You are on page 1of 13
IcE-CONTAMINATED- TAILPLANE STALL*~~- If ice builds up on your horizontal stabilizer, will you recognize the symptoms in time to avoid an accident? Securely strapped into t NASA's Tail (TIP) de Ha we set up for landing. The stabilize carefully crafted We si descent. We made the usual small, crisp heading and descent rate changes to stay on azimuth and glide- path. Something definitely was wrong with this airplane. The yoke tugging Usk stall ‘Malaga sa rocovery vshniaies SiSh ti isereh Comers helo ireratt operations. craft in pitch trim. Small nose-down pitch changes seemed to make the ‘wheel tend to jump forward. The Twin Otter seemed to be uncomfortably close to stalling. “Go around” directed our simulated approach controller We added thrust and the control column started to buf- fet wildly. The column moved forward ‘and the nose pitched down. Was this a full wing stall? Our gut reaction was to relax all back pressure on the yoke, ‘add maximum power and let the air- ‘raft recover on its own. ‘That reaction could have proved fatal. We had just encountered an impending ice-contaminated-tailplane stall (ICTS) and the recovery tech- niques are exactly opposite of those associated with a wing sta. Successfully recovery from ICTS requires holding or increasing pitch attitude, regardless of stick-pull force, reducing flap extension position and even possibly easing off on the power “If you use wing stall recovery tech niques to recover from a tailplane stall, it could kill you,” emphasized Richard Ranaudo, NASA Lewis Research Center's chief of aircraft ‘operations who was in the right seat ‘when we flew the maneuver. Ranaudo pointed out that the tailplane is an airfoil that is flying upside down—one that is designed to produce lift in a downward direction. Many of the tactile cues associated with tailplane stall are similar to the symptoms of a wing stall; however, mistaking one for the other can be deadly. The symptoms of tailplane stall are well docu- mented and ICTS has been the focus of several reports in B/CA. Until four ye ‘ago, though, no one in the United States had attempt- ‘ed to actually measure the degradation in stability and control performance caused by tailplane ice contamina- tion. In mid 1993, NASA and the FAA embarked upon a four-year ICTS study program, aimed at jeeeee virtually every performance parameter. (See “Quantifying the Effects of Tailplane Icing” sidebar.) In late October, with a Ranaudo as our guide, we flew a series of five ICTS demonstration maneu- vers. They culminated in the simulat- ed balked landing approach to which wwe referred in the above section. Dur- ing those maneuvers, we deliberately pushed the aircraft, step by cautious step, outside of the AFM limitations for flight in severe icing conditions in order to explore the symptoms of sta- bility and control degradation associat- ed with ICTS. In a nutshell, it was a real eye opener. At least 139 people have lost their lives in airplane crashes caused by ICTS, according to icing experts at NASA and the FAA. Indeed, one government official estimated the actual number at greater than 160 fatalities. Doubtless- ly, pilots of those aircraft encountered dramatic stability and control changes caused by ICTS. Accident investiga- tors, for example, determined that ele- vator forces exceeded 400 pounds of stick pull in one extreme case. Based on documented accidents, incidents or instances of degraded sta bility and control caused by tailplane icing, the FAA has issued ADs for six business or commuter aireraft, accord- ing to John Dow, an FAA aviation safety engineer. If you fly a Beechjet 400-series, Cessna T303 Crusader, de Havilland Dash 6 Twin Otter, Embraer EMB-110 Bandeirante, Jet THREE ICE SHAPES TESTED ON NASA's TWIN OTTER stream 81-series or Nihon Aeroplane YS11 aircraft, undoubtedly you're well aware of these caveats. This is just the edge of a storm of new icing ADs. As noted in our November issue (page 17), the FAA hhas issued 24 proposed icing ADs that would apply to 12,000-plus airplanes, with particular emphasis on aircraft fitted with pneumatic deice boots and ‘manually operated flight controls. ‘Those rules probably won't be issued until after the winter of 1998. More icing ADs are likely as the FAA reviews its icing certification criteria, particularly in view of real-world icing conditions, And with the weather guessers inundating the public with predictions of an extremely wet win- ter, spawned by El Nifio, the odds are in favor of your being exposed to the rmost-severe icing conditions