2006 VT Motorsports Design of Experiments
October 30, 2005
Brandon Evans
Chris Barbour
Paul Landers
Sam Vernatter
Tim Johnsrud
Introduction
Testing is a vital part to creating a competitive race car. This process is especially crucial when a new design is being developed every year. The cars from the previous years need to be tested with proper data acquisition systems to help detect any areas that may need improvement for the coming year. The team has been offered the use of a Pi System for testing days. This will greatly improve our analysis of the 2005 car and allow for us to better improve upon this design. This design of experiments will break down what type of experiments will be performed using this system. In order to optimize the use of the Pi system, multiple tests will be conducted in the same day. Most of these tests may be carried out at the same time.
Testing with Current Sensors
The team currently owns a limited supply of sensors. These sensors are compatible with the system except for their connections. These connections will need to be converted to a deutsch connector. Our current supply of sensors consists of four linear potentiometers and several wheel speed sensors that may be compatible with the Pi System. The team currently owns a Hot Lap system which will be able to take lap times around a continuous course but, will not be able to take times during acceleration runs without purchasing another transmitter. A collection of strain gauges that can be used to measure the strain/force in each of the suspension members is available to the team. Table A1 in the appendix lists the specifications for each of the different types of strain gauges that we have. This is just in terms of what we can use for suspension testing.
Since an accelerometer is built into the Pi MCU, lateral and longitudinal accelerations will be able to be measured. Shock displacements will be measured at each shock using linear potentiometers. This data can give us shock and wheel displacement, velocity and acceleration. Front and rear body roll angle, as well as roll gradient, can be calculated from this data. Using this data combined with steer angle, camber angle can be calculated using camber gain coefficients measured from physical testing. Wheel loads can also be calculated using data from the shock dyno curves and shock displacement and velocity.
Loads in the arms will be measured using strain gauges on each of the suspension members. This data will be used to validate the design calculations. Several different loading scenarios will be run with the strain gauges on the car. These loading scenarios can be run independently or they can coincide with other tests to save time. Full acceleration, full braking, steady-state cornering, full braking and cornering and full acceleration and cornering. These tests will be run a number of times with varying radii and length in order to get the best data. Six strain gauges are required at each corner with each taking up one channel. This limits the number of strain gauges that can be run at one time to twelve, which means that either the front and the rear or the left and the right need to be tested separately. In order for the strain gauges to work with the Pi System, a custom connector block will be build to interface between the two. This block serves as a signal conditioner and amplifier which will allow the gathering of much more accurate results. With the help of Randy Smith, the ME Department expert on strain gauges, the strain gauges will be correctly installed and configured in order to correctly measure the loads in the A-arms.
Possible Measurements
The problems with the current list of sensors are the lack of integrated timing equipment, steering angle sensor, thermocouples, and brake system sensors. The timing equipment problem will be an issue for accurate acceleration testing. A steering angle sensor is necessary to monitor the oversteer/understeer behavior of the car during the skidpad and endurance/autocross testing. This would be one of the sensors that would be necessary to buy for testing this year. Infrared thermocouples would be used to measure the surface temperature of the tire during testing. Data of this type will be important to help adjust the static setup to get the temperature profile to match what is needed. For the braking system, infrared thermocouples and inline pressure sensors would be used to monitor the brake system during testing. This would be beneficial to know the heat characteristics of the rotor during the course and separate specific maneuvers, as well as the force from the caliper acting on the brake rotor.
In order to measure the speed at each wheel a set of non-contact wheel speed sensors will be used. These will be measured using one of the digital channels on the system. This will measure vehicle speed and can be used to calculate slip ratio. The data received from these sensors will show when the wheels are slipping under acceleration and when the wheels are locking up under braking. Wheel speeds will be important in all testing and will be measured during each testing day.
Tire temperatures are vital data that directly indicates the handling characteristics of the car. The tire temperatures will be measured using non-contact thermocouples at the outside, center and inside of the tire. This will allow for the temperature profile of the tire to be measured throughout the entire course. Analysis will be done on the data and necessary changes will be made to the setup to heat the tires more evenly. Changes that will be made based off tire temp data include camber, tire pressure, wedge, and front to rear weight bias.
Steering angle is a crucial measurement that will be used to evaluate the handling of the car. The angle will need to be measured with a rotary potentiometer on the steering column or with linear potentiometers measuring the travel of the steering rack. Using known ackerman values this data can be used to get the steering angle at each wheel. This data will be useful when plotted against lateral acceleration. From this plot the understeer/oversteer balance of the car can be determined throughout different sections of the course.
Another important area that will need to be measured is the brake system. Brake line pressure will be measured inline before each caliper. From this data, brake bias, force generated by the caliper on the rotor, and left/right brake balance can be calculated. The brakes rotors will also be monitored with infrared thermocouples. This data will give real time brake temperatures and can be used to monitor the heating and cooling of the brake rotor throughout the course. Brake testing can be run on braking specific test courses or on regular endurance/autocross setups.
