The
Collins-class submarines experienced a wide range of problems during their construction and early service life. Many of these were attributed to the submarines being a new, untested design, and were successfully addressed as they were discovered. Most systems and features worked with few or no problems, while the boats' maximum speed, manoeuvrability, and low-speed submerged endurance were found to exceed specifications. The ship control system, which during development had been marked as a major potential problem, functioned beyond positive expectation: for example, the autopilot (which aboard
Collins was nicknamed 'Sven') was found to be better at maintaining depth during
snorting than most helmsmen. However, problems with the combat system, excessive noise, and engine breakdowns were recurring and appeared across the entire class. These and other shortcomings were often made harder to solve by disagreements between Kockums, ASC, Rockwell, the RAN, and the Australian Government over the nature of problems, their causes, and who was responsible for solving them. Media reporting of the problems during the mid-1990s was often negative and exaggerated, creating poor public perception. This was aided by politicians, who used the shortcomings to politically attack the Labor Party and Kim Beazley, particularly after Labor was defeated by the
Liberal-National Coalition in the
1996 federal election, and Beazley became
Leader of the Opposition. During the mid-1990s, it was recommended on several occasions that the submarine project be abandoned, and the completed submarines and incomplete hulls be broken up for scrap. Following the
McIntosh-Prescott Report, which indicated the long-term faults with the class that still required solving, successful efforts were made to bring the submarines to operational standard. This same period saw the dispelling of the idea, widely held within the RAN, that the
Collins-class boats would be like any other vessel previously ordered by the RAN: in service with another navy, well tested, and with all the problems solved before they entered Australian hands. The RAN began to realise that as the parent navy for the class, they had a greater responsibility than normal in ensuring that the boats were at an operational standard.
Welding of Collins During assembly of
Collins bow and escape tower sections in Sweden, multiple defects in the hull welding were discovered. Different reasons were given by different parties for the problems: To speed production, Kockums employed welders who were not qualified to work on high strength steels; the Qualified Welding Procedures developed by Kockums for these steels were not followed in production; the
steel alloy used for the hull required different welding techniques to those normally used by Kockums; the Swedish navy always requested partial penetration welds for their submarines, while the RAN wanted full penetration welding, but had not made this clear; delays in delivering the steel plates to Kockums resulted in rushed work and a resulting drop in quality. Repairing these welds quadrupled the time
Collins spent in dock.
Noise signature The noise made by the submarines, which compromised their ability to stay hidden, was another major problem with the design. In the original requisition, the RAN guidelines for the noise signature of the new submarines were vague; for example, asking that they be "twice as quiet" as the
Oberons. Expectations and operational requirements also changed between the 1987 contract signing and when the submarines began operating in the late 1990s. The major element of the noise signature for the
Oberon class was machinery noise transmitted through the hull; this was successfully avoided during construction of the
Collins class by mounting machinery on platforms isolated from the hull. The shape of the hull was the main cause: although a scale model of the design had been tested during the funded study and was found to have a minimal signature, the hull shape was changed after the contract was signed, primarily by a lengthening of the submarine and a redesign of the bow dome to accommodate the larger-than-expected main sonar and reduce its blind spot (the
baffles). The design had not been retested, as who would pay for this could not be agreed on. Cavitation had not been a problem with earlier Swedish submarine designs or during early testing of the Type 471 design, but the propeller had to be redesigned late in the process to provide more power, and like the redesigned hull, was not retested. During the year 2000, an unusual meeting took place with a next door neighbor (Francis 'Frank' Smith) of the then HMAS
Stirling Naval Base commander. He was an Aircraft Maintenance Engineer (originally trained at Government Aircraft Factories Fisherman's bend) who had been aware of the fluid dynamics issues of the Collins class for some time, purely by interest and observation on television. After a lengthy discussion, he was invited to discuss and demonstrate where possible, his observations at the
Stirling Naval Base with Navy and
Defence Science and Technology Organisation (DSTO) staff who were there at that time as part of an investigative group. He showed on a white board, the aerofoil issue with the Dorsal – Sail conning tower structure showing that the aspect ratio (span (height) to chord (width)) was too short and that severe turbulence / cavitation would be generated by such a design. This was demonstrated again on the white board using aircraft aerofoil wing shapes as a basis for the discussion. That the turbulence / cavitation generated would, by natural rearward flow, move down the rear upper surface deck of the hull and be drawn into the propeller. He was also able to demonstrate that the design of the bow section would not pass a flow test for generated turbulence / cavitation, with the change in shape from circular bow section to long hull, being ill-conceived. He made several recommendations during the lecture that would be cost-effective and possible. 1) To lengthen and taper the dorsal fin and create a more streamlined integration of the dorsal to flat upper Hull deck section. and 2) To 'fill in' the hollow section of hull aft of the bow curvature. Both these could be achieved with Carbon Fibre or Fibreglass covers as no load bearing strength would be required. Subsequent studies by the DSTO showed that the submarine's hull shape, particularly the redesigned sonar dome, the
fin, and the rear of the submarine, focused the displaced water into two turbulent streams; when the seven propeller blades hit these streams, the propeller's vibration was increased, causing cavitation. These problems were fixed by modifying the casing of the submarine with fiberglass fairings.
