Taxi testing starts with runs at low and then moderate speed. Once these are successfully completed, the testing moves to higher speeds. Careful planning is vital as the target speed approaches rotation speed. Once the airplane exceeds stall speed, it can generate enough lift to take off. As weight transfers from wheels to wings, things can change very quickly and the pilot must be well prepared. It’s particularly important to have a clearly planned sequence of actions at target speed and to be alert to ensure the airplane does not exceed the target speed for each run.
High-Speed Taxi Options
The amount of runway needed to accelerate to condition, test and then safely stop increases with the square of the speed. As the runs get faster, runway length becomes critical. There is a decision point in the test program when taxi test speed approaches the stall speed or rotation speed of the airplane.
There are three possible fast-taxi test plans, and all can be used successfully.
The choice of which plan is appropriate depends on the details of the test program. The right way to proceed depends on the speed and wing loading of the airplane, the amount of runway available, the control system and the type of engine powering the airplane.
The higher the liftoff speed, the more runway will be needed to accelerate to speed and then brake safely to a stop.
Jet airplanes typically take more runway both to accelerate to liftoff speed and to stop on the runway. Jet engines generate a significant amount of residual thrust, even at idle, so the brakes must not only absorb the momentum of the mass of the airplane but also fight the residual thrust. Airliners and transports typically have thrust reversers, which dissipate the residual thrust and can also produce significant retro-thrust to help slow the airplane. Fighters and other military airplanes rarely have reversers, so the residual thrust must be taken into account.
The type of control system is also significant. Unmanned aircraft are entirely controlled by the flight-control laws programmed into their computers. Special laws might be needed to control the airplane properly during transition from rolling to lifting off or prolonged running with the nosewheel off the ground.
Three test sequences are commonly used, as follows.
Liftoff & Hop in Ground Effect
In this approach the highest-speed “taxi” tests are actually low-altitude hop flights down the runway.
First, the speed is increased up to the speed for initial rotation. Once at rotation speed, the pilot will pull the throttle to keep the speed from increasing further and then apply up elevator to lift the nosewheel. The goal is to establish rotation speed and ensure the pilot has good control of the airplane in pitch during the rotation and can capture and hold pitch attitude after the airplane rotates. It is critical that this initial rotation attempt be done at a speed below stall speed to eliminate any chance of an inadvertent liftoff.
Once the pilot has good control stabilizing the pitch attitude, the next run will include the first liftoff. The sequence for this run is:
- Accelerate to rotation speed.
- Reduce throttle, rotate and stabilize the airplane in the takeoff attitude.
- Increase power, and without changing pitch attitude let the airplane accelerate until it lifts off.
- Reduce power as soon as the airplane lifts off, and hold attitude until it touches down.
- Decelerate and stop.
In subsequent runs, the pilot will establish level flight in ground effect (about one wingspan altitude) and evaluate the controllability of the airplane. Once these low straight hop flights are successfully completed, the airplane is ready for its first up-and-away flight.
This approach has several advantages. The controllability of the airplane can be established before committing to full flight and the pilot can become familiar with the control feel and sight picture for the initial takeoff. The primary disadvantage is the amount of runway needed to do the whole accelerate-hop-land-stop sequence.
Nosewheel Liftoff, No Flight
For some test programs the risks inherent in low hop flights over the runway may be higher than the risks of committing to an up-and-away flight after the first liftoff. It may also be that there is not enough runway available to execute the accelerate-hop-land-stop sequence of a runway-hop flight. This is common for faster, higher-wing-loading airplanes and almost exclusively true for any jet-powered airplane.
If this is the case, the airplane will not reach stall speed during even the fastest taxi-test runs. The fastest taxi speed is above rotation speed but below liftoff speed. The goal on the fastest taxi test is to establish rotation speed by using the elevators to rotate to a takeoff attitude, but not to lift off. The fastest runs allow the pilot to confirm that the elevators can rotate the airplane for takeoff and that it is possible to control the airplane in pitch during rotation to capture and hold the takeoff attitude.
During these tests, it is important for the pilot to set the throttle to prevent the airplane from accelerating beyond target speed before initiating rotation. Exceeding target speed can lead to an unintentional liftoff, at which point the pilot must make a very quick decision whether to pull the throttle and try to stop on the runway or commit to a premature first flight.
No Nosewheel Lift Until Ready to Fly
On some programs, the nose gear does not come off the ground during taxi testing. The first rotation is intended to be for takeoff for the first flight and the airplane does not exceed stall speed during taxi test. This approach is the least common for piloted airplanes but is quite common for autonomous unmanned aircraft (drones or UAVs).
Things are changing rapidly in the period between the start of rotation and liftoff. Weight is shifting from the wheels to the wings and the effect of the gear loads and wing lift on pitching moment are changing. Before rotation, the attitude of the airplane is set entirely by the landing gear interacting with the ground. After liftoff, pitch trim and control are entirely aerodynamic. During rotation, both are in play.
The idea behind delaying rotation until you are ready to fly is to minimize the need to deal with the transient effects during rotation. Automated flight-control systems have flight-control laws that mode-switch between ground and air.
Taking off with a rapid rotation can get the airplane through the rotation transient quickly enough so that it does not need to have special control laws just for the rotation phase. UAVs are more suited to this approach. Human pilots have to limit the pitch-up rate during rotation in order to stay oriented and capture the liftoff attitude. The flight-control computers on an autonomous vehicle can handle a much higher pitch-rotation rate, and therefore the vehicle can move very rapidly through the rotation transient and transition from “ground” to “air” control laws without special control laws to handle the transient during rotation.
Very important: If the airplane does not rotate as expected at the predicted rotation speed, abort the run and do not penetrate to higher speeds or try to force rotation.
Going faster sets up a possible situation where when the airspeed gets high enough the elevators will rotate the airplane at a speed well above stall speed. The airplane can suddenly rotate and jump into the air with a rapid nose-up pitch rate. If this happens the pilot may not be able to maintain control and push the nose down before stalling in a very nose-high attitude low to the ground. This “hop-off” phenomenon has caused several nasty accidents and in the ones I know of less than half of the pilots involved survived.
It is critical to proceed slowly, thoughtfully diagnosing why the airplane would not rotate and coming up with appropriate modifications to fix the issue. Make one change at a time. Avoid the temptation to go for the quick fix. Evaluate the effect of each change, not only on takeoff rotation, but on the stability and controllability of the airplane overall. I am aware of two accidents where the changes made to make it easier to rotate for takeoff also destabilized the airplane in pitch so much that it was not controllable in the air.
Quick fixes kill. Any time an airplane is not behaving as predicted is a time for lots of analysis and great caution, particularly on an initial test program where we are exploring the as-yet-unknown characteristics of a new machine.
I am a firm believer in high speed taxi testing especially in a tail dragger. Not so much for nose draggers. They are much safer in the sky, use a long runway in the right conditions and you can tell within seconds whether to go or not.
I was taught this way and have graduated probably a dozen pilots into tail draggers this way.
Slow taxing up and down the runway moving back and forth, left and right, gradually increasing the speed.
Up to a certain speed after which it would become uncontrolllable.
Graduation is running down the runway , lifting the tail of the ground , throttle back to idle and then letting the tail settle down.
Do a few circuits and remember,,,,peddle peddle peddle.
My friend Hal taught me this way after another guy didn’t and I ground looped my brand new airplane.
I have flown some of the mosst vicious tail draggers and never even come close to a ground loop.