by Peter Yule
submarine and Kockums had not designed such a big
submarine before – but neither had anyone else. Size in itself
was not the difficulty, but as a submarine gets bigger there
are more systems and more complexity, which means more
specs, more drawings, and more work.
Further, almost all of the equipment that went into the submarines
was specific to purpose and very little ‘off-the-shelf’ equipment
134
T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y
was used.3 Kockums, even where it did not design the equipment
itself, was responsible for supervising the sub-contractor and for
integration. The company had 25 000 drawings to do, as well as
drawing up parts lists and preparing the procedures for production
and testing.
Kockums’ partners in ASC felt that the biggest risk of the
project was the design process, and feared that this might be
beyond Kockums. Although Kockums’ risk management evalu-
ation convinced the partners that the potential problems could be
successfully overcome, most people involved believe that Kock-
ums underestimated the size of the design task. Mark Gairey, a
naval architect with the Australian navy team in Malm ö, says that:
Kockums didn’t think the design effort would be more than
20 per cent more than on the Swedish submarines, but it
turned out to be about double. While doing Collins they were
building one class of Swedish submarines and in detailed
design of the next one and doing detailed research for the one
after that. And while they were doing this they were finding
our project took more work than they expected. To do the
work on Collins as well took more resources than they had
and they are not the sort of resources you can just go out and
buy. They robbed Peter to pay Paul to fix the crises as they
emerged.
While those at Kockums would think this is an overstatement,
members of their design team agreed that the design process took
longer than they expected, largely because they were dealing with
a new customer and they had to get used to responding to new
questions and more formal ways of working. An additional com-
plication was the need to rely on others. Gunnar Öhlund pointed
out that:
ASC was responsible for procurement and Kockums relied
on information from suppliers to be delivered on time and of
high quality. Kockums was slowed by others being slow – the
structure of the organisation might not have been the best.
Vendor furnished information was a real challenge. Kockums
should have been stronger explaining to ASC the need to
deliver information in a timely manner and this led to
problems settling details for the design.
D E S I G N I N G T H E C O L L I N S C L A S S
135
About 18 Australian engineers and designers joined ASC and were
sent to work with Kockums in Malm ö. The intention was that
they would work closely with the Kockums design team to learn
skills and to equip them for management roles during the con-
struction phase. Martin Edwards was one of this group, and he
recalls that the Kockums people were obviously fairly cautious to
begin with and would allocate one or two very small or simple
design tasks or engineering activities. They would delegate more
responsibility after the Australians had demonstrated knowledge
and ability. His impression was that after about four to six
months a number of ASC people were taking on more significant
roles within the design activity but being mentored by Kockums
designers:
I worked on designing aft control surface arrangements,
integration of the masts, garbage ejector and compactors,
main propulsion motor installation, some of the more
significant mechanical tasks, their integration into the
submarine.
Edwards found that in designing Collins, Kockums used the prin-
ciples from the V ästerg ötland and other earlier Swedish sub-
marines, but most of the detailed design was completely new,
partly because of the great difference in size and partly because
much of the equipment specified or chosen for Collins was new.
So, while the hull outline was virtually indistinguishable from the
V ästerg ötland and elements of previous designs could be used for
mounting equipment and so on, most of the detailed design work
had to be done from scratch. For example, ‘the masts were com-
pletely new to the Swedes and it [was] an Italian design that was
used, so they had to be designed and integrated from first prin-
ciples’. The Swedes aimed to build on previous work but it was
often found that only minor details could be used.
As well as the ASC engineers and designers working with Kock-
ums, there was also a navy team of about 20 people, headed by
Captain Rick Canham, based in Malm ö to supervise the design
work, clarify the navy’s requirements and report back to the
project office in Canberra. Mark Gairey was given the title of
‘senior technical representative, Europe’ and recalls that the team
was ‘on the ground in Sweden before ASC’. He felt that Kockums
136
T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y
had a pretty good idea what they had to do. They’d been
designing and building submarines for 70 odd years.
However, one of the problems we encountered was that they
kept looking at things in a Swedish frame of reference. There
were some interesting and prolonged discussions which
ended up with us saying: ‘That’s fine but we simply can’t live
with that solution in an open ocean environment. It’s all very
fine for you guys where, if it goes wrong, you’re sitting in a
little puddle of water and can fix it, but we can’t.’ We had
several discussions of that type on and off because ultimately
we had a specification but there were still gaps in it.
