Jetstream 31 national flying laboratory: Lift and drag measurement and modelling

https://doi.org/10.1016/j.ast.2016.11.001Get rights and content

Abstract

Lift and drag flight test data is presented from the National Flying Laboratory Centre, Jetstream 31 aircraft. The aircraft has been modified as a flying classroom for completing flight test training courses, for engineering degree accreditation. The straight and level flight test data is compared to data from 10% and 17% scale wind tunnel models, a Reynolds Averaged Navier Stokes steady-state computational fluid dynamics model and an empirical model. Estimated standard errors in the flight test data are ±2.4% in lift coefficient, ±2.7% in drag coefficient. The flight test data also shows the aircraft to have a maximum lift to drag ratio of 10.5 at Mach 0.32, a zero lift drag coefficient of 0.0376 and an induced drag correction factor of 0.0607. When comparing the characteristics from the other models, the best overall comparison with the flight test data, in terms of lift coefficient, was with the empirical model. For the drag comparisons, all the models under predicted levels of drag by up to 43% when compared to the flight test data, with the best overall match between the flight test data and the 10% scale wind tunnel model. These discrepancies were attributed to various factors including zero lift drag Reynolds number effects, omission of a propeller system and surface excrescences on the models, as well as surface finish differences.

Introduction

In the field of full scale aircraft aerodynamics, many challenges still exist in predicting the aircraft performance, particularly at high Reynolds numbers and transonic Mach numbers [1], [2]. More specifically, if modelling an aircraft in a wind tunnel, model scale and tunnel conditions generally result in one of the non-dimensional variables of interest being incorrectly scaled and methods such as boundary layer transition trips are required [3]. Numerical or computational fluid dynamic (CFD) modelling schemes of full scale aircraft have advanced substantially since the 1970's but assumptions must still be made. For example, if a complete solution of the Navier Stokes equations was to be obtained, a direct numerical simulation (DNS) would be required where the mesh density scales with Reynolds number Re9/4 [4]. Therefore, given large civilian aircraft have Reynolds numbers based on wing chord in excess of 10 million, three dimensional mesh sizes in excess of 1×1015 cells would be needed for a DNS model. To address this mesh limitation, turbulence models and other advanced numerical schemes have led to significant developments in alternative numerical methods, based on the Reynolds Average Navier Stokes (RANS) equations and more recently Detached Eddy Simulation (DES) and Large Eddy Simulation (LES) methods [5], [6]. However, there is still a need for numerical model validation through either flight test or wind tunnel testing.

In flight test it is possible to correctly scale both Mach and Reynolds number to measure an aerodynamic system of interest on a full scale aircraft [7], [8]. This approach, however, has challenges in terms of finding a suitable flight test technique, but also in finding a suitable flight test platform, which can modified within an available budget and to the satisfaction of the certification authorities [9], [10]. Furthermore, the flight test facility is likely to need specialist instrumentation [11], [12].

The nature and associated costs of research flight testing has historically limited its use to national facilities, funded by organisations such as NASA, DLR or ONERA [13], [14], [15], [16], or by large aerospace companies such as Airbus [17]. In all these cases, regular research campaigns using flight test platforms have allowed the study of challenging flow regimes such as laminar flow transition [18], [19] and the development of complex aircraft systems [13], [14]. These advantages and limitations of flight test, however, must also be considered against the advantages but also limitations of both numerical and wind tunnel test campaigns. For example, the increasingly complex numerical methods, although allowing detailed studies of the flow field, require increasing computational resources. In wind tunnel testing, the increasing demands of wind tunnel size and instrumentation for data fidelity, must also be balanced by the accuracy of the data required and the associated cost.

The objective of the following paper is to compare full scale lift and drag flight test data to other comparable models, including a numerical RANS CFD model, an empirical model based on Engineering Science Data Unit (ESDU) methods [20] and data taken from two wind tunnel models. One of the wind tunnel data sets is historical data from tests by the original equipment manufacturer (OEM), Handley Page Ltd [21]. The flight test data presented in this paper has been recorded from over 1000 flights and includes methods of data validation for both the drag and angle of attack. At the time of writing, the authors believe this amount of flight test data in combination with the thrust validation analysis, has not been published before, from an equivalent turboprop aircraft. Hence this data presents the aerodynamic community with valuable reference material for future modelling and validation. To this end, the basic surface model will be made publically available through the corresponding author who can be contacted directly. In the paper, from general comparisons of the different models with flight test, the authors discuss and explain the differences between the model data sets and the flight test data.

