Is the future real – or virtual?

There are a number of voices on all sides. Tom Cox of Lloyd’s Register says: “I am a computer geek and I say ‘you never trust, never believe a computer until you have built the real thing’.”

On the other hand, Karl Randle of Ulstein says he has been working on the Ulstein X-Stern for the SX175 “and the results we are getting with the tank tests are incredibly close to what we’d predicted with the CFD”. Further, he points out “there’s no issue with the scaling effects that you get with model tests”. So, will it begin to erode the ground traditionally taken by tank tests?

There are a couple of balancing elements to consider. Of course with CFD there’s a lot of work up front, and with a departure like the X-Stern, it took time even though Ulstein has a wide library of numerical models. “There were several different variations under comparison with changes to the keel angle, the width on the deck, the slope of the transom showing where we could make compromises between deck area, load, stability, motion, comfort and fuel consumption,” says Mr Randle. His colleague, Ivan Schrooyen, added since the X-Stern is a very new concept, “the biggest challenge has been proving the concept works by using numerical tools”, especially as it meant getting rid of the transom completely and venturing into “unknown territory”.

But the CFD underscored the potential for a number of intuitive benefits says Mr Randle: “The superstructure is quite large at the forward end of the ship and this provides a natural weather vane thus enabling stern operations for lower FOC. We’ve also noticed that by modifying the aft end it’s lowered wave excitation and there is not as much stern slamming in quartering seas due to the newer lines design.” However, the numeric model also pointed to a surprise: a 25% drop in calm water resistance when travelling ahead, and less than 50% stern first, almost comparable to ahead steaming.

So it maybe that more and more will be developed by the mathematical model – though no-one seems comfortable with the idea that tank tests will be abandoned completely.

There’s a number of reasons for this: Angus Webb, Research Fellow in Southampton University’s Fluid Structure Interactions Research Group says: “Certainly we are at that stage where CFD can answer a lot of questions but there are areas where either it’s too expensive to tie up the computing power or the physics isn’t quite there yet.” He adds that “while you can get ‘trends’, they don’t tend to be accurate enough at the extremes: “It’s probably best to get the direction of your development, then pin down the values with the model tank”. Mr Schrooyen agrees with this: “We have plenty of different hull forms in CFD and we can compare them, then we decide which one we take forward to the absolute check-up in the basin test.”

There were further insights to be gained: Mr Randle explains that you can even look into the moonpool via CFD: it showed that the water lost half its energy just by being inside the column – the research we’ve made and tank validations enabled us to look into new developments with confidence. When Ulstein’s Chief Designer Øyvind Kamsvåg challenged the team to thinking about an ‘add on’, that again was fed back through the CFD, which lead to the recently released ‘Extender’ add-on. This is an innovative concept that effectively brings a moonpool into the vessel, making life much easier than working over the ship’s side or ends, especially in bad weather.

Further, it’s not part of the ship’s structure and not counted in the ship’s length – a factor when costs are in focus.

But even having got a good idea of the shape of the behaviour of these design modifications through applying the CFD, Ulstein went back to a couple of days of tank testing to make sure the mathematical predictions were right – especially that huge water resistance reduction that accrues to the Xstern: the CFD was spot on.

Although tank time seems expensive it’s still worth it, says Mr Randle and where money is concerned, comparisons aren’t straightforward.

“Once you’ve made the model everything else seems to be a lot more cost-effective as you can carry out over fifteen or twenty runs inside a day or two, rather than weeks tying up computer power to calculate the effect of waves,” says Mr Randle. However, he points out the models themselves can be as much as £30,000 each – and you often need two, one for the resistance and propulsion (which is determined often by the scale of the off-the-shelf propeller model) and the second for the seakeeping (which needs to be on a large scale as possible), though Ulstein usually tries for a compromise – the SX175 relied on a 1/15 scale model.

And it’s not so easy to change a model, so, the bringing together of ABB’s new azipod design with another existing hull – the PX121 – meant a mixture of tank tests and CFD. “To physically change the tilt angle would mean drilling new holes and resetting all of the motors which have to be set into a flat bed on the bottom. It’s the same if you want to move the shaft or propeller location of the shaft angles – you can do it but it takes a big chunk out of the tank test time.”

The ease of change yields another practical advantage: Mr Randle admits it allows the team to work all night on a project against a deadline without waiting for tank time “then change everything again when industry shows an interest”.

Elemental

The water is, quite literally, another element says Karl Randle: “Waves in CFD are challenging all by themselves and obviously have a big impact especially when you’re looking at a lot of time in dynamic positioning mode. Usually the CFD tends toward particular design waves tuned to a certain size and height, but not many people have taken on looking at the effects of irregular waves simply because of the time it takes: you have to generate all the small frequency and longer waves in the same domain.” But there’s no doubt that this approach gives a certain realism. “There are some interesting developments out there we aim to try out in the near future, offering to cut computation time 10-fold, so the future is exciting”.

However, there’s another, human element and that is ‘seeing is believing’. “The numbers that everyone trusts and respects come from the model tests,” says Mr Randle. However, things are gently changing: “It’s interesting to note that once you start creating videos from the CFD, people start believing the CFD a lot more too.”

He adds that creating videos has become part of the tool set, “sometimes we want to look at the bow wave or the stern wave, or even look down into the moon pool to see what’s happening there in various kinds of sea states”. And it’s also been useful to show owners what exactly running with certain types of trim, moving the ballast from fore to aft or even just forgetting to empty a tank “will do to efficiency”. Again, seeing is believing.

Wind, not just water

While Tom Cox of LR explains that the big new 186m towing tank that is has been built on the Lloyd’s Register-Southampton University site is sized to get the best out of model tests there’s another resource right next door.

Dr Webb says that the wind tunnel could be utilised for a lot more marine research than at present. And he points out that while with the towing tank you get a picture of the effects directly on the hull, there’s actually not much information on the whole water flow, that is, the way the body of water reacts as a mass.

However, the wind tunnel can look at this missing element. Dr Webb explains that while you can’t see wind, you can see smoke – although there’s just too much information in the particles to make sense of that alone. “However, if a cross section is illuminated by a ‘laser sheet’, the camera can only register the smoke particles in that one plane,” he explains.

“The camera takes pictures 25 microseconds apart, and when you flick between the pictures, you can see the particles move very slightly. The computer takes this and calculates entire vectors from it,” says Dr Webb. “If you place the laser sheet horizontally you can see the resistance patterns, if it’s placed vertically, you can see the other effects: it’s actually a very good validation technique.”

By Stevie Knight