Thought to Hydro to Cad to Paper and Beyond.

What works for you? I see a fairly consistant production thread for most design work, at least for the modern designer.

The thought goes on the napkin, that is realized in some form of hydro software, that gets exported to a CAD system for detailing, that gets dissected for drawing generation.

My usual progression has been to create the drawing from the the hydro package, because low end (read: piss-ant) CAD systems that I can justify, don't model boat hulls well. There could be some operator error here, but I seem to have the same problems with my new piss-ant software that I have with my old piss-ant software. The newer software appers to be more capable. but it's the free form nature of boat hulls the quickly separates the MAN software from the BOY software!!! I have reached a designing impass of monumental proportions that has me living in the thinking chair trying to move beyond the modeling of a suitable hull in ANY software.

I do hydro work and hull design in Freeship. We all know there are various form of output from the wonder gift to the boat designing community. Would it not be great to take this output, move it to a CAD package as a surface and start playing ball, immediately. I think there are users of Freeship that are using the IGS export function. I've not been able with any CAD packages I own. Please speak up here with the the software that you are exporting to.

What works for you?

What hydro packages? What modeling/CAD software? How do you make them interface? What drawing package do you use? Is this separate from your modeling software?

Sorry for the rant. I need to stop spinning my wheels and start making progress. Modeling a hull shouldn't be this difficult. I am very curious about what will come back on this thread.

 

Freeship plays sort-of well with Rhinoceros using IGES as the medium of exchange... as long as your Freeship models have a control grid topology that lends itself to B-spline representation. If you do weird things with control points and edges in Freeship, you get some rather... er, interesting IGES coming out. It sometimes takes a fair bit of manual cleanup to get nice clean NURBS out of Freeship-generated IGES (mergesrf, removecontrolpoint x2, repeat, oh crap now the edges don't match, ^z ^z ^z, try a different edge merge order....). On complex models, I often just join all the shards into one polysurface and forget about trying to change the hull in Rhino.

My usual workflow at the moment is:
Preliminary: Paper, Excel / LibreOffice
Basic hull shape: Freeship+, exporting to IGES
Design details: Rhinoceros, calcs in Excel / LibO
Hydrostatics: Preliminary in Freeship+, detailed in ArchimedesMB
Construction drawings: Extracted in 2D from Rhinoceros, finished in AutoCAD

Given an unlimited budget and a laser-like focus on boats only, I would likely switch to Formsys/Maxsurf. But I can't justify its cost at the moment.

I would really like to get away from AutoCAD, and would welcome any suggestions for simple 2D paper-space CAD programs that aren't horrifically bloated in size and cost.

 

What you're describing is ostensibly, a method to make it easy to transfer "data" that has been generated in one system to another and then another and so one, seamlessly. There are software programs out there that can do this, but you talking telephone number figures for cost.

I'm not overly fussed about transfer data from one software to another, since the only data that matters, is correct lines, production faired lines that is. Then the software package you use for structure/production dwgs, can it take the lines with minimal errors in 2D or 3D. If the errors are minor, ie less than plate thickness, then its a done deal. Beyond that...whatever problems you think exits, only appear so on your monitor or the numbers of decimal places by the software output. The reality of building has a far greater error range.

Try this Matt:-

http://www.progecad.co.uk/

Based around AutoCAD, by wayyyyyyyyy cheaper.

 

I loved maxsurf back in school, but like Marsh I can't justify the cost for my personal use and none of the companies I've worked for have used it.

My workflow isn't linear and neither is my software use. I bounce back and forth between different products. Sometimes I might be on autocad for weeks and weeks and other times or other projects I might not touch it. I don't strictly follow a design spiral but that idea is definitely prevalent.

