Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.600923
Title: The wind tunnel simulation and effect of turbulent air flow on automotive aerodynamics
Author: Mankowski, Oliver Andrew
Awarding Body: Durham University
Current Institution: Durham University
Date of Award: 2013
Availability of Full Text:
Access through EThOS:
Access through Institution:
Abstract:
This thesis presents the research completed to design, commission and evaluate a turbulence generation system for Durham University’s 2m wind tunnel and the development of a method to simulate on road turbulence and measure its effects on a vehicle. The objective was to develop a test approach for simulating and analysing a vehicle’s response to unsteady airflows. This approach focussed on simulating the overlap of the range of turbulence frequencies which exist both at significant energy in the on road environment and the frequencies at which a significant vehicle response is seen. The frequency range where both conditions exist was seen to be between 1 – 10Hz. Confirmation of this transient frequency range was through the use of an admittance technique developed in this thesis which compares unsteady effects to quasi-steady effects. The technique was also developed to account for the component of unsteady pressure self excitedness that exists, effectively the noise component in an admittance analysis. The approach concluded with the operation of a new turbulence generation system (TGS), which simulates the wind characteristics experienced by vehicles as they move through the on road wind environment. The design was informed both by previous works and an on road investigation of environment and vehicle response. An on road study consisting of 8,800-seconds of on road measurements was completed to record incoming flow velocities and passenger sideglass static pressures (a region noted in studies to show a notable response to yawed flow). The on road environment was shown to have significant energy in the 0.1 10Hz range (reduced frequency K = 0.1 10 for a vehicle driving at highway speeds). Yaw angles ranged between ±20o, but with the vast majority within ±6o. Correspondingly, the turbulence intensity range was 0.5 15%, but with the majority below 8%. The challenges of generating turbulent length scales in the order of size of a vehicle’s length, whilst also at reasonable turbulent intensities were assessed to be beyond the capability of a passive device. Through a series of iterative CFD tests, an active “lift based” TGS was designed, based around two oscillating yaw aerofoils, which also encompassed additional inlet and outlets controlled by shutter panels. These ensured that the jet shear layer did not interact with the test model and helped to achieve higher peak yaw angles and good flow uniformity. A full aerodynamic design of the TGS was completed from the CFD studies, from which a high level mechanical design was specified including target aerofoil displacement and acceleration rates, control system requirements and the linkage design. The construction and installation of the TGS was undertaken by an external contractor. Due to its numerous configurable control parameters, a significant commissioning project was required and completed to determine the system’s optimum configuration. The system is capable of operating up to 10Hz at ±10o flow yaw angle and in a programmed arbitrary mode. The system also has the capability to generate pitch and longitudinal turbulence effects (Cooper et al (1989)). A 40% scaled model of the vehicle studied on the road was placed into the wind tunnel and a range of cases were generated including wind conditions previously recorded on road. The results showed that the technique of using both a roof mounted probe and the TGS system are able to take on road flow conditions and accurately recreate their effects on vehicles in a wind tunnel. Multiple aspects of the work (on road, CFD and wind tunnel) showed that below K = 0.3 pressure fluctuations behaved in a quasi steady manner. Admittance greater than unity was observed near the A pillar, but admittance was generally below unity and reduced progressively for K > 1. Self excitedness was seen to decrease in unsteady tests (in comparison to quasi steady) tests in the A pillar region, but increase between unsteady to quasi steady tests in the mirror wake region.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.600923  DOI: Not available
Share: