Aviation Accident Report: Piedmont Airlines Flight 349

Aviation Accident Report: Piedmont Airlines Flight 349  (1961) 
by Alan S. Boyd, Chan Gurney, G. Joseph Minetti, Whitney Gillilland for the Civil Aeronautics Board

SA-348 File No 1-0065


ADOPTED April 18, 1961 RELEASED: April 24, 1961



About 2040, October 30, 1959, Piedmont Airlines Flight 349 crashed on Bucks Elbow Mountain located about 13 miles west of the Charlottesville-Albemarle County, Virginia, Airport. The crew of 3 and 23 of 24 passengers were killed; the sole survivor was seriously injured. The aircraft, a DC-3, N 55V, was demolished by impact.

From the available evidence it is the determination of the Board that this accident occurred during an intended instrument approach. More specifically, it occurred during the inbound portion of the procedure turn which was being flown S to 11 miles west of the maneuvering area prescribed by the instrument approach procedure.

The Board concludes that the lateral error resulted from a navigational omission which took place when the pilot did not turn left about 20 degrees in conformity to V-140 airway at the Casanova omni range station. Consequently, when the pilots believed the flight was over the Rochelle intersection it was in fact 13 miles northwest of that position. As a result of this position, when the pilot turned left and flew the heading normally flown from Rochelle intersection, the path of the aircraft over the ground was displaced 6 to 11 miles west of the prescribed track. The Board further concludes that the error was undetected because tracking and other instrument approach requirements were not followed precisely.

From information regarding the personal background of Captain Lavrinc and expert medical analysis of this information, it it the Board's opinion that preoccupation resulting from mental stress may have been a contributing factor in the accident cause.

Investigation (See Attachment A for reference.)

Piedmont Airlines Flight 349 was a regularly scheduled flight between Washington, D.C., and Roanoke Virginia, with intermediate stops at Charlottesville and Lynchburg, Virginia. On October 30, 1959, the assigned flight crew consisted of Captain George Lavrinc, First Officer Bascom Haley, and Purser George Hicks.


The aircraft departed wWshington at 1949. [1] It was about 20 minutes behind schedule, the result of traffic, ground, and ramp delays. The flight was according to an instrument flight plan and clearance which specified a routing over airway V-140 and a cruising altitude of 4,OOO feet.2/ [2] At departure the gross takeoff Weight of DC-3, N 55V, was 25,346 pounds, which was also the maximum allowable weight. The load was properly distributed.

The flight made a Georgetown departure and then proceeded to the Springfield, Virginia, radio beacon where it entered V-140 airway. It followed the airway from Springfield to the Casanova omni range station. The centerline of the airway over this segment is defined as the 260-degree radial of the Casanova omni. At 2012 [3] Flight 349 reported that it was over Casanova at 2010. It also reported it was at 4,000 feet and estimating the Rochelle intersection at 2024, with Charlottesville next.

At Casanova, V-l4O airway turns left and from Casanova to the Rochelle intersection the airway is defined as the 239-degree radial of Casanova omni; the distance is 31 miles. [4] The Rochelle intersection is the 239-degree radial of the Casanova omni and the 335-degree radial of the Gordonsville omni.

About 2018 Flight 349 made a routine inrange report to the Piedmont ground radio station located at the Charlottesville Airport. The flight informed the company that it was in range, IFR (instrument flight rules), and would have 250 gallons of fuel on board when it departed Charlottesville for Lynchburg. The radio operator furnished the flight the latest altimeter setting, 30.47, and the current surface wind which was calm. Normally, the operator would also furnish the latest ceiling and visibility observation. This, according to his testimony, he failed to do. The current observation at this time was Ceiling, measured 1,500 feet broken, 4,000 feet overcast; visibility 10 miles.

Immediately after 2025 the flight reported to the Washington Air Traffic Control Center that it was over the Rochelle intersection at 2025 at 4,000 feet and estimating the Charlottesv111e Airport at 2030. Washington Center acknowledged and cleared Flight 349 for an instrument approach. About one minute later the flight informed the center as follows, "You can put us out of four thousand." This was the last transmission from the aircraft.

Investigation revealed the weather observations for Charlottesville were, at 1900, 1,700 feet scattered, ceiling 4,000 feet, overcast, visibility 10 miles; at 1930, ceiling 1,700 feet broken, 4,000 overcast, at 2000, ceiling measured 1,500 feet broken, 4,000 feet overcast, visibility 10 miles, wind calm; at 2100, ceiling measured 1,500 feet broken, 2,400 feet overcast, v1s1bility 10 miles, wind calm. A study of the weather observations from other stations in the area showed the Charlottesville conditions should have been substantially as reported. The Situation, however, showed that conditions would be much worse in the area near and parallel to the mountains west of Charlottesville. In this area the mountains were obscured and Visual flight would not have been possible.