in more than a decade. Most pilots encounter ICTS when they're close to the ground on final, in IMC icing conditions and just after extending flaps for landing. ICTS usu- ally takes them by surprise and, by then, there’s no margin for recovery. We flew NASA's ICTS demonstra, tion maneuvers under much more con- trolled conditions, resulting in a fat ‘margin of safety. Ranaudo thoroughly briefed each maneuver to be flown. He insisted upon a 5,000 foot agl “hard deck” during all tests. Because of the ice shapes on the tailplane, he prohib- ited extending the flaps below 5,000 feet, even for the final landing. He was ready to take control of the aircraft in case we encoun- tered difficulties. The five ICTS demonstra- tion maneuvers were designed to be flown in cumulative succession. We first explored the effects of elevator inputs, then pro- gressively added in flap, speed and power changes. ‘The balked landing maneu- ver was the last in the series. First, we performed a series of push-hold, pull- hold pitch input maneuvers known as elevator doublets to check the aircraft's sta- ity and control character- isties. We pulled the yoke ~ back sharply about one, inch, held it for three sec- onds and then pushed for- reprint compliments of FightSafety Canada - de Havilland Learning Centre a ward two inches and held it for | "®*? five seconds. With the flaps up in typical ‘eruise, the pitch control forces and pitch stability characteris- ties were nearly normal, as illustrated in Figure 2. Special instrumentation in the passenger compartment of the Twin Otter, along with live video of tufts on the bottom side of the tailplane, enabled us to monitor the airflows for- ward of and near the surface of the horizontal tail (Figure 3). ‘They indicated thatthe airflow over the tail was slightly dis rupted because of the ice shapes, but there was plenty of EFFECT OF PITCH INPUTS — ot attack of the tailplane, because it’s flying upside down through the airflow aft of the wing. ‘The same elevator dou- blets that produced very lit- tle control-feel change with + tas Devt Bato flaps up, caused ro ane rma! PS UP, more pi Prenferereet | nounced effects with flaps 10 degrees. Elevator buffet, as evidenced by mild shak- ing of the yoke, clearly told us that the tailplane stall margin was starting to SS +Mewteeeredet | Pulling back on the yoke Semberpeceeases —“ atacton | increased the overall cam- NS nme | ber of the tailplane airfol ‘This increased the stali tailplane stall margin. The stall mar- ‘in increased slightly with nose-up ele- vator and it decreased slightly with nose-down inputs. Next, we extended the flaps to 10 degrees. There was mild elevator buf- fet and a softening of pitch control feedback, especially with nose-down elevator inputs. With each notch of flaps, the test equipment showed a decrease in tailplane stall margin and more rows of tufts began to dance. ‘This was caused by an increase in downwash over the wings associated with extending the flaps. The down- wash increases the negative angle-of- margin by improving laminar flow. ‘The tufts lay down and the yoke buffet decreased. Pushing forward on the yoke resulted in less tailplane camber. ‘More tufts came alive and the yoke buffet increased, indicating more air- flow separation. In addition, the yoke force was spongy and inconsistent QUANTIFYING THE EFFECTS OF TAILPLANE ICING Four years ago, the FAA and NASA ‘embarked upon a program to quan- tify the actual aerodynamic perfor- ‘mance and handling characteristics associated with ice-contaminated tailplane stall NASA's veteran de Havilland Dash 6 Twin Otter icing-research sireaft was called into service for the tailplane icing program (TIP). A team led by Thomas Ratvasky, tech- nical program manager of NASA Lewis Research Center's Icing Technology Branch, installed an elaborate monitoring and recording system with hundreds of data chan- nels to record pitch, roll, and yaw angle and rate; pressure distribu- tion over the upper and lower sur- faces of the tailplane; control forces and control surface deflection angles; plus engine thrust output. They also mounted yarn tufts on the bottom surface of the tailplane and installed video cameras to mon- itor the tailplane, tape pilot and without ice contamination and to develop analytical methods to predict an aircraft's sensitivity to tailplane icing and stall. “We really are trying to understand the aerody- namies and physics that somebody will apply to a future airplane design,” commented Richard Ranaudo, chief of aircraft opera- tions at NASA Lewis. Ranaudo’s