Who Benefits from Airworthiness Directives?

AirworthinessDirectives have been with us since the 1940's, butit seems as if the industry's attitude toward them has changeddramatically in recent years. Twenty years ago, manufacturerswould often brag in their sales literature that their airplanesor engines "have never had an AD issued against them."Issuance of an AD was considered a black mark against a manufacturer,suggesting that the firm's engineering or quality control wasnot up to par.

In the last decade, that attitude appears to have made a 180-degreeturn. Nowadays, it seems as if most ADs that come out of the FAAare being issued at the request of the manufacturers. Sometimesfirms ask the FAA to issue an AD at the advice of their lawyersin an effort to limit their liability exposure. In other instances,their requests are prompted by honest-to-goodness safety concerns.

But we're troubled by an increasing incidence of manufacturerrequests for ADs that look suspiciously as if they're intendedto enhance the company's revenues and/or hurt their competition.And what's really disturbing is that the FAA seems to be goingalong with these requests, issuing ADs without first doing independentdue-diligence to establish that they are justified by the facts.

That's what apparently occurred with two recently-proposed crankshaftADs. In two rulemaking actions last year, the FAA declares thatthey've identified an "unsafe condition" affecting thecrankshafts of 60,000 TCM and Lycoming engines in many of theworld's most popular light airplanes, and says it wants at least20,000 of those crankshafts to be scrapped and replaced, at acost to owners of $50 to $100 million. Ouch!

Now, before taking this kind of drastic action, you'd think theFAA would need to be pretty doggone sure that there is a realsafety problem with the existing crankshafts and that the replacementcranks actually solve the problem. But it sure looks to us asif the FAA simply took the word of TCM and Lycoming that thereis a problem...and that neither the FAA's own statistics on crankshaftfailures nor independent studies support the manufacturer's claims.To make matters worse, we don't think it's an accident that TCMand Lycoming have adopted pricing policies on replacement crankshaftsthat will compel most affected owners to exchange their enginesfor factory remans, to the severe detriment of field overhaulshops who nowadays are the principal competitors of TCM and Lycoming.In short, we smell a rat.

TCM has used two different processes to cast the steel billetsused in forging their crankshafts. For decades, the cranks weremanufactured using a conventional "airmelt" process.But in the late 1970s, TCM became concerned about metallurgical"inclusions"-tiny air bubbles trapped inside the steelas the billet is poured-which cannot be detected through ordinarymagnaflux inspection and which might weaken the cranks. So between1978 and 1980, TCM gradually switched over to a different processcalled "vacuum arc remelt" (VAR) to eliminate the possibilityof inclusions. They also increased the crankshaft diameter slightlyon their 520-series engines. However, they continued to re-useserviceable airmelt crankshafts in their factory-rebuilt engines.

It is very rare for a crankshaft to fail because of a materialor manufacturing defect. According to FAA Advisory Circular AC20-103,the vast majority of crankshaft fractures are the result of undetectedor unreported prop-strike damage, out-of-balance propellers, spunbearings, excessive oil temperatures, contaminated oil, and otherlubrication failure situations.

Nevertheless, TCM became convinced that a small number of crankshaftfailures in their 360- and 520-series engines were due to "unexplainedsubsurface origin fatigue cracks initiating below the intermediatemain bearing fillets" which could not be detected throughmagnaflux inspection. So in 1987, TCM issued Service BulletinM87-5 which recommended (but did not require) that crankshaftsbe inspected at overhaul using ultrasonic test equipment capableof detecting subsurface cracks. Magnaflux inspection continuedto be required.

TCM also persuaded the FAA to issue an Airworthiness Directive,AD 87-23-08, requiring ultrasonic inspection of IO-520 and TSIO-520crankshafts to detect subsurface cracks, and replacement of thecrankshaft if a crack is found. The impact of this AD was relativelysmall: the extra inspection added about $200 to the cost of amajor overhaul or post-prop-strike teardown inspection, but fewerthan 1% of the crankshafts flunked the ultrasonic exam.

Five years later, TCM decided they wanted to get all the olderairmelt cranks out of the system. In September, 1992, they issuedMandatory Service Bulletin M92-16 which required that all airmeltcranks (or cranks whose manufacturing process could not be determined)be scrapped and replaced with a VAR crankshaft at the next majoroverhaul or teardown. About 25,000 airmelt-crank engines wereaffected. They also announced that all factory-new and factory-rebuiltengines would be equipped with VAR cranks.

