Not a member? Join today. Already a member? Please login for an enhanced experience. Login Now
Menu

Engine-Out Booby Traps for Light Twin PilotsEngine-Out Booby Traps for Light Twin Pilots

Flight Instructors' Safety Report

Engine-Out Booby Traps for Light Twin Pilots

By Melville R. Byington, Jr.

Professor of Aeronautical Science, Embry-Riddle Aeronautical University (From the AOPA Air Safety Foundation, April 1993.)

BACKGROUND

Previous research into the engine out flying characteristics of light twins has identified the critical effect of pilot technique, particularly with respect to the proper bank angle. The "Five Degree Forever Syndrome", defined in Reference 1, continues to die slowly. Most multiengine pilots have accepted the reality that, for engine-out operations, the once popular "five degree bank angle into the operating engine" PROVIDES NEITHER BEST CLIMB PERFORMANCE NOR BEST DIRECTIONAL CONTROL.

Best performance demands minimum drag, corresponding to zero sideslip flight, which corresponds to bank angles much less than five degrees. "Best" directional control, presumably meaning least rudder pressure, requires bank angles greater than five degrees, by which climb performance suffers severely. The "magic" five degree value simply is a holdover from FAR 23.149, which permits the manufacturer's test pilots a bank angle of up to five degrees for the purpose of determining the nominal Vmc. The manufacturer desires a low nominal Vmc and employs the full 5 degree bank for this test.

The manufacturer also conducts engine out rate of climb tests under conditions designed to cast the results in the most favorable competitive light. New airplanes and engines are flown by factory test pilots under optimum conditions, INCLUDING ZERO SIDESLIP. MINIMUM DRAG CONDITIONS OF FLIGHT, These measured data form the basis for engine-out climb performance estimates found in the pilot operating handbook (POH).

THE PROBLEM

Typical pilots fly airplanes several years and thousands of hours old. The airframes and engines no longer are new, but the POH data is of original vintage. Moreover, the POH recommended engine-out climb bank angle is likely to guarantee performance inferior to the optimum available. Even though precluded by physical law, the POH may predict a positive climb rate for prevailing weight and ambient conditions. The conscientious pilot who obtains and naively accepts this positive POH climb estimate is in an unenviable predicament. Under such circumstances, attempts to achieve a positive climb rate are futile. Worse yet, unless correctly recognized and managed, loss of flying speed and a fatal stall/spin crash may follow.

Figure 1 depicts typical light twin engine-out characteristics under nominal conditions of equipment age and condition, weight, and density altitude (DA). See Reference 1 for details of three models. The strong influence of bank angle is evident in this "rooftop" curve, with peak performance corresponding to zero sideslip (ZS) flight at point B. Both zero bank (A) or five degree bank (C) degrade climb performance from the optimum (ZS) condition. Bank angles beyond 5 degrees (D) may preclude climbing even under "nominal" conditions.

Tests of four light twin models, two for "critical" and two for non-critical engine flight, showed that the average optimum bank angle at point B corresponded to 2.1 degrees. The average slope of the curve between B and D was thirty feet per minute loss per degree of "overbank."

Figure 1 (point E) also depicts the typical POH climb prediction and recommended bank angle for engine-out climb. The POH predicted climb performance (E) is unachievable at five degrees bank and usually even under the most favorable conditions (B, at zero sideslip). In addition to the unwelcome drag addition from excessive bank angle, Reference 2 warns against the threat of unporting the fuel tank feeding the operable engine.

Rate of climb (ROC) measurements were conducted for four models and seven engine-airframe combinations. Table 1 summarizes results. Actual climb rates were compared to POH predicted rates for existing weight and DA, both under optimum (zero sideslip) and under POH recommended conditions (usually at the 5 degree bank angle). Differences between POH ROC estimates and measured performance are tabulated. Comparisons between the POH estimates and optimum conditions correspond to Figure 1, point E minus point B ROC values. Likewise, comparisons to ROC when flown as recommended corresponds to points E minus C. For example, consider case #4. The best available RCC at ZS was 138 FPM inferior to corresponding POH forecasts. When flown as recommended at 5 degrees bank, RCC deteriorated an additional 85 FPM and became 223 FPM inferior to POH predictions.

In no test was POH performance achieved when flown as recommended by the POH. In one case (#I), involving relatively new equipment, the POH figure was exceeded marginally when zero sideslip was used. The identical airplane was retested as shown in cases 2 and 3. The original performance was not reproducible, and the decrements appear related to airframe and engine time.

While Figure 1 depicts "nominal" conditions of equipment, weight, and (DA), it is instructive to consider conditions more and less favorable than nominal. Figure 2 shows more favorable conditions resulting from some combination of lower weight, lower DA, or better equipment condition. This causes the curve to move vertically upward across the full scale. Climb performance is improved at all bank angles, and this may mask sloppy airmanship. Despite greater margins of error, compared to Figure 1, it is important to note no significant change occurs to:

  1. optimum bank angle (still approximately 2 degrees).
  2. slope of the curve (approximately 30 FPM per degree).
  3. difference between POH forecast (point E) and achievable ROC, for any bank angle.