I'd describe a sample workflow as follows. At my previous job Catia and solidworks might be mixed into this rotation instead of or alongside autocad and Rhino but the principle is the same:

1. GA in autocad
2. Hull form in Rhino
3. Guess at weight based on similar vessels and experience, put info in Excel
4. Figure out draft with that weight in Rhino
5. Put preliminary numbers into NavCAD (or spreadsheet) to estimate required engine HP
6. Spec out major equipment
7. Structural arrangement drawn in pen over a printout of the GA, possibly drawn over the GA in autocad
8. Produce scantlings for midship section of vessel in Excel
9. New weight estimate based on midship scantlings
10. Compare draft with new weight in Rhino
11. Draw midship section in Autocad with appropriately sized stiffeners
12. Adjust GA to reduce spans with additional bulkheads to reduce size of stiffeners if required (almost always is)
13. Draw new structural arrangement in Autocad (very sketchy still, no details)
14. Adjust scantlings for new layout
15. Draw new midship section
16. Etc, etc... Repeat as necessary
17. Adjust hull in Rhino to support weight with reasonable draft
18. Adjust other aspects if required
    
    That would be the end of the preliminary design stage.
    
    
19. Produce full hull scantlings in Excel
20. Produce first crack at full weight estimate in Excel
21. Run navcad powering again with current weights and hull form
22. Check preliminary hydros with proper weights and centres in Rhino or GHS, find out hull trims heavily forward
23. Adjust hull volume forward to support weight
24. Run powering in Navcad, find out new hull takes 20% more power
25. Make a big decision - slower speed, cutout a stateroom, spec bigger more expensive engines, etc
26. Adjust GA, models, structure, weights, equipment to fit changes
27. Model simple structure and interior in Rhino
28. Find out that what looked good in 2D doesn't actually fit in 3D so go back to step 11 and repeat as necessary
29. Continue adjusting hull form, scantlings, weights, arrangement, powering, etc until you've run out of money or time. While adjusting these aspects your drawings and models will be continually refined, so when done this step you should be ready to submit to class or to yards for a cost estimate or to the owner for showing his friends

That would be the end of the contract design stage. Detail design would be next. It is very similar except the ratio of time spent in programs switches. So the time spent in stuff like Navcad and excel go way way down and the time spent in autocad, shipconstructor, solidworks, or potentially even rhino go way way up. Teams of people pull their hair out when the hull form or arrangement changes during detail design. In a perfect world there would be zero decisions to be made at this stage and it would be a simple matter of drafting up bracketing or joinery details. But inevitably things never work like that. The time spent in each stage is substantially longer than the previous. I hate detail design by the way.

 

LP like most I jump around a bit but at the moment its Maxsurf for hull modelling and hydro(been using it since beta testing) but happy to use Rhino for some hull modelling. For deck and super structure and interior its Rhino at the moment and sometimes, some things in maxsurf or XCAD. Then 2d drawings from model some rhino most in DesignCad (over twenty years). I have over many years used a lot of different software, some have just come and gone with time or bad updates but usually I gravitate to the ones that work for me at the time and work required. The most important thing for me with working with multiple software is the import/export and what information you can get out of the computer.

 

A very interesting and informative discussion.

DraftSight by Dassault is an alternative to AutoCad for creating and editing DWG and DXF files. http://www.3ds.com/products/draftsight/overview/ DraftSight is free but there is a charge for support. Its relatively simple with the emphasis on creating drawings.

 

This is the hardest, yet most import aspect of design, the weight estimate. Cannot be emphasized enough.

...along with this

If I recall correctly, a friend of mine tried this sometime ago and said it's not free. It is only free for a short period of time, like 30days. Omitted from the front page of blurb. Whether this has now changed, I am unaware.

In addition to the CAD link I posted above, there is another, which is the same as above but different, not 100% what the differences are, since they are the same, yet different webpages/sites:

http://www.progesoft.com/en/

 

This is all great information. I really appreciate everyone taking time to contribute. I will try to add my own bits in the near future.

 

I've had DraftSight loaded on my computer since soon after it was announced two years ago and it's always been free. Same for my brother.

 

I gave up on IGES and other formats so long ago I can barely remember why. But, they did cause me grief, and I did settle on 3D DXF meshes and 3D DXF polylines..

One thing I learned the hard way from Delftship and Freeship is that I must remember to export my 3D DXF/pl items in High or Highest precision, not low or medium. In Low or Medium, I get surfaces and curves that don't look quite right under close-up visual inspection.

I also have to remember in ViaCAD Pro to change the line types such that they produce cleaner starting points from extruding or sweeping surfaces from station to station.