Timing beacons must be purchased in order to break up laps into segments. With the course broken into segments, each section can be evaluated separately to determine what the handling characteristics are for each. Multiple beacons will allow for acceleration runs to be measured. In order to accurately record acceleration times a beacon should be placed at the beginning and end of the track.
Testing Days
Course Setup
The acceleration course will be a 246 foot (82 yard) straight line. Beacon transmitters will be placed at start and finish of the straightaway to record the time it takes to complete the distance. The car will start from a stop positioned one foot behind the starting gate. This testing will be run the same way as the acceleration event at competition. Elapsed time, rear shock displacement, longitudinal acceleration and wheel speeds will be measured for this test. Elapsed time will quantify the performance of the car so it can be compared to previously compiled data. Shock displacement can be used to calculate how much the rear of the car squats under full acceleration. The data can also be used to estimate rear wheel loads. Longitudinal acceleration will give real life value to compare to our design values. It may also be possibly to estimate the CG height based on the wheel load estimates and the measured lateral acceleration. From the wheel speed, the speed at the last gate will be known without the use of a radar gun. Relative wheel speed from the front and rear will also show the amount of wheel spin the car is producing at launch. The rear of the car will also be instrumented with strain gauges to measure the forces in the suspension arms under full acceleration. This data will be compared to the design data to validate our design assumptions.
Skidpad testing can either be setup in a circle or a figure eight. The figure eight is the setup used at competition and will be more beneficial for driver preparation for competition than testing data. The skidpad will have an inner radius of 25 feet and an outer radius of 35 feet. Each lap around the circle will be timed to compare runs. Average lateral acceleration can also be calculated from this time and will be confirmed by the actual measurement from the accelerometer. From the shock displacements front and rear vehicle roll angle can be calculated. Vertical tire loads can also be estimated from this data. Steering angle at each wheel combined with the data from the shocks can be used to calculate camber at each wheel throughout the course. Steering angle graphed against lateral acceleration will also display the understeer/oversteer behavior of the car. Wheel speed will be measured to know the speed of the car around the circle and to calculate slip ratio. Tire temperature at three points on each tire can be used to help bias the setup to get more heat in the tires and to know the relative temperature across the contact patch. Strain gauges will be used to measure the suspension member loading during the test runs. Knowing the maximum lateral acceleration and the loads in all the members we can correlate this data to our design calculations to verify the assumptions made during the design process.
Brake testing will be done using several different course setups. Straight line brake testing will be done as well as brake testing on sections of endurance/autocross courses. Wheel speeds front and rear, brake temperature, tire temperature, brake line pressures, strain in a-arms and longitudinal acceleration will all be measured. Wheel speeds will give test starting speed and will also show whether the front or rear wheels are locking up. Brake temperature will help with design and tell if the rotors are in the optimal range for the pads. Measured tire temps will tell how much heat is being generated in the tires by braking and will help bias static setup to get more heat in the tires under braking. Brake line pressures can be used to calculate the amount of force the caliper clamps the rotor with. Strain in the a-arms is used to calculate member loading and can be used to verify design calculations. Longitudinal acceleration can be used to aid in design and member loading calculations.
Full endurance/autocross tracks will be setup. The courses will be designed using the specifications from the rules book on corner radii, straightaway length and slalom cone separation. The specifications are listed in Table 1. Split times can be taken at certain sections to examine a smaller part of any given course design. Potential course designs will be similar to the endurance and autocross tracks at the past several year’s competitions as well as original tracks created using the data in Table 1. During these tests all parameters will be measured.
Table 1. Course specifications from rules. Courses will be designed using this data.
Examples of certain corners can be designed for specific testing. There will be constant radius turns for steady-state cornering. These turns can also be used to test corner entry and corner exit transient behavior. The radius will be varied for the minimum and maximum radii that will be seen at the endurance and autocross courses at competition. A slalom will be setup to further investigate transient behavior. The cone separation is specified in the rules. Straightaways into specified radius corners can be setup to test the corner entry behavior and exit behavior. Radius and straightaway length will be varied within the range that will be seen at competition. More course sections can be designed and tested based on test experience to gain better results. A few sample course designs can be seen in Appendix A.
Suspension parameters to vary
Each of the four testing tracks will have a specific list of variable that will be changed during the test. These variables are listed in Table 2. Every test will have the same five parameters that will be varied. The parameters to be changed at every test include camber, toe, tire pressure, damper settings, and springs. Parameters such as sway bars and sway bar settings will be varied according to the particular test being conducted.
Table 2. Suspension variables that will be varied for each type of testing
Acceleration / Skidpad / Braking / Endurance/AutocrossCamber / Camber / Camber / Camber
Toe / Toe / Toe / Toe
Tire Pressure / Tire Pressure / Tire Pressure / Tire Pressure
Damper Settings / Damper Settings / Damper Settings / Damper Settings
Springs / Springs / Springs / Springs
Sway Bars / Brake Bias / Sway Bars
Sway Bar Settings / Sway Bars Settings
For each adjustment the range and amount of adjustment will be set before the testing. Some of these ranges and amounts of adjustment may be changed during testing to gain better results. The preliminary values are listed in Table 3.