymmetric drag coefficient polar are seen, with more deviation at higher yaw angles.
Studied three different dynamic pressure correction methods for CFD WT simulation, out of that modified nozzle method showed good agreements with the WT test. Nozzle and plenum methods are offset of 3% and 16% lower normalized drag coefficient than the WT test.
There is significant shift in the total value of [C.sub.DWAn] between the CFD WT simulation and the OR and also the shape of the yaw curve. The WT reduces the model's sensitivity to yaw angle. Drag effects of yaw are less pronounced in the scaled wind tunnel because of increased base pressure due to the pressure build up on the WT collector wall. The drag discrepancy between the open road simulation and scaled tunnel simulation increases with Yaw angle.
When compared to CFD OR result, CFD WT simulation show similar normalized drag coefficient reduction for the TTGF configuration and 0.6% less reduction for the TTGF with TS variant. Even though higher back pressure on the trailer back face which lead to lower trailer back face drag coefficient, overall higher total normalized drag coefficient in the WT environment is due to a slightly wider shear layer around the vehicle, because of the open jet type flow in the wind tunnel result in a smaller trailer underbody wake, more exposed trailer bogie which leads to increased surface pressure on the bogie than in an OR environment. In addition to this, WT environment were run at lower Re [1.1M] compared with the OR [5M] which will increase the skin friction in the WT environment and also affect the underbody flow.
1. The current study doesn't compare the CFD WT and OR case at the same Re, future study can be performed to match the Re.
2. We don't have any correction details for the WT test and also it's beyond the scope of this current paper, future studies can be performed to collect these data.
[1.] "Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles," Federal Register, Vol. 76, No. 179, pp. 57106 - 57513, September 15, 2011
[2.] "Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles," Federal Register, Vol. 80, No. 133, pp. 40137 - 40766, July 13, 2015
[3.] U.S. Environmental Protection Agency, "Proposed Rulemaking for Greenhouse Gas Emissions and Fuel Efficiency Standards for Medium-and Heavy-Duty Engines and Vehicles-Phase 2 Draft Regulatory Impact Analysis," EPA-420-D-15-900, Jun. 2015
[4.] U.S. Environmental Protection Agency (EPA), "Greenhouse Gas Emissions Model (GEM) User Manual," EPA Phase 2 proposed, EPA-420-B-15-076, Jun 2015,
[5.] SAE Surface Vehicle Recommended Practice, "Road Load Measurement Using Onboard Anemometry and Coastdown Techniques," SAE Standard J2263, Rev. Dec. 2008.
[6.] SAE Surface Vehicle Recommended Practice, "(R) SAE Wind Tunnel Test Procedure for Trucks and Buses," SAE Standard J1252, Rev. Jul. 2012.
[7.] SAE Surface Vehicle Recommended Practice, "Fuel Consumption Test Procedure - Type II," SAE Standard J1321, Rev. Feb. 2012.
[8.] U.S. Environmental Protection Agency (EPA), "Greenhouse Gas Emissions Model (GEM) User Manual," EPA Phase 2 proposed, EPA-420-B-15-076, Jun 2015
[9.] U.S. Environmental Protection Agency, Peer Review of the Greenhouse Gas Emissions Model (GEM) and EPA's Response to Comments, "EPA-420-R-11-007, Aug. 2011
[10.] U.S. EPA, "Final Rulemaking to Establish Greenhouse Gas Emissions Standards and Fuel Efficiency Standards for Medium- and Heavy-Duty Engines and Vehicles Regulatory Impact Analysis,", document EPA-420-R-11-901, Aug. 2011 U.S. Environmental Protection Agency, "Greenhouse Gas Emissions Model (GEM) User Guide," EPA
[11.] Horrigan, K., Duncan, B., Sivakumar, P., Gupta, A. et al., "Aerodynamic Simulations of a Class 8 Heavy Truck: Comparison to Wind Tunnel Results and Investigation of Blockage Influences," SAE Technical Paper 2007-01-4295, 2007, doi:10.4271/2007-01-4295.
[12.] The University of Washington Aeronautical Lab (UWAL), "UWAL on the Web; 1/5th scale model wind tunnel," https://www.aa.washington.edu/AERL/KWT, accessed July 2016.
[13.] Horrigan, K., Duncan, B., Keating, A., Gupta, A. et al., "Aerodynamic Simulations of a Generic Tractor-Trailer: Validation and Analysis of Unsteady Aerodynamics," SAE Technical Paper 2008-01-2612, 2008, doi:10.4271/2008-01-2612.
