How to effectively model oceanographic surface buoy dynamics
Originally published on dsaocean.com
Flying sheep may help you get to sleep, but they can save your life, too. In 1934, the Italian army had a huge logistics problem: crossing one of the most inhospitable deserts in the world. Resupply for the trip was vital, and the army had to consider their options carefully.
One option was for soldiers to carry their supplies. But weighing down soldiers wasn’t possible: they would move too slowly under the oppressive sun and wouldn’t survive heat stroke.
Another option was trucking in supplies. But that wasn’t going to work either, as the stifling desert heat was already spoiling their food.
The army was left with only one viable option: using aircraft. Soldiers were able to travel quickly through the desert on foot, and supplies were sent by airdrop. Everything, including live sheep, was sent in by parachute! It was the only effective option.
In contrast, there’s often more than one effective option in modelling surface buoys. Surface buoys are a critical component of oceanographic mooring analysis and need to be considered by mooring designers carefully. Without careful consideration of dynamics caused by harsh ocean waves, the buoy and mooring won’t survive. In this article, we’re going to cover a few useful approaches for modelling surface buoys:
- point mass
- simple rigid body
- complex rigid body
First, we’re going to look into point mass approach.
The point mass model is one of the most straightforward approaches to use
They’re straightforward because they ignore all rotational effects of the buoy: a single point represents the hull. It can still account for the linear buoy motions: heave, surge, and sway. But it also means only the most basic shapes can be used to represent the buoy hull, like a sphere or in some exceptional cases, a cylinder.
It’s natural to underestimate simple models
While simple, point mass models can be very extremely useful. Often, a point mass model is more than capable enough to analyze an oceanographic mooring response in currents and waves. This level of detail may be all that’s needed to finalize a robust mooring design.
There are only a few critical parameters needed to get started with a point mass buoy model. Fundamentals like buoy mass, net flotation, and basic hull shape are required and sufficient for many simple systems. But in certain circumstances, buoy rotational effects can be significant. This takes us to the second point on modelling surface buoys: using a simple rigid body approach.
Like a point mass model, a rigid body model accounts for all linear motions
These are heave, surge, and sway. But the difference is that rigid body models also account for rotational effects: roll, pitch, and yaw. These rotation motions may be vital because they can have a significant impact on the equipment used on the buoy. Of course, this means more information is needed to get the model set up correctly.
But some of this information is not easy to find. For example, most buoy spec sheets provide mass, but usually don’t contain information about what the buoy rotational inertia is. This lack of information is not a surprise because the rotational inertia can change depending on the specific buoy loadout.
This is where the simple rigid body model really shines
With a few assumptions and approximations, mooring designers can estimate many of the additional missing parameters. The first thing is to approximate the hull shape with something simple like a sphere or cylinder. The rotational inertia is easily calculated from these basic shapes and the mass of the buoy.
This is a significant improvement on the point mass model because there is some basis for the rotational effects. A rigid body model can also handle more detail like an offset mooring connection point below the hull. However, it’s still an approximation to the actual buoy shape.
Regardless, it’s certainly a starting point to understand rotational motion. Like the point mass approach, it doesn’t need a lot of information to set up the model. But the simple rigid body model relies on a lot of assumptions to help fill the gaps in knowledge. So what can you do if you have a lot more details on the buoy on hand and want to make use of it? This brings us to the third and final section on modelling surface buoys: using the complex rigid body models approach.
There are a lot of assumptions used in the simple rigid body approach
Complex rigid body models are about reducing some of those assumptions. Using a complex rigid body approach will undoubtedly require a lot more information. In some circumstances, a detailed CAD software model of the buoy may be available. This CAD model can help provide the specific buoy hull geometry. The CAD software can also offer more accurate details like rotational inertia and the location of the centre of mass.
However, it can take a lot of time to set up a detailed CAD model. On top of this, generating a custom mesh of a buoy hull can be a challenging task in its own right. Nevertheless, it is the most flexible and powerful approach to modelling surface buoys.
What about the computational cost of these approaches?