New VAR crankshafts have list prices which vary from $6,700 to$8,200. However, TCM announced a special exchange program wherebyold airmelt cranks could be traded for brand new VAR cranks fora mere $2,200. This was presumably TCM's way of trying to persuadeoverhaul shops to comply with M92-16.

But then a funny thing happened. TCM quietly established a policyof accepting runout airmelt-crank engines as cores in exchangefor factory-rebuilt engines with no up-charge. This meant thatthe owner of an airmelt-crank engine would pay a penalty of atleast $2,200 on a field overhaul, but no penalty at all on a factoryreman. Needless to say, the overhaul shops and their trade association,the Aeronautical Repair Station Association (ARSA), cried "foul!"

TCM responded by going to the FAA and asking them to put "teeth"into M92-16 by turning it into an Airworthiness Directive. TheFAA agreed, and in June 1993 issued a proposed AD (58 FR 39748)that would supersede AD 87-23-08, eliminate its ultrasonic inspectionrequirements, and mandate replacement of airmelt crankshafts withVAR cranks anytime the case was split. The proposed AD would alsoexpand the affected population of engines to include 360-seriesengines as well as 520-series.

The FAA was inundated by negative responses to its NPRM, mostlyattacking the FAA's lack of data justifying the need for the ADand disputing the FAA's estimate of economic impact as unrealisticallylow. The principal commenter was ARSA, who claimed the FAA's crankshaftfailure data was anecdotal and not representative of the entirefleet.

The wheels of FAA rulemaking grind slowly. Two years later inAugust, 1995, the FAA issued a supplemental NPRM (60 FR 43995)to address the criticisms by ARSA and others by publishing additionaldata in support of the proposed AD. In the supplemental notice,the FAA dismissed its own Service Difficulty Report (SDR) databaseas inaccurate and elected to rely on TCM-supplied failure analysis.According to this analysis, there are approximately 18,000 airmeltand 25,000 VAR cranks in service in 520-series engines as of 1994,and that between 1986 and 1994 there were 40 failures of airmeltcrankshafts but only 3 failures of VAR cranks. The implicationwas that airmelt cranks were 15 times more likely to fail thanVAR cranks. For 360-series engines, the numbers quoted were 5,000airmelt cranks and 10,800 VAR cranks in service, and 8 airmeltfailures but only 1 VAR failure during the same period. Again,a failure ratio of more than 15 to 1. Pretty convincing, wouldn'tyou say?

Not so fast, said ARSA. Two weeks after the supplemental noticeissued, ARSA submitted a Freedom of Information Act request tothe FAA, asking for the source information (primarily TCM warrantyclaims) on which the FAA justification was based. The FAA declinedto provide this data, indicating that it was proprietary to TCM.

In October, 1995, ARSA joined forces with three other alphabetgroups (the Aircraft Owners and Pilots Association, the CessnaPilots Association, and the National Air Transportation Association)to file a response in opposition to the supplemental NPRM. Theycriticized the FAA for relying on data that has not been madeavailable for public scrutiny, and suggested that TCM's data mightbe "presented in such a way as to favor the commercial goalsof the party that compiles it (such as selling more crankshafts)."

They also pointed out that the FAA elected to ignore its own SDRdatabase (which did not support the need for the proposed AD),and also ignored the analysis submitted by ARSA in February, 1994.ARSA's analysis was based on the results of ultrasonic testingof crankshafts required by AD 87-23-08 by TCM-authorized inspectionfacilities throughout the U.S. The ARSA data showed that companieshad detected 27 subsurface fatigue failures out of 3,821 airmeltcrankshafts subjected to ultrasonic test, for a failure rate of.707%, and had detected 5 failures out of 488 VAR cranks tested,for a failure rate of 1.025%. ARSA pointed out that these failurerates are statistically equivalent.

Furthermore, the joint Association response points out that theFAA's own supplemental notice "exhibits a lack of confidencein the superiority of VAR crankshafts." The proposed AD requiresthat VAR crankshafts undergo ultrasonic testing at every removal,so that "clearly the FAA recognizes that VAR crankshaftsare subject to the same unsafe condition as airmelt crankshafts.The FAA has offered no evidence to conclude that unsafe conditionsexist in airmelt crankshafts that do not exist in VAR crankshafts.The two types of crankshafts should be treated in a comparablefashion."