Studies of the Winds aloft by the U. S. Weather Bureau showed that at altitudes used by Flight 3h9 they were predominantly southerly and averaged about 15 knots.

When the flight did not land as anticipated a radio search was made, which proved futile. A ground search was begun as quickly as pos51ble and supplemented by an air search the next day. Throughout that day both were seriously hampered by bad weather. On November 1, about 0800, the wreckage was Sighted from a helicopter on the southern slope of Bucks Elbow Mountain, which is located about 13 miles west of the Charlottesville Airport. It was almost hidden by dense tree cover.

Investigation at the scans showed the aircraft crashed where the upslope of the 3,100—foot mountain was nearly 30 degrees. It crashed against the rocky slope on a magnetic heading of 3h0—350 degrees and at an elevation of approximately 2,600 feet. Initial contact occurred when the right Wing of N 55V struck and cut through several trees which progressively tore off the right Wing outboard of the landing light. At initial impact the right wing was down about 10 degrees from level and the aircraft was descending slightly. The aircraft was yawed to the right and rolled to the right when, about 180 feet farther, it crashed against the upslope.

In the final impact the forward fuselage rearward to the center section was destroyed. The fuselage was broken from the center section and displaced to the left so it rested on the left wing outboard of the attach angle. The fuselage was also cooked to the left about 15 degrees relative to the center section. Most of the fuselage from over the center section rearward to the cabin door was destroyed or badly crushed. All of the passenger seats were torn from the floor, including the one in which the surviving passenger remained fastened when he was thrown clear of the fuselage.

The structural investigation determined that the landing gear was extended at impact and the flaps were fully retracted. Also, although the aircraft sustained great damage, it was reliably determined that there was no malfunction or failure of the aircraft prior to impact.

Examination of the engines and propellers disclosed no evidence of malfunction prior to impact. Evidence showed that both engines were operating and developing appreciable power when the aircraft struck the mountain, the specific amount of power, however, could not be determined. Evidence also Showed that at that time both propellers were capable of normal operation. Examination and bench checks where required showed the engine accessories were capable of normal operation before the impact.

Mr. Phll Bradley, the sole surv1vor, stated that the flight seemed perfectly normal untll the crash. H13 descrlptlon of the weather 1ndlcated that except for the 1n1t1al portlon, the flight was made 1n 1nstrument weather cordltlons. He noted thls when he periodically peered through the cabin wlndow from hlS seat. He also noted reflected llght from the antlcollls1on llght. N . Bradley stated the fllght was a llttle rough and except for a short perlod following takeoff the “fasten seat belt” Slgn was on until the crash.

The surv1vor stated there was no 1ndlcatlon of any dlfflculty. He sa1d the cabin llghts were on and the englnes sounded normal. He sald that a few mlnttes befo‘e and at the tlme of the crash the cabln was qulet and all pas— sengers were seated. He Bald both pllots were ;n the pllot compartment and the purser was attendlng to file dutles.

Mr. Bradley sald the plane made several turns, although he was not sure of the amount or dlrectlon. One such turn, and seemlngly the largest, occured a few mlnutes before the crash. qe, at flrst, thought the turn was about 180 degrees but later felt it was more 1n the order of 90 degrees.

The watness sald he had Just folded hlS arms and was looklng at hls wrlst watch when the crash occurred, 1t was BOhO. He sald hlS watch was an accurate tlmeplece and he had checked it agalnst an alrport clock earller that day. In thls connectlon, other watches were recovered and some had been impact stopped. Toese showed warlous tames whlch bracketed 20hO.

The approach procedure for the Charlottesv1lle Alrport 13 an ADF (auto— "aIlC direction flnder) procedure performed on a Pledmont-owned and operated comer beacon feelllty. The nlght landlng mlnlma for Piedmont DC-3 alrcraft are. Celllng hOO feet, Vlslblllty l mlle.

The current Federal Av1atlon Agency (FAA) approach procedure as applicable to Fllght 3h? would begln at the Rochelle 1ntersect10n. Accordlng to the pro— cedure 1n effect at the tlne of the acoldent, when Rochelle was reached the flight would tran51tlon off V—lhO alrway in a left turn to a headlng of 212 de— grees. It would then estaollsh and fly a 212—degree track to the Charlottesv1lle {CEO} homer beacon, whlch transmlts on 28h kcs. As an addltlonal ald, though not requlred, the company owns and operates another hmner beacon, Earlyv1lle CEVL}, whlch transmits on 266 kcs. Charlottesv1lle, the outer homer, 15 located h.3 nautlcal mlles from the approach end of runway 3 and about 15.5 mlles from the Rochelle 1rtersect10n.

