Dispelling the Myth of the MV-22
By Lieutenant Colonel Kevin Gross, U.S. Marine Corpsv
Proceedings, September 2004
In 2000, two tragic accidents grounded the Marine Corps’ newest and most innovative aircraft, miring it in controversy and casting doubt on its future. Rigorous testing since then, however, has resulted in measures that should prevent the unusual aerodynamic condition that caused the first accident from happening again. (Photo by Mike Jones, U.S. Navy)
The
MV-22 Osprey is a very capable medium-lift military transport aircraft the
Marine Corps has needed for a long time. It is twice as fast, can carry three times as much, and goes six times farther than the
CH-46E, the aircraft it is replacing. It is no reach to say that if the MV-22 continues its current run of success in testing and is fielded as planned, it will change everything about how maneuver warfare is conducted.
The Osprey, however, also is an airplane with an image problem, primarily resulting from two highly publicized mishaps that killed 23 Marines four years ago. The investigation into these accidents resulted in the discovery and subsequent resolution of an aerodynamic condition affecting all rotorcraft but unduly linked with the MV-22: vortex ring state (VRS).
On the evening of 8 April 2000, a flight of four MV-22s was conducting a night assault mission to a small airfield in Marana, Arizona, when the second airplane (or “Dash 2”) rolled nearly inverted on short final and crashed, killing all on board. During the subsequent investigation, it was discovered that the lead aircraft was almost 2,000 feet higher than planned at the initial point (the location where the conversion from airplane mode to VTOL [vertical take-off and landing] mode for landing begins). The lead aircraft entered a steep approach profile with a high rate of descent while it rapidly decreased speed for landing. During the rapid deceleration, Dash 2 no longer could remain in trail as briefed but came abeam of the lead’s right side. To return to the trail position, Dash 2 flew slower and with a higher rate of descent than his lead. At approximately 300 feet above ground level, with a more than 2,000-feet-per-minute rate of descent and with less than 30 knots forward airspeed, the mishap aircraft started a right roll that could not be corrected by the pilot.
The mishap investigation, having ruled out all other possibilities, soon focused on the extremely high rate of descent at low altitude as the primary cause of the accident. It was concluded that during the descent, the aircraft entered an aerodynamic condition called vortex ring state.
Vortex Ring State
Our search for the VRS boundary started with a thorough review of analytical and wind tunnel research and slow-speed, high-rate-of-descent flight testing. We soon discovered that the body of known actual flight-test data for VRS in other rotorcraft was very small. There were only two other rotorcraft flight research projects known to us at the time we began our initial flight testing, one by NASA Langley in 1964 and one more recently by the ONERA organization in France with a Dauphin helicopter. There was a larger amount of theoretical data available, however, from the private and academic sectors of flight-test research. The principal work we reviewed was from a paper published in 1965 by Kyuichiro Washizu and Akira Azuma of the University of Tokyo.
[1]
VRS is an aerodynamic condition in which the tangential airspeed at the rotor is small (associated with low forward airspeed) and the airspeed perpendicular to the rotor is high (associated with powered rate of descent). VRS typically becomes a concern below 40 knots forward airspeed at high rates of descent. To reach this condition, power must be applied during the steep descent. Some might think that during VRS the rotor stalls, but that is not the case. VRS is a reingestion state, not a stalled state. Rotor lift creates a down flow of air, called induced velocity (Vi), and the up flow created by the nearly vertical rate of descent is called vertical velocity (Vv). When the induced velocity equals the vertical velocity, VRS may occur, causing a reduction in rotor lift or increased sink rate. VRS can occur as a rotorcraft settles down through its own vortex field at slow forward airspeeds.
This condition is not peculiar to the tiltrotor; in fact, every rotary winged aircraft is susceptible to it. A helicopter pilot flying a single rotor system exits VRS by lowering his collective (reducing power with his left hand) and pushing forward the cyclic (tilting his rotor disk forward to accelerate) with his right hand. A tiltrotor pilot has another, more-effective, option: he can move the nacelles into clean air.
Test Objectives
In developing the plan to understand VRS, Naval Air Systems Command directed the MV-22 Integrated Test Team at Patuxent River, Maryland, to conduct flight-test exploration of the Osprey tiltrotor’s high-rate-of-descent/low-airspeed boundary. The initial test plan was developed to conduct partial- and minimum-power descents to investigate the low airspeed descending flight characteristics and determine the effects of the thrust-control lever and cockpit-control inputs on handling qualities in this flight regime.
