A Novel Undersea Cable Design with Plastic Loose Tube, Low Excess Fibre Length, and Fixed Fibre Termination for Bend-Sensitive Fibres
By: TyCom Ltd, USA


This article describes a design for an undersea optical cable that successfully resolves the attenuation issues of highly bend-sensitive fibres with large-mode field diameter used in ultra-high speed broadband transmission. To meet the requirements of the rigorous environment of an undersea cable system, this new design combines a loose tube with low fibre excess length, and uses a new technique of fibre termination in a cable splice box. This approach minimises the microbending of the fibre in the cable and reduces local accumulation of fibre along the cable and near a splice box. Plastic loose tube provides economical protection to the fibres in a traditional cable structure. The design of the fibre termination in the splice box demonstrates that the fibres can be terminated firmly without significant attenuation change.


An undersea optical fibre cable traditionally accommodates either tight-buffer (tight-fit) or loose tube design. Due to the increased demand for ultra-high speed broad band transmission, fibres with larger-mode field diameter (or effective area) are needed. These fibres are usually highly bend-sensitive. To resolve this issue, some have proposed coating traditional fibres with a layer of larger-diameter UV acrylate [1], while others have moved away from the loose tube design and utilised a tight buffer[2] to control the loss in the cable.

Figure 1: Fibre attenuation with different MFD in various cable designs

In our experience, the cabling process usually generates some degree of microbending within the fibres in a tight buffer structure. A large-mode field diameter fibre with high bend sensitivity is more vulnerable to increased loss in the tight buffer design. Figure 1 gives the fibre attenuation results comparison between a loose tube cable and a traditional tight buffer cable.


Core Structure
The core of the undersea optical fibre cable is a loose tube of a traditional thermoplastic tube material. Gel in the tube serves as a filling compound to prevent water ingress and to gently support the fibres to reduce the impact of microbending. The fill ratio of the gel is designed to be as high as possible to reduce voids and increase water-blocking effectiveness. The fibre count varies: up to 16 for a 17mm cable, and 24 for a 21mm cable. Fibres are laid in the tube straight without helix.

Cable Structure
The basic design is the same as that of a tight-buffer cable, but the core is replaced by a loose tube. Outside the tube, 24 stranded wires form a tight package surrounding the loose tube (Figure 2). Next, copper tape is welded and swaged down on the wires to serve as an electrical conductor and a protective sheath. Then a polyethylene jacket is extruded over the copper to serve as insulation. This structure is designated Light Weight (LW) cable. For additional strength and abrasion capability, armour wires can be added to the outside of the LW cable to create different types of protected cable.

Figure 2: Loose tube cable (LW) structure

Loose Tube
There are several important advantages in the use of a thermoplastic material such as polybutylene terephthalate (PBT) in an undersea loose tube construction. Among these are:
(1) A plastic tube is more economical than a metallic tube as to raw material cost and capital costs of the processing equipment;
(2) Unlike a metallic tube construction which requires a longitudinal jointing technique (thus introducing seams), plastic can be extruded, thus eliminating concerns about seam reliability.
(3) PBT affords excellent dimensional stability and surface morphology to minimise optical impact on enclosed fibres.
(4) PBT has proved to be very stable for the operating environment of undersea cables and has no corrosion issues, thus is unlikely to generate hydrogen in the presence of sea water.
(5) In addition to material stability, the shrinkback characteristics of PBT are utilised to effectively control the excess fibre length in the loose tube.

An extensive material qualification program was carried out to evaluate PBT, including its physical and chemical properties. Specific qualification items included hydrolytic stability, melt flow characteristics before and after processing, water content, degree of crystallisation, and compatibility with other cable construction materials.

Filling Compound
The cable filling compound, or gel, must also be chosen carefully. In evaluating different sources of gels, emphasis was placed on the rheological properties, specifically viscosity and critical yield stress. On the basis of analytical modelling and cable attenuation testing, critical shear stress of 0.003 psi to 0.010 psi and cone/plate viscosity of 20,000 cps to 40,000 cps were established to be the optimum value ranges for loose tube construction. The choice of such rheological property ranges allows us to control the coupling of fibres to the loose tube so that sufficient excess fibre length (EFL) is maintained to allow fibre movement during the loading and unloading of tensile strain, while preventing water ingress in the event of a cable cut.