In normal executlon of the lustrument procedure most Pledmont pllots use both homer feellltles, tunlng one ADF to the Charlottesv1lle homer, 28h kcs., and the other to the Earlyv1lle beacon, 266 kcs. In addltlon, many of the pllots also cneck passage of the Charlottesv1lle homer by nelng the Gordons— Ville omnl, set to the 287~degree radlal whlch passes through the homer. Some, at the same time, also check passage of the Earlyv1lle feelllty by nelng the second own; set, tuned to the BOl-degree radlal of Gordonsv1lle.

Ehen Board 1nvest1gators attempted to plot the Rochelle intersectlon and alrport locatlons, 1t was noted that the magnetlc heading from Rochelle to the Charlottesv1lle homer, as depleted on the ADA Form Ell and thus on the approach plate, was in error. rI‘he correct heading should have been 201 degrees instead of 212. The error resulted from not amending the heading when, several months earlier, the course of V—lhO airway was shifted slightly, In this acc1dent the error loses Significance because, according to the fllght plan for Trip 3h9, the correct heading was used. Further, the uSe of tracklng procedures in flight would eliminate the effect of the erroneous heading. Hevertheless, the attention of both the FAA and the company was immediately directed to the error for correction.

According to the instrument approach procedure, upon reaching the CharlotteSVille homer beacon an outbound track of 20? degrees should be flown, normally for l to 1—1/2 minutes. This is followed by a standard procedure turn on the southeast Side of the track. USing the standard procedure turn, the outbound heading is 162 degrees and the inbound heading is 3h2 degrees. The final approach track to the airport is 02'? degrees.

Descent below 3,000 feet is not authorized prior to the final approach; then a descent lS permitted to not less than 2,200 feet before reaching the Charlottesnlle homer. Thereafter, descent may be continued to the authorized comm altitude of 1,039 feet, or [LOO feet above airport elevation.

From the Rochelle intersection the entire instrument approach to landing in the DC-3 takes apprOXimately 15 minutes. From Rochelle to the inbound heading of the procedure turn about 10—11 minutes1 time 18 normally required. It appears important to note that based on the reported time of Flight 31;? over Rochelle, 2025, and the crash time indicated by TI . Bradley, 201:,0, the Eelapsed time was approxzunately 15 minutes.

Although the radio and nangational equipment from N 551? was badly damaged and some portions were destroyed, information which was important to the investigation was available. Examination of the omni equipment dis— closed that both receivers were tuned to 115.3 H1ch the Gordonsnlle omni range frequency. It was also learned that the right omni bearing indicator was set to select the 30l—degree radial or, as prevxously indicated, the radial which passes through the EarlyVille homer. The radial selected on the other unit could not be determined

Examination of the Various components of the red and green ADP units disclosed both receivers were peeitioned to select hand 1, the ZOO—hit) kc. range. It was determined that the red ADF unit was tuned to 265 kcs., oI' appromately the frequency of the Earlymlle homer, 266 kcs. Impact damage to the green ADF made the most reliable indication of the frequency setting the peeition of the frequency selector. This , slightly damaged and locked 19— peeition, was set on 286 kcs., or close to the frequency of the Charlottesulle homer, 28h kcs.

Examination of the ADF loop assemblies revealed that the position of the red ADF loop was equivalent to a bearing of 91 degrees on the cockpit needle: 0n the same 138.513 the green ADF was pOSitioned on a bearing of 112 degrees. The loop pOSitions are most reliable as to the pOSitions of the needles 011 the ass mstrument; however, this cockpit instrument, except the face, was osgtfob’ed' It is Significant that With the aircraft on the crash heading of about 3:47 09'" grass the bearings of 91 and 1.12 degrees extended from the crash Site pass very close to the Earlyvllle and Charlottesv1lle homer locatlons, respectlvely. Further, the 21—degree angle between the loop bearlngs extended from the acc1dent s1te to the radio feellltles subtend an arc equal 1n males to the dlstance between the homers.

Both alrcraft altlmeters were found. The rlght 1nd1cated an altltude of 9,200 feet and was set to a barometrlc settlng of 30.h3. The left 1nstru, ment showed about 2,520 feet and a barometrlc settlng of 30.h0. The elevation of the accldent locatlon was 2,600 feet. The last altlmeter settlng glven the fllghp was 30 h7. Other fllght lustruments were damaged to such an extent that no useful 1nformatlon was obtalned.