[14.] NASA Ames Research Center, "wind tunnel details on the Web; pressure wind tunnel," http://www.nasa.gov/audience/forstudents/9-12/features/F_Wind_Tunnels.html, accessed July 2016.
[15.] Auto Research Center (ARC), "wind tunnel details on the web, wind tunnel testing," http://www.arcindy.com/wind-tunnel.html, accessed July 2016.
[16.] Reynard A., Camosy M., Marinko F., Kowalczyk H. and Jennings T. "In Depth Cd/Fuel Economy Study Comparing SAE Type II Results with Scale Model Rolling Road and Non-Rolling Road Wind Tunnel Results" The Aerodynamics of Heavy Vehicles III, pages 287 to 301
[17.] PowerFLOW Users Guide, Release 5.0, 2014. Exa Corporation, 55 Network Drive, Burlington, MA 0103
[18.] H. and Chen S. Y. and Matthaeus W. H., "Recovery of the Navier-Stokes equations using a lattice-gas Boltzmann method", Phys. Rev. A, 45, R5339-R5342, 1992.
[19.] Chen S., and Doolen GD., "Lattice Boltzmann Method for Fluid Flows", Ann. Rev. Fluid Mech., 30, pp. 329-364, 1997.
[20.] Chen, H., Teixeira, C., & Molvig, K. "Realization of fluid boundary conditions via discrete Boltzmann dynamics", International Journal of Modern Physics C, 9(08), 1281-1292, 1998.
[21.] Chen, H., Kandasamy, S., Orszag, S. et al, "Extended Boltzmann Kinetic Equation for Turbulent Flows", Science, Vol. 301, pp. 633-636, 2003.
[22.] Kotapati, R., Keating, A., Kandasamy, S., Duncan, B. et al., "The Lattice-Boltzmann-VLES Method for Automotive Fluid Dynamics Simulation, a Review," SAE Technical Paper 2009-26-0057, 2009, doi:10.4271/2009-26-0057.
[23.] Kandasamy, S., Duncan, B., Gau, H., Maroy, F. et al., "Aerodynamic Performance Assessment of BMW Validation Models using Computational Fluid Dynamics," SAE Technical Paper 2012-01-0297, 2012, doi:10.4271/2012-01-0297.
[24.] Duncan, B.D., Fischer, A., and Kandasamy, S., "Validation of Lattice-Boltzmann Aerodynamics Simulation for Vehicle Lift Prediction." ASME Conf. Proc. 2010, 2705 (2010), ASME Paper No. FEDSM-ICNMM2010-30891
[25.] Cyr, S., Ih, K., and Park, S., "Accurate Reproduction of Wind-Tunnel Results with CFD," SAE Technical Paper 2011-01-0158, 2011, doi:10.4271/2011-01-0158.
[26.] Lounsberry, T., Gleason, M., Kandasamy, S., Sbeih, K. et al., "The Effects of Detailed Tire Geometry on Automobile Aerodynamics - a CFD Correlation Study in Static Conditions," SAE Int. J. Passeng. Cars - Mech. Syst. 2(1):849-860, 2009, doi:10.4271/2009-01-0777.
[27.] Duncan, B., Kandasamy, S., Sbeih, K., Lounsberry, T. et al., "Further CFD Studies for Detailed Tires using Aerodynamics Simulation with Rolling Road Conditions," SAE Technical Paper 2010-01-0756, 2010, doi:10.4271/2010-01-0756.
[28.] Fischer, O., Kuthada, T., Widdecke, N., and Wiedemann, J., "CFD Investigations of Wind Tunnel Interference Effects," SAE Technical Paper 2007-01-1045, 2007, doi:10.4271/2007-01-1045.
[29.] Fischer, O., Kuthada, T., Wiedemann, J.: CFD Validation Project on Vehicle Scale Models under the Constraint of an Open Jet Wind Tunnel. EUROMECH Colloquium 509 Vehicle Aerodynamics, Berlin, 2009.
[30.] Fischer, O., Kuthada, T., Mercker, E., Wiedemann, J. et al., "CFD Approach to Evaluate Wind-Tunnel and Model Setup Effects on Aerodynamic Drag and Lift for Detailed Vehicles," SAE Technical Paper 2010-01-0760, 2010, doi:10.4271/2010-01-0760.
