Seakeeper & Hullspeed

I would like to know if by using seakeeper and hullspeed, model testing could be avoided??

I have never used them no i have no idea of their capabilities.

Is it worth investing in them? Has anyone compared the results obtained from them to that of tank testing?

thanks

 

Hi KRL,

Hullspeed and seakeeper have been around a while, seem to be stable and are accepted by many people/professionals/companies. So, they do what they say they’ll do quite well.

They are basically algorithms that link and calculate various equations, assumptions and approximations together with an approximation of your vessel design. If you truly understand the equations and assumptions etc and (importantly) you understand how well your design fits the limitations of said equations and assumption etc – then you will have a reasonable idea of how accurate your results are. For example: How valid is the wall sided expression for your design? And so on…

They are not virtual test tanks. However if you were experimenting with a candidate design and you had full scale or tank test data for a similar vessel/s then you should be in a position to validate (or not) your hullspeed/seakeeper results –however again you must understand where error will be creeping in.
Indeed the software and tank testing are complimentary. You might map out a design space using the software and then validate it by tank testing strategic point/s within the design space. or you could map out the design space with the software and go on to build without tank testing. In both cases you must try to understand where your assumptions are and what impact they have on the results for both the software and the tank testing.

In short: hullspeed and seakeeper (like most software) are tools used to leverage the operators understanding – they are not experts within themselves. Thus the greater the operators understanding, the more powerful the results. The same can be said for tank testing.

To tank test or not is dependent on (but not limited to) how conventional your design is, how well it fits various formulas/approximations and your budget.

Rob

 

Thanks, Rob.
Does your explanation also hold true for hullspeed's wave predictions? I can't find any comparison with tank tests for that.

 

Hi guys,,

I totally agree with all that Rob said. I would say all maxsurf related packages are fairly relaible and robust and you can definitely present your results very formally.

However, hullspeed incorporates different semi-empirical methods, mostly based on regression formulae that were created after a series of model tests by different researchers / universities and so on. This is NOT a first-principles analysis. If you want first-principles hydrodynamics you should use some potential flow code like ShipX. CFD methods are out of question anyways. But if you cant have a tank testing, then you could create your own semi-empirical code based on a well-known series or method (Delft series, NPL series, series 60, Savitsky and so on) in an excel spreadsheet (macro-enabled) or other programmable code and see if the results are fairly similar. Hullspeed has perhaps more sophisticated versions of these methods, so dont expect to come up with the same results, but they should be similar.

Also you have to bear in mind that hullspeed wont be able to grasp minor changes in the geometry especially changes in local areas. This is because small changes (for example in the keel, in the bulb, in the bow, in the transom, in the skeg and so on) do not necessarily affect the basic dimensionless parameters (ratios) with which regression methods work.

It can also be quite a tricky task to choose the appropriate resistance prediction method for your problem. If you are using Hullspeed just to compare different hulls with each other (or within hull shape/size optimisation routines), it's ok and actually pretty convenenient, but if you're using Hullspeed to precisely define your power requirements, then it can be quite tricky...

Finally, as Rob mentioned, your hull has to be within the validity range of the parameters set for the series that generated each resistance prediction method. Therefore, the more conventional the hull, the more accurate results you are to expect. Conversely, the more extreme the hull, the less reliable the results.

As far as Seakeeper is concerned, it is really convenient as it also incorporates regulations criteria and so on.

Overall, I would definitely recommend using these programs without real need for verification of the results from tank testing, but always being aware of the validity range of the parameters for each method and that this is not a first principles calculation.

 

Not unless it's a stock standard hullform. You still need to know which series to run with, sometimes none of the resistance curves come close to the scale model results.

I've never checked model RAO against seekeeper but I'd presume it would be reasonably close and the animation looks good .