T-AKE_CLASS_NOTICE_15_-_MAIN_DIESEL_GENERATOR_COUPLINGS.pdf

T-AKE CLASS NOTICE

Notice # 15 Date: 28 Jun 2010

Issued by N43C __________________ Francis W. Cunningham

T-AKE Class Manager

TITLE: MAIN DIESEL GENERATOR COUPLINGS

1 PURPOSE:

The purpose of this notice is to address issues regarding Main Diesel Generator couplings on T-AKE Class ships.

2 BACKGROUND:

2.1 On 7 February 2010, there was a catastrophic failure of #3 Main Diesel Generator coupling on USNS LEWIS AND CLARK (T-AKE 1). Initial inspections raised concerns with apparent lack of compression between membranes internal to the coupling.

2.2 After the #3 Main Diesel Generator coupling failure, magnetic particle testing was accomplished on the outer clamping rings and accessible portions of membranes on MDG #1, #2, and #4 MDG couplings on USNS LEWIS AND CLARK (T-AKE 1) and all couplings on T-AKE 2 through 8. An indication that showed during the MT inspection on the USNS LEWIS AND CLARK (T-AKE 1) MDG #2 coupling was found to be a tool mark after recent disassembly and inspection. No other cracks or indications were found on any other couplings in the class.

2.3 The failed #3 Main Diesel Generator coupling components were sent to IMR Laboratories for a failure analysis. Dimensional checks were accomplished that confirmed the membrane dimensions were not in accordance with design drawings and the membranes were not compressed when the coupling was in operation. (Reference IMR Test Lab Report Number 100341 dated 16 June 2010 and N7 Engineering Task Report dated 26 June 2010).

2.4 Vulcan submitted a statement attributing the coupling failure to a fatigue crack in the outer clamping ring caused by fretting fatigue accompanied by tribochemical attack.

2.5 During the investigation, Vulcan stated that three membrane couplings were installed on the first three T-AKEs in the class and Vulcan changed to two membrane couplings on USNS RICHARD BYRD (T-AKE 4). MSFSC has not received official feedback from Vulcan on why this configuration change occurred.

2.6 On 21 April 2010, visual inspection found a fastener loose in the inner clamping ring on the #1 MDG coupling on USNS SACAGAWEA (T-AKE 2). See photo of SACAGAWEA #1 MDG Coupling Fastener. During the subsequent inspection and repair

MAIN DIESEL GENERATOR COUPLINGS NOTICE # 15

28 Jun 2010

T-AKE CLASS NOTICE Page 2 of 2 with a Vulcan rep, three other adjacent fasteners in the #1 MDG coupling were found to be only hand tight.

2.7 During a June 2010 Voyage Repair Period (VR), the #1, #2, and #4 MDG couplings on USNS LEWIS AND CLARK were disassembled and rebuilt with two membrane assemblies. Alignment readings were taking between the main diesels and associated generators in way of all four MDG couplings. Additional concerns were raised with alignment and original coupling assembly.

3 ACTIONS:

3.1 MSFSC is working on a MDG coupling inspection and rebuild schedule with FME and Vulcan. It is hoped that the coupling work will be coordinated with the MDG exhaust transition piece inspections/replacements. The schedule will be provided when finalized.

3.2 It is anticipated the inspection and rebuild the MDG couplings will include:

- Temporary removal of rubber elements, inner clamping ring, and membranes,

- Accomplishing a magnetic particle test of outer clamping ring, inner clamping ring, and membranes to ensure they are suitable for reuse,

- Accomplishing a hardness test of outer clamping ring,

- Replacing the inner clamping rings and three membrane assemblies installed on USNS SACAGAWEA (T-AKE 2) and USNS ALAN SHEPARD (T-AKE 3) with new inner clamping rings and two membrane assemblies,

- Measurement of the removed membranes and the fabrication and installation of shims to ensure membranes are “crushed” during reassembly,

- Ensuring the #1 cylinder is at top dead center position and the keyway on the generator shaft is at the 12 o’clock position at the start of reassembly process,

- The installation of a CuSn6 shim between the rubber elements and outer clamping ring during reassembly to address fretting fatigue and tribochemical attack concerns,

- Reassembling to ensure lifting holes in outer clamping rings do not align with gaps in rubber segments,

- Ensuring rubber segments are installed in the proper orientation,

- Ensuring fasteners are installed in correct locations and torqued to correct values.