The objectives of the test effort were to define the boundary of VRS, derive a fleet operational envelope, define the recovery technique from VRS, determine applicability for warning systems, and document the condition in pilot training ground school simulations and the NATOPS (Naval Air Training and Operating Procedures Standardization) flight manual for the tiltrotor. The first two objectives initially were approached as two phases, to validate the current fleet low-airspeed rate-of-descent limit and document the VRS boundary, and to evaluate the capability to expand the current fleet limit. As the testing progressed, however, we soon realized that breaking the test effort into two phases was not necessary, and we added multiaxis control inputs and dynamic maneuvers to our test effort.
For the third objective, we knew that recovery from VRS, like the poststall departure recovery of a fixed-wing aircraft, requires proper and timely procedures. For the Osprey, the most powerful flight control is the pilot’s ability to change nacelle angle (thrust vector) at up to 8° per second through a thumb switch on the thrust-control lever. This ability to change nacelle angle (and hence inflow angle at the rotor disk) would lead to an effective and immediate recovery tool.
For the fourth objective, several crew-alerting methods were discussed to determine the feasibility of active VRS avoidance. A few of the mechanical and tactile methods considered were a mechanical stick shaker, seat shaker, or rudder pedal shaker, but these were quickly rejected because of complexity, weight, and systems integration concerns. Instead, a visual and aural warning system was developed to alert the aircrew when the aircraft exceeded the NATOPS flight manual’s rate-of-descent limit.
For the fifth objective, data obtained from flight testing were used to develop pilot training courseware, update the simulation model to replicate VRS if deemed necessary, and update the NATOPS flight manual with a comprehensive narrative description of VRS to include pilot cueing to aid avoidance and emergency procedures for recovery should VRS occur.
Our first six-month period of flight testing started within two and a half months of the Marana mishap, and only weeks after preliminary mishap board results had been reported. On 11 December 2000, a second MV-22 mishap not related to vortex ring state occurred, taking the lives of all on board and resulting in the grounding of the Osprey fleet. Several investigative bodies evaluated the MV-22, the most significant of which was the blue ribbon panel.[
2] The cause of the second accident was found to be the combination of a software anomaly and a hydraulic line failure. The grounding allowed the test team to analyze flight data collected to date, refine the test plan, and develop instrumentation, including a new low-airspeed measurement system. After evaluating several low-airspeed sensors, the test team selected the R. M. Young Model 81000 Ultrasonic Anemometer to provide the desired airspeed to as low as ten knots forward velocity. When integrated into the aircraft, this sensor provided the low-airspeed confidence required to complete our flight test. In addition to the normal aircraft display of flight information on the pilot primary flight display, we added two flip-down liquid crystal displays that indicated sideslip in one-degree increments and calibrated airspeed with one-knot accuracy from the ultrasonic anemometer. The addition of the flip-down digital displays, within the pilot’s central field of view, greatly reduced workload and increased test-point efficiency.
Flight Test
Our test-plan strategy was simple: approach the unknown boundary from higher air speed and down from above with increasing sink rate increments. For the purpose of pilot build up and standardization, only one pilot, chief test pilot Tom Macdonald of Boeing, flew each test point with a small group of copilots assisting. The aircraft was configured with nacelles full aft at 95°, flaps auto, and landing gear down. Each descent track was completed at the same target airspeed, with a package of data points at each 500-feet-per-minute descent increment, where we checked stability, handling qualities, ride quality, aural signatures, and descent arrest and recovery effectiveness.
Our testing was conducted at altitude with a target test band of 8,000- to 7,000-feet altitude to allow for recovery well before the hard deck of 3,500 feet above ground level. Each aircraft input at the target airspeed and rate of descent required one dedicated descent. With the addition of the Young Model 81000 low-airspeed system, we gained flight efficiency by omitting the sawtooth climbs to determine true winds in the test band for the next descent profile. By the end of our VRS test effort, we had flown 62 flights for 104 flight hours and exceeded 5,600-feet-per-minute rate of descent and flew as slow as ten knots calibrated forward airspeed.
During our testing, we experienced 12 roll-off events, 8 to the right and 4 to the left. The direction of roll off was not predictable from the cockpit. In fact, the cockpit characteristics approaching VRS were not as well defined as in single-rotor helicopters. We noticed a slight increase in vibration, rotor noise, and flight control loosening that would not in every instance foretell of an impending roll off. Each roll off, however, was characterized by a sudden sharp reduction of lift on one of the two proprotors, resulting in an uncommanded roll in that direction. We also noted that roll offs required nearly steady-state conditions to trigger them. Any dynamic maneuvering tended to delay or prevent a roll off from occurring. On many occasions, we entered the VRS boundary during dynamic maneuvers and then exited the boundary without encountering a roll off.
