Center of mass & center of buoyancy analysis

buoyancy

#1

Hello,

I’m currently working on characterizing the dynamics of the OpenROV 2.7 and thought you guys would find these results interesting.

I have found the locations and magnitudes of the centre of mass and centre of buoyancy in my model using CAD. The centre of mass was found by incorperating the relevant weights and densities of all components into the model; and the centre of buoyancy through finding the centroid of the submerged volume. The magnitude of mass is the weight and the magnitude of buoyancy is the mass of displaced water.

In my model, I measured the mass as 2519g and the estimated total weight from CAD is 2513g, this agreement gives confidence in the results. The calculated buoyancy force is 24.47 N whereas the weight is 24.71 N. So in this case, the ROV is slightly heavier than water though in cases where water is denser (saliter, colder) it may be neutrally buoyant or positively buoyant.

In equilibrium state the centre of mass lies below the centre of buoyancy. The intensity with which the ROV returns to its equilibrium from a disturbance is dependant on the relative positions of the COM & COG, the inertia properties of the body and also the drag characteristics. From this analysis the COM & COG are seen to lie fairly close to one another meaning the restoring force is less than if they were further apart.

Another interesting thing is that the thrusters axes coincide with the center of mass, which is ideal to prevent a moment about the COM while thrusting. I’m not sure if this was a design consideration or a coincidence, does anyone know?


#2

For the COB calculation you used freshwater (1000 kg/m3), right?


#3

Hey Roy, yes indeed, 1000 kg/meter cubed for the buoyancy force - the location of the COB shouldn’t chage for other densities, only the magnitude.


#4

Great work @zacmacc!

Interesting thought, that an enlargement of the vertical distance between the COM and COG would increase the ROVs stability. Did you have look at the horizontal distance between the two? From your image it looks like the ROV would be slightly pitched, which matches my experience from the field.

When talking about the thruster axes in correlation with the COM, it needs to be considered that hydrodynamical drag has a significant influence on the stability while thrusting. Due to the asymmetrical layout my ROV 2.8 does a significant dive during thrusting. So it might be helpful, to move the COM a bit, so that the moment resulting from thrusting counteracts the hydrodynamical moment. I will take this into consideration when adjusting the final balance of my ROV. Thanks!


#5

Thanks @Fe3C. Yes, as you said I think this does explain the slight pitch that I have also noticed. The COM & COG more or less lie on the vehicles plane of symmetry. Yes, very good point about the hydrodynamic drag and this is surely the main cause of moments while thrusting. As part of the dynamics characterization I’m currently also working on a CFD simulation, I intend to find the center of pressure for forwards motion. Regarding moving the center of gravity to counteract the pressure I think that this could bring some improvement, though, the drag force is a function of velocity so the effect would not be stable at all speeds. Perhaps through modifying the geometry with some sort of spoiler the center of pressure could be moved so it’s also in line with the thrusters. If I have time I may play around with this and give some recommendations. It would be interesting if implementing this in reality would bring noticeable improvements to the stability of the vehicle.

More to come soon!


#6

That is correct as long as the ROV is fully submerged :wink:


#7

I would assume the moment caused by the thrust force is also indirectly a function of the speed, as the force needs to be higher for greater speeds. However, both the drag moment and the static moment from thrusting do not correlate linearly with the amount of thrust. So maybe the counteract each other for a bigger variety of speeds.

However, in field I have only little problem with unstable pitch, as straight forward movement is only a small part of the entire exploration. And apparently all these problems have been tackled with the Trident (mostly likely due to its improved symmetry drag wise). So for explorers that prefer straight forward movement, they could use a Trident.

Finally, I feel that ROV is very sensitive to only small changes in the water properties (salty, temperature etc.) that it is very hard to find the perfect configuration for a wide variety of applications.


#8

I’ve been watching this thread with great interest. I’m afraid my exploration has not been good so far. I don’t really understand the technicalities of it but I do know that it is almost impossible to steer the ROV in any sensible way.
Most of what we have been looking at is on the seabed. When you go forwards, it just dives and hits the bottom hard. When you go up it goes astern, when you go astern it goes up. When you turn, it doesn’t really do much, presumably because there’s too much drag from the tether.
I’m really getting quite despondent, it just seems to be pot luck what you actually see, there’s no way of steering towards something and stopping.
We went out to dive on a serpulid reef nearby and it was just careering about all over the place. I have checked the thrusters and they are all working in the right way. What are we doing wrong?
Andrew


#9

@e4andy, what is your pitch angle without thrust? (Maybe use the IMU to determine the actual degree?) Is it a strong nose down pitch? The ROV does dive slightly with forward drag to due the above discussed hydrodynamical drag but it should not be so bad, that you can not avoid hitting the bottom. And when going slow this issues should not be that big of problem, especially with depth hold.

When i use the smaller thruster settings (e.g. 1 or 2) i find the ROV to be very sensitive in steering, in all dimensions.

You could try to do some balance work on your ROV, so reducing some of the weight at the front and attaching more to the back (to tackle the go-forward-and-dive issue), maybe also in field.


#10

Hi Lukas, I will try adjusting the pitch. It is a little nose down but not much. If I have it on low settings it doesn’t really move much, I have to use 3 or 4 to get any significant movement. Maybe the currents are too strong.
It would be really good to try it in a swimming pool where I can see what is happening.
Have you done any interesting dives?
Andy


#11

Here’s some preliminary analysis on the fluid induced forces and moments, here are the findings:

The test was done at a speed of 0.5 m/s.

Notable forces include drag force (against axis of movement) at 2.247N (229 g) and also a downward force of 0.2203N (22.4 g). This could have an overall downwards moving effect though I’d expect it to be marginal.

Of greater importance when talking about stability are the fluid induced moments, the center of pressure is a point where all of the drag forces effectively act on the body. In this case it was found to lie about a centimeter below the center of gravity, which could explain the observed downward pitching. Its important to note that in my fluid simulation I excluded the stabalizing bars underneath the battery tubes for simplicity reasons, this would have the effect of lowering the COP even further.

In trying to understand this behaviour I have a few hypotheses, one candidate is the region of low pressure underneath the main tube caused by the high velocity flow through this region, though looking at figure 4, it looks like this should be equally balanced by the high pressure region at the front of the tube. Possibly another large contributor to this moment is the flat battery cover at the front, the perpendicular distance from the COG is large and it may be significant.

In terms or remedies I intend to try a few things, I will add an upwards slant to the front battery cover and see if the moment can be corrected that way. Another thing I may try is to add a spoiler along the stern of the ROV (at the top) and tilting it upwards.

Figure 1 - Surface pressure distribution

Figure 2 - Surface pressure distribution


Figure 3 - COP Shown in red with the direction of equivilant force shown in blue

Figure 4 - COP Shown in red with the direction of equivilant force shown in blue


#12

hi zac,

what software u used for this study?