The characterisation of gels involves an assessment of both static and dynamic rheological properties, which are related to both tube processing and the resultant fibre properties. In addition, extensive chemical and physical evaluations of gel properties, such as volatilisation of gaseous species (including hydrogen) and gel compatibility with tube and fibre coating materials, were carried out.

Cable Properties, Performance, and Qualification

Excess Fibre Length (EFL)
To minimise the interaction of fibres in a loose tube, the excess length of fibres must be tightly controlled, especially at the last stage of the manufacture process. During tube manufacture, the fibre excess length is controlled to be less than 0.20 per cent. In the process, which adds the steel strand wires and copper sheath, the tube strain is controlled to make the EFL in the strand wire package 0.05 per cent or less.

During the polyethylene extrusion process, the fibre strain in the cable is generally unchanged. Figure 3 shows the change of fibre attenuation after various process steps. The fibre attenuation is significantly improved from loose tube to cable, which correlates to a significant reduction of excess fibre length. In the final cable stage, the fibre is almost straight in the tube.

Figure 3: Fibre EFL and attenuation relationship
Fibre Count
The fibre count for 17mm cable, with tube inner diameter of 1.88mm, was tested for up to 16 fibres. The average attenuation of fibre relative to fibre count is shown Figure 4. The slightly higher cabled fibre losses for the four- and six-fibre counts are believed to be the result of unstable EFL during the early stage of production. The results show that fibre count does not significantly affect the fibre attenuation in a 17mm loose tube cable up to 16 fibres, which gives a packing density of 29 per cent.

Figure 4: Cabled fibre loss versus fibre count

Gel Fill
The gel fill is a critical parameter affecting water ingress. This value is defined as the average ratio of actual gel volume in the tube to the total volume available. The gel fill is controlled to approximately 95 per cent. It was found in water ingress tests with cable sample lengths of 250 meters that the gel fill could be as low as 87 per cent and still pass the test. For a 1km test length the cable must withstand a pressure of 8,000 psi for two weeks without water leak. In fact, the actual water penetration in this test was less than 10m. From the gel fill, the critical shear stress of the gel, and the length of the cable sample, it was calculated that a 1km cable could withstand well over 8,000 psi without water ingress through the full length.

A series of qualification tests subjected the cable to the most severe in-service environmental conditions of tension, bending, temperature, and pressure expected in the field. This design has met the requirements without any issues. A primary area of focus was the coupling and the interaction between the fibres and the loose tube gel medium. Test procedures were developed to permit measurement of the EFL on 100m long test specimens after the tube manufacturing stage and in the finished cable stage.

Other testing focused on the shear interaction between fibre and gel. A number of trial lengths of loose tube with different gel types were manufactured and tested. Figure 5 gives a comparison of the shear adhesion properties of different gels in cabled loose tube with the fibre unrestrained at the specimen ends. Gel A was chosen for its better coupling with the fibres.

Once this most favourable gel material was characterised, cable manufacturing parameters could be specified to provide the tight control of EFL desired in the finished cable.

Figure 5: Shear adhesion of fibre in different gels

Using optical phase delay test methodology for fibre strain measurement, testing on finished cable lengths was conducted to determine the final EFL for the product. Figure 6 shows fibre slack measured during tensile testing on an LW cable specimen.

Joint Design, Performance, and Qualification

Joint Design
The cable-to-cable joint and cable-to-repeater coupling anchor assembly are required to withstand the forces of deployment and recovery and to withstand ocean environments including pressure of up to nearly 15,000 psi.

Figure 6: EFL test of a 17mm LW cable

They must provide reliable connections for fibres which transition from the cable structure, and for splicing. Joints and couplings must also be capable of withstanding full system voltage for the system lifetime requirement of 27 years.

When the cable core design was changed over from one using an extruded elastomer with a central wire to the loose tube, a new method for preventing the fibres from moving into the cable structure from the joints became necessary. A ferrule and holder were developed. The fibres are splinted and potted within the ferrule. A commercial product originally designed as a mass fusion splice holder provides adequate strength and shows low loss under tension, which recovers when the tension is removed.

The ferrule does not damage the fibre coating during severe load and temperature exposure tests. The ferrule retainer assembly has a trough that contains and restrains the ferrule. Tension forces applied to the fibres are transferred to the joint through the retainer assembly.

The PBT tube is reliably fixed to the cable structure and does not require restraint at the joint. Other features of the joint required slight modification to accommodate the new fibre trajectory and placement in the joint. The details of the joint are shown in Figure 7.