An exhaustlve fllght check was made of all of the varlous nav1gatlonal ground fac1llt1es pertlnent to Fllght 3h9. The checks were made to determlne what, 1f any, condltlon existed whlch mlght have led the fllght 1nto the moun— tain, or 1f the fa01llt1es were functlonlng wlthln FAA operating standards.

To thls end nearly 25 hours were flown on the fa0111t1es. The checks were flown shortly after the acoldent and at varlous tlmes, nlght and day, there— after. hhth Board 1nvest1gators aboard, they were flown by the FAA Flight Inspectlon Branch us1ng an espe01ally equlpped alrcraft for the purpose, by the FAA Alr Carrler Branch, and by Pledmont 1n 1ts own equipment. No d1s— crepanc1es were found.

In addltlon, at the request of the Board the Federal Communlcatlons Comm1551on entered the lovestlgatlon. Wlth Speclal equlpment and expert per— sonnel, 1nvest1gatlon was made to determine 1f there ex1sted any ground phenom— ena, 1nclud1ng the operatlon of electronlc equlpment 1n the local area, whlch could adversely affect the normal operatlon of the facllltles. Slgnal strength for proper reception was measured.1n approprlate areas, a search for a reported; unauthorlzed homer was conducted, and the p0551b111ty of spurlous radlo algnals was lnvestlgated. After the work was completed, the spokesman for the FCC team sald nothlng was feund which would preclude or serlously 1mpa1r the normal opera— tlon of the approach feellltles.

An 1ntense search resulted in the flndlng of a serles of groundwitnesses who had heard a low-flylng alrcraft. Because of weather condltlons, con31st— ently described as cloudy and foggy, none had seen the airplane but a fllght pattern based upon the alrcraft englnes‘ sound was revealed. For several reasons the soundpath was attrlbutable to Fllght 3L9. Most 1mportant of the reasons was that the sound proceeded to and stopped abruptly 1n the scoldent area. Other reasons were the 001n01dence of tlme when the alrcraft was heard wlth the estlmated progress of Fllght 3h9, the knowledge that no other known alrcraft operated 001nC1dent wlth the soundpath, and to some degree the corre- latlon between the sound movement and Mr. Bradley‘s recollectlons. Because the path was 8-11 miles west of the alrport, a flnal reason was added when at least three persons on the alrport Spec1f1cally llstenlng for Fllght 3L9 stated they did not near 1t.

The alrcraft was heard by the serles of wltneSSes between 2020—20h5. The flrst of the serles were located 8—10 miles northeast of the acoldent locatlon. One of them, located on high terraln near Glbson Mountaln, stated the aircraft passed over on a westerly headlng and it was so low "1t rattled the trees.” Other wltnesses were peeltloned along a south—southwest line thich W35 appronmately parallel to but 8—11 miles west of the normal approach path from Rochelle to the alrport. Most of these mtnesses sad

the englnes sounded normal but as 1f the aarcraft were low. Several in the area of Whlte Hall, Crozet, and Afton heard the plane approachlng from the northeast. They mdlcated that from the sound the aircraft then made a turn from the southwesterly headlng to a northwesterly heading and proceeded in the directlon of Bucks Elbow Mountain. The dlrectlon of the turn was uncer— tain. One watness who described the turn and that after the turn the dimin— lBhJJlg englne sound stopped qulckly but that she heard no sound llke a crash. Another, closer to the 1:10th, heard the alrcraft proceed toward the mountam and at 20115 heard a sound llke an explosmn. Stlll another stated she had not heard the aircraft but did hear a sound like thunder of short duratlon. She placed the tme after 2035 and before 20140.

Analysis and Conclusions

Exammatlon of the wreckage of N 55 V revealed no endence of malfunctlon or failure of the alrframe or powerplants. There was no indicatlon of an :m«  fllght fire, all magor components of the aircraft were located an the immedlate crash zone, and 1t was clearly endent that both engmes and propellers were capable of normal operatlon prlor to Jmpact. There was nothing found indlcat— 1ng that an emergency ex1sted before the acc1dent. These fmdlngs, reached by examnatlon of the avallable physchal endence, were substantlated by the observatlons of Mr. Bradleyo

For reasons hereinbefore enumerated, the Board 15 of the oplnlon that the soundpath developed from the descrlptlon by several groundwltnesses was made by Fllght 31:99 From the soundpath 1nformation 1t 15 apparent the air- craft approached the acoldent locatlon on a southwesterly course approxamately parallel to the prescrlbed :Lnstrument approach path from the Rochelle anter- sectlon to the Charlottesnlle homer, but 8 to 11 miles west of the normal track. The mformat10n shows that the southwesterly course was maintained to the Crozet area 10cated 8 to IL males abeam of the dealgnated area for the mstrument approach procedure turn. It 15 clearly apparent that m the Crozet area the fllght executed a turn from 1ts southwesterly headmg to a northwest— erly heading. By 1ts amount and 1ts north-south orientatlon, this turn was comcldent w1th the turmng portlon of the procedure turn mediate-1y prior to the inbound heading of 3,42 degrees.

