images/wallpaper.png

Chassis of Chimera Evoluzione

images/design.png
images/designChassisgrafico

Engineering Specification

QFD diagram to understand parameters needed for an initial design, based on adversary teams.

images/designChassis1.png

Chassis Design Workflow

images/designChassisgrafico

Configuration

3 different layout candidates were taken into FEM analysis to define a starting point. The one with better torsional stiffness compared to the mass has been chosen as the basis for the following optimization process. Configuration 3 has been selected and taken into successive development.

images/designChassis2.png

Configuration 1: Ks_torsion = 50.35 [Nm/° Kg]

images/designChassis3.png

Configuration 2: Ks_torsion = 69.29 [Nm/° Kg]

images/designChassis4.png

Configuration 3: Ks_torsion = 99.31 [Nm/° Kg]

Design Exploration

Input parameters:

  • Structure geometry variables: - angles on side view - lengths on side view - widths on top view
  • Tubes cross section variables: - Section Moment of Inertia - Area cross Section
images/designChassis5.png images/designChassis6.png

Output parameters:

  • Torsional stiffness: main parameter for chassis design, crucial for vehicle performance.

    images/designChassis/designChassis7a.png
  • Bending stiffness: chassis under load with a force directed downward in the driver cockpit area.

    images/designChassis/designChassis7b.png
  • Braking stiffness: a rigid link simulates the contact point of the wheel on the ground. A force is applied on the center of gravity to simulate the braking phase.

    images/designChassis/designChassis7c.png

Parameters Correlation

Parameters correlation study allows to:
  • Determine which parameters have higher influence on the design
  • Find relationship grade between different parameters
Rank order clustering algorithms:
  • Assign the binary weights with its help determination of the weight in decimal for each row and column
  • Rearrange the rows in the order of descending
  • Repeat steps 1 and 2 for every column. And the reiterate from step
  • The procedure has to be repeated till there reaches a point that the elements cannot be further changed from their positions.
images/designChassis7.png

3 macro families found and taken into further development process:

images/designChassis8.png

First Group:

Variables: 7 (position)

Target: maximize Ks_bend and Ks_break

Main constraints: mass =< 30 Kg

Optimization method: MOGA (Multi-Objective Genetic Algorithm)

images/designChassis9.png

Second Group:

Variables: 13 (lengths and angles)

Target: maximize Ks_torsion > 2000

Main constraints: mass =< 30 Kg

Optimization method: NLPQL (Nonlinear Programming by Quadratic Lagrangian)

images/designChassis10.png

Third Group:

Variables: 17 (moment of inertia and tubes cross sectional area)

Target: maximize Ks_torsion > 2000

Main constraints: mass =< 30 Kg

Diameters and thickness from rules Optimization method: NLPQL (Nonlinear Programming by Quadratic Lagrangian)

Optimization Results

The black chassis represent the first design layout, before optimization. The yellow chassis is the result after optimization

images/designChassis11.png

Design for Performance

FEA Models

Torsion

images/designChassis12.png

Accelleration

images/designChassis13.png

Bending

images/designChassis14.png

Final Design Results

images/designChassis15.png

Performance Evaluation

  • Torsion Analysis
  • Modal Analysis
  • Transient Analysis on circuit

Torsion analysis

k_torsion = 2054 Nm/° with rigid wishbones

images/designChassis16.png
images/designChassis17.png

Transient Analysis on Circuit

The trajectory plotted below is calculated to avoid the wheel slip with the highest possible speed. To find the ideal path and the corresponding accelerations, a Matlab code has been developed. The data relating to the accelerations and therefore the forces acting on the structure were transferred into the Ansys APDL car model.

images/designChassis19.png
  • Colors:
    • Track Limits: Green
    • Tire Trajectory: Red
    • Car Centreline: Blue
images/designChassis20.png
images/designChassis21.png
images/designChassis22.png
images/designChassis23.png

Loads on Tires

To understand how the total structure is stressed the load history must be known. The starting point are the reaction forces that the car receives from the ground through the wheels. The plot below represents the tires reaction forces during the skid pad test. The weight car force was considered in the dynamic load transfer. Vertical loads results:

images/designChassisgrafico

Grip limit values found during the tire analysis model:

  • 1700 [N] for the front wheels
  • 3800 [N] for the rear wheels.

Observing the lateral loads results can be seen that the rear wheels in the points of higher acceleration slip. In fact, the springs that simulate the wheels give a reaction force higher than the real grip condition. This can lead to have a peaks load on the structure higher to the actual ones.

Stress analysis

images/designChassisgrafico

The stresses peaks occur during trajectory change. This happens when the machine moves from one circle to the other during the skid pad test.

Chassis Torsion

To find the chassis torsion in Ansys, has been calculated the difference between the front and rear rotation of the axes with respect to the roll center.

images/designChassisgrafico

As can be seen from the chart above, the chassis behaves like a torsional spring due to the weight distribution difference. The advantage of having such small deformations is that at each elastic return of the chassis in the curve exit, there are less oscillations, resulting in a better drivability

Results Validation Model

Final results from the two software are plotted below:

images/designChassisgrafico
images/designChassisgrafico
images/designChassis25.png
images/designChassisgrafico

Additive Manufacturing Integration

The idea of 3D printed joints

Once the X, Y, Z coordinates of the kinematic points have been defined, is mandatory to maintain the maximum precision after the welding. Adding the brackets “by hand” using a complex support structure is expensive, less precise and incline to misalignments and errors.

The possibility given by additive manufacturing allows to overcome many of those problems. The final version of the joints has unique features:

  • Integrated suspension brackets
  • Topology optimized with internal ribs and variable thickness
  • Nesting guides for precise assembly, torque transmission, easier welding process
  • Allowing better electrical grounding across the entire structure
  • Head-to-head welding with tubes due to 4 spacers for each terminal

Design process illustration:

images/designChassis26.png
images/designChassis27.png
images/designChassis29.png
images/designChassis29.png
images/designChassis30.png

FEA Validation

images/designChassis31.png
images/designChassis32.png
images/designChassisgrafico
images/designChassis34.png

NODES REACTION FORCES (EXPRESSED IN NEWTONS)

images/designChassisgrafico

Conclusion

In the first and the second plot the lines are essentially overlapped. This proves the validity of the Ansys model. In the central zone of the curves, that corresponds to the work range, the lines are the basically the same. The only difference can be seen in the final parts when the antiroll stiffness rear is low. This low value generates a non-linearity in the Ansys model not expected in Matlab. This non-linearity can be observed in the plot of the structure torsion in respect with the front antiroll distribution stiffness:

images/designChassisgrafico

An assumption done in the Matlab analytical model is that the torsional stiffness of the structure is linear. On the other hand, as can be seen from the above plot the structure behaviour coming from the Ansys model it isn’t linear changing the front antiroll distribution stiffness. While calculating the stiffness of the structure with shock absorbers infinitely rigid the stiffness response is linear.

images/designChassis36.png
images/designChassis37.png

Analysis before and after joints integration. With the same loads and constraints applied, torsional stiffness has been increased by almost a factor of 3

Manufacturing Process

The addition of joints allows a much easier assembly and welding process, the structure can support itself without requiring a complex tooling structure. Weld bead is simpler and with more controlled thermal distribution. Tubes laser cutting in easier, with less waste material.

images/designChassis38.png
images/designChassis39.png