Mike Harris, DERA Malvern, U.K.

At the second WakeNet Workshop in Munich (October 1999), several speakers discussed the best method of analysing lidar data from field tests, in order to best characterise the vortex hazard. Frank Holzapfel made a very welcome comparison between lidar and simulation data, and suggested a common definition to aid comparison of different work. There were further valuable contributions from Pilar Vela (BAe) on re-analysis of DERA lidar data from Toulouse, and from Friedrich Koepp on the analysis methods used at DLR.

A common (or "unified") method of vortex characterisation is an essential part of future studies (including work planned under the C-Wake programme where inter-comparison is required), and for any future aircraft certification scheme. In this Job Report, I present a brief appraisal of some possible approaches; this is intended to promote discussion and I invite comment and criticism via the website.

For simplicity, a single-parameter characterisation is attractive (does everyone agree?); this parameter should fulfil two basic requirements:

1. It must directly reflect the level of vortex hazard
2. It should be defined so as to minimise the level of experimental uncertainty

There seems general agreement (any comments?) that the total circulation represents a good measure of the hazard. Alternative parameters, e.g. peak velocity, are generally considered less appropriate and fail to fulfil condition 1 above. Two suggested methods for evaluating the circulation are as follows:

1. (As proposed by Frank Holzapfel in Munich) - Evaluate the circulation averaged over a range of 5-15 metres from the vortex core, i.e. <G (R=5-15m)> where the brackets denote the mean value, and G (R) = 2p R´ V(R). To be valid, this region should exhibit a 1/R velocity dependence giving circulation independent of R; the cut-off at 15 m is intended to minimise errors that occur when the velocity V becomes too small. Possible difficulties include the elimination of the contribution from the other vortex, and the fact that for large aircraft (A340, B747), the velocity is not proportional to 1/R in the region between 5-10 m from the core – that is, these regions still exhibit some level of vorticity (see e.g., J S Greenwood & J M Vaughan, Measurements of aircraft wake vortices at Heathrow by laser Doppler velocimetry, ATC Quarterly, Vol. 6 179-203 (1998)).
2. (As used by DERA Malvern in their analysis of data from Airbus/Toulouse trial) - Evaluate the down-draught (V_m) at the mid-point between the two vortex cores, averaged over an appropriate distance (usually several metres). The circulation is then given by G = p V_m S₀, where 2S₀ is the separation of the two cores – about 40-50 m for large aircraft. The method has the attraction of simplicity and robustness; it does however limit flexibility for lidar positioning. Note that the summation of the contributions from the two vortices doubles the measured velocity, thus reducing the uncertainty. V_m lies typically in the range 4-8 m/s; the measurement uncertainty is of order ~0.2 m/s, giving a likely uncertainty in G of order 3-5% in ideal conditions. This could be reduced by some more sophisticated processing of the raw data.

The similarities between the two techniques must be stressed. Both involve processing of velocities of very similar magnitude, and some averaging over a range of vortex radii. Both require an absolute length scaling (best derived from measurements of core separation) for extraction of quantitative circulation data. The experimental uncertainties are likely to be similar, and both methods are equally susceptible to the effects of atmospheric turbulence. The best choice may depend to some extent on the available measurement geometry – i.e. whether the vortices are viewed from beneath or from the side. Further consideration is needed on the possible uncertainties resulting from crosswinds etc.