pilots performed a series of calibration flights in the ‘Twin Otter in unmodified form, fing Ratvasky’s team with baseline performance data. Based on preliminary flight tests, Ratvasky's team mounted a full- scale Twin Otter tailplane in NASA's Icing Research Tunnel CURT) at an angle of attack repre- sentative of typical cruise speeds. ‘Then, they fired up the IRT, chilled the airflow to freezing temperatures and sprayed in an icing cloud. Their recipe for water content, tempera- ture and droplet size came right out of FAR Part 25, Appendix C, which is the FAA's inflight icing certifica- tion regulation Ratvasky’s flew two accretion profiles in the IRT. One simulated the inter-boot, or eyele-to- cycle ice accumulation on the tions and wateh the horizon sequent deice boot expansion cycles. through the windshield (Figure 3). taperg dh Won Sern, PhD. wi The other simulated 22.9 minaes of ‘The goals were to collect precise section of xe shops derfied flight in icing conditions with a data on tailplane aerodynamics with failed deice boot. 82 ‘eprint compliments of FlightSafety Canada - de Haviland Leaming Centre Fond ‘TWIN OTTER TAILPLANE FLOW TUFT LAYOUT (ene ‘Air Data Probes sexs: NASA Bottom View of Lat Side of Talplane Fens with forward pitch-control inputs. Each successive notch of flaps exac- erbated the effects. Pulling back the ‘yoke was accompanied by less buflet- ing and more pitch control. Pushing forward resulted in stick force redue- tion or control force reversal, both of which were more erratic with more flap extension. Those are two classic symptomatic differences between impending tailplane stall and wing ‘gente echeaes + en eater ae Sento stall. Notabl; =n although the ele- vator buffet and EFFECT OF FLAP INPUTS = 2 (Sten Fores Uterine pitch force anomalies could be felt through the controls, there wasn’t a trace of airframe buffet, which is s0 typical of an impending wing stall. Later, when we rode in the back of the Twin Otter as a passenger, we noted how smoothly the aircraft rode up to the point of tailplane stall. That was another differentiating cue between impending tailplane and wing stall At flaps 40 degrees, the yoke buffet unmistakably indicated disrupted air- flow under the tailplane. It was diffi- cult to keep the aireraft in pitch trim. Pulling back on the yoke only partially reduced the buffet. Pushing forward ‘When each test was complete, the team took an impression of each ice buildup using a special, low-temper- ature molding compound. The molds were used to form artificial, plastic ice shapes. Two more traditional ice shapes were built. One was modeled using NASA's Lewis Ice, or Lewice, computer code and the other was a stability and control (S & C) ice shape that previously was used in ‘Twin Otter flight test. All four shapes were flown on the tailplane in Ohio State determ tailplane stall margin and control forces, The results? The standard test shapes, used for FAA cer- tification, produced worse effects on’ tailplane aerodynamics than the ice accretion shapes. That's good news for pilots flying in actual icing conditions in icing-cer- tificated airplanes. ‘The wind tunnel indicated that tailplane perfor- mance degradation caused by the Lewice and S & C shapes was so close that Ratvasky elected to flight test only the S & C ice shape, along, with the eycle-to-eycle and failed- boot ice shapes. Forty-two test flights were conducted between 1995 and 1997 in the unmodified configuration and with the three ice ‘Thomas Ratvasky and team will months reducing data before shapes. The intent was to quantify the aerodynamic characteris with various combinations of flaps, thrust, angle of attack and sideslip. ‘The maneuvers included push-pull elevator doublets, steady sideslips, power transitions, a balked landing profile and flap extension/retraction. In addition, Ratvasky’s team mea- sured the tailplane aerodynamics during a series of reduced and zero “g” pushover maneuvers. The final analysis will require several more months of computer number crunching. Ratvasky's team and Ranaudo would like to begin a second series of tests using NASA's Learjet 25. The air- craft's low wing and T-tail configuration endow it with much different aerodynamics. Its wing leading edges have an evaporative, bleed-air heat. anti-ice system, and its hor zontal tail is fitted with elec- trothermal anti-ice blankets. Using ICTS data from the TIP Twin Otter tests and the Learjet 25, the FAA and NASA believe that airframe manufacturers might be able to develop computer programs that would predict adverse