The right wing separated on the seaplane shortly after takeoff from Miami, and the crash into the coastal water killed all 20 aboard. The wing was recovered from the water on Dec. 20 and National Transportation Safety Board (NTSB) engineers and metallurgists quickly observed that the wreckage indicated fatigue fractures (see photos this page). As an immediate result, Chalk's grounded its remaining fleet of Gulfstream G-73 airplanes. The Federal Aviation Administration (FAA) did not order grounding, preferring instead to advise operators of what happened (see box page 24). This December 23 advisory was followed a week later by an emergency airworthiness directive (AD) ordering inspection of the affected structure and reporting the results to the FAA in order for the agency to have a better appreciation of the extent of the problem (see box page 24). While Frakes Aviation (formerly Grumman American) is developing a special detailed (i.e., nondestructive) inspection within 45 days, the AD must be complied with immediately if any flying is to be done.

Examination of the wreckage

The wreckage from the accident aircraft reveals that the failure likely started at the location where two offset drill holes are located. Inspecting other airplanes for such holes seems a prudent first order of business. The term "offset drill holes" is perhaps a gentlemanly way of referring to a carelessly drilled second hole that ruins the first. When this occurs, the appropriate step is to open the hole to a larger size so that only a clean new hole is left. This can be a tedious procedure with little forgiveness and the potential for ruining any of the holes in the mating surface.

There are limits to this stratagem. If the standard engineering criteria for hole edge distance to the edge of the structure (usually three to four times the fastener diameter), or to any next fastener hole edge, cannot be met, and a doubler cannot be utilized, normally the fitting must be replaced.

When the fasteners (bolts, Hi-Locs or rivets) are reinstalled, the sloppy repair ("offset drill holes") is hidden until cracks develop and migrate outwards. From the shading in the cross-sectional view on the NTSB photograph, it appears that the cracking existed for some time, as evidenced by the large area of darker colored metal. The emergency AD calls for inspection of the entire area.

The drill holes

The fracture photographs suggest also that stress was raised as a result of the offset drill holes, and salt-water corrosion would exacerbate the structural weakening. Corrosion can occur anyplace, but it can manifest especially in drilled holes after production, where the surface may have a coarse texture (as in an improperly drilled out hole).

The two different sides of the hole shown in the NTSB photographs appear to have two different types of fatigue initiations at their surface, which could possibly constitute abusive machining or corrosion on one side and secondary initiation on the other.

What is evident are two separate drill holes, not parallel to each other and apparently merging together. It would seem that such an action would create a stress concentration.

The inexpert drill work should raise an obvious question in the engineering community's mind, specifically, why are there two offset drill holes in this particular location which have removed maybe 15 percent or more from the horizontal leg of the spar cap?

The resulting fatigue induced failure (whether corrosion-related or not) would seem to have inevitable consequences in just a matter of time and flight cycles after those offset drill holes were inexpertly made.

Not an isolated event

Recall that there have been spar problems recently on the Beech T-34, as evidenced by the fatal crash November 19, 2003, of a Texas Air Aces T-34, when the right wing separated during aerial maneuvers. Also, several firefighting aircraft were found to have cracked or broken spars. As in the case of the G-73, these aircraft were designed many years ago, before the awareness of metal fatigue that prevails in the design community today.

Although this case represents an amphibian operating in a salt water environment, the case is significant for revealing that, like it is said about rust, corrosion and its attendant metal fatigue "never sleeps," it must be guarded against constantly, and offset drill holes can accelerate the inevitable consequences.

Commentary:
Arm Chair View - Chalk G-73
Grumman Mallard Fatal Accident

Currently, 27 G-73 and five G-73T (turboprop version), for a total of 31 total, Frakes Aviation (formerly Grumman American) amphibian Mallards are registered in the U.S. One of the five G-73Ts, all operated by Flying Boat, Inc., doing business as Chalk's Ocean Airways, crashed at Miami, Fla., on Dec. 19, 2005, shortly after an afternoon water take-off on a scheduled airline FAR 121 flight, number 101, to the Bahamas.

All indications point to a structural failure of a right wing in flight with the resulting loss of all 18 passengers and a flight crew of two. Most likely the wing failure originated from stress corrosion and/or fatigue site(s) somewhere along the wing structure. Why the operator's inspection program did not find this problem during earlier inspections is the big question for now.

Long history

In the mid-1940s the aircraft was designed by Grumman originally for use as a 10-seat utility amphibious transport but few G-73s were ever used in airline service.

For many years Chalk's has been the only airline operating the G-73T in scheduled service. The other G-73/G-73Ts, those still in flying condition, are in private use and don't accumulate many of flight hours and are probably not as "aged" as the Chalk's fleet.

A total of 59 Mallards were built between 1946-51 and the worldwide total surviving number is probably well under 30. The accident aircraft, registration number N2969, was built in 1947 and had accumulated over 31,000 flight hours and over 39,000 flight cycles.

These remaining aircraft are now all approaching 60 years in age. The G-73 original type certificate (design standard, CAR-4a, Transport) of Grumman was approved in 1947 by the FAA (then known as the Civil Aeronautics Administration). Some years ago Grumman transferred its type certificate to Frakes Aviation of Cleburne, Texas. In effect, this could have slackened the product support of the Mallards. Any service or aging technical, mechanical, or structural related problems with the Mallard may not have reached Frakes with the same accuracy, frequency or urgency as before.