Similarly, Figure 3 depicts less favorable conditions resulting from some combination of higher weight, higher DA, or poorer equipment conclition. In this challenging situation, positive climb is possible only near zero sideslip conditions despite POH indications to the contrary. Only precise airmanship permit a positive climb rate.

Table I (Measured vs POH Rate of Climb)
# MODEL A/F hrs ENG hrs Optimum ROC Per ROC RMKS
    (E-B) (E-C)    
1 Cessna 303 2130 199(L/R) +13 -78  
2 Cessna 303 6400 1004(L) -77 -168  
3 Cessna 303 6400 1005(R) -47 -138  
4 Beech A65 16937 1684(R) -138 -223  
5 Beech A65 11534 162(R) -28 -83  
6 Beech 58 3769 1450(R) -150 -150 Note 1
7 Piper PA 44 Unknown Unk(L/R) -88 -142 Note 2
AVERAGE     -74 -140  

Note 1: No bank angle specified in POH. Best case assumed.

Note 2:'Tbree to five degrees" recommended 4 degrees assumed.

LOSS OF CONTROL HAZARDS

Reference 1 reported an analysis of light twin accidents following engine failure in the initial climb after takeoff. It showed that 57% of these accidents and 75% of fatal/serious injuries resulted from situations in which control was lost. Low airspeed results from vain attempts to climb or hold altitude under conditions which render this impossible. Loss of control typically results from low airspeed combined with too much or too little bank angle, i.e., either more or less than that corresponding to zero sideslip.

Analysis was conducted of light twin engine-out stall/spin accidents in the initial climbout after takeoff. A 28-month period (1 /84 through 5/86) was studied. Analysis revealed:

  1. at least 7.7 accidents and 24 fatalities per year on average, and
  2. a 95% fatality rate for occupants in such crashes.

Reference 2 reports data covering engine-out fatal stall /spin accidents over a six year span (1983-89). Fatal stall/spin accidents comprised 10% of all multiengine fatal accidents. Half of these accidents occurred during training situations.

Causes of the frequent and lethal engine-out stall /spin accidents apparently deserve better understanding by light twin pilots. Low airspeed, combined with bank angles greater than those for zero sideslip (approximately two degrees), "sets the trap" for the unwary. 71w reason is illustrated by smoke tunnel photographs in Figure 4. Flow conditions are depicted by the smoke wisps and the yawstrings on the nose.

Figure 4a depicts the airplane in zero slip flight with well behaved airflow over both wings. Figure 4b depicts the airplane in a (right) bank greater than optimum for climb with the left engine failed and propeller feathered. This case requires somewhat less rudder deflection. The airflow on the lowered (right) inner wing is no problem, since it is not blanked by the fuselage and benefits from the propeller's high energy, slipstream-induced airflow. Conversely, the raised (left) inner wing suffers from disturbed, low energy airflow. This arises from the combination of the nacelle and propeller disturbance plus the fuselage's blanking effect. Consequently, the raised wing will stall at a substantially higher airspeed than would occur in zero sideslip flight.

The fuselage's blanking effect also will distort airspeed indicator accuracy, giving a false high or low reading. The amount and type of error depends on system characteristics and which engine is operating.

Reference 2 reports engine-out stall speeds 9-20 knots IAS greater than power-on stall speed for the Baron B55. During the author's flight tests of the Queenair A65, engine-out stall buffet was encountered 25 knots above the published Vs under severe sideslip conditions at 9.5 degrees bank.

With five degrees or greater bank, the pilot must hold a large amount of (right) wheel to maintain lateral control. Aileron deflections add to drag and further increase the left wing's angle of attack, while reducing that of the right wing. The trap is ready to snap!

Reference 2 asserts that "a single-engine stall in most (multi-engine) airplanes will result in loss of control." Indicated airspeed may read well above Vs when the trap snaps shut. The airplane begins a sudden, uncommanded left roll. The roll motion, coupled with any remaining "corrective" aileron deflection, further increases the left wing's angle of attack, deepens the stall, and produces a spin. Readers can probably recall headlines of fatal crashes following this scenario.

Knowledgeable pilots avoid such traps by understanding the relationships between engine-out climb performance and control. Optimum performance and minimum airspeed system errors both correspond to zero sideslip. Directional control should be quite adequate at zero sideslip, unless the airspeed decays below recommended values. Low speed rudder effectiveness can be improved at increased bank angles, but only at substantial performance penalties.

Figure 5 depicts the tradeoffs between climb performance, rudder effectiveness, and directional control. Condition 3 combines optimum performance with adequate directional control. Other bank angles offer no competition. Figure 6 depicts the tradeoffs between climb performance and insurance against the stall/spin hazard. Again, condition 3 combines the best performance achievable with greatest stall margin. Bank angles beyond ZS only aggravate this grave hazard.