The Cessna Pilots Association (well known for its activism inopposing oppressive ADs) analyzed Cessna production figures, anddetermined that Cessna delivered aircraft with approximately 30,000airmelt crankshafts versus only about 8,500 VAR crankshafts. Assumingthat roughly the same percentages hold true for Beech aircraft,CPA concluded that the TCM-supplied population estimates thatthe FAA relied upon (18,000 airmelt vs. 25,000 VAR) "areeither greatly in error or indicate that the airmelt crank willdisappear without the need of regulatory action."

That's where things stand as this article is being written. Thenext move is up to the FAA. They may issue an AD, issue yet anotherNPRM, or drop the matter altogether.

Realistically, what the FAA does may not make any real difference.TCM's Mandatory Service Bulletin M92-16 already has the forceof law on Part 135 operators. Furthermore, no reputable engineshop would fail to comply with such a mandatory service bulletinbecause of potential liability concerns. So with the AD or withoutit, airmelt crankshafts are going on the scrap heap.

TCM has no obligation to hold the line on its $2,200 special trade-inprice, and in fact has already increased it once to $2,330. TCMcontinues to accept airmelt cores in exchange for factory remanswith no up-charge, making the field overhaul shops the big losersin this affair.

A strikingly similar scenario is now playing out with respectto four-cylinder Lycoming engines that drive fixed-pitch props.At Lycoming's request, the FAA has proposed an Airworthiness Directivethat would require the inspection and possible replacement ofthe crankshafts in an estimated 46,000 U.S. aircraft.

The story began in October 1993, when the British CAA receiveda report of a Piper PA-28-161 Warrior that made a forced landingdue to a crankshaft failure that caused the propeller to separatefrom the aircraft. Analysis of the crankshaft indicated that thefailure was caused by "a high cycle reverse torsional fatiguemechanism that had initiated from a number of corrosion pits inthe crankshaft bore." Ya just gotta love those Brit engineers!The Warrior's Lycoming O-320-D3G engine had 1,950 hours SMOH andthe crankshaft had 4,429 hours and 15 years since new.

In plain (American) English, here's what investigators discovered.Crankshafts designed for use with constant-speed propellers arehollow in the front, allowing oil pressure from the governor toreach the propeller hub and adjust the blade pitch. Lycoming electedto use the same sort of hollow-front crankshaft for most of itsfixed-pitch prop installations, simply installing an expansionplug in the front of the crankshaft to retain the oil. That'sapparently where the problem lies: in engines with the plug installed,the oil in the crankshaft hollow is stagnant, and crankshaft rotationcentrifuges this trapped oil and can create a buildup of acid-richsludge on the inner surface of the crankshaft bore. The sludgepromotes internal corrosion, and the corrosion pits can becomestress-risers that ultimately lead to crankshaft fracture nearthe propeller flange.

As a result of the British Warrior incident and several otherreports, Lycoming issued Mandatory Service Bulletin 505A in October,1994, affecting all Lycoming 235- and 290-series engines and mostLycoming 320- and 360-series engines with fixed-pitch propellers.The SB requires inspection of the plugged-up crankshafts immediatelyif they are more than ten years old, or within ten years of shipdate if they are recent-vintage. The mechanic must remove theprop and the plug, clean the forward four inches of the crankshaftbore, and inspect the surface for corrosion. Surface corrosionmay be removed, but only if the inner diameter of the crankshaftremains 1.91" or less. If further corrosion or pitting isfound, the SB calls for the crankshaft to be removed from serviceimmediately (which would require a preemptive engine teardown).New crankshafts for these engines cost between $4,500 and $6,500apiece.

Lycoming also asked the FAA to issue an Airworthiness Directiveagainst the crankshafts in these engines. The FAA agreed, andpublished a proposed AD (60 FR 58580) in November, 1995. The proposedAD is somewhat more lenient that Lycoming's SB, allowing crankshaftsto remain in service for five years after pitting is found, providedrepeated fluorescent penetrant inspections reveal no cracks. Butfive years after pitting is detected, or anytime FPI reveals cracks,the AD would require that the crankshaft be replaced.

The proposed AD requires initial inspection within six monthsor 100 hours for engines with at least 1,000 hours, or withinten years if that is reached first. Thereafter, visual inspectionsare required every five years if no pitting is found, or FPI every100 hours after pitting is detected.