When the navy team and Kockums could not agree on an issue,
they would report back to the project office in Canberra outlining
the problem and saying what they thought should be done. It then
became a contractual matter back in Australia between the project
office and ASC.
To help ASC achieve 70 per cent Australian content Kock-
ums set up a design office in Adelaide. This office was manned
by Swedes and Australians, with most of the Australians who had
been working in Malm ö returning to Adelaide by 1990 or 1991. At
its peak this office had about 40 Australian and 20 Swedish staff,
headed first by Jan Hansson and later by Martin Edwards. The
fact that an Australian took over from a Swede was in itself one of
many vindications of the project’s aspiration to develop Australian
skills. As design manager at ASC Edwards was responsible for all
the detailed design groups and some of the engineering support.
The design groups covered hull design, mechanical design, out-
fitting design, weapons handling, electrical design and electrical
installation, fo
cusing on finalising the detailed design components
and supporting production.
With design work proceeding on opposite sides of the world,
communication between the design offices was vital. There were
many issues that required discussion and debate between Adelaide
and Malm ö and, with e-mail only coming into common use about
halfway through the project, they had to be resolved by fax or
telephone discussion. It was vital that both design offices have the
same drawings and documentation, so a dedicated link was set up
for daily transfers. Drawings done in Malm ö would be encrypted
and sent by satellite, navy to navy, to Adelaide.
D E S I G N I N G T H E C O L L I N S C L A S S
137
Gairey recalls that the Australian navy team at Malm ö had
many debates with Kockums’ designers, but he does not recall
any issues that could not be resolved. Significantly, he says: ‘the
problems that arose later caught us by surprise. They were things
we didn’t expect. Whether we should have been better prepared
is a good question.’
During the controversies that raged around the submarine
project in the late 1990s, many aspects of the design were ques-
tioned. Hindsight is a wonderful thing, but how were these issues
looked at by the Kockums designers, their Australian protégés,
and the project team in the late 1980s?
The shape of Australia’s new submarines was taken directly
from that of the V ästerg ötland. Looked at in profile the two classes
are almost identical. The hull shape of the V ästerg ötland, as with
all Swedish submarines, was developed jointly by Kockums, the
Swedish navy and SSPA Maritime Consulting AB (now SSPA Swe-
den AB) in G öteborg. Hans Peder Loid, who was managing direc-
tor of SSPA when the V ästerg ötland was designed, recalls that all
the design activity for Swedish submarines was between Kockums,
FMV and SSPA and no changes would take place unless all three
were united. It was never a rigid supplier/customer relationship.
If there was anything not completely necessary on the hull, SSPA
would reject it, as a tiny protrusion or an extra hull penetration
could make a big difference to noise levels.
Tank testing of models for new classes was done as a normal
procedure, and he was involved with the tests of the V ästerg ötland
and of the Australian design up to the signing of the final contract.
At that stage the testing done at SSPA had revealed no problems.
Loid emphasised that the shape of the Australian submarine at
that time was the same as the V ästerg ötland except bigger, so
there was no reason why its hydrodynamic signature would not
be identical to a Swedish submarine unless things were added to
it later during the design or construction.
However, the hull design of the Collins class did alter from
that put forward at the end of the project definition study. The
submarine became even bigger – although this in itself should not
have affected the water flow characteristics – and other changes
were made that had the potential to affect the submarine’s noise
level. For example, it was discovered that the sonar suppliers,
Thomson-CSF, had underestimated the required clear angles for
138
T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y
the sonar and the design had to be stretched one metre forward
to accommodate the sonar accessories and one metre aft to retain
balance, while the sonar bow dome shape had to be drastically
re-designed. The modified design was not tank tested, mainly due
to disagreement about who should bear the cost of building a new
model to test.4
Mark Gairey says that in retrospect both the Swedes and
the Australians were looking at potential noise problems from
the perspective of the Oberons, where the noise came mainly
from the machinery. Machinery noise was overcome in Swedish-
designed submarines by construction using resilient platforms
with rubber isolation elements. This link between the design and
the modular construction techniques had been one of the great
attractions of the Swedish bid. However, while concentrating
on eliminating machinery noise, neither the Swedes nor the
Australians anticipated problems with hydrodynamic noise.