changes in stability and control characteris- tics caused by tailplane icing. The ‘computer programs could make air- craft safer to fly in icing conditions and reduce development costs. However, funding for a second series of tailplane icing tests remained uncertain at press time. ‘eprint compliments of FightSafety Canada - de Haviland Learning Centre ‘on the yoke caused much more buffet ing, accompanied by a reduction and then a sudden, but never predictable, reversal in control force. In plain terms, the yoke seemed like it wanted to pull out of our hands without warn- ing. The tailplane stall margin col- lapsed to near zero. ‘Ranaudo pointed out that the combi- nation of positive control force feel ‘when pulling back on the yoke, cou- pled with a reduction or reversal of control force when pushing forward on the yoke can lead to pilot induced oscillations (PIOs). The dynamic rock- ing pitch motion of a PIO can further ‘erode tailplane stall margin due to the changing angle of attack caused by pitch rate, a: ‘The third maneuver involved vary- ing the airspeed between 65 and 85 knots. As the speed increased, the air- craft's nose attitude decreased. That reduced the wing’s angle of attack. It had the opposite effect on the tailplane, which is designed to produce downward lift. Its negative angle-of- attack increased, with the adverse effects shown in Figure 5. With the flaps extended, faster speeds produced lower tailplane stall margins, more elevator buffet and more erratie pitch control feel. Each notch of flaps aggravated the effect of speed change with respect to decreas- {ng tailplane stall margin. ‘The fourth maneuver involved changing engine power (Figure 6). It produced even more memorable results, Before starting the maneuver, Ranaudo cautioned against making robust elevator inputs because they can lead to a possible full tailplane stall and complete loss of pitch control. The instrumentation onboard the aircraft showed that an increase in prop thrust increased tailplane nega- tive angle of attack and decreased stall margin, as illustrated in Figure 3. EEE, Coanda effect is the increase in laminar flow caused by forcing higher ve neat air over an airfoil. Ranaudo said it’s too early to say for certain why this occurs. However, two industry test pilots with whom we spoke said they wouldn’t rule out ‘Coanda effect. ‘Coanda effect is the increase in lami- nar flow caused by forcing higher ‘velocity air over an airfoil. Airflow sep- aration is delayed because the higher velocity airflow adds energy to the boundary layer. This, too, may increase the downwash off the wing and thus the negative angle-of-attack of the airflow at the ‘When we added full power with full flaps, we experienced imminent stall. The yoke shook strong- ly enough to scramble eggs if it had been linked to a whisk. It also errati- cally pulled forward with a force of 75 to 90 pounds, according to onboard instrumentation. If we had mistaken the symptoms for a wing stall and relaxed back pressure on the yoke, the tailplane certainly could have stalled completely, resulting in a stick-pull force in excess of 195 pounds, based on ‘Ranaudo's experience in the fight test program. “Imagine trying to recover the airplane close to the ground on short final under those conditions,” Ranaudo mentioned. ‘The final two tests involved more dynamic maneuvering. The first involved a series of pitch-down maneu- vers that simulated the FAA's recently imposed, zero “g” stability and control test with simulated ice shapes on the tailplane, The test, now required for flight-into-known-icing certification, specifies that the aircraft must remain positively stable and its control forces must not reverse. In essence, it’s an inflight check of the tailplane’s stall ‘margin when contaminated with ice During three, progressively more robust, pushover maneuvers, we expe- rienced pitch control force reduction and partial reversal that was propor- tionate to nose-down pitch rate. How- ever, the adverse effects associated with’ such pitch over maneuvers in the ‘Twin Otter weren't as strong as those we experienced during the thrust change demonstration or elevator dou- Dlets with full flaps The simulated balked landing ‘maneuver was the culmination of the Fawe 5 IPFECT OF SPEED ads EFFECT OF PROP THRUST (Possibly Configuration Dependent) ie === 8 SS = & = soe = om 3.2 @| ss. = EE oe See ee ‘Siow eoanares, sa tcncen tee s moon” nner reprint compliments of FlightSafety Canada - de Havilland Learning Centre demonstration. It combined virtually all of the individual maneuvers in an approach to landing and go-around sce- nario; however, in the interest of safety, Ranaudo restricted flap extension to 30 degrees, rather than using the 40 degrees of landing flaps. At the “go around” point, still more than 5,000 feet above the ground, we added full takeoff power. With flaps 80 degrees, the nose pitched down sharply. ‘The pitch con- trol force fully reversed, requiring 60-plus pounds of force to maintain nose-level attitude. Tt was difficult: not to enter a PIO with so much back Powe? EFFECT OF C.G. arTc. = Ss Lest Down Force Reged Loner tm coomcert [ATTACHED Arbo FORWARD C0. 