Figure 7: Cable-to-cable joint

A standard set of qualification tests is performed whenever joint designs are altered significantly. For the loose tube joint, bench tests were done to verify the fibre retention and optical loss characteristics of the ferrules. These included temperature cycles and simulated loading during cable tension with a design factor (multiplier). The testing concluded with tests of joints on cable samples within 100 meters. Criteria included no visible damage or excursive or residual loss greater than 0.05 dB times the number of splices in the joint. Testing was done for:

  • Storage temperature;
  • Shock and vibration;
  • Applied torsion;
  • Sheave laying and recovery;
  • Tensility (with and without swivel);
  • Cyclic tension;
  • Hydrostatic pressure;
  • High voltage.
Figure 8 gives tension test results for the joint. The fibres show slightly higher but stable loss during peak load holding, and losses coming back to normal after loading. Separate tests were done to qualify the insulation moulding process. The optical loss was monitored during the moulding process.

Sea Trial

The purpose of the sea trial was to study the performance of the fibre in the loose tube cable and in the cable-to-cable joints during deployment and recovery. The sea trial was conducted in November 1998 in the western Atlantic at an ocean depth of 5000 meters. Three different cable joints and cable joint designs were integrated into the cable.

Figure 8: Tensile test results of cable-to-cable joint

The designs that were studied employed various methods of fibre restraint: an unrestrained fibre, fibre restrained with a drum, and fibre restrained with heat-shrunk ferrule. A cable repair operation was simulated and a shipboard cable joint with unrestrained fibres was constructed. The shipboard joint was then deployed and recovered.

Deployment and Recovery
The sea trial cable system was deployed at an average payout speed of 2 knots. During deployment the behaviour of the fibres was quite stable. No significant attenuation changes were observed in any of the fibres, and attenuation in most of the fibres showed no change. The maximum increase in attenuation in one fibre was less than 0.01dB/km. Cable recovery tensions are greater than tensions during deployment, and larger changes in measured optical attenuation were seen during recovery. The average loss increase in the fibres ranged from 0.0047 to 0.016dB/km. After complete recovery, the residual average loss increase was 0.005 dB/km on all the cables.

Cable Repair Operations
The cable-to-cable joint deployment and recovery cycle was followed by cable redeployment, cutting of the cable mid-span, grappling of the two ends, and a shipboard cable-to-cable joint. The shipboard joint was then deployed and recovered. The cable was grappled and, although the grappling hooks secured the fibres and prevented any fibre movement, slight movement of the fibre was observed after the cable end had been brought to the jointing area.

Repair Joint
The change in fibre attenuation in the cable joints with restrained fibres was insignificant. The optical attenuation measurements at the cable joint with unrestrained fibres indicated unstable fibre pull-in and loss increase. After dissection of the cable joint, the fibre pull-in was measured to be in the range of 5.5 in. to 6.4 in. from deployment to complete recovery. Due to this fibre pull-in, two fibres showed a higher local loss increase in the range of 1.3 to 5dB. This confirmed that the unrestrained fibre joint design was inappropriate for this cable design.


This cable design is now in steady production. Several systems have been successfully deployed, and their excellent performance demonstrates that this design is adequately benign to bend-sensitive fibres with larger-mode field diameter. In addition, this simple and robust design with economical yet reliable materials makes this cable system easy to manufacture, handle, and repair.


C-S Ma, S. Bernstein, Q. Gong, T. V. Chute, R. J. Rue, C. E. Murphy, and G. Gull,
Tyco Ltd., Eatontown, New Jersey and Newington, New Hampshire - USA


[1] K. Mitsuhashi, T. Hayano, T. Shimomichi, T. Abiru, K. Oohashi, and M. Miyamoto, "Development of Submarine Optical Cable Unit Using a Large Effective Area Fibre," IWCS (1999), p. 312

[2] 0. Nagatomi, K. Yamamoto, H. Wakamatsu, R. Kanda, and R. Morikawa, "Development of New Tight-Fit Type Submarine Cable with Large Core Fibres for a WDM System," IWCS (1999), p. 317

This paper was delivered at the 49th seminar IWCS
Atlantic City, USA - November 2000
Printed by courtesy of IWCS - © IWCS 2000

TyCom Ltd
Eatontown, New Jersey, USA
Newington, New Hampshire, USA
+1 732 578 7830