Another factor possibility impeding the flow or impact of bad news to a type certificate holder is that individual owners don't carry as much clout as an airline would. It is doubtful the product liability for the G-73/G-73T still rests with Grumman. A few years back a U.S. federal law set a "statute of repose" (18 years) concerning aircraft product liability of general aviation manufacturers but it is questionable if it applies to a transferred type certificate holder, Frakes.

Previous modifications

In 1984, a modification was made to the accident aircraft increasing its passenger capacity to 17, adding an emergency exit, and changing its floor. Earlier in the 1970s another modification was made replacing the aircraft's two radial piston engines with more powerful and reliable PWA PT-6 turboprop engines. Both these modifications, called supplemental type certificates (STCs) and held by Frakes Aviation, were FAA approved.

In the history of the G-73T Mallard the FAA has issued only two airworthiness directives (ADs), in 1998 and 1999, requiring mandatory changes to the aircraft flight manual. Four other ADs were issued against the G-73 from 1948 through 1997 and affected the G-73T but, again, these ADs did not involve the wings. So few ADs speak well of the G-73/G-73T design.

Corrosion Conundrum

Amphibious aircraft are much more prone to corrosion, especially in warm salt-water environments, than other types of aircraft. A near-60 year old G-73/G-73T in frequent amphibious service (read Chalk's) would definitely require an unrelenting, first-rate maintenance/inspection program carried out by a dedicated, capable, informed and vigilant maintenance staff in order to insure a continued safe, airworthy operation.

Reportedly, Chalk's changed ownership around 2000. It is not known if the maintenance personnel of the previous owner stayed on. There are probably only a handful or less of well-experienced G-73/G-73T maintenance/inspection program management personnel in the world today. It is to be seen if Frakes Aviation was up to the task.

Without a strong manufacturer's product support effort and given a relatively young or inexperienced operator maintenance staff, an "aging or well-aged aircraft" will likely eventually meet a catastrophic end. Once the NTSB provides the exact site location of the apparent wing failure and the remaining G-73/G-73Ts are inspected, now with greater care, it will only be then that the true nature of failure will become apparent--either (1) a new emerging structural distress area not known by the industry before, or (2) poor or superficial inspections were carried out previously, or (3) the inspection program was deficient or outdated. My guess is the cause will have elements of all three.

In the meantime, the FAA issued emergency AD-2006-01-51, requiring before further flight an interim intensive visual inspection of the wings and reporting the results to the FAA. Frakes Aviation is expected to soon develop a nondestructive inspection procedure and then likely a revised AD will be issued.

Whatever the eventual determined root cause of the accident, it will most likely fall into the category of aging aircraft. Perhaps the title of this article should be changed from armchair view to "old rocking chair view."

Bart Crotty is an aviation airworthiness and maintenance consultant, and a former FAA inspector. e-mail bjcrotty@verizon.net.

Not Sure What To Do

Extracts from FAA's Special Airworthiness Information Bulletin, No. NM-06-18, of Dec. 23:

"On Dec. 19, 2005, a G-73 modified by STC [supplemental type certificate] ... to increase the passenger count [from 12] to 17 experienced a wing separation on take-off ... The wing separated between the fuselage attachment and the engine attachment. Preliminary indications ... reveal occurrences of fatigue failure of a wing spar. This particular airplane was operated in passenger service and in a salt-water environment.

"At this time, the initiating cause of the failure is unknown. The FAA is closely monitoring the investigation to determine the existence of a condition on the accident airplane that may be manifest on other G-73 airplanes in the existing fleet.

"At this time, the FAA is not able to identify an inspection or other action that can be taken to determine whether other aircraft may have the same condition.
FAA is considering the need for mandatory detailed inspections of parts or all of the wing structure on the fleet of G-73 airplanes." Source: FAA

More Sure of What To Do

The FAA issues an emergency airworthiness directive (AD 2006-01-51) calling for inspections and reports of the findings (extracts):

"[P]reliminary indications from the investigation reveal occurrences of fatigue cracking of a wing spar, skin cracking, and a broken z-stringer.

"The loss of the lower skin capability, or the spar and stringer capability, will likely lead to wing failure. This condition, if not corrected, could result in structural failure of the wing and loss of control of the airplane. ...

"Therefore ... this AD requires (1) a detailed visual inspection to detect repairs, cracking, or corrosion of the wings from wing station (WS) 77L to WS 77R, front spar to rear (main) spar; (2) removal of repairs, if found, to allow for inspection of the wing structure underneath the repairs; (3) removal of sealant from the interior of the wet bays to allow for inspection of the skins, stringers, and both spars; and (4) repair of any crack or corrosion.

"This AD also requires sending the inspection results (both positive and negative) to the FAA. ...

"This AD is considered to be interim action. The inspection reports that are required by this AD will enable the FAA to obtain better insight into the nature, cause, and extent of the cracking, and eventually to develop final action to address the unsafe condition." Source: FAA