Improved Vehicle Fuel Economy Through Optimization Aaron Godfrey - - PowerPoint PPT Presentation

Improved Vehicle Fuel Economy Through Optimization

Aaron Godfrey – CD-adapco Michael Elmore – CD-adapco

Building an optimized car has difficult tradeoffs:

– Seal the front end to lower drag – Open the front end to increase cooling

Car must give the lowest drag while keeping the engine cool Drag is analyzed at high speed on flat ground The engine is at maximum load moving uphill

The Challenge of Tradeoffs

We encounter the largest drag at high speed We encounter the largest heat load uphill A particular low drag design must be capable of rejecting a large amount of heat in a scenario when drag is not important! We must determine the car’s HR needs based upon how it’s driving to determine if a design is feasible Consider these scenarios from the perspective of power usage This description will allow us to calculate the HR for a scenario!

Coupled Analyses

𝑬 𝑰𝑺

Uphill High-speed

Power Budget

𝑭𝒐𝒇𝒔𝒉𝒛 𝒋𝒐 𝒈𝒗𝒇𝒎 𝒖𝒋𝒏𝒇 𝝑 = 𝒘(𝑬𝒔𝒃𝒉 + 𝑺𝒑𝒎𝒎𝒋𝒐𝒉 𝑺𝒇𝒕𝒋𝒕𝒖𝒃𝒐𝒅𝒇 + 𝑮𝒑𝒔𝒅𝒇 𝒆𝒗𝒇 𝒖𝒑 𝑯𝒔𝒃𝒘𝒋𝒖𝒛) 𝑭𝒐𝒇𝒔𝒉𝒛 𝒋𝒐 𝒈𝒗𝒇𝒎 𝒖𝒋𝒏𝒇 𝟐 − 𝝑 = 𝑺𝒇𝒓𝒗𝒋𝒔𝒇𝒆 𝑰𝒇𝒃𝒖 𝑺𝒇𝒌𝒇𝒅𝒖𝒋𝒑𝒐 𝑰𝑺 = 𝟐 − 𝝑 𝝑 𝒘(𝑬 + 𝑺𝒔 + 𝑮𝒉) The engine provides power to move the car: But due to inefficiency, it must also be cooled: So our budget for power must balance: How do we calculate the components of the RHS?

Equations for Each Component

𝑰𝑺 = 𝟐 − 𝝑 𝝑 𝒘(𝑬 + 𝑺𝒔 + 𝑮𝒉) 𝑬 = 𝟐 𝟑 𝝇𝒘𝟑𝑫𝒆𝑩 𝑺𝒔 = 𝑫𝒔𝒔𝒏𝒉 ∗ 𝐝𝐩𝐭(𝛊) 𝑮𝒉 = 𝒏𝒉 ∗ 𝐭𝐣𝐨(𝛊)

We try to minimize this component Crr is the rolling resistance coefficient θ is the incline (hill) angle for Rr and Fg

Now we have a series of equations that describes our car in any scenario For a given Cd, we can determine the required HR!

1. Determine Cd 2. Calculate the required HR 3. Impose the required HR on the radiator 4. Modify the radiator inlet temperature until this HR is achieved This procedure is carried out for both (high speed and uphill) cases for each design If the output of 4 is too high (boiling coolant), the design is considered infeasible Goal: Minimize drag while maintaining a feasible coolant temperature

Solution Technique

Grill can be modified in many ways

– Change the vane angle – Change the horizontal baffle length – Change the vertical baffle length

Optimization Variables: Grill

Lower bumper can be modified in many ways:

– Change the open area width – Change the open area height

Optimization Variables: Lower Bumper

One way to increase HR is to use a fan We do not want this fan to be too loud As a secondary objective, we will minimize the fan RPM and maximize its radius Optimate allows us to say these variables are not as important

Another Variable - Fan

Surface wrapper used to close geometry Trimmer with prism cells volume mesh – 5.2M cells

Grid Generation

Dual-stream heat exchanger utilized to calculate radiator inlet temperature A single-stream heat exchanger (a condenser) is in front of the radiator Radiator, condenser treated as porous media

Physics

• Steady, 3D, RANS
• Ideal gas
• Realizable k-ε Turbulence
• Two-layer wall treatment
• 50/50 Glycol coolant

Crea eate e and d conver nverge ge base selin line Setup up Optim imate ate Evaluat aluate e Desig sign Crea eate e New w Desig sign SHERPA PA determ ermines ines desig sign n perform rforman ance ce STAR-CC CCM+ + build ilds s CAD and d mesh sh Drag ag case se Tow case se

For each design: For all designs:

Best Designs

Design Comparisons

Baseline Optimized

Design Comparison

Baseline Optimized

Design Comparison

Baseline Optimized

Design Comparison

Baseline Optimized

Design Comparison

Baseline Optimized

Design Comparison

Baseline Optimized

With 60 designs run, Optimate has shown large improvement Drag reduced by 12.7%

By the Numbers

Baseline case Parameterize in 3D-CAD, Mesh (5.2M cells) Converge baseline run Man Time: 4 hours Machine Time: 4.5 hours Optimization Set up Optimate and submit for 70 design iterations 4 simultaneous jobs x 32 cores each = 128 cores Man Time: 10 minutes Machine Time: 157.5 hours/ 4 runs = ~40 hours

8.5 Hours 40 Hours