Flight testing and Race car testing - a short abstract

Hello Everybody,

Flight Testing is a very extensive and demanding phase of development of an Aircraft. There is no way to see if the predicted models of the Aircraft used at development stages of the aircraft is accurate unless you perform Flight Testing. Before the construction of the prototype, data gathered from simulations (Computational Fluid Dynamics, simulations of 6 degrees of freedom of aircraft, etc), wind tunnel testing, bibliography and isolated system testing is all that the Engineering team has to predict how the Aircraft will behavior when operating.

Fig. 1 - An Aircraft properly instrumented for Flight Testing

However, there must be flight testing to get the real data of the aircraft, in order to refine/update aircraft models and simulations. This is possible due to planned aircraft maneuvers and the extensive use of instrumentation at the aircraft. Instruments like DGPS, accelerometers, rate gyros, anemometric system, instrumented flight control system, etc, must be installed into the aircraft in order to acquire data enough to perform aircraft motion reconstruction into simulators. Because Flight Testing is very expensive, those instruments normally have redundancies in order to not lost data.

Fig. 2 - Accelerometers and rate gyros and its use at an aircraft.

Fig. 3 - Differential GPS (DGPS) on an aircraft. Due to triangulation between aircraft antenna, satellite and base antenna, a very precision location of the aircrat with respect to antenna is possible.

In addition to, data measured must be treated accordingly, adjusting delays of measurements, bias and scale factor of each instrument, noise treatment, etc. There are many ways to make this treatment, varying from simple individual signal treatment to methodologies that treats those signals using Flight Path Reconstruction techniques for identification of biases and noise treatment, using techniques of simple linear regression to sophisticated Non-linear filtering methods. As an example of Non-linear filtering, we have the variations of Kalman Filters (like the Extended, Unscented and Cubature Kalman Filters) and Particle Filter (which is more suitable when the model is highly non-linear). The usage of each technique depends of time of processing available, data accuracy needed, where the data will be used, regulations for certification of the data identified, etc.

With the techniques above, it is possible to identify from a simple brake coefficient to a full aerodynamic data of the aircraft.

All of those techniques could be also used at automotive industry, in order to obtain aerodynamic, tire and vehicle dynamics data from the car (for example), recreating the car motion and identifying the parameters using the same techniques from flight testing. These tests are very important in order to verify if the prototype has the same characteristics than its computational model. With this, very accurate simulations for a race team can be made, which could lead to sucess on preparing the correct strategy in order to complete a grand prix or even win the race!

Fig. 4 - Racecar properly instrumented for race track testing. With the instrumented car, race engineers alongside trackside aerodynamics engineers and vehicle dynamics engineer can identify racecar parameters.

Did you like the article above? Do not forget to subscribe in order to know when a new article is available. Thank you for yout attention!