3S ‘More own Fore Required heer cuctt tailplane-stall__recovery™, requires opposite controi, flap and thrust inputs Veterans of inflight icing encounters also warn pilots to be wary of the symptoms of ice accretion and to be ready to immediately undo configuration changes. For example, if extending the produces elevator buf- feting or abnormal stability and control changes, then retract them. Standard procedures for flying into known icing con- ditions are especially impor- tant to protect against ICTS. Change your course pressure needed to stabilize the aireraft. If the Twin Otter hhad been configured at flaps 40 degrees, the pitch control forces could have been more than three times greater. SOBERING LESSONS ‘The biggest reason why none of the ICTS maneuvers put us at risk was that we knew what to expect. If we hadn’t been briefed by Ranaudo, the results might have been far different, if not potentially catastrophic, in the event of an actual ICTS. As shown in Figure 8, the symptoms of tailplane icing are increased eleva- tor buffet, elevator or pitch force light- ening and reversal, reduction of tailplane lift and a reduction of pitch A forward c.g. reduces ice-contominated-tailplane stall margin. stability. Some aircraft also exhibit mild yawing tendencies. The illustration also shows the five factors that individually, or together, reduce ice-contaminated-tailplane stall margin. Avoiding a tailplane stall in icing conditions boils down to the following: Elevator buffeting, especially if the aircraft is experiencing little or no tur- bulence, may indicate ice contamina- tion of the tailplane. Control force reduction or rever- ssal, especially with nose-down pitch inputs, is symptomatic of tailplane ice contamination ‘Compared to wing-stall recovery, Figwe 8 FACTORS THAT DECREASE ICE-CONTAMINATED-TAILPLANE STALL MARGIN and altitude as soon as pos- sible to fly out of the area of exposure without delay. Use the ice protection sys- tem to buy time to fly out of the area of exposure. Don’t continue to fly through it. Use the guidance provided by the AFM, if ice contamination is suspected “ during landing approach, and abrupt control inputs. Avoid using the autopilot when operating in icing conditions so that you can sense abnormal control feel changes in your hands that may be symptomatic of sta- bility and control changes due to ice accretion. ‘Other aireraft with ice-contaminated tailplanes may not exhibit the same symptoms as NASA's Twin Otter. However, NASA, the FAA and air- frame manufacturer icing experts all agree that there's no such thing as an aircraft that’s immune to severe or extreme icing. Flight into known icing approval only means that the aircraft passed a series of tests required by FAR Part 25 Appendix C. If you encounter icing conditions in excess of those theoretical standards, you may be at risk for tailplane icing. Tailplane ice contamination can be insidious. The penalty for not recog- nizing it can be fatal, particularly when configuring an airplane for landing. However, if you're on the lookout for telltale symptoms and take corrective action, the risk exposure all but dries up, Then, you can add your tail- ~ plane icing experience to your list of \ hangar flying stories instead of becom- ing an ICTS accident statistic. reprint compliments of FlightSafety Canada - de Havilland Learning Centre Reprited trom Business & Commercial Aviation, December 1997, copyright by The MeGraw-Htil Companies nc. with al rights reserved QYEATION “WEEK iio _NASA Safety Study Focuses “On Tailplane Icing Stalls DAVID M. NORTH/CLEVELAND, he test ig being Lo done here ac NASA's Lewis Research Center shows that a pilot responding to what is believed to be a wing stall during an approach to landing could be applying the wrong recovery | procedure—what the plot may be ac tually encouncering ina stall due to iplane icin “Fhenormal recovery procedure froma wing stall requires the pilot to lower the nose