It’s reasonable to expect the computational cost to increase if the model complexity increases. This is always on the top of my mind because more computational cost means more time needed to compute a mooring design. However, in this particular case, it’s not so obvious.
An oceanographic mooring simulation tends to require a fair number of elements in the mooring line. The extra equations introduced between a rigid body and point mass model does not make much of a difference in the computational time. That said, some buoy hulls can have intricate shapes. These intricate shapes mean a lot of geometric detail needs to be used in hydrodynamics calculations to resolve the hull forces. In practice, a complex rigid body model with a detailed hull might take something like twice as long to compute a mooring solution when compared to the other models. But it all depends on how much detail is in the hull.
Let’s look at an example
We’ve looked at the Southern Ocean Flux Station (SOFS) system in previous articles. In 4km deep water off the coast of Tasmania, it uses a 3m diameter buoy at the surface. We set up a detailed configuration of the entire mooring and surface buoy in ProteusDS. The ProteusDS model allowed us to examine what happens to the mooring loads and buoy motions when using the three different model approaches.
The general environmental conditions used for this example were 0.5m/s surface current dropping to zero after a few hundred meters. The wave conditions were 3m significant wave height and 10 second spectrum peak period.
We set up separate ProteusDS projects using the three different surface buoy models configured as a point mass, simple rigid body, and complex rigid body models.
A ProteusDS ExtMassCylinder was used to represent the simplest approach – the point mass surface buoy. The diameter and length of the hull were set to encompass the flotation volume of the SOFS buoy. The mass and maximum wet weight were set equal to the SOFS buoy mass and maximum reserve buoyancy, respectively.
The simple rigid body approach was represented in ProteusDS by the Rigid Body model along with a Cylinder Mesh hull. Just like the simplest model, the hull geometry was set to encompass the flotation volume of the SOFS buoy. The mass was set equal to the SOFS buoy mass. The rigid body inertia was approximated using a solid homogenous cylinder with the centre of mass right in the middle of the flotation volume.
The complex rigid body model was represented using a ProteusDS Rigid Body model along with a Custom Mesh hull. The hull mesh was formed based on the shape of the SOFS buoy hull. The rigid body inertia and centre of mass location were computed by a detailed CAD model provided by CSIRO mooring engineer Pete Jansen.
Each model approach produced 3 hour peak mooring tensions of 23kN, 24kN, and 25kN, respectively, in the set environmental conditions. This shows how well the simplest model did at driving the peak loads in this storm condition. But what about buoy motions?
The simple point mass model tells us nothing about the rotational motion of the buoy. But there are results from the simple and complex rigid body models. In the 3m significant wave height conditions, we compared standard deviation and maximum buoy tilt. Each rigid body model produced 5.8deg and 5.7deg standard deviation. The maximum tilts were 37deg and 39deg, respectively. Since they were so close together, it shows for this kind of buoy and mooring combination, the simple rigid body approximation does a reasonably good job.
The Italian army only had one option when crossing a desert to succeed: pass as quickly as possible and airdrop supplies (live sheep included). Fortunately, you have more than one option when analyzing buoys in your oceanographic mooring. Now it’s time to review them.
The first approach is the most direct and straightforward way using a point mass. This approach is the fastest to set up and easiest to use. It often does the job for mooring design. But it does not account for any buoy rotational effects.
When these rotational effects may be significant, it’s time to look at the second approach: the simple rigid body model. The simple rigid body approach is excellent when there’s limited information on the buoy available. It uses underlying approximations and assumptions to fill in gaps in details like those for rotational inertia.
If you have a lot more information on hand, and likely a CAD model to help out, the third approach can be useful: the complex rigid body model. Not for the faint of heart, it can take a lot more time to set up. But you can use more detailed buoy hull geometry, and the rotational inertia computed from a CAD tool for much more accuracy.
Next step: check out this video tutorial comparing point mass and rigid body models
DSA develops ProteusDS as a software tool to help evaluate oceanographic mooring designs. We post a variety of materials online, including video tutorials on our YouTube channel. Check out a video tutorial below showing more detail on a comparison of point mass and rigid body models in the software here.