After the turn the fllght flew northwest for, as near as can be deter— mined, a distance of two to four rules and crashed agaanst the alde of Bucks Elbow Mountain. It crashed on approxmlately the headlng of the mbound por- tion of the procedure turn w1th the landmg gear extended. The elevation of the crash, however, was appromately 1400 feet below the altltude spec1f1ed for the procedure turn.

From all thls evidence 1t 15 most apparent to the Board that the accldent took place whlle the general maneuverlng requlrements of the mstrument approach were being flown 8 to 11 males west of the de51gnated maneuvermg area pre- scrlbed for the approach.

Because it was apparent that the fllght flew a ground track well west of the deslred track, a major mvestlgatory effort was centered on the possibility that faulty operatlon of ground nav1gatlonal and 1nstrument approach fac1lit1es caused or contrlbuted to the erroneous fllghtpath. This effort 1ncluded exhaustlve fllght and ground checks. The company and the Alr Carrler Branch of the FAA made several lnSpectlon fllghts, each nelng 1ts reSpectlve alrcraft. In addltlon, the Flight Inspection Branch of the FAA made evaluatlons cf the alrway and approach feellltles nelng spec1f1c pro- cedures and an alrcraft equlpped for the purPOSe. ThlS work revealed normal operatlon of the feellltles.

As part of thls phase the Board called upon the Federal Communicatlons Comm1ss1on to a551st. hhth epe01al equlpment and quallfled personnel the FCC team made a comprehen31ve 1nvest1gatlon coverlng a w1de range of con51dera— tlons relatlve to the performance, rellablllty, and 1nstallat10n of the homer beacons. Included were a search.for an unauthorlzed homer rumored to be 1n operatlon, an investlgatlon of electronic equlpment used by local manufactur- ers, and an evaluatlon of Slgnal strength w1th1n the operatlonal range re- qulrements spec1f1ed for the homer beacons.

Thls phase of the 1nvest1gatlon revealed no condltlon wnlch would 1mpa1r fllght conformlty along V-lhO or the executlon of a normal 1nstru- ment approach on the homer facllltles.

As prev1ousLy descrlbed, the airborne nav1gatlonal equlpment was determlned by physlcal ev1dence to have been properly tuned for an 1nstru— ment approach utlllzlng the ADF equipment on whlch the Charlottesv1lle approach was based. Because of thls 1t was of prlmary concern to the 1nves- tlgatlon whether or not the.ADF cockplt presentatlon was accurate. Most 1mpcrtant in this determinatlon were the peeltlons of the ADF loops relatlve to the crash headlng and locatlon. The extended.bear1ngs of the red and green ADF loop peeltlons passed nearly through the locatlon of the respectlve homer beacon to whlch each was tuned. Also, the angle formed between the bearlngs subtended an are at the homers equal to the dlstance between them. The Board does not belleve these loop peeltlons to be a matter of c01nC1dence but rather dlreot evldence the ADF's were functlonlng normally at the tlme of the acoldent. Furthermore, the elapsed tlme between reportlng Rochelle to the crash apparently exceeded the normal elapsed tlme from Rochelle to the 1nhound headlng of the procedure turn by several mlnutes. Bellev1ng the fllght operated 1n 1nstrument weather condltlons and made the flnal turn abeam of the procedure turn area, 1t 1s probable the turn was started wath reference to the ADF 1nd1catlons. The Board comelders 1t 1mprobable that the ADF presentation would be accurate in show1ng the alrcraft abeam of the fac111t1es and 1naccurate shortly'before this 1nd1catlon. It 15 equally 1mprobable that such 1naccuracy would be followed by an accurate presentatlon at the tlme the crash occurred.

Based on the work performed and the ev1dence found, 1t 15 the oplnlon of the Board that this accldert occurred for Operatlonal reasons. Consequently, the Board sought a determinatlon 1n tnls area whlch would aCCOunt for the

llghtpath of the aircraft oelng parallel to, but 8 to ll miles west of, the proper trace. It sought a Sltuatlon whacn could develop easrly and, because 1t 15 probable that Captaln Lavrlnc was flylng, one wrlch escaped observation by Copllot Haley. It also sought a eltuatlcn 1n Wthh the precarlous lateral error would not be readlly detested as such by either pilot. Because of a number of unxncwn elements and the 1nnerent 1ntang1bles of the operatlonal situation, it is doubtful that any analysis can determine the sequence of events with complete definitiveness. Nevertheless, the Board believes it reached a determination which best satisfies the aforestated requisites and the known factors.