Improvements to the Osprey’s pilot display include an expanded rate-of-descent scale and a red line added behind the vertical sink scale. Both features should help prevent future encounters with vortex ring state. (Photo couresty of author)
Pilot recovery procedures from a VRS roll off are easy, immediate, and effective. The pilot fixes the thrust-control lever and simultaneously pushes the nacelle-control thumb wheel forward for two seconds. This two-second beep forward moves the nacelles 12° to 15° lower, which immediately takes the rotors out of the VRS condition as the aircraft accelerates rapidly. If required, the pilot then uses lateral stick to level the wings and then adds power to stop the rate of descent.
Avionics Improvements
Now that we knew where the VRS boundary was located, how the aircraft responded during VRS, and the proper recovery procedures and techniques, our focus turned toward avoidance of VRS. Boeing and Naval Air Systems Command avionics and crew systems engineers developed a simple yet eloquent method of keeping pilots away from VRS. We made two changes to our avionics displays that increased the pilot’s situational awareness during low-speed, high-rate-of-descent flight. First, we expanded the rate-of-descent scale from 1,000 feet per minute to 2,000 feet per minute in 200-feet-per-minute increments. Second, we added a red line behind the vertical sink scale at the rate-of-descent limit along with a visual and aural “sink” warning when the airspeed and rate of descent exceed the limits. The flight display used by the pilot as the primary performance instrument in the Osprey is shown above. The scale on the right side is the vertical speed indicator. The indicator’s arrowhead is pointing to an 800-feet-per-minute rate of descent. The airspeed box is on the left of the display and indicates 39 knots. With this combination of airspeed and rate of descent, the pilot has exceeded the existing rate-of-descent limit and now hears “sink rate, sink rate” in his headset and sees the red “sink” warning in the display near the top and to the right of center. It is important to note that this warning system is not a predictor of VRS, but a rate-of-descent limit to keep the pilot away from VRS.
Where We Go from Here
Our ultimate goal for this flight-test effort was to understand fully the aerodynamic effects of vortex ring state on the tiltrotor, to define the recovery procedures should the pilot encounter VRS, and, most important, to develop warning signals to keep pilots away from this condition. Pilot awareness of VRS and avoidance with rate-of-descent limits are the only tools available to prevent a high-rate-of-descent mishap from taking more lives. Our current rate-of-descent limit is 800 feet per minute below 40 knots calibrated airspeed (KCAS), increasing linearly to 1,600 feet per minute at 80 KCAS. Presently, Naval Air Systems Command is evaluating the potential to expand the rate-of-descent envelope in the 30-to-50-knot airspeed range to provide the user communities more capability during flight tests and approaches to landing.
So what have we found in our 14-months of low-airspeed, high-rate-of-descent testing in search of VRS?
- Above 40 KCAS, VRS will not occur, regardless of sink-rate magnitude.
- The lower the disk loading (the ratio between an aircraft’s weight and rotor size), the smaller the sink rate where VRS might occur. Conversely, in the case of the MV-22 with higher disk loading, VRS may occur at a much larger sink rate.
- VRS requires a nearly steady-state condition. Any maneuvering tends to delay or prevent a roll off.
- As both sink rate increases and airspeed decreases, periodic rotor thrust fluctuations increase.
- As the VRS boundary is approached, handling qualities degrade because of unsteady flows at the rotor(s).
- Entry into fully developed VRS may be characterized by a sudden, sharp reduction of net thrust at the rotor.
- Recovery from VRS is immediate and effective using two seconds of forward nacelle tilt.
- There is a large margin of safety between rate of descent limit and VRS boundary below 40 KCAS.
With our test effort behind us, the Integrated Test Team at Patuxent River is confident we fully understand the location of the VRS boundary for the tiltrotor, the aircraft roll-off characteristics during steady maneuvers within the boundary, and the immediate and effective recovery procedures. We have developed avionics warnings to aid pilots in avoiding high rates of descent at low airspeed. The fleet now has a better understanding of the capabilities of the MV-22 and will be confident to fly in harm’s way knowing vortex ring state never will be encountered again.
- Kyuichiro Washizu and Akira Azuma, “Experiments on a Model Helicopter Rotor Operating in the Vortex Ring State,” University of Tokyo, presented at the American Institute of Aeronautics and Astronautics Symposium on Structural Dynamics and Aeroelasticity, Boston, MA, 30 August-1 September 1965.[back to article]
- “Report of the Panel to Review the V-22 Program,” Memorandum for the Secretary of Defense, John R. Dailey, Chairman, 30 April 2001. [back to article]
Lieutenant Colonel Gross was the government flight test director for the MV-22 program from August 2002 to August 2004 and participated in several test flights. He currently is assigned to the V-22 Joint Program Office at Patuxent River, Maryland. He would like to acknowledge Tom Macdonald’s and MV-22 lead government engineer Ray Dagenhart’s contributions to this article.