I also have to decide whether or not to create surfaces from compartment to compartment or frame to frame (frames being, in my case, of whatever distances I set up in the Stations interface in Freeship/Hydronship). But, then I have to choose whether the stations/frames will be coincident to or different than the FS/HS stations. This is because I prefer to have vertical creases representing locations of stations in FS/HS, not the buttocked curves. This is because in order to know the actual weight distribution per compartment of the hull, I absolutely cannot have compartment 2's material sweeping aft into compartment three at the keel area and compartment 3's material sweeping forward into compartment 2' topside.

It really would be nice if some sort of XML or relatively platform-neutral data representation could exist and could be edited via an intervening API panel. This, I think, would make it possible to start in either a CAD app ORRRR a hydros app and nudge the concept toward the direction the user wants.

So, when I go from paper, to hydros, to CAD, I could discover in CAD that I did not leave enough clearance between the hull and GTMs. To fix that discrepancy, I could edit the CAD model, where greater finesse and precision exist at the scrolling and editing level, then have those edits drive the hydros package to tweek itself.

Then, as I make changes, the hydros package could -- based on library and target constraints -- alert me when I start to encroach upon library and target constraints, and help me not waste time unscrewing my over-tweaking sessions. It could presumable tell me that since I shaved .1 of a knot off due to an 24 cm beam increase at the midships point, I need to widen the transom area or some point halfway to the transom or narrow some wetted surface halfway to the bow, giving me options, but leaving the choice with me (since greater automation could and should start to cost money....).

It would be nice if CAD developers began to include basic mirror constraints in the code so that minimal CGs/stability/other info can be obtained in ways newer than traditionally endowed in CAD apps.

Curves of any type from a source app need to actually EQUAL themselves whether imported or exported once in the target app. Subtle nuances that wreck productivity are asinine and infuriating, especially when they they Spock-like explanatory or trouble-shooting language to a level that would stupefy Spock.

I also have to remember to do a number of things. Some are fun.

As for ditching ACAD, if you haven't seen Encore Software's Punch ViaCAD Pro, you may want to give it a whirl. It doesn't have multi-letter shortcuts, something that defies my logic. But, with at least 99 keys assignable or reassignable, it is easy enough to assign keys to jump to port, stbd, fore, aft, plan, and keel views. Isometric views can be jumped to the same way. Hiding and unhiding geometry as well as interface palettes can be a one-key affair, too.

The most important thing for me in VCP is being able to have surface AND solid models in the same file, and have surf and solid props information (centers, moments/inertia) that I can export in csv or other formats and not have to strip out horizontal presentation from vertical presentation just to start making sensible use of the spreadsheet. In other words, VCP output is columnar, not a mix-match/mish-mash duo-presentation that is convenient for Q&D (quick and dirty) glancing but counter intuitive and forces one to use some coding to unfrack that presentation choice imposed on the user. This might be no longer a problem in ACAD 2013, but I somehow doubt it.

Layers in VCP are handled somewhat differently than in ACAD. In VCP, layers management is kinda crude, and even the drag and drop of them is haphazard, and layers CANNOT be dragged outside of their tree. But, one can create a new target tree/branch and select and move those items to the new destination.

It isn't perfect, but I would never trade in VCP for ACAD. If CAD designers had a minimum agreed file type that would not behave in wonky ways when imported from or exported to other apps, things might be great. DXF supposedly IS that, but it doesn't help at all when my favorite 3D CAD app won't let me directly edit the mesh as a surface and retain the crisp, clean model that exists in Delftship. At any rate, those minor diffs probably would not impact the speed or resistance by .05%, but visually, they can be annoying. Wonky lines have made VCP crash on occasion, but I still find it more pleasing to use than trying to deal with the way ACAD displays (not talking about rendering).

Also, VCP has a neat "logicursor", or smart cursor that identifies what is beneath the cursor when superimposed geometry would be clicked on.