It is the opinion of the Board that the laterally erroneous flightpath developed from an initial navigational error at the Casanova omni. It is believed it occurred as an omission in that the flight did not turn in conformity with the V—l4O airway from the inbound radial of 260 degrees to the outbound radial of 239 degrees, a left turn of about 20 degrees. It is believed that the flight continued on the 260-degree radial until it reached the 335-degree radial of the Gordonsville mum at a location approximately 13 miles northwest of the Rochelle intersection. It is the Board's opinion that at this location, which was believed by the pilots to be Rochelle, the flight turned left to and flew the approximate heading indicated by the flight plan and log to be flown from Rochelle, 200 degrees.

The Board has reached its opinion as to the sequence of events based upon several factors. The first was the results of an analytical time, distance, and groundspeed plot. It is probable that the flight flew about 15 minutes after reporting Rochelle until it crashed. This time, being considerably longer than the time normally required to fly from Rochelle to the inbound heading of the procedure turn, Shows a greater distance must have been flown. A plot of the probable flightpath in reverse was therefore prepared using the time flown, a reciprocal of the soundpath, and the estimated groundspeed of the DC-3. This showed that 15 minutes before the accident the aircraft would have been over an area about 13 miles northwest of Rochelle intersection.

The second phase of this work was a radius of action plot from the Casanova omni. Based on the elapsed time between the Casanova and Rochelle reports, 15 minutes, it was determined that the flight would have flown 33 miles. A line of position with a radius of 33 miles from Casanova was found to intersect the initial plot at a location which was approximately 15 minutes from the crash or again about 13 miles northwest of Rochelle.

At the completion of this work two additional Significant factors were apparent. The point of intersection of the plots was closely coincident with the 335—degree radial of Gordonsville. Secondly, the heading to the location of intersecting plots from Casanova was the Zoo—degree radial of Casanova and the same as the inbound radial to Casanova from Springfield.

At Casanova Copilot Haley made the position report and most probably recorded it in the flight plan and log. Thereafter he would be expected to tune his omni set to the Gordonsville frequency and select the 335-degree radial in order to identify the Rochelle intersection. Considering the small amount of turn required at this time, the first actions could nave diverted his attention for the period during which the captain would normally have made the turn. Tuning his omni to Gordonsville, though necessary, would also reduce his opportunity to observe by omni indications the relative position of the aircraft to the course of the airway. Additionally, there was indication that Captain Lavrinc flew with a lower than average level of instrument panel illumination. In the Board‘s opinion these factors are valid reasons in this instance for the copilot not having detected the navigational omission.

After reporting Rochelle and turning to the southwest heading it is likely that both pilots believed the aircraft was describing a groundpath west of, but only a short distance west of, the normal track from Rochelle to the Charlottesville homer. It is the Board's opinion that at this time the flight was, in fact, 13 miles northwest of Rochelle. While this position was only two to three miles farther than Rochelle from the Casanova omni, the position placed the southwest course of the flight eight to nine miles west of the specified track. The location also positioned the flight about 10 miles farther from the homer facilities than from the Rochelle intersection.

It is believed that this latter factor could work to obscure the lateral error which existed during the southwest portion of the flightpath. The greater distance from the Signal source would reduce the angular displacement of the ADF presentation. Thus, if the aircraft was pos1tioned 10 miles farther from the Signal source than it was believed to be by its pilots, the angular deflection of the ADF presentation caused by the lateral course error could be obscured considerably by the greater distance. For example, the ADF presentation 2h miles from.the Signal source and eight to nine miles west of track is not alarmingly different than the presentation 12 miles from the Signal source and three miles west of the prescribed track. Similarly, the ADF presentation 19 miles from the Signal source and eight to nine miles west of track is not alarmingly different than the presentation seven miles from the Signal source and three miles west of the track desired. In addition, as the flight progressed toward the facilities but from a greater distance than believed by the pilots, the increa51ng angular displacement of the ADF needles showing lateral error could be interpreted as a closure on the Signal source. The Board believes the foregoing discussion to be a valid consideration in the reason that the pilots were not alerted early in the approach to the large lateral track error.