3.3 Until the coupling inspections and rebuilds are accomplished, ship’s force will accomplish a monthly inspection of the outer clamping ring and T-AKE port engineers will arrange for a magnetic particle test of the outer backing ring and accessible portions of the membranes every six months. T-AKE Port Engineers will also arrange for an annual inspection of couplings by a Vulcan Representative. These inspections will be added to the Shipboard Automated Maintenance Module (SAMM).

T-AKE MAIN DIESEL GENERATOR COUPLING

MSC LIST OF QUESTIONS AND CONCERNS

8 MARCH 10

ITEM 1: “AS FOUND” CONDITION OF LEWIS AND CLARK MDG #3 FAILED

COUPLING

The LEWIS AND CLARK (T-AKE 1) failed MDG #3 coupling configuration had three

(3) membrane plates separated by four spacers at the inner ring to hub area.

- Each spacer was measured at 0.25 mm. This is per Vulcan Dwg. 4G56R5005M, Membrane Part List (old 3 Membrane), piece 3.

- Each membrane plate was measured at 1.72 mm. This is not per Vulcan Dwg.

4G56R5005M, Membrane Part List (old 3 Membrane), piece 1. The drawing lists the membrane thickness as 1.95 mm.

- The stacked membrane and spacers are (4 X 0.25mm) + (3 X 1.72mm) = 6.16mm thick.

- The inner ring lip recess dimension was measured at 7.31 mm. This is per Vulcan Dwg.

3G65B5010M, Innerdecking (Inner Clamp Ring).

- The Hub landing was measured at 14.00 mm. This is per reference Vulcan Dwg.

4G56B0014M, Nabe ( Hub)

-When the inner membrane ring is hard against the hub there is gap created where the membrane assembly fits. The gap is 14.00 mm - 7.31 mm = 6.69 mm.

- The 6.69 mm gap is larger than the membrane stack (6.69mm- 6.16mm = 0.53mm).

Question 1: Why are the three failed membrane plates' thickness not per the referenced drawings ?

Question 2: Due to the above stated dimensions, there appears to be a lack of clamping force between the membrane plates and the hub. It appears as if all the torque is transmitted from the membrane plates to the generator hub via the bolts. Would this arrangement cause failure of the coupling?

Question 3: How was the hub assembled during new construction? What measures were taken during new construction to verify the coupling hub/membrane/inner ring dimensional fit-up? Are “as released” readings available from initial installation?

ITEM 2: “AS FOUND” CONDITION OF LEWIS AND CLARK MDG #3 FAILED

COUPLING – ALIGNMENT CHECKS

During LEWIS AND CLARK MDG #3 alignment checks, the “as found” axial, radial, and angular alignment readings were found to be above new installation criteria. Vulcan found the alignment readings satisfactory to support the installation of a replacement coupling. Please find new installation tolerance vice as found tolerances below.

AXIAL

Tolerance = ± .8 mm Findings = -1.78mm Δmmc = 0.9825 mm (out of tolerance – too near flywheel)

RADIAL

Tolerance = + 1.0mm -2.5mm Findings = - 2.2 mm (Gen. shaft is higher than engine – OK)

ANGULAR

Tolerance = ± .6mm Findings = -1.15mm

Question 1: Are “as released” alignment readings available from initial installation at

NASSCO?

Question 2: What is the maximum allowed misalignment permitted (axial, radial, and angular) to allow for the installation of a replacement coupling?

Question 3: What is the permanent or continuous misalignment that the coupling is designed to withstand ? What is the expected lift at this misalignment and what is the limiting component or expected mode of failure ?

ITEM 3: HOLES IN CLAMPING RING/GAPS IN MEMBRANE SEGMENTS

Vulcan indicated that the drilled and tapped lifting holes in the clamp ring are not supposed to align with gaps in the rubber elements. Vulcan Drawing 2G56G0000M shows the subject holes staggered away from gaps in the rubber elements. The failed MDG #3 coupling had the subject holes aligned with gaps in the rubber elements.

Question 1: What coupling parts could be stressed if the drilled and tapped lifting holes in the clamp ring are aligned with gaps in the rubber elements? Could this have caused the MDG #3 coupling failure?