 

My approximate work

My approximate work flow:

ON PAPER:

After mass recollection, re-sorting, re-reading of papers, books, mags, and sketches of stuff related to what I want to work on, I "silhouette" my notional drawging against real-world analogs or competitors, then I I:

1. define some preliminry general particulars: LOA, LWL, bow rake/flare, BOA, BWL, T, D, lengths of exposed weather deck, shaft angles, etc.

2. sketch, on paper, the profiles/elevations, cross sections, and preliminary WT TVBHD locations according to distribution of major machinery and clearances I want for maintenance/extraction/damage control response

3. sketch general arrangements, longitudinal, transverse, and vertical access and passages, distribution of firefighting/DC terminals, pumps, generators, and fuel bunkers, sea bag lockers, dry, chill, fresh, and frozen stores, boat decks, weapons points and space for rounds/shots to be carried, helicopter bay space sized to accommodate (though not necessarily allow comfortable maintenance on) MV-22 and AV-8 (only for area, not reality), so that SH-60s and scout birds can be stowed abreast each other, and lay out the passageways to allow for port and starboard traffic, athwartships passage around 40-50% of the length of the particular compartment, and consideration for space for a number of per-compartment and per-deck "citadels" for refuge from fire, chemical agents, and live-fire-fight security breaches.

4. lay out operations, mechanical, electrical, habitability, and stowage items in terms of functionality, ease of access IN-HULL, ease of ingress/strike-down into the hull, and redundancy againsts single-hit loss of certain (but not all) major items

5. clean up linework to a level of personal satisfaction

6. tally/catalog all my major items into Lotus Approach, so that I can run reports on the ship weight, mass distribution, and more (I initially in 2003 created my own database in Lotus Approach because to me a spreadsheet is not appropriate nor constraining enough for me);


blurb: this database contains all the compartment locators for each compartment I design into the ship, and helps to identify forgotten items, overlapping designators, space volumes per compartment, per deck, and more. It does not do route planning and the like, but it is more than enough to whet the appetite of anyone wanting to (when I release the physical database) use forms, charts, views, reports, crosstabs, and such the way I do. Basically, it is a highly-souped-up CCOL (Compartment Check-Off Lis) I used to envision in the 80s when I was DDCPO (Divisional Damage Control Petty Officer), but has the ability to extend to PMS (Planned Maintenance) and billets assignments/duty assignments, and damage control reporting, too. Such are the powers of database programs when bent to the imagination. No, I am NOT afraid of "Ship Weight" the company. My work and time stamps go all the way back to around 1993, created in the curmudgeoness of my own nerdy bedroom after Lotus Development managers gave me legit copies of Lotus SmartSuite for Windows 3.0 and for OS/2 Warp. Thanks to all those dozens of types Computer World, PC World, et mags I had daily access to in the 90s as an office temp, thousands of spreadsheet and database interfaces from fuel flow to music writing to medical and financial view ports gave me a billion ideas of what to mix and match to make interesting stuff. If the APP can let you make it, then nobody can claim a PATENT on the process; they only deserve a COPYRIGHT on the very, very specific interface design they originate.)

blurb 2: also there is something of a CCOL or Joe-something a USN individual or team made around 1999, but it is built in MS Access, and is pretty powerful. But, it almost requires a developer to maintain it or modifiy it. I feel that my own tool is easier to maintain, but not necessarily extend, since I don't delve into Lotus Script, despite a huge jump in capability/flexibility LS might provide. I just want scripts and macros, and joins to do my work, to guard against too much risk of failure. The Access-based app may or may not still be available, and may be markedly different than the 1998/1999 version I explored, and may be standard across the fleet.

7. calculate in Lotus 1-2-3 (I can always export to OOO or LIbreOffice, et al) the fuel flows, according to number of engines on the line, RPMs of each, SHP of each, artificial loads or material degradations assigned, and so on according to published information (naval, professional, and student papers, all UNCLAS)

All this includes the superstructure GA as well. This gives me a paper-based map to guard against forgetting what I want to do, and is a guard agains losing the friendly feel of paper before the digital model is born.

Much of the internal GA or other optional arranging will be based on a list of 200+ items i know to be needed in the hull, but in general, the major spaces, the typical rattle-off items, but not clips, brackets, hangers, bolts, nuts, light switches, and the like.