Gairey suggests that:
In part it was our inexperience and in part because of the
way the Swedes operated their submarines, they had never
seen these problems. I’m sure that their submarines would
have the same problems if operated by our people, but they
didn’t so it was a surprise for them as well.
The Swedes operated at very short range. They’re very
much a coastal defence force . . . They went four hours out of
harbour and sat there. They weren’t running at any sort of
speed for any sort of distance . . . Whereas we had big
distances to transit to an operating area and we wanted to
transit as quickly as we could, while remaining undetected. It
was a completely new set of requirements . . . Maybe they
could have done more model testing but they were looking at
Collins through a Swedish frame of reference. They didn’t do
with their submarines what we did so they didn’t have the
problems and they didn’t anticipate that we would. They
didn’t realise our operating environment was so different. In
retrospect they should have done more testing but at the time
it seemed that what they did was fine. We were totally
surprised by the problems we had and the Swedes were just
as surprised.
Critics of the Collins class submarines have argued: ‘The shape of
the submarines in the water was a disaster because of the insistence
D E S I G N I N G T H E C O L L I N S C L A S S
139
on six torpedo tubes across the front so that the submarine has a
front like a circumcised dick.’5 The designers agree that the snub
nose of the Collins class creates turbulence when on the surface,
but the submarine spends most of its time submerged, where the
bow shape works well, although the area aft of the bow where
the sonar dome joins the casing was later found to be a source of
turbulence.
Gunnar Öhlund, who worked on the bow design, says the
arrangement of tubes side by side provides compact weapons stor-
age and enables fully automated handling and loading. He argues
that this had no consequences for the hull shape and there are ben-
efits for firing in keeping the torpedo tubes below the boat’s centre
line. This design has worked well with the Swedish submarines.6
SSPA in G öteborg was closely involved in the design of pro-
pellers for Swedish submarines and also for the new Australian
submarine. In the 1980s the Swedish submarines had a problem
with a ‘singing’ propeller but a lot of research and then a few small
marks on the propellers were enough to fix this. However, cavita-
tion was always the main concern with Swedish propellers. SSPA
developed a computer program
that prevented the crew operating
the submarines in a way that caused cavitation and this essentially
solved the problem on Swedish submarines before the Australian
project, although it could emerge if the design was changed or the
propeller was poorly manufactured.
Generally the requirements for the submarines were stated in
performance terms, but in some areas design requirements were
specified. One of these was that the propellers be made from
Sonoston or a similar highly damped alloy because it had been
used in recent American submarine propellers.7 Unfortunately the
Australian navy did not known that the alloy was liable to crack
and had been abandoned by the Americans. Mark Gairey com-
ments that: ‘In hindsight, this was a mistake and we should have
concentrated on the “no cavitation” requirement and left it to the
designer to solve.’
A controversial aspect of the design of the new submarines
was the decision to have flat rather than round bulkheads in the
bow and stern sections. The Swedish navy had used both round
and flat bulkheads and decided that either worked. Kockums saw
flat bulkheads as robust, saving space and allowing the torpedo
tubes to be positioned identically. They have been used in Swedish
submarines since the Sea Serpent class of the 1960s, which are still
140
T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y
in service with the Singapore navy, indicating that hulls built with
flat bulkheads can have a very long life.8
Some significant design changes were initiated by the Aus-
tralian navy team at Malm ö which, while beneficial, often caused
contractual difficulties. One such change occurred with the air
purification system. Greg Stuart, the senior engineer with the
project, recalls that the Swedish system in the original design
was ‘heavy and inefficient’. Stuart contacted Wellman Processing
Engineering of Birmingham, England, which owned the technol-
ogy used for air purification on some British and American sub-
marines. Working with Wellman, Stuart calculated that changing
the air purification system could save length and weight on the
submarines, and he developed a proposal for a ‘mini-scrubber’
to remove carbon dioxide and other acid gases. The first unit
was called ‘Kylie’, ‘the little Aussie scrubber’, and the second was