ITEM 4: VULCAN INSTALLATION GUIDANCE

Robert Burkhardt, the American Vulcan Corporation rep onboard LEWIS AND CLARK during the recent investigation mentioned that Vulcan recommends that the main diesel engine and generator be in particular positions when the coupling is assembled (i.e. No. 1 cylinder should be at top dead center).

Question 1: Was the main diesel engine and the generator at the recommended position when the MDG #3 coupling was assembled?

Question 2: The generator rotor thrust bearings had to be removed and the rotor jacked back approx 0.900" to allow the membrane ring assembly to fit past the flywheel and hub. Is this standard installation procedure?

ITEM 5: COUPLING MATERIAL QUESTIONS

The new membrane assembly installed has two membrane plates vs. the old membrane assembly that had three plates.

Question 1: Why was the design changed? Was there a design problem or was the design changed for a less complicated fabrication process? Can both style membranes be ordered? Which style should be used?

Question 2: Can new membrane plates be purchased separately, or does the entire membrane assembly have to be purchased as a unit? Or does the entire coupling have to be purchased?

Question 3: If the membrane plates can be purchased separately, is there a special alignment procedure to align the plates with the outer ring?

Question 4: Could the rubber sections of the coupling fail without being detected by visual inspection? And could rubber failure cause the membrane ring to fail?

Question 5: New bolts and washers had to be procured by the Vulcan rep to mount the coupling to the fly wheel. The bolts/washers did not come as part of the new coupling assembly. Should these bolts/washers be included in the shore based spare assembly?

ITEM 6: FUTURE INSPECTION CRITERIA/EXPECTED SERVICE LIFE OF

COUPLING

The Vulcan manual recommends an annual visual inspection. The MDG scheduled maintenance is to perform an alignment check of the coupling at 6000 hrs.

Question 1: Should there be additional NDT checks at a specified period? Should there be any reason to disassemble and inspect at a specified period?

Question 2: Is the 6000 hr coupling alignment check to be repeated at every 6000 hrs?

Question 3: What is the expected service life of the coupling? Should we plan on replacing certain components (i.e. the rubber elements) after a specific number of operating hours ?

ITEM 7: COUPLING SELECTION

Question 1: What were the design specifications or requirements for the coupling selection ? Can a copy of the component selection calculations, selection criteria, purchase specification, etc. be provded ? Can information be provided to MSFSC that validates that the installed coupling is correct for this application ?

1/ 2

Statement concerning fracture of the membrane clamping ring of RATO-S 531T coupling on TAKE -1 MV "Lewis & Clark"

With this statement VULKAN Couplings obliges the enquiry of AVC concerning more detailed information about the failure of the clamping ring of VULKAN coupling RATO- S 531T on TAKE -1 MV "Lewis & Clark". The statement is based on submitted photos only.

Cursory observations

Apparently the fractured surface shows characteristics of a fatigue fracture (figure 1).

The initiation area of the crack can probably be located where wear (fretting fatigue accompanied by tribochemical attack) took place between the clamping ring and its counter part (figure 1). This wear can be a significant notch for a high strength material and thus the originator of severe stress concentrations. Crack initiation must not necessarily be located in the drilled hole.

Figure 1: Tribochemical fretting corrosion considered to be the origin of the fracture

FE-calculations show bending as the dominant load for the clamping ring. The Equiva-lent Stress caused by the centrifugal force and the torque TKN is quasi-static under sta-tionary operating conditions. The dynamic safety, which is relevant for fatigue, is con-siderable. However initiation of the fatigue crack is a significant notch effect which was caused by wear namely fretting fatigue accompanied by tribochemical attack. The ra-dial thread in bending area of the clamping ring might have abetted micro slip between the surfaces. In order to illustrate the possible effect of notch effects caused by fretting corrosion: Winterfeld [Win] reports for polygonal shaft hub connections, that the notch effect amounts to kcb = 6,1 under static torsion and rotating bending.

Fretting Corrosion: a tribochemical phenomenon

Fretting corrosion refers to corrosion damage at the asperities of mating contact sur-faces. This damage is induced under load and in the presence of repeated relative surface motion. Pits or grooves and oxide debris characterize this damage, typically found in machinery, bolted assemblies and ball or roller bearings.