THAT WAS BEFORE I FOUND OUT ABOUT FREESHIP, DELFTSHIP, AND VIACAD, and PolyCAD

IN HYDROS:

In Freeship/Hydronship,

1. I lay out the shape of the ship by profile, then switch alternately between plan and ortho and cross sections

2. enter stations distances, waterline elevations, and sometimes buttocks and diagonals.

3. tweek the hull to close in on target speed without sacrificing internal volume at the midships area (particularly maintaining enough beam to make it a no-brainer capability of choosing either LM-2500 OR WR-21 OR MT-30)

4. work out the skeg then develop any rudder placements/shaft alignments/angles, and stern flaps and stabilizer fins (but not any bossings or fixed keels)

5. optionally, play around with gondola arrangments to visualize positive or adverse impact on hydros


IN VIACAD:

1. Import the model

2.a. for the hull, separate the lines types (edge, waterline, stations, buttocks, diagonals) by layers according to watertight compartment

2.b. for the superstructure by types and layers do something similar

2.c. keep setting up the model to cater to some peculiarities of ViaCAD, so that I can quickly and easily turn off/on/hide/unhide or render transparent the various types of geometry

RANDOM, non-Design-Spiral Activities

During all this, I randomly export from Freeeship/Hydronship parts of if not the entire model -- up to that point -- as 3D DXF mesh and 3D polylines and:

1. do preliminary mirroring, testing of line quality

2. extrude from waterlines the surfaces needed for deckplates,

3. sweep a curve or profile along stations to form side shell stiffeners

4. sweep profiles along waterlines to form stringers

5. sweep added curves/lines for making deck stiffeners

6. solidify the sideshell, decks, and stiffeners/structural supports, and superstructure and mast work to obtain CG, mass/weight, moments, and balancing information.

Once I get this far, I generally feel okay with building ladders, adjusting angles of the GTMs and reduction gear or generators/motors, tweeking the shaft positioning, and more. I also do initial joiner bulkhead placements, and then start finding other things I didn't initially fix.

Then, almost back to square one when an update to Hydronship means using an updated Hydronship/Freeship to compenste for corrections to erroneous performance or resistance info.

OR, if I stumble in ViaCAD Pro, I have to sometimes explore the problem in Freeship/Hydronship and re-export the model and re-extrude and re-sweep things since there is no parametric nor other facility to automatically clean up the shortfalls.

.......

So far, this is THE shortest "workflow" description I have to date written about how I do things. I have not included PolyCAD in any detail as it is not able to replace Freeship/Hydronship simply because Freeship/Hydronship provide a plethora of reports that even the free or paid versions of Delftship cannot replace.

SOME EXPERIMENTS:

Some experiments I did last year and earlier this year in Freeship include manually rebuilding all 15 or 16 watertight compartments as models in separate files, then importing each into a new parent file, all just to be able to make Freeship/Hydronship give me per-compartment informtion the way I can get in PolyCAD when I subdivide the model. In PolyCAD, this, to me, is awesome information to obtain. But, in Freeship/Hydronship, doing this has the grave price of having huge spikes in the SAC presentation since there is a doubling of points and surfaces and solids being accounted for. If Victor T could circumvent this, further breakdowns in the FS/HS reporting could be had at a per-compartment basis. Also, it might mean that I could create modules that might be interchangeable with other future models, and have file-independent stats on modules I might substitute. It could speed up mix-match prototyping and hydros testing in the case of FS/HS modelbuilding.

 

Same here. As far as I know, the only requirement is that you register when you download it, and re-register once a year.

 

Interesting, thanks for the update. I'll inform my freind too.

What is it like compared to the usual 2D CAD programs?

 

Not enough experience to compare it. Online reviews claim it is very similar to AutoCAD LT. It loaded without any problems so anyone interested should be able to easily try it. Available for Windows, Mac and Linux.

Perhaps the software your friend tried was AutoCAD LT rather than DraftSight. AutoCAD LT offers a 30 day free trial, then it's over $1000.

 

As a company, given the ridiculous high costs of auto cad and maintenance as well, we have progressed to zwcad.
It's 99.99% like auto cad at 0.1 of the cost.