Best Marine Design Software for Hull Design? (2001-2005)

[Admin]

If you could only use one marine design software package for hull design, which would it be?

Autoship (Autoship Systems Corporation)

DefCar (DefCar Engineering)

Fastship (Proteus Engineering)

HullCAO (HullCAO)

Hull Form (Blue Peter Marine Systems)

Maxsurf (Formation Design Systems)

MultiSurf (Aerohydro)

Prolines (Vacanti Yacht Design)

ProSurf (New Wave Systems)

Rhino (Robert McNeel & Assoc.)

Naval Designer (US Sales by Forum Marine

SeaSolution (SeaSolution)

TouchCAD (Lundström Design)

[Stephen Ditmore]

I'm a bit surprised to see how well Prolines is doing in the pole so far. Many years ago it struck me as a very crude imitation of AutoYacht. Has it dramatically improved since?

I don't think Prolines has changed that much in the last 2-3 years since the change from version 6.0 to 98 (hence the name Prolines 98 still).

Overall I think Prolines might be a little easier to get started with and have less of a learning curve than AutoYacht. In fact I would say that Prolines is intuitive enough that most users probably won’t have to look at a manual before getting started with a design. For example you can start with a parametric hull shape in Prolines which generates a variety of basic hull shapes (no chine power, 1 chine, 2 chine, round, or canoe, complete with stern) or you can start from a table of offsets. And a lot of people might get started with the Basic version which doesn't include any file input/output capabilities (e.g. no dxf out, etc.) but which is only $350. (the full version is $1500 though which gives you import/export!)

The icons and menus are mostly "one click does it" in Prolines, whereas they are a bit more involved in AutoYacht. In Prolines you're working with "the hull" whereas in AutoYacht you are involved directly with adding and shaping surfaces 0 it's more technical and more time consuming, but of course the user will probably learn more about surfacing and the actual drawing process. But for someone designing a limited number of boats, Prolines might let you focus more on the project at hand than learning to use a high-powered cad program. The calc menu is also pretty easy to use in Prolines – one click and you have a graph.

The one thing that I do not like about Prolines is that the control points are not on the actual surface.

Note that I do not use either program right now, but I have looked at both recently.

[Nomad]

The best 3D Marine Design software is Solidwork!
I have been using it for over a year now and I have yet to see a program that can compare.

[tspeer]

What are people looking for in way of the capabilities of their design software?

Just about any CAD program will handle NURBS surfaces, import and export common CAD file formats, generate offsets, and render 3D objects. Marine CAD programs have at least a basic hydrostatics capability. So all these can be taken as given, although we could have lots of discussion about the merits of different programs in each of these categories.

But in the end, the program is just a digital substitute for a drawing board and the result is just a pretty picture. Unless the program also has the ability to contribute to other engineering calculations.

One of the key items I was looking for in my CAD system was programmability. In that respect, I've found Multisurf's support of object linking and embedding (OLE) to be outstanding. It's fairly straightforward to call Mulisurf functions from Excel's Visual Basic for Applications. This lets me combine the flexibility of the spreadsheet with the power of the CAD program.

For example, here's how I've been designing the hull shapes for my sailing trimaran design. I start by entering the desired characteristics into the spreadsheet. These include the Cp, Lcb, displacement, etc., plus section parameters like the beam/depth ratio, angle at the waterline, deadrise, etc., at each end and the middle. The spreadsheet generates a complete curve of section areas and interpolates the section parameters at each station. From these it generates a set of provisional control points to define the hull.

Here's where the OLE capability comes in. The spreadsheet treats Mulisurf as a slave to generate all the control points in the CAD model, create surfaces, generate offsets, and compute the hydrostatics. Once it gets back the actual hydrostatics from Multisurf, the spreadsheet scales all of the sections to match the design curve of areas, updates all the control points, and has Multisurf return the revised hydrostatics.

The result is a hull which precisely meets the desired characteristics in every respect. Something like 200 control points are used to design the hull, including topsides and deck, so there's no compromise on the shape, yet the hull is very fair due to the analytical nature of the interpolating functions.

But the big payoff is the hydrostatics and any desired geometric data are placed back into the spreadsheet where they're available for other engineering calculations like determining sinkage and trim, performance, and detailed weight amd mass properties.

I've found this flexibility and programmability to be cruicial to my whole design process.

[Stephen Ditmore]

Tom,

Check out my post at
http://boatdesign.net/forums/showthr...=1756#post1756
It seems to me you could use this formula in your spreadsheet to optimize longitudinal distribution of volume (i.e. the curve of areas). Could be slick!

S

[tspeer]

It looks to me like the formula is for the potential flow about an axisymmetric body using Rankine sources distributed along the centerline. The problem is determining the source density distribution, sigma. You might be able to determine what you wanted for the velocity or pressure distribution along the length, say at some distance from the centerline. Then solve for the source strengths that matched the velocity distribution and then see what kind of volume and hull shape that resulted.