[31.] Fischer, O., Kuthada, T., Wiedemann, J., Dethioux, P. et al., "CFD Validation Study for a Sedan Scale Model in an Open Jet Wind Tunnel," SAE Technical Paper 2008-01-0325, 2008, doi:10.4271/2008-01-0325.
[32.] SAE Surface Vehicle Recommended Practice, "Guidelines for Aerodynamic Assessment of Medium and Heavy Commercial Ground Vehicle Using Computational Fluid Dynamics," SAE Standard J2966, Rev. Sep. 2013.
[33.] Mercker, E. and Wiedemann, J., "On the Correction of Interference Effects in Open Jet Wind Tunnels," SAE Technical Paper 960671, 1996, doi:10.4271/960671.
Devaraj Dasarathan, Matthew Ellis, Surya Chinnamani, and Ray Ayala
EXA would like to thank Laydon Composites (a WABCO company) for providing the results for the wind tunnel experimental data and supporting the paper
SAE - Society of Automotive Engineers
EPA - Environmental Protection Agency
NHTSA - National Highway Traffic Safety Administration
GHG - Greenhouse Gas
OEM - Original Equipment Manufactures
[C.sub.D] - Total drag coefficient
[C.sub.Dn] - Normalized drag coefficient
[C.sub.DWA] - Wind Averaged total drag coefficient
[C.sub.DWAn] - Normalized wind averaged total drag coefficient
[C.sub.p] - Coefficient of Static Pressure
[C.sub.Pt]. - Coefficient of Total Pressure
ARC - Auto Research Center
CFD - Computational Fluid Dynamics
CAD - Computer-aided drafting
GCM - Generic Conventional Model
UWAL - University of Washington Aeronautical Laboratory
GEM - Greenhouse Gas Emissions Model
Re - Reynolds number
VLES - Very Large Eddy Simulation
EMF - Electromagnetic fields
TTGF - Tractor-Trailer Gap Fairing
TS - Trailer Skirt
WT - Wind Tunnel
OR - Open Road
BR - Blockage Ratio
[P.sub.N] - Average Static pressure at nozzle, closed to exit section
[P.sub.sc] - Average Static pressure at settling chamber
[P.sub.p] - Average Static pressure at plenum
Table 1. Summary of aerodynamic data for the WT test (All [C.sub.D] values are normalized wrt Baseline 0[degrees] yaw wind tunnel result; Delta = (x-Baseline)/Baseline) [C.sub.Dn] 1/8th Scale Wind Tunnel Test Yaw ([degrees]) Baseline TTGF TTGF Delta [C.sub.Dn]% +TS TTGF TTGF +TS -9 1.261 1.225 1.029 -2.8% -18.4% -6 1.127 1.105 0.967 -1.9% -14.2% -3 1.028 1.017 0.901 -1.0% -12.3% -1 1.001 0.991 0.876 -1.0% -12.4% 0 1.000 0.994 0.873 -0.6% -12.7% 1 1.007 0.997 0.884 -1.0% -12.2% 3 1.050 1.034 0.923 -1.5% -12.1% 6 1.167 1.139 0.998 -2.4% -14.5% 9 1.304 1.246 1.055 -4.5% -19.1% [C.sub.DWAn] 1.091 1.073 0.947 -1.6% -13.1% Table 2. Summary of aerodynamic data for CFD WT nozzle method (All [C.sub.D] values are normalized wrt baseline 0[degrees] yaw WT result; Delta = (x-Baseline)/Baseline) [C.sub.Dn] 1/8th Scale CFD Yaw ([degrees]) Baseline TTGF TTGF Delta [C.sub.Dn]% +TS TTGF +TS -9 1.229 1.191 0.986 -3.1% -19.8% -6 1.094 1.068 0.929 -2.4% -15.1% -1 0.977 0.960 0.861 -1.7% -11.8% 0 0.977 0.960 0.853 -1.7% -12.7% 1 0.980 0.973 0.870 -0.8% -11.3% 6 1.123 1.088 0.943 -3.1% -16.1% 9 1.241 