2/ 2

Damage can occur at the interface of two highly loaded mating surfaces which are not designed to move against each other. The protective film on the metal surfaces is re-moved by the rubbing action and exposes fresh, active metal to the corrosive action of the atmosphere. Reduced fatigue life is a result of stress concentrations produced on the metal surface.

Prevention and remedial action

As an remedial action, VULKAN recommends the installation of shims (CuSn6) be-tween the clamping ring and the counterpart (flexible elements) in order to avoid any wear reliably. Metal surfaces of the clamping ring and the segmented spring are then separated by the CuSn6 shims. This design is state of the art and has proven to be ef-fective as far as the membrane side of the coupling is concerned. Furthermore the segments should be mounted on the clamping ring in such a way that the radial threads are not in the area where two segments join.

Figure 2: Installation of CuSn6 as remedial action against fretting corrosion

[Ber] Berthold Schlecht; Maschinenelemente 1; Pearson Studium 11/2006; ISBN:

978-3-8273-7145-4 Berthold Schlecht; Machine elements 1; Pearson Studium 11/2006; ISBN: 978-3-8273-7145-4

[Win] Jens Winterfeld: Einflüsse der Reibdauerbeanspruchung auf die Tragfähigkeit von P4C-Welle-Nabe-Verbindungen; Dissertation TU Berlin; 2001 Jens Winterfeld: Influences of Fretting Corrosion on the Load Capacity of P4C-Shaft-Hub- Connections; ; Dissertation TU Berlin; 2001

i. V. Dr. Gunnar Gödecke Chief Design Engineer VULKAN Kupplungs- und Getriebebau B. Hackforth GmbH & Co. KG

Herne, 2010-04-29

CuSn6 shim

-----Original Message----- From: Harold, John S CIV MSFSC N43C [mailto:john.harold@navy.mil] Sent: Monday, June 14, 2010 4:19 PM To: Black Christopher D SOSBSD Cc: Hoyt Rebecca R SOSBSD; Diaz Arturo A SOSBSD; Whaley Barbara A SOSBSD; Carr Thanh Q SOSBSD; Cunningham, Frank CIV MSFSC N43C Subject: LEWIS AND CLARK COUPLING WORK - I NEED SOME HELP Importance: High

Chris, Brad Van Vleck is keeping a daily log for the ongoing LEWIS AND CLARK coupling work that includes discrepancies discovered.

Discrepancies to date:

1 - The bolts used to secure the rubber elements to clamping ring and the fly wheel were mixed up. 10.9 bolts belong on the clamping ring and

8.8 bolts belong on the fly wheel.

2 - Some link plates seem to have been opened up with a "P" grinder so the bolts will fit. (Similar condition found on RICHARD BYRD (T-AKE 4)).

3 - One rubber element was installed backwards. The element deformed from being operated in this condition. We have a spare set of rubber elements.

Will replace all the elements on the effected coupling.

4 - MSHS (a MAN Service rep out of Florida) is doing alignment checks for us. It appears there is significant misalignment in way of the #1, #2, and #3 MDG couplings. I have heard that the T-AKE main diesels and generators are installed out of alignment intentionally since the diesel grows more and moves more than the generator during normal operation. During operation, the different movement of the diesel and generator bring the two machines into alignment.

I need to know:

1 - What was NASSCO shooting for wrt to an intentional misalignment (in the cold condition) to account for calculated diesel engine and generator movement ?

2 - Did NASSCO accomplish hot checks to verify the anticipated diesel engine and generator movement occurred ?

3 - The LEWIS AND CLARK crew found "as released" alignment readings for all 4 main diesel engines/generators. Would like to receive the "as released" readings from SUPSHIP for the LEWIS AND CLARK (to make sure the ship has the right data) and for the rest of the class. MSFSC is planning to do thorough alignment checks as part of the coupling repairs on all the ships.

Need the above information as soon as possible.

mailto:john.harold@navy.mil

Currently our plan on LEWIS AND CLARK is to reassembly the couplings, run the engines individually at as high a load as possible, then shut down and immediately take alignment readings to see if the misalignment we are seeing on MDG #1, #2, and #3 will lessen when the engines warm

up. We will do the same for MDG #4.

I'll try to call you tomorrow morning.

r/ John

T-AKE Engr Type Desk Military Sealift Fleet Support Command (757)443-0966 (Office Phone) (757)777-6769 (Cell Phone) (757)443-1499 (Office Fax) email: john.harold@navy.mil