![]() |
07 - Systems, Subsystems and EquipmentNONLINEAR BEHAVIOR IN AIRCRAFT FUEL GAUGE READINGS: AN EXPLORING ANALYSISM.A.D. Di Marzo¹, O.A. Gonzatto Junior¹, H.N. Najafabadi, Linköping University – LiU, Sweden; J.H. Bidinotto¹; ¹University of São Paulo - USP, Brazil Fuel level gauge in aircraft is a challenging matter, once the aircraft’s movements affectrnits measurement. In this context, this study addresses a detailed analysis of simulatedrnresults from a computational tool specifically developed to study aircraft fuel systems.rnThe tool simulates fuel sensor readings across various aircraft attitudes and fuel levels.rnThis analysis highlights the high nonlinearity of these signals in different. The objectivernwas to explore the relationship between the sensor readings, aircraft orientation, and fuelrnvolume. The dataset consisted of simulated readings from eight capacitive sensors placedrnat different locations inside an aircraft fuel tank. Each row in the dataset represented arnspecific combination of aircraft attitude (yaw, pitch, roll) and a “cut” (or cross-section)rnof the fuel tank, simulating a particular fuel volume. The sensors’ readings at each cut,rncorrelated with the simulated fuel volume and aircraft attitude, provided valuable insightsrninto the relationship between these variables, revealing a complex nonlinear interaction.rnContour plots were generated to visualize each sensor’s readings at different pitches andrnrolls, and a polynomial regressions surface was fitted to understand the trends in thernsensor readings. A detailed statistical analysis was conducted to quantify the coefficientsrnand significance of pitch and roll angles and their interactions, as well as a NormalizedrnSensitivity Matrix (NSM) was used to analyze the sensitivity of the multiple sensors tornchanges in attitude angles, providing insights into how sensors respond to the attitudernvariations and their relative sensitivity. Each sensor responded differently to pitch, roll,rnand yaw variations, providing essential information for sensor positioning and systemrndesign. This analys |