Section snippets

Jetstream 31 flying laboratory

The flying laboratory is a modified British Aerospace Jetstream 3102, twin turboprop aircraft. This commuter category aircraft has 19 passenger seats, with 18 of these seats in six rows of three. The aircraft was acquired by Cranfield University in 2002 and underwent a major modification to become a flight test laboratory.

The aircraft modification included the installation of seat back displays for each flight test observer and the fitting of new instrumentation, including a P.C., a National

Comparison models

The following section describes the models used for comparison with the flight test data. These include a computational fluid dynamic (CFD) numerical model, a semi-empirical model and two wind tunnel models, one based on historical data published by the first manufacturer of the Jetstream, Handley Page Ltd. The following shorthand labels will be used when presenting data in the remainder of the paper:

  • i)

    flight test data – Flight Test

  • ii)

    computational fluid dynamic numerical model – CFD

  • iii)

    empirical ESDU

Results and discussion

In the following, the flight test data is compared to results from the CFD model, the wind tunnel testing including selected Handley Page results [21] and the empirical model. The comparisons are in three parts including the lift characteristics, the drag polar and specific examination of the drag components of the aircraft. The discrepancies and possible improvements to the different datasets will also be discussed.

Conclusions

This paper has presented flight test data taken from the National Flying Laboratory Centre, Jetstream 31 twin turboprop aircraft. The aircraft is a commuter category aircraft with 19 seats and has been set up to demonstrate flight test techniques to aerospace engineers. Over the period of 10 years of operating the aircraft, over 1000 lift-drag data points have been analysed. The data presented here has allowed analysis of the basic lift and drag aerodynamic characteristics of the aircraft, with

Conflict of interest statement

The authors have no conflicts of interest in the work presented in this article.

Acknowledgements

The authors would like to acknowledge the staff of the Applied Aerodynamics wind tunnel workshops for their support in this work.

References (43)

  • A. Jameson et al.

    50 years of transonic aircraft design

    Prog. Aerosp. Sci.

    (2011)
  • B.E. Launder et al.

    The numerical computation of turbulent flows

    Comput. Methods Appl. Mech. Eng.

    (1974)
  • C.-T.E. Lan et al.

    Airplane Aerodynamics and Performance

    (2003)
  • J.B. Barlow et al.

    Low-Speed Wind Tunnel Testing

    (1999)
  • M. Gharib

    Perspective: the experimentalist and the problem of turbulence in the age of supercomputers

    J. Fluids Eng.

    (1996)
  • P.R. Spalart

    Topics in detached-eddy simulation

  • B. Galperin et al.

    Large Eddy Simulation of Complex Engineering and Geophysical Flows

    (2010)
  • D. Stinton

    Flying Qualities and Flight Testing of the Aeroplane

    (1996)
  • D.T. Ward et al.

    Introduction to Flight Test Engineering

    (2001)
  • G. Gratton

    Initial Airworthiness – Determining the Acceptability of New Airborne Systems

    (2015)
  • A.W. Burner et al.

    Aeroelastic deformation: adaptation of wind tunnel measurement concepts to full-scale vehicle flight testing

    (2005)
  • F. Boden et al.

    IPCT wing, deformation measurements on a large transport aircraft

  • L.E. Wallace

    Airborne Trailblazer: two decades with NASA Langley's 737 flying laboratory

    (1994)
  • M. Wusk, ARIES: NASA Langley's Airborne Research Facility, AIAA paper 2002-5822, AIAA's Aircraft Technology,...
  • S.V. Baumgartner et al.

    Usability of LTE for transmitting radar data from DLR's research aircraft DO 228–212

  • J. Maestrati et al.

    Laminar flow for business jets: Falcon 900 HLFC demonstrator

  • J. Reneaux

    Overview on drag reduction technologies for civil transport aircraft

  • C.C. Lee et al.

    Flight-measured streamwise disturbance instabilities in laminar flow

  • A. Patzold et al.

    Sensors and actuators for laminar flow flight experiments

  • A.K. Cooke

    A simulation model of the NFLC Jetstream 31

    (2006)
  • Cited by (0)

    View full text