of the aircraft and add power while leaving the flaps in their present position ‘The recovery from a tailplane stall calls ‘bP te pilot to reduce power, pull back on the yoke and raise laps. In many in- stances, when a pilot encounters unpre- dicted flight characteristics, there is time to analyze the situation and recover. But this might not be the situation for a pi- Jot making an approach in adverse weath- rand ata low alticude Following a history of 16 accidents be- lieved caused by tailplane icing, the Lewis Research Center atthe request ofthe FAA, embarked on a four-year program todeline thei tors that lead to a tailplane stall and 10 demonstrate ef- fective tailplane stall recovery. Dur- ing the program, NASA pilots and researchers have de veloped a compre- recite tailplane aero-per formance by the wse ‘of wind cunnel test ing, numerical sim- ulation and flight Phase one ofthe program was init ted in September 1994 and included ice shape selection, sai [NASA Lewis’ de Haviland Twin Otter has ‘been used forthe Fight testing segment of © four-year study of tiiplone icing. The icing study was done atthe request of he FAA. aero-wind tunnel time, analytical code de- velopment and flight testing the baseline DHC66 aircraft. A total of 16 baseline tests were flown in 1995. The second phase of the program included icing wind unnel tests, analytical code development and fight testing with ice shapes, with some shapes developed and flown inthe aero-wind tune Artficiol ice castings were bonded to the let horizontal stabilizer ofthe Twin Otter. video «camera mounied onthe lower fuseloge mon- itored yorn tus for low separation. nel at Ohio State TD Uses. Some 26 fighes were made in Is; ee program. The ice thapes flown during that yar ice thos repeating residual oo, ict formed by a flled deicing boot and an ice shape whose cross section was By calelared from the ‘ADS-4 manual, “The light esti Berni cconluedlinhcrpoe atte Edd several media Highs { was frou nate to fy in the centers de Hivilland DHC-6 Tin Ore wth projec tex poe Rich Ranaudo and NASA lead project en- inet Torn Rarvashy, Jy Van Zante an- ther NASA project engineer, took pert in the prefigh briefings but didnot tipar in our Aight I chared the left a flying with ancther Ranaudo occupied the ‘The NASA Twin Otter, NOO7NA, had «tice shape installed on the full-span ead ing edge of the horizontal st flow angle probes and stabilizer and ele- ‘ator pressure belts on the lft side only ‘The research instrumentation inside the aircraft includes an inertial data system recording three-axes acceleration, rate and angle displace ment. The aircrafts flight parameters are recorded and ‘monitored, as well as uilplane angle of axtack, yaw and air- Al of the con: trol surface deflec- tion angles, as well asyoke and rudder force and tail sur- face pressure data, are shown on the rear eabin monitor on the right side of the aircraft. Two vido reorders have been installed to Recrntes rom Aviation Wook & Space Technology, February 9, 1998, copyright by The MeGraw-Mil Companies, nc. with al ight reserved, AIR TRANSPORT take data. One has been mounted in a housing on the aircraft's lower fuselage, and the other is located behind the pilot to record pilot actions and the horizon during maneuvers. The tail-mounted video records airflow and the deflection of the yarn cufts mounted on the un- derside of the horizontal stabilizer. THE TWIN OTTER WAS PICKED 2s che pro- peller-driven candidate for the tailplane icing tests because it was in the NASA in- ‘ventory and its flight characteristics were well-known, Ranaudo said. He also said thatthe aircraft gives warning cues tothe pilor when encountering tailplane and air- frame stalls, not unlike many other pro peller-driven aircraft. The NASA pilot was quick to point out that we would be flying the aircraft with flap configurations thac are not authorized by de Hav- illand in their operation man- tals for flights in known icing, The DHC-6 is limited co 10 deg. of flaps in known icing by the manufacturer. ‘Once in the Twin Otter, I ‘occupied a scat in the cabin on the lefe side that included a monitor of many of the para- meters being shown on Rat vasky's research instrumenta- tion system across the isle The first set of flight ma- neuvers was designed co give the gucs pilot some confidence in che light characteristics of the Twin Otter. My observa- tions ofthe numerical data por- trayed on the interior monitor when the maneuvers were flown by the other guest pilot were consistent with what I observed while flying feom the lft