Mentor Matters: Simulator versus airplane training

Some events—specific systems failures in particular—must be trained in a simulator simply because they can’t safely be reproduced in an aircraft. An engine fire, for example, requires branching actions of engine shutdown and fire extinguishing agent deployment, which leads to other system changes as electrical, hydraulic, and/or bleed air sources are lost. A pilot training in a simulator will see the changes occur as they actually would with a fire present, while in an aircraft the procedures can only be talked through—with a good bit of imagination required.

Simulator training also is preferred for procedures that feature either a very steep or very shallow learning curve. Many of the former are unsafe to perform in an aircraft, doubling the importance of experiencing them in the simulator.

For example, most light jets feature a back-up braking system that uses pneumatic (trapped gas) pressure to directly actuate the wheel brakes, bypassing the normal hydraulic system as well as the computer-controlled antiskid process. Because no antiskid protection exists, the first time many pilots practice emergency braking, they apply too much force to the control handle. It’s common for the initial training attempt (and many attempts in real life) to result in blown tires and a loss of directional control on the runway. However, the second attempt is often completely acceptable, as pilots can quickly recalibrate the force they use.

This is a textbook example of a steep learning curve: Skill rapidly increases with a small amount of practice (note that the colloquial use of the term “steep learning curve” actually means the opposite of the technical meaning). Only a few practice applications will dramatically lower the risk of bad outcome. As this procedure is unsafe to attempt in-aircraft, a pilot who has received only in-aircraft training misses out on the rapid skill increase potential of just one or two practice events in a simulator.

A maneuver with a shallow learning curve is one in which much repeated practice is needed for skill to increase to the necessary level; sometimes these maneuvers also are best practiced in a simulator. Consider the single-engine go-around—the most procedurally complex maneuver performed during an initial or recurrent checkride. Combining precise (and often strenuous) rudder control with demanding lateral and vertical aircraft control, flight management system changes, and auto-flight mode programming, the single-engine go-around can be performed perfectly only after relentless drilling.

While a creative in-aircraft instructor can simulate single-engine go-arounds at altitude, allowing for more repetitions in a given space of time, the maneuver isn’t quite the same as when it’s performed out of an actual approach. In a sim, however, the instructor can reposition the aircraft to just above minimums time after time, allowing for a massive number of practice maneuvers in the amount of time only one could be practiced in flight. Particularly for the first-time jet pilot, the repeatability of tasks in the sim is of high value here.

As good as the visual and graphic systems on simulators have become, tasks that require visual maneuvering often can be better trained in real flight. The prosaic visual approach, for example, is one that can be surprisingly difficult for a new jet pilot to execute well. Approaching an airport from an unusual angle, especially to a runway without an instrument approach, requires a high comfort level when interpreting the sight picture to derive cues for distance, alignment, and relative height. Yet visual approaches in a sim tend to be very scripted: takeoff, make closed traffic to a five-mile final, and land—not at all the way that 99 percent of real-life visual approaches are flown.

Because of similar factors, circling procedures also are of higher training value when performed in a real aircraft. For reasons of FAA certification of the simulator and training course, three domestic circle-to-land procedures—at Memphis International Airport, John F. Kennedy International Airport, and Wichita Dwight D. Eisenhower National Airport—are used for nearly all initial and recurrent simulator training. Not surprisingly, after teaching the same circling approach hundreds of times, sim instructors have developed tips on executing the procedure (“turn to this heading when acquiring Runway 27, turn base to the landing runway when over this building…”) that, while helping teach the student to fly this approach nicely, are in fact counterproductive to the much more important task of teaching the student to fly any circling approach safely.

In aircraft, the student will be presented with circling procedures with any possible combination of acute or oblique angles between the runways, and MDAs from as low as 300 feet agl to as high as 2,000 feet above the airport. Each requires a very different set of steps to execute safely, ensuring the student truly learns how to assess the relationship in space between the aircraft and the landing runway, not just how to execute the same five steps at the same five trigger points every time.

Landings round out the visually based procedures that are more effectively trained in aircraft. All runways in the sim end up looking the same, and depth perception just isn’t quite up to a real-life level. As a result, it’s relatively simple in the sim to just pull the power to idle at 50 feet agl and wait for the airplane to contact the runway. Combined with the ubiquitous steady 10-knot crosswind, the landing becomes as formulaic as the visual and circling approaches.

Training in the airplane, the pilot is confronted with different runway widths and slopes, obstacles such as trees or buildings obscuring the runway along short final, and distractions of other aircraft moving on the surface and competing for visual attention. Skill progression can occur in ways not possible in a simulator.

A last set of maneuvers lend themselves to periodic training in the aircraft: those air-work maneuvers, such as stalls and steep turns, where the kinesthetic feeling in the aircraft will be dramatically different than in a simulator. Because of the limits of a simulator’s motion system, sustained change in any axis can’t be replicated, nor can continuous high or low G loadings. Being aware of hearing and feeling the changes in sensory cues as a stall is approached is as important, if not more important, than executing the recovery properly. Pilots who have availed themselves of in-aircraft training opportunities (especially specialized upset prevention and recovery training) can be considered more “innoculated” against an inflight upset, from having experienced the warning cues that a stall or other event is imminent.

Neil Singer is a master CFI with 9,500 hours.