Actually, the formula I originally started with for my cross sectional area distribution was based on something very much like this. In supersonic flow, if you assume a slender body that has very weak shock waves, you can use line sources to come up with the body shape that has the least wave drag for a given volume and length. This is known as the Sears-Haack body, and it has the area distribution

X = L/2(1+cos(theta))
S = 4*V/pi/L * (sin(theta) - 1/3 sin(3*theta))
0 < theta < pi

X = distance along body
L = length of body
V = volume (for whole body of revolution)
theta = independent parameter used to make things easier

I thought, "Wouldn't be a hoot to see what kind of hull would result from using a Sears-Haack area distribution?" So I tried it, and it didn't look too bad. Then I found a paper concerning a whole series of tow tank tests done in Japan to determine minimum drag shapes for ships. The area distribution they used reduced to the Sears-Haack formula when you chose the right prismatic coefficient! And that value, around 0.589, is pretty much the optimum Cp for a Froude number of 0.4 - right near hull speed where the wave drag is greatest. Hmmm....

The Sears-Haack body has its maximum thickness at the midpoint, however, and for a boat you want the maximum area to be somewhat behind that. So I generalized the formula some more to be able to place the lcb where I wanted it to be.

If you want more details on the formulae I used, see http://www.basiliscus.com/CaseStudy/geometry.html.

Cheers,

[Leo Lazauskas]

Tom Speer wrote:
...And that value, around 0.589, is pretty much the optimum Cp for a Froude number of 0.4 - right near hull speed where the wave drag is greatest. Hmmm....

Tom, maximum wave drag occurs, in general, for Froude number = 1/sqrt(pi) which is approximately 0.564.

Cheers,
Leo.

[vacanti]

Vacanti Yacht Design LLC is pleased to announce that we are in the process of large changes in our design software. We will roll out new versions of PROLINES, WINGS and FOIL.

In recognition of the continuing reduced costs of computers, we are lowering our prices to more closely reflect the cost of computing in general these days.

We have also completely revamped our web site and will shortly post several new design articles on Keel and Rudder Design.

We appreciate comments made in this forum about our products and will use them to help upgrade and improve them dramatically this coming fall and winter.

David Vacanti
http://www.vacantisw.com

[Leo Lazauskas]

Are there any plans to include resistance estimates based on physics (e.g. thin ship theory, Kelvin-Neumann, Neumann-Stokes)as opposed to empirical methods (e.g. Holtrop Delft, etc)? If so, will your product be able to handle multihull resistance?

Good luck with the new line!
Leo.

[vacanti]

Leo-
Thanks for the question - We don't have plans just yet for Physics based drag models. Do you have a source where I might find means to implement such analysis methods? Keep in mind the methods need to be reasonable and compliment the use of NURB surfaces.

I am always open to ways to improve our software, assuming the interface to the user is straight forward and easy to understand.

Dave Vacanti
http://www.vacantisw.com

[Whaler4me]

Is there one software package that does everything, ie, is there software that does a good job at rendering as well as calculations for hydrostatics.

Anyone using Pro engineer ??

[Stephen Ditmore]

Dave:

Great to hear you're still upgrading the foil software! I'd like to know when you feel it's really ready. FYI it looks like you may have a new competitor at http://www.hanleyinnovations.com/

[vacanti]

Stephen-
I have known about Hanley for some time - but it appears he has a new keel design program. We have implemented something far less costly and more flexible in LOFT 97 ($95). In LOFT 97 you can design keels and rudders with curved trailing edges, compute lift and drag forces, compute lift loading and ballast loading as well. All NACA foils and a few others are also included, more can be added.

WINGS 32 will far surpass LOFT 97 in terms of design flexibility. We have a few bugs in the demo on the web just now but that will change soon. I should have a new version up in a day or two. This is a full NURB surface keel and rudder design program with full bulbs etc possible. Allows placement of up to 50 different foils (not advised but possible).

FOIL 97, will also be revised to directly export WINGS 32 version foils.

Dave Vacanti
http://www.vacantisw.com

[tspeer]

I must be missing something here. Why would any CAD system need to have a separate program for doing foils? A foil is just another solid object or surface, typically created by interpolating between given stations or sweeping a section shape along a control curve. If you can define a surface using NURBS', why not just specify the coordinates from published section shapes? I don't understand why this wouldn't be an inherent part of any CAD system intended to precisely loft a given shape.

As for programs to design the section shapes themselves, the key capability is the boundary layer calculation. And in particular, the ability to handle some amount of flow separation. The only low-cost code that can handle separation is XFOIL. Predicting transition is also very important, and XFOIL uses the modern e^n method which calculates the stability of the laminar boundary layer, while older codes like Eppler's use a simpler empirical correlation.

Take a look at a low Reynolds number pressue distribution from Eppler, Handley, Dreese, etc. Do you see the tell-tale flat spot that indicates the presence of a laminar separation bubble? No. Eppler tells you an excessively large bubble has probably formed, but can't tell you what it's doing to the drag or if it's burst and the foil is stalled. From what I've seen of their promotional material, Handley and Dreese don't even address the issue.

The only codes more capable than XFOIL are ISES/MSES (also by Drela and using a similar boundary layer calculation), and various 2D and 3D Navier Stokes solvers. These are all intended for the professional aerodynamicist and priced accordingly.

Aside from the section data, what you'd really like to know to design a foil is what the induced drag and angles of attack are so that you can design the planform shape and twist (for an asymmetrical shape), and get the 3D lift and drag coefficients. Typically, these effects will outweigh the differences between competing section designs. It's not that difficult to get a first cut at the 3D effects using a lifting line or vortex lattice approach (see the spreadsheet at http://www.tspeer.com/DesignTools/vortex95.xls) and this is something that might make it worthwhile to have a separate foil program.