Propulsion system During trials of the first submarines, the propulsion system was found to be prone to failure for a variety of reasons. Most failures were attributed to the fifteen-tank diesel fuel system: the tanks were designed to fill with salt water as they were emptied to maintain
neutral buoyancy, but water would regularly enter the engines due to a combination of poor design,
gravity separation of the fuel and water being insufficient, and operator error resulting from poor training. The fuel-related issues were solved by installing
coalescers, improving training and operational procedures, and adding
biocides to the fuel. Although designed to allow for a leak of per hour, during trials it was found that the seals would regularly misalign and allow hundreds of litres per hour into the boat—during one deep diving test the flow rate was measured at approximately a minute. The propellers themselves were also found to be poorly manufactured, having been shaped by hand, with at least one cast at the wrong
pitch. This was rectified by using a five-axis
milling machine for future shaping work and replacing the miscast propeller. The material used for the propellers was also found to be weaker than expected, developing
fatigue cracks after only a few years of use. Despite the Americans fixing the problems with the propeller design, resulting in significant performance improvements, the Swedish company was dissatisfied with the Australian actions; the dispatch of the propellers was one of the points of contention in the company's legal action in the mid-2000s against the Australian government over ownership of the
intellectual property rights to the submarine's design. Other propulsion problems included excessive motor vibrations at certain speeds which damaged various components (which was attributed to the removal of a
flywheel and to corrosion caused by the fuel problems), and excessive fuel consumption in
Collins at high speed (found to be caused by manufacturing problems with the turbines and turbochargers). The propulsion system was also found to be a secondary source of noise: poor design of the exhaust
mufflers, weight-saving measures in the generator mountings, and an incorrect voltage supply to the battery compartment exhaust fans were noise-creating factors found and eliminated during studies by the DSTO. In March 2010, the Department of Defence revealed that the generators in five of the submarines were flawed and had to be replaced. The three generators aboard each of the five submarines are to be replaced in the submarines as they come in for their next
maintenance docking. They were not streamlined; raising a periscope while moving would create enough
drag and
turbulence to shake the entire submarine. As with many elements of the submarine, there were disagreements as to who was responsible for the problem. The periscopes also had problems with their optics: periscope users reported difficulty in refocusing after changing magnification, duplication of images, and bands across the field of vision. The problems had started during the funded study, when Singer Librascope and
Thomson-CSF, who were partnering with Rockwell to develop the combat system, refused to release their intellectual property or their software code for Rockwell to sell. It was proposed that Computer Sciences of Australia, a division of
Computer Sciences Corporation and a minor partner in the consortium, take over the role of writing the software for the combat system, although this meant that Singer Librascope, which had prior experience in creating submarine combat systems, was reduced to a minor role in the project. Australian Submarine Corporation was made responsible for the delivery of the Rockwell combat system, but had little ability to enforce this. Rockwell was contracted to deliver the combat system by 9 September 1993, but was unlikely to do so. ASC's management board voted to issue a default notice to Rockwell as the American company had defaulted on the contract, but was ordered by the Department of Defence to retract the default notice and accept gradual delivery of partially completed versions of the combat system—referred to as 'releases' and 'drops'—until the complete system had been delivered. By March 1994, the combat system had become the major area of concern for the submarine project: assembly of the system was almost nine months behind schedule, and at least 20% of the software had not been compiled. The combat system continued to be a problem during the next few years, with progressive drops offering little improvements in performance over the previous version, and the completion date of Release 2—the designation for the full contractual realisation of the combat system software—was continually postponed. In 1996, Rockwell sold its military and aerospace division, including responsibility for the
Collins combat system, to
Boeing. Boeing attempted to produce a workable combat system, but believed that this could only be done if the changes in technology were accounted for in a contract alteration, which the RAN and the Australian Government initially refused to do. Boeing sold its naval systems division to Raytheon in May 2000, making the latter company solely responsible for completion of the combat system. Because there was not enough time to evaluate the replacement system to include it in the
"fast track" program,
Dechaineux and
Sheean were fitted with the old Rockwell combat system, which was enhanced by the addition of sub-systems developed during the early 1980s for the
Oberon-class mid-life upgrade and
commercial off-the-shelf components. Even with the enhanced system, it was believed that the capabilities of the fast track
Collins boats was at best equivalent to the
Oberons. observing the Collins Weapon System Trainer Facility at HMAS
Stirling in August 2007
Lockheed Martin,
Thales,
STN Atlas, and Raytheon were approached to provide tenders to design and assemble a new combat system for the submarines, with all four submitting proposals during early 2000. In May 2000, after the DSTO tested operational versions of the proposed combat software packages, the Lockheed and Thales tenders were eliminated, despite the Thales proposal being rated better than Raytheon's. After indepth testing of the remaining systems and observations of the systems in action, the German STN Atlas ISUS 90-55 aboard an Israeli
Dolphin-class submarine and the American Raytheon CCS Mk2 aboard a USN
Los Angeles-class submarine, it was decided that the STN Atlas system was the best for the class. The replacement program received Australian government approval in September 2002. The second combat system development program proceeded with far fewer problems, and took the tactical and fire control components from the CCS Mk2 system, and the sonar interface component from the fast track program. The system is the AN/BYG-1 that was developed for the new USN
Virginia-class submarine and has since be retrofitted to the whole USN fleet. The system can receive new software releases and hardware can be upgraded with new versions of the system regularly released with the version operated by a boat dependent on its full cycle docking schedule. As of the launch of the first submarine, the project cost had increased from A$3.892 billion in 1986 dollars to A$4.989 billion in 1993 dollars, which corresponded to the rate of
inflation during that period. By 2006, A$5.071 billion had been spent to build the submarines (excluding the fast track program); after taking inflation into account, the project had run less than A$40 million over contract. Of the A$1.17 billion allocated to the fast track program, only A$143 million was required to fix problems where the submarines did not correspond with the original contract: the rest was used to update components that were
technologically obsolete and make changes to the submarines beyond the contract specifications. ==Characteristics==