1.206 0.994 -2.8% -19.9% [C.sub.DWAn] 1.062 1.038 0.916 -2.2% -13.7% Delta [C.sub.Dn] wrt WT Yaw ([degrees]) Baseline TTGF TTGF +TS -9 -2.6% -2.8% -4.2% -6 -2.9% -3.4% -3.9% -1 -2.4% -3.1% -1.7% 0 -2.3% -3.4% -2.3% 1 -2.6% -2.4% -1.6% 6 -3.7% -4.5% -5.5% 9 -4.8% -3.2% -5.7% [C.sub.DWAn] -2.7% -3.3% -3.3% Table 3. Summary of aerodynamic data for CFD WT plenum method (All [C.sub.D] values are normalized wrt baseline 0[degrees] yaw WT result; Delta = (x-Baseline)/Baseline) [C.sub.Dn] 1/8th Scale CFD Yaw TTGF Delta [C.sub.Dn] % ([degrees]) Baseline TTGF +TS TTGF TTGF +TS -9 1.065 1.034 0.856 -3.0% -19.6% -6 0.949 0.927 0.806 -2.4% -15.1% -1 0.848 0.833 0.748 -1.7% -11.8% 0 0.848 0.833 0.740 -1.7% -12.7% 1 0.851 0.844 0.755 -0.8% -11.3% 6 0.975 0.944 0.818 -3.1% -16.1% 9 1.076 1.046 0.862 -2.8% -19.9% [C.sub.DWAn] 0.921 0.901 0.795 -2.2% -13.7% Delta [C.sub.Dn] wrt WT Yaw TTGF ([degrees]) Baseline TTGF +TS -9 -15.6% -15.7% -16.9% -6 -15.8% -16.1% -16.6% -1 -15.3% -15.9% -14.7% 0 -15.2% -16.2% -15.2% 1 -15.5% -15.3% -14.6% 6 -16.5% -17.1% -18.0% 9 -17.5% -16.0% -18.2% [C.sub.DWAn] -15.5% -16.0% -16.1% Table 4. Summary of aerodynamic data for CFD WT modified nozzle method (All [C.sub.D] values are normalized wrt baseline 0[degrees] yaw WT result; Delta = (x-Baseline)/Baseline) [C.sub.Dn] 1/8th Scale CFD Yaw Baseline TTGF TTGF Delta [C.sub.Dn] % ([degrees]) +TS TTGF TTGF +TS -9 1.277 1.237 1.024 -3.1% -19.8% -6 1.129 1.102 0.958 -2.4% -15.1% -1 0.996 0.980 0.879 -1.7% -11.8% 0 0.995 0.978 0.868 -1.7% -12.7% 1 1.000 0.992 0.887 -0.8% -11.3% 6 1.159 1.123 0.973 -3.1% -16.1% 9 1.289 1.253 1.033 -2.8% -19.9% [C.sub.DWAn] 1.091 1.066 0.941 -2.2% -13.7% Delta [C.sub.Dn] wrt WT Yaw Baseli TTGF TTGF ([degrees]) +TS -9 1.2% 1.0% -0.5% -6 0.2% -0.3% -0.9% -1 -0.4% -1.1% 0.3% 0 -0.5% -1.6% -0.5% 1 -0.7% -0.5% 0.4% 6 -0.7% -1.4% -2.5% 9 -1.1% 0.6% -2.1% [C.sub.DWAn] 0.0% -0.6% -0.7% Table 5. Summary of drag coefficient data for Open-Road, WT test and CFD WT (All [C.sub.D] values are normalized wrt baseline 0[degrees] yaw wind tunnel result; Delta = [x-Baseline]/Baseline) Yaw Baseline TTGF ([degrees]) [C.sub.Dn] [C.sub.Dn] Delta [C.sub.Dn] 0 0.899 0.886 -1.5% OR 2 0.918 0.907 -1.2% 6 1.082 1.047 -3.2% [C.sub.DWAn] 1.005 0.983 -2.2% WT [C.sub.DWAn] 1.091 1.073 -1.6% CFD WT [C.sub.DWAn] 1.091 1.066 -2.2% TTGF+TS [C.sub.Dn] Delta 0.804 -10.5% OR 0.814 -11.3% 0.896 -17.1% 0.862 -14.3% WT 0.947 -13.1% CFD WT 0.941 -13.7%
|Printer friendly Cite/link Email Feedback|
|Author:||Dasarathan, Devaraj; Ellis, Matthew; Chinnamani, Surya; Ayala, Ray; Haws, James|
|Publication:||SAE International Journal of Commercial Vehicles|
|Date:||Oct 1, 2016|
|Previous Article:||Aerodynamic performance of flat-panel boat-tails and their interactive benefits with side-skirts.|
|Next Article:||EPA GHG certification of medium- and heavy-duty vehicles: Development of road grade profiles representative of US controlled access highways.|