seat. Once in the train- ing area, and at an altitude of 8,000 ft, wwe started the maneuvers associated with the tailplane icing program. THE INITIAL MANEUVER WAS to fly sever: al pitch doublets in the flaps-up config- uration, followed by both left and right rudder kicks. This was to provide famil- iarity with nominal control response and ‘damping characteristics and establish the face thatthe Twin Oxcer has good longi tudinal static stability without unusual characteristics. Ranaudo then had me es tablish an 80-kt. cruise speed with the laps retracted. The tailplane stall margin was 19 deg., which is the difference berween the stall angle found in the wind tunnel and the real-time tail angle of attack Dropping the flaps to the 10-deg. posi- tion reduced the stall margin to 12-deg, angle of attack. At 20 deg. of flaps, the tailplane stall margin was 6 deg., with a contaminated leading edge of the hori- zontal stabilizer. I also found that the DHC-68 control forces were lighter in this configuration With the flaps set at 30 deg, the stall margin was 5 deg., and I could discern a very slight buffet in the control yoke but not in the airframe. Dropping the flaps to their full 37.5-deg. sertng gave a stall m jn of 3 deg. and a more pronounced buf- FE in the Fin Oters yoke, The contol column vibration would be only one ication vo pls that they ar expren ing onset of a tailplane stall and nor an ee Seely mere Ratvasky said that the first four rows of tufison the underside ofthe stabilizer were showing flow separation. There also was Lewis project engineer Judy Van Zante ob- serves fight dota ot the console inthe Twin ‘ter. Tellplone icing program findings ore ‘expected to be completed later this year. a tendency of the control yoke to want to go forward during the flight with full laps, and Ranaudo said that pushing the yoke forward, the nose would want to go far- ther than expected. He said thatthe in- creased flap deflection increases negative tail angle of artack due co wing downwash ‘Once the flaps were retracted to less than 25 deg., the yoke butt ceased The next maneuver was a speed transi tion to show the effect of speed on tailplane stall. The power was reduced to above idle, the laps were extended to full and a 60-kt. airspeed was established. I twimmed the aircraft for level ght at that speed. Power was added to the cruise set- ting, and the nose was lowered to attain 85 ke. The contro force increased as the speed increased, and I found myself fight- ing the yoke as the angle of attack on the tailplane became more negative. Daring the power transition maneuvet the flaps were lowered to 30 deg., and ale 85-kt. airspeed was flown at low power a gradual descent. Once trimmed at 85 added the power slowly until reaching maximum engine torque. Nearly 50 Ib. of steady force and a maximum peak force of close to 90 Ib. was required in the at tempt to keep the aircrafts nose up and maintain 80 kt. Almost all the cufts showed flow separation as we neared the ‘maximum amount of force I could use for the 30-deg. lap position. THIS MANEUVER VERY EFFECTIVELY cemon- strated the adverse effect of engine pow- cron heirs begining stability characteristics and el- evator control hinge moments. Ranaudo said that if chat same maneuver is flown with full flaps, the tailplane will sal ‘One such maneuver required 180 Ib. of force on the yoke to recover from the stall ‘The next maneuver I flew was a “pushover,” similar to those required by the FAA for aircraft certification. The ma neuver was flown with a con- stant low cruise power settin tri ape 10 deg etch at 20 deg. In both maneuvers. —~ the nose of the aircraft was raised 15 deg. and then was nosed over to achieve about a S-deg./sec. pitch rate. At zer0-G and as the aircraft’ nose was dropping through the hori- wots asta caves lighter and I experienced con. ol force reversal. As Ranaudo said, the Twin Oxter with tailplane icing would nox have passed the certification requirement for this maneuver at 20 deg, of laps, which is why the aircraft is restricted to 10 deg, of flaps in icing conditions. ALL OF THESE MANEUVERS had been per formed to lead up tothe next, which was to show the combined effects of lap po- sition, power setting and speed on the flight characteristics ofthe aireraft wich tailplane icin Ratvasky, acting asa ground control approach controller, gave me a number of instructions on heading and altitude changes that required fairly aggressive power and control inputs to comply. I started at 70-kt. airspeed with flaps at 30 deg. for the approach. Again, Ranaudo