RUDDER SPAR CRACKS

HISTORY OF THE RUDDER SPAR CRACKS

In 1989 the Airline Training Center in Tucson contacted Beech with the news that one of their fleet of fifteen F33As had a broken rudder. That is, the upper hinge had torn out the portion of sheet aluminum spar to which it was attached. Later the ATC found 5 more rudders in their fleet that were progressing to failure, with cracks under the upper hinge and some under the middle hinge. Two more were reported to the FAA. Thus a total of 8 partial failures had been discovered out of a sample of 17, and there were about 4,300 Bonanzas out there with the same rudder design!

The Service Bulletin

Beech performed a "statistical" study of the 17 inspected aircraft. The 8 cracked rudders had flown between 1,200 and 5,000 hours each. This tiny sample became the basis for Mandatory Service Bulletin #2333 issued in October 1989 to owner/operators of about 4,500 model 33 and 36s that were currently flying with the same rudder hinge supports. For aircraft with more than 1,000 total flight hours it called for dye-penetrant inspections of the rudder spars within 50 flt. hrs. and at every annual thereafter or at 500 flt. hrs, whichever comes first. The lowest flight time among the sample of 8 cracked rudders was 1,200 hrs. and from this it was reasoned that the other 4,500 flying rudders would be safe for the first 1,000 flight hours! What do you Flying Physicians think of this epidemiology?

The Follow-up AD

Eight months later the FAA published an AD, with the words of the MSB re-arranged; they repeated "500 hrs." but with a different meaning, and they left out the annual inspection requirement. (Compare the MSB2333-1 with the current AD93-24-03; I think you will agree). To illustrate an example of the danger in these errors, an aircraft that averages 100 hrs. per year must be inspected for spar cracks in 11 years since new, and again 5 years after that!

Beech and the FAA relied heavily on the faulty assumption that flight hours cause the rudder cracks. They are really caused by a magnitude- frequency spectrum of aerodynamic forces on the rudder. Full rudder deflections at maneuvering speed, any number of them per hour are permitted by the pilot's handbooks, yet the AD states, in effect, that there is no danger for the first 1,100 hrs. of flight, or for 500 hrs. after each inspection!

The Production Line Mod

Shortly after the discovery of the rudder problem Beech devised a reinforced spar for rudders not yet assembled. This later was adapted to "Kit No. 33-6001-1S, intended to fix rudders in the field. To install this kit on an assembled rudder, labor and material would cost about $2,300.

This fix can be identified on newer aircraft by 1.5 inch inspection holes (covered with steel caps) just aft of each hinge opening. Annual inspection for possible further cracking is required by Beech, using a small mirror and a flashlight.

Raytheon maintenance shops are installing DW replacement hinge brackets instead of their kits: read on - - - -

The Only Solution

When I received MSB 2333 in October of >89 I designed the add-on SMP Brackets primarily to fix my own F33A. Later I developed the much-improved and lower priced DW-1 (upper) and DW-2 (middle) hinge brackets which replace, instead of straddle, the original Beech brackets. The DWs are approved by letter from the FAA, instead of by AD revision. This letter is included with the kit shipments.

The figure below illustrates the problem, and the DW cure:

Added Man Hours

The DWs required the drilling of 4 holes for 8/32 screws (replacing rivets) at the top hinge, and 2 at the middle. The web no longer carries the load, safely wiring the 2 original attaching screws (with surgical forceps and some cussing ) is avoided.

So ask your mechanic how long it will take to drill the 6 holes.

About 65% of the aircraft are now equipped with SMPs or DWs. They are free of the costly inspections, and approved for use on previously cracked spars. The installations have been proven to the FAA, by computer analysis and fatigue tests, to be at least 5 times stronger than the original ones.

UNDER-STRENGTH RUDDER ATTACHMENTS

Suppose the Forest Service issued this warning to hikers: If a grizzly bear is following you, look back to see if he is getting closer; (a) every hour, or (b) every mile; whichever occurs first.

All of the replacement brackets, SMPs and DWs, were designed by an Aero Engineer (me) after doing this sort of work at Douglas for 36 years. They are STCd, PMAd and AD-relieved; all the government certificates you could hope for. They are flying on about 2,800 Bonanzas. All were purchased at retail, and installed by Raytheon, Elliott, Woodland, Piedmont, Stevens, Tradewind Turbine, Ultimate Engines, American Beechcraft, and Executive Beechcraft, as well as hundreds of other FBOs and independent IAs.