How many crankshafts will exhibit pitting and need to be replaced?It depends who you ask. In the cost analysis portion of the proposedAD, the FAA estimates that 10% of the cranks will be pitted andrequire replacement at each overhaul. But the Lycoming factoryclaims that most of the cranks they see coming back ascores are pitted and need to be scrapped. If Lycoming is right,then the cost impact of the proposed AD will be an order of magnitudebigger than what the FAA claims it is.

For rulemaking action with a potential cost impact of $15 to $50million (depending on whom you believe), the proposed AD is shockinglydevoid of supporting data. The only justification cited by theFAA is the British Warrior forced landing, plus ten additionalreports allegedly received by the FAA but about which absolutelyno details are offered.

The proposed AD is opposed by AOPA and Cessna Pilots Association,among other commenters. AOPA filed a FOIA request demanding toknow what data the FAA used to justify the five-year limitationon crankshafts which are found to be pitted but not cracked. Failingto get a substantive response, AOPA requested a two-month extensionto the original January 29, 1996, deadline for public comments.

The Cessna Pilots Association (CPA) got involved because the Cessna152s and 172s represent the single largest group of aircraft thatwould be affected by this AD. CPA filed a FOIA request askingfor details of the "ten additional reports" that theFAA says they used to determine that an unsafe condition existsthat justifies the AD. The FAA replied that they could not providethis information because it was proprietary to Lycoming! (Soundfamiliar?)

CPA also asked the FAA to provide SDR reports filed during thepast five years that involved flange-end crankshaft failures inLycoming O-235 and O-320 engines (as used in the C152 and C172,respectively), and the FAA did so. The SDR database revealed someinteresting facts. There was not a single report of a flange-endcrankshaft failure in O-235 engines, although there were somefractures back in the solid portion of the cranks (probably dueto lubrication failure).

For O-320 engines, there were five reports of flange-end crankshaftproblems in five years, a very modest number. Only three of thosereports involved actual crankshaft failures; the other two involvedoil leaks that were ultimately traced to crankshaft cracks. Oneof the three crankshaft failures involved an aircraft that hadobviously had a prop strike. In a second failure report, the reportingA&P stated that he could not rule out the possibility thatthe aircraft had an unreported prop strike (making it sound asif he suspected as much). Even discounting all of this, five reportsof failures in five years from a population of more than 20,000engines would make the O-320 crankshafts among the least failure-pronein the fleet.

Most significantly to CPA, not one of the five SDRs involved Cessnas!Three involved Piper PA-28-161 Warriors just like the Britishaircraft that started the whole thing, one involved a Piper PA-18Super Cub, and the last one involved a Socata TB-9 Tobago. Yetthe largest group of owners that would be affected by the proposedAD are Cessna owners.

Keep in mind that Lycoming is not offering a redesigned solidcrankshaft to replace the hollow ones that they claim are corrodingaway from the inside. The new cranks will be subject to the samecorrosion problem as the ones they replace. Also, Lycoming isoffering no "special deal" on crankshafts. Owners whoneed a new crank will have to pay $4,500 or more to get one. Butjust as with the TCM situation, Lycoming is not charging any up-chargefor pitted-crank cores that are exchanged for Lycoming factoryremans. So it would appear once again that the field overhaulshops are the big losers here, and Lycoming stands to be the bigwinner.

When the general aviation market imploded in the early 1980s,both TCM and Lycoming went through a period of denial, waitingfor their OEM market to come back to life. But by 1990, both companiesrecognized that this just wasn't going to happen. They realizedthat they were now basically in the overhaul business, and thattheir competition was the field overhaul shops.

Both companies responded to this revelation by slashing the pricesof factory-rebuilt engines while simultaneously hiking parts prices,putting field overhaulers at a tremendous disadvantage and drivingmany of them (including such well-known names as Mid-States, Schneck,and Western Skyways) out of business. The overhaul shops thatmanaged to survive reacted by buying cylinder assemblies and everythingelse they could get from PMA firms like Superior and ECI, andby reducing their need for new parts by reworking old ones wheneverpossible.

All of this is arguably in the great American tradition of a laissez-fairefree-market economy. But when manufacturers like TCM and Lycomingstart using a government agency like the FAA as a weapon to gaincommercial advantage over their competition, it's time to blowthe whistle. Perhaps you can't blame the manufacturer's for trying.But the FAA has an obligation not to issue an AD without firstindependently verifying that there's a genuine threat to safetythat the AD will rectify. From what we see, the FAA isn't doingits job. Toot toot!