As far as is possible, any test of this nature should flow in a logical sequence so that the candidate can apply realistic management which should include liaison with cabin crew and ATC. Base exercises are not a good way to evaluate this, as the candidate and examiner all too readily fall into simulator mode and take short cuts with some of the management aspects. Headsets should be worn. The point is, to do a proper evaluation of what is expected of an ATPL, the examiner has to prevent the candidate from making any assumptions simply because it is a simulator.
I would run it over a short route - such as Launceston to Melbourne or Canberra to Melbourne - and brief that it is a winter operation, icing conditions for the entire flight except when in level flight and clear conditions. The departure should be low visibility and the rejected takeoff could be dealt with right at the beginning, to get it out of the way. Although it would have been briefed as being possible at any time. The RTO need not be for something that would 'ground' the aircraft - as that introduces an unrealistic scenario for another immediate departure. Something like a false configuration warning would do, with emphasis on clear communication to cabin and ATC. By doing it first, if the candidate stuffs it up, he fails right there with no further expenditure of time or money.
After the low visibility takeoff, SID and climb to clear conditions, normal operation. Then do the steep turns and a stall in landing configuration with bank angle applied. Climb to cruise and do a high altitude stall.
Somewhere en route a TCAS event. Once all that is dealt with, the sim examiner could dial in enough of a tailwind to hurry things along, but reduce it to a sensible speed for descent.
Continue to destination via a STAR with engine fire or damage on descent with enough time to secure it - hold if necessary (candidate's decision). Raw data manually flown approach and miss to divert to somewhere nearby like Avalon for another approach using another navigation aid without radar vectors. Use of autopilot allowed for remainder, subject to aircraft type limitations.
Land, reset to night, with weather to preclude an immediate return but something more reasonable at a nearby alternate (e.g. Essendon). Failure at V1, back to Essendon with reasonable radar vectors to allow time to manage the failure to a logical conclusion and a third approach to a more limiting runway.
To give the candidate value for money, ask a couple of sensible questions from the form so that you can sign off his instrument rating for another 12 months at the same time.
Job done easily in two hours.
Last edited by Mach E Avelli; 6th May 2015 at 01:49.