The following letter went to the FAA by certified mail on 9/26 (with the photo links printed)

--------------------------------------------------------------------------------------------------------------------------

The history of Bonanza rudder hinge deficiencies and resulting MSBs and ADs are discussed in detail above, where I pointed out that the current AD 93-24-03 does not adequately address that airworthiness problem. In 1999 I wrote to your office, presenting arguments for a revision to the AD. In this letter I will discuss some recent developments that further illustrate a need for a revision.

On 11/12/2001 an Airbus A300 crashed after T.O. from JFK. (Go to NTSB , insert accident no. DCA02MA001). The NTSB accident report contained the following probable cause:

"...the in-flight separation of the vertical stabilizer as a result of the loads beyond ultimate design that were created by the first officer's unnecessary and excessive rudder pedal inputs."

The first officer evidently combined a sideslip load on the stabilizer with a rudder deflection that increased that load. I suspect that "unnecessary and excessive" rudder pedal inputs at sufficient speed might cause a failure on any aircraft if large yaw angle and large (unnecessary) rudder deflection combine adversely. (A comparable wing-loading condition would be an abrupt aileron deflection during a high G pullout.)

On Sept. 8, 2002, an A36 Bonanza N8261J crashed 20 miles W. of the Grand Canyon. (Go to: NTSB , enter accident no. LAX02FA278). The NTSB report contained the following details:

The airplane was mostly intact with the wing tip tanks broken downward. The vertical stabilizer broke over towards the right, and the rudder severed from the vertical stabilizer in trail. The entire airplane displayed high vertical crush and collapse of the structure. All ground scars and wreckage debris were found immediately around the fuselage. Based on an examination of the ground scars and the wings and fuselage, the left wing was moving rearward and the right wing forward when the airplane impacted the ground.

Apparently the A36 descended vertically to the surface in a flat CCW spin.. Assuming a forward CG (pilot only), I believe this would be very unlikely if the airplane, specifically the rudder and vertical stabilizer, were intact during the descent. The web site AirCombat explains the elaborate cross-control details of how this maneuver is intentionally performed in an airshow.

I recently learned that Dodson International in Kansas had the A-36 wreckage, and I bought the rudder from them to examine the details in the hinge areas. It was shipped to me on May 31. I sawed off the balance horn to get a clear view of the upper hinge. Photo 1 shows the upper flange of the bracket bent to the side about 0.3 inches. The lower flange behind it is approx. 0.4 inches off center. Photo 2 shows the badly warped bracket removed from the 2-screw attachment to the web. The web of the spar between the screws is twisted 14 degrees. So the sidewise displacement of the hinge bolt is due to a combination of bracket warp and twisting of the web. During load application this web twist must have been substantially larger, since the web is a relatively soft spring when subjected to out-of-plane bending. Note the strange failure of the bearing arm attachments to the vertical stabilizer in Photo 3, tattered shards bent backwards. This suggests that the "tattered shard" failure might have occurred progressively due to side-to-side oscillations of the rudder, perhaps flutter, during the descent. This would not be expected in a "single stroke" ground impact.

The Bonanza rudder hinge problem, discovered in 1989, is caused by the very improper, weak and flexible cantilever mounting of the hinge bracket to the 0.032 thin spar web. Starting with ser. E-2519 (the crashed A36 was ser. E-2807) the A36s were "fixed" at the factory and are not subject to the SB or the AD. But at the upper hinge this fix consists only of a 1" x 3" aluminum strip on the back of the spar web. The 4 nut plate rivets are attached to the strip, removing those 4 stress-concentrating holes from the spar web but this strip can have almost no effect on the flexibility of the bracket mount. To make matters worse, the hinge bolt hole in the bracket is about 17% farther from the spar web than on the original (to realign the 3 hinges after a change at the middle hinge). This increases the spar web twist by that percentage. On E-2807 the 2 screws twisted the strip, along with the web, about 14 degrees (permanent set, after spring-back).

Refer to Figure 1. If the rudder was "aggressively" deflected left as shown the resulting aerodynamic force (represented by the vector) could bend the hinge bracket to the left of the rudder C/L, causing the rudder nose to move right and wedge against the stabilizer skin as shown. (The E-2807 rudder nose is partially dented in along that line). This wedging or prying could have locked in left rudder, or initiated a failure of the hinge arm attachments in the stabilizer, or both in sequence. At the same time the stabilizer front and/or rear spars might have been in the process of bending or failing.