CONCLUSIONS

Pilot Operating Handbooks commonly misinform pilots regarding available engine-out climb performance and how to obtain it. Perhaps the POH should contain "warning labels" similar to those on tobacco and alcohol products! A naive pilot, convinced by POH calculations that he can climb, likely will attempt to do so. If the pilot follows POH recommended bank angles greater than those corresponding to zero sideslip, the performance deficit increases. Contrary to the pilot's expectation, combinations of equipment condition and "overbank" guarantee that even the best climb rate is negative. See Figures 1 and 3.

Such attempts to achieve the unachievable climb will result in airspeed decay until control is lost. If the bank angle is greater than approximately 2 degrees, a stall/spin is the likely result. If bank angle is too low, directional control loss is probable when the vertical fin loses effectiveness before the dead engine wing stalls.

Figures 1 through 3 and Table I show the influence of bank angle on climb performance. Reference 1 provides amplification and shows the performance difference between ZS and 5 degrees is equivalent to weight changes in the 6-9% range.

Flight tests have shown that performance data in the POH are consistently optimistic, since service airplanes seldom achieve charted climb performance. Table I is a small but sobering sample of actual deficits. The data hint that airframe and engine time correlates with performance deterioration. Performance differences also were observed between equal time engines on the same airframe. With airplanes flown per POH recommendations, overall climb performance degradations as large as 181/6 gross weight equivalent were measured. Fortunately, the degradations can be reduced approximately 90 FPM simply by maintaining the ZS bank angle.

RECOMMENDATIONS

The pilot must realize his airplane and POH probably have characteristics similar to the samples cited. Once the nature of the threat is grasped, the remedy is apparent. The informed pilot rejects the "Five Degree Forever Syndrome" to climb at zero sideslip and achieve the best available performance. The prudent pilot will install a yaw string to determine ZS bank angle and ball deflection for his airplane. Unless equipped with counterrotating propellers fl-ds should be done for both engines. See Reference 1 for details. For airplanes used frequently for instruction, why not leave the yaw string installed?

Compare measured and POH-charted engine-out ROC for each engine. This can be done concurrently with the ZS determinations above. One must determine airplane weight at the time of each test run, so a careful calculation is necessary just before or after the test flight. ROC measurements should be made at zero sideslip in smooth air. Measure altitude change over a 3-4 minute period at maximum continuous power on the operating engine. Do not rely on the VSI. Conduct at least two runs per engine, preferably on reciprocal headings. Record mid-test temperature and pressure altitude (29.92" set).

Compare measured results to the POH predictions for corresponding weight, temperature, and pressure altitude. The differences represent "deltas," likely negative, which represent the particular airplane's climb performance signatures. Each engine will have its individual delta, which can be expected to increase as flight hours accumulate. The larger delta should be applied routinely to future POH climb tabulations as the minimum safety factor applicable to ROC estimates.

As discussed in Reference 3, informed pilots also will recognize the likelihood of encountering conditions rendering it impossible to hold altitude with a failed engine. So alerted, the pilot will sacrifice altitude as required to conserve airspeed and maintain control. The only viable option in such cases may be a controlled ditching. As one wag put it, "pick something soft and cheap and go for it." Compared to a 95% fatality rate in stall/spin crashes, controlled forced landings of light twins are virtually 100% survivable. The 28-month sample cited earlier showed 41 controlled ditchings, including several into trees, with a total of 106 occupants involved. There were no fatal injuries and only 9% serious injuries. The remaining 91% experienced minor or no injuries.

SUMMARY

Regardless of the number of engines involved, the pilot must remember the .maintain flying speed" maxim learned before his first solo. Placing blind faith in the engine-out POH climb performance data and recommended bank angles is hazardous to the health of pilots and passengers. Regular practice of engine-out (feathered or zero thrust) conditions provides invaluable training. In such drills, marginal thrust-to-weight conditions should be simulated by reducing power until performance conditions depicted in Figure 3, rather than Figure 2, are achieved. One such experiment will demolish the Five Degree Forever Syndrome!

REFERENCE 1

Byington M.R. Jr. (1989, April) "Principles to Bank On." AOPA Flight Instructors Safety Report. Reprints available in the pamphlet "Flying Light-Twin Engine Airplanes" for $1.75 from the AOPA Air Safety Foundation (1-800-638-3101).

REFERENCE 2

Kelly, W.P. (1989) "Multi-Engine Stalls." The Aviation Consumer, June 1, 1989.

REFERENCE 3

Kelly, W.P. (1992) "When Twins Turn Nasty, Part L" Aviation Safety, Oct. 1, 1992.

Editors Note: Embry-Riddle Aeronautical University is offering for sale a videotape titled "Optimized Engine-Out Procedures for Multi-Engine Airplanes" which parallels the information contained in Reference 1, price $25 plus $3 for postage and handling (Florida residents add 6% sales tax). Send check or money order to Embry-Riddle Aeronautical University, University Distribution Center, Daytona Beach FL 32114. Credit card orders (VISA or MasterCard) accepted, call (904) 226-6484.