On the other hand, the Board 15 aware that as the flight proceeded on the southwest course the rate of progression of the ADF needles to the left 90—degree steam indication would have been much slower as the result of the greater distance and time to be flown. At the 90—degree position the angular spread between the needles would.have been much narrower 9 to 10 Miles west of the homers than three to four miles west of the homers. Further, 90 degrees abeam of the Signal source, a 20-degree relative bearing change on a flightpath three to four miles west of the homer, would take 30 to ho seconds as contrasted to approximately 1—1/2 minutes on a flightpath about nine miles West of the homer. In addition, the ADF presentation during the period the flight turned from the southwest heading until it struck the mountain would have been incompatible with a close-in position. The Board believes that these factors should have served to alert an attentive pilot that the lateral course error Was of considerable magnitude.

During the course of the investigation the aeronautical history of Captain Lavrinc was reviewed. His training, qualifications, and proficiency reports were satisfactory. His history showed that he had progressed normally to captain and had served in that capacity Since may 1957. It also showed that he had flown in and out of Charlottesville and over the route involved for several years on a regular basis. Captain Lavrinc had flown a total of 5,101 hours, of which 4,77l were in DC—3 aircraft.

To the Board there were numerous factors which were obViously inconSistent with Captain Lavrinc's record. Some were. The apparent naVigational omission, a nonadherence to precise tracking procedures, and a descent below the authorized procedure turn.altitude. Others were The failure to note that the time for station passage was in excess of that commensurate with a close-in peeition, and that ADF indications were not compatible With the normal procedure turn presentation. Still others were a failure to request the latest Charlottesville weather when the Communicator did not furnish it, and not uSing the altimeter setting given in response to the inrange report. The Board believes these factors were not only inconSistent with Captain Lavrinc's reputation as an exacting pilot but were indicative of a serious departure from the high standard and quality of performance expected during an instrument operation. Because of these factors a comprehenSive investigation was made into the personal background of Captain Lavrinc. This was done to search for reasons which could seriously impair his normal piloting ability. Durnng this work reasons were found which could result in his preoccupation.

Captain Lavrinc had, for several years, been under severe emotional strain. The Board conSiders that disclosure of detailed information relating thereto would adversely affect the interests of certain persons and is not required in the interest of the public. A resume of the Board's significant findings and certain recommendations, however, are in the public interest and are set forth below.

Captain Lavrinc received psychotherapy in l953—l954; he obtained further psychiatric counseling in 1957; intenSive psychotherapy was resumed in May 1959, which he underwent several times a week thereafter; his last appointment was the night before the acc1dont. This latter treatment involved the services of two psychiatrists. In the course of this treatment the first psychiatrist prescribed certain psychotropic drugs. After trials on CompaZine, Presine, Sparine, and ThoraZine, Drosine was prescribed 1r Augusr 1959 in a dosage of three or four times Jail; and was reicsued on September 18, 1959. This prescription specif- ied an amount which, if taken as directed, would have been sufficient to last urtil tro days before the acCident.

On September 23, 1959, however, Captain Lavrinc commerred psychotherapy urder the second psychiatrist who prescribed no drugs. The Board has been arable to determine whether or not Captain Lavrinc continued to take the medicine in the prescrioed manner curing the latter treatments, although there is eVioence that he took the earlier trial prescriptions.

The Board has evaluated the background and history of Captain Lavrinc, including data set forth above, In afldltlQi, it submitted all the availaole information covering Captain Lavrinc to particularly qualified medical experts for evaluation as to its Significance with respect to this acc1dent.

The consensus is that Captain Lavrinc was so heavily burdened with mental and emotional problems that he should have been relieved of the strain of flight duty while undergoing treatment for his condition. Tris condition was such that preoccupation With his proolems could well have lowered.his standard of per— formance during instrument flight. Furthermore, with respect to this acCioent the consenSie 18 that the emotional and mental problems were of far greater importance in cauSing the preoccupation that could have the use of psychotropic medication.

The Board belleves that the facts dlsclosed by th1s 1nvest1gatlon demonstrate the adverse effects of serlous emotional and mental stress on alrman prof1c1ency and performance. It further belleves that the early recognitlon and correction of such condltlons Wthh mlght tend to 1mpa1r an alrman's prof1c1ency and performance would be benef101al to flight safety. Accordlngly, the Board recommends that the Federal Av1at10n Agency, approprl- ate segments of the avlatlon 1ndustry, and the medlcal profe551on 1n1t1ate exploratory studles 1n thlS fleld.

The Board also Con51ders that the 1nvestigatlon of thls acoldent demenstrates the need for reexaminatlon of the use of drugs Wthh may affect the facultles of a fllght crew member 1n any manner contrary to safety.