An interesting point here, with regard to vertical stabilizer/rudder loads: I previously mentioned adverse combination of yaw angle and rudder deflection. At a given speed the rudder hinge load and bending at the stabilizer attachment to the fuselage are both maximized by a large right yaw angle with a large left rudder angle (or vice versa). If the airplane had previously yawed to the right (CW) by turbulence or right rudder or both, then the pilot might have rapidly applied (unnecessary and excessive) hard left rudder to correct that yaw. This dangerous concurrence of high yaw angle and high rudder angle would depend not only on how rapidly the rudder is deflected, but also on the "authority" of the rudder. Pilots who have flown Bonanzas with tip tanks know about the slow rudder response when the tip tanks are full. This effect is caused by a large increase in Moment of Inertia in the yaw plane. E-2807 had tip tanks.

In March of 1995, in the process of obtaining AD approval for my DW-1 and DW-2 hinge bracket STC SA00044LA, I submitted report DW Aero 1 to your office. (On 4/5/99 I sent a second copy to Mr. Baktha at his request.) That report contains results of static tests I performed that showed the original Beech upper bracket side deflection was 4 times greater than for the DW-1 replacement, for the same load. The ratio for the DW-2 middle bracket was 7.5. The side flanges of the steel DW brackets bolt directly to the skin and spar flanges, eliminating the out-of-plane twisting of the spar web. Links to comparison photos of the original and the DW Brackets are TOP_BRACKETS and MIDDLE_BRACKETS showing the vast differences in load paths.

I believe there has been a change in FAA policy regarding cracks since AD 93-24-03 was issued. In the T-34 meeting last Feb. Mike Reyer of the Small Airplane Directorate stated the maximum allowable limit of "one chance in 1000 that a structure will have a fatigue crack". In a Staff Study dated 9/25/91 RAC reported that 5 rudder spars in a fleet were inspected, all were cracked and one upper hinge bracket had completely broken free. The latter was a total failure of primary structure. Since then at least 200 SMP and DW replacement brackets have been installed primarily because cracks were found. However it is interpreted, certainly the stated statistical limit of "1 in 1000" has been grossly exceeded

Another problem with AD 93-24-03 is that the SMP rudder brackets referred to have not been available since June '97, Spacecraft Machine Products is out of business and their PMA has not been valid for 8 years. The current DW-1 and DW-2 brackets have only been authorized as "substitutions for the SMP brackets" by a letter sent to me 9 years ago from your office. That authorization has been questioned by a number of shops and owners.

On April 5, 1999 I sent a letter to your office recommending that AD 93-24-03 be revised, to at least conform to the more stringent requirements of the Beech Mandatory Service Bulletin #2333... I believe that the above discussion of the E-2807 accident shows a much elevated urgency for an update of the AD.

I would be happy to meet with FAA or NTSB personnel in the LA area, to show them the hardware and to discuss the above reasoning. In any case, a response, detailing any technical disagreements with the above, would be greatly appreciated.

Eual Conditt Subject: Rudder spar cracks in Beech models 33 and 36
Wichita Aircraft Certification Office
1801 Airport Road, Room 100
Wichita, Kansas 67209

On September 26, 2005 in a letter to you I made another request for a reconsideration of the inspection requirements in AD 93-24-03. Specifically, I recommended that the 500 hour inspection requirement (after the first at 1000 hours) should be changed to annual inspections, as specified by Raytheon MSB-2333.

(1) Start by inserting the crack discoveries by the Air Training Center that originated the MSB and 4 subsequent ADs. See attachment 1 which your office sent to me in 1999. In web site click on StaffStudy. This Raytheon Staff Study reported 5 cracked spars plus 1 complete hinge separation in the ATC fleet of 15 aircraft; a 40% failure rate.. They also stated that the FAA told them of 2 other cracked rudders (see second paragraph of their study). Apparently these were the same two cases recorded in the 32 year-old database that you cited in your recent letter.

(2) Raytheon has sold many replacement spars to replace cracked spars. Please ask them for the number, of both original and reinforced designs, and add that number to your database.

(3) Approximately 2700 of the 4300 affected aircraft have been repaired with my SMP and DW hinge brackets. A large percentage of those brackets were reportedly ordered because cracks were found during inspections. One example: on 5/25/99 a rudder spar web was reported to have "two long cracks and a triangular piece missing" under the top hinge bracket.

To summarize, the "two occurrences of cracks in 32 years" as recorded in your database should be increased to an estimate of as many as a thousand. Then the airworthiness of the many aircraft not yet fixed should be carefully reconsidered..