said that in a Twin Oceer, an instrumer approach would not normally be flowin with flaps at 30 deg, The 30-deg, setting was used for the purpose of demon- strating how aircraft handling qualities are affected by flap configuration, flight ition and power effect, Ravasey initially for a 500-fe./min. descent AL 4 S-deg. heading change, then a 17000. /min. rae of descent and an- other 5S-deg. turn in the opposite direc- tion to level off ‘Other heading and descent changes fol- lowed at 20-sec. intervals until the final fra 1,000-u/min. descent was by a command from Ratvasky to sound. I advanced the power levers to Il and maintained speed until I saw a positive rate of limb, bur the aireraftini- tially did not want o respond as T used both hands to pull the yoke back to climb, 1 did not stall the tailplane during the re- ‘covery; however, as Ranaudo pointed out, Iwas briefed for the event and knew the potential dangers. Imagine, Ranaudo said, trying this = actual weather, in turbulence, on a proach at low altitude while starting high an the approach and attempring to get down to the glideslope? All ofthese fac tors could lesd to an accident with the 3¢ iced up, he said. The beginning ‘of elevator buffet isthe best warning pri- ‘orto al stall, he added. THE NEXT TWO MANEUVERS were co demonstrate some conditions that allevi- ate the effects of tailplane icing. Ranaudo hhad me pick out some section lines on the ground and line up at 75 kt. with fall Sit ually “i ing left rudder and tight leon to hsp lined up on the eo- tion line, the initial response of the Twin Onter was a buffet in the control yoke at 12 deg. of sideslip. Once past 14 deg., the butfer disappeared, indicating improved longitudinal characteristics with tai icing. A sideslip co the opposite side re- sulted in the same flight characteristics. Ranaudo said chae these results were an “Onter-ism.” ‘Much the same results were encountered during ‘bank turn at 100 kt with flaps at 30 deg. botfeting decreased as the turning g,forces increased to main- tain altitude—demonstrating the effect of the ge eter sim de acteristics caused by tailplane Then had hed ira ‘Twin Otter back to the cilicy ac the Cleveland airport, and found the aircraft to be sable, pectic and not difficult toland with ape up [NASA officals expect data analysis and reporting on tailplane icing will continue into the third quarter of this year. anaudo, Ratvasky and Van Zante said they filly realize some of the Twin Otters longitudinal flight characteristics with tailplane icing are unique to the type, but that other aircraft types may have the same tactile warning cues tothe pilot of an im- pending ailpane sal “ings high roy in viaon sae ty issues, and tailplane icing is an im- porrant element in the equation,” Ranau- do said. "Once we truly know what we learned here, we will look at how best to transmit that knowledge to pilots in their training,” “The information gained from the NASA research program aso could be used to set new FAA certification requirements, as well as having a positive influence on tailplane design, Ranaudo said, NASA already is having a with Bombardiers Learjet in Wichita, icing studies of the more modern airfoils found in the companys business jets. “T regret that U.S. export laws make it difficult for us to cooperate with foreign aircraft manufacturers,” Ranaudo said. “I believe that we have something to share with them that would benefit aviation safery” ° Reprint compliments of FlightSafety Canada-de Havilland Learning Centre Shaking in my boots A new twist to in-flight icing encounters. tn normal in-flight bore antice/deice system(s). By Stuart Lau ATP/CFIl. Canadair Regional Jet, Saab 340. URING initial flight training, Doreen esos pilot's career, much atten- tion was given to airframe icing and the negative impact it has on the aerodynamic performance of the aircraft’s wing. The crux of this information was that an airfoil con- taminated with ice added weight, caused a loss of lift and decreased the stall margin. In recent years, many new prob- lems with in-flight icing have been discovered, primarily focusing on lateral control problems caused by prolonged periods of flight in a ‘weather phenomenon called super- cooled large droplets (SLDs). For those of you who are in the need of a review, an SLO is the freak of the airborne water droplets. SLDs remain in a liquid state at wemperatures well below freezing and are very large, up to 4000 microns in diameter, which is 100 times larger than the droplet size ing conditions, airborne water

You might also like