Slnce world War II there have been great advances 1n pharmacology and whole new familles of drugs have become ea51ly'avallable to the publlc, elther over the counter or by prescriptlon. Slnce l953-195h one of the most Slgnif1- cant advances has been 1n the fleld of psychopharmacology. There has been a prollferatlon of drugs whlch 1nfluence the state of mand, are employed 1n the treatment of mental dlsorders, or are used as psychlc energlzers. Withln thls group of drugs the so—called tranqulllzers are belng uldely used by the publ1c.

The basic question which the Board believes must be resolved, therefore, is how does the use of these drugs relate to the safety of flight. For example, within the framework of the present Civil Air Regulation covering the use of drugs, [5] should these drugs be classified as "…drug which affects his (crew member) faculties in any manner contrary to safety…" The Board is of the opinion that the answer to the question is a qualified "yes." In great part this decision is reached from review of military research into the relationship of drugs to the flying profession. The basic conclusion derived from this research can be stated quite simply: If a flight crew member's personal situation demands tranquilizers he should be removed from flying status while on the drugs.

Probable Cause

The Board determines that the probable cause of this acc1dent was a navigational omission which resulted 1n a lateral course error that was not detected and corrected through precision instrument flying procedures.

A contributing factor to the accident may have been preoccupation of the captain resulting from mental stress.


Robert T. Murphy, Vice Chairman, did not take part in the adoption of this report.


Investigation and Hearing

The Civil Aeronautics Board was notified of the accident shortly after the flight was presumed to have crashed. Board investigators were immediately dispatched to Charlottesville and participated in the search and rescue activity until the aircraft was located. Thereafter, an investigation was initiated and conducted in accordance with the provisions of Title VII of the Federal Aviation Act of 1958. A public hearing was held in conjunction with the investigation in Charlottesville, Virginia, December 10–11, 1959.

Air Carrier

Piedmont Airlines is the Airline Division of Piedmont Aviation, Inc. The company is incorporated in North Carolina with its principal offices in Winston-Salem, North Carolina. The Piedmont Airlines Division of the company was established in 1947. It operates under a currently effective certificate of public convenience and necessity issued to the company by the Civil Aeronautics Board and an air carrier operating certificate issued by the Federal Aviation Agency. These authorize the company to transport by air persons, property, and mail over various routes including the one involved in the accident.

Flight Personnel

Captain George Lavrinc, age 32, was employed by Piedmont Aviation January 23, 1950. He was employed in the radio department and transferred to the Airline Division November 24, 1951. He was promoted to reserve captain on May 1, 1957. Captain Lavrinc held a valid FAA airline transport pilot certificate with a DC-3 aircraft rating. Company records showed he had flown 5,101 hours, of which 4,771 were in DC-3 aircraft. His last first-class medical mas satisfactorily accomplished October 13, 1959. His last semiannual proficiency flight check was satisfactorily accomplished May 26, 1959.

First Office Bascom L Haley, age 27, was employed by Piedmont Airlines on May 2, 1957 He held a valid FAA commercial pilot certificate with an instrument rating. According to company records he had accumulated 2,858 hours, of which 1,678 were in DC-3 aircraft. His last first-class medical examination was satisfactorily accomplished May 12, 1959.

The Aircraft

DC-3, N 55V, was manufactured June 2, 1944, and procured by Piedmont Aviation on January 24, 1955. It had been flown a total of 26,339 hours. The aircraft had been flown 83 hours since the last No. 4 inspection. The engines were Pratt and Whitney, model R-1830092, equipped with Hamilton Standard model 23E50 propellers.

CAB Accident Report, Piedmont Airlines Flight 349.pdf

(Upload an image to replace this placeholder.)

This work is in the public domain in the United States because it is a work of the United States federal government (see 17 U.S.C. 105).

  1. All times herein are eastern standard based on the 24-hour clock.
  2. Altitudes are mean sea level unless otherwise stated. Weather reports of ceiling and cloud levels are feet above the ground.
  3. Reporting times were determined by timing the recordings of the transmissions and are accurate within one minute. Voice identification from the recordings showed the radio transmissions were made by First Officer Haley, thus indicating that captain Lavrinc was flying the aircraft.
  4. Distances are nautical miles.
  5. Section 43.45. Use of Liquor, Narcotics, and Drugs. No person shall pilot an aircraft or serve as a member of the crew while under the influence of intoxicating liquor or use any drug which affects his faculties in any manner contrary to safety. A pilot shall not permit any person to be carried in the aircraft who is obviously under the influence of intoxicating liquor or drugs, except a medical patient under proper care or in case of an emergency.