logo
NOTICE:  This is the new PunchCAD forum. You should have received an email with your new password around August 27, 2014. If you did not, or would like it reset, simply use the Lost Password feature, and enter Answer as the security answer.
Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

Options
Go to last post Go to first unread
brandgw  
#1 Posted : Thursday, August 2, 2012 8:03:32 PM(UTC)
brandgw

Rank: Junior Member

Joined: 2/15/2012(UTC)
Posts: 7

Here is a task for ZLC or anyone else that delves into boat design with the help of Freeship.

ZLC,
I was reading this thread...

http://forum.punchcad.com/showt...7&highlight=mesh+surface

I haven't covered it all, yet, but you are doing things that elude me. A very nice task would be converting the imported DXF meshes into a real surface. I've been playing with it a bit, but the current feeling is that it is VERY labor intensive. Unless of course, I'm not tasking VCP properly.

First question:
How are you converting Freeship meshes?

I've been using a kayak model for my explorations into VC. It's a hardchined with a V-bottom and a curved upper deck. Very similar to this one.

http://www.boatdesign.net/forum...ak-loggerhead-42099.html

The garboard panel were easy to model. A simple Cover-Surface command along the chine and keel were sufficient. The sheer strake was an easy model also. A Net-Surface command worked nice using the sheer and chine for one direction and the bow and stern edges and intervening stations in the other.

The trouble comes in with the deck. It tapers to a point and has contour. I have tried to work the Net-Surface approach with little success. The problem is that both M and N components converge at the bow and stern and VC doen't like this. I may have to settle for a minutely small M or N component at those places, but this messes with my internal sense of order.

Question 2:

How you deal with this situation?
ZeroLengthCurve  
#2 Posted : Saturday, August 4, 2012 7:59:21 AM(UTC)
ZeroLengthCurve

Rank: Senior Member

Joined: 5/15/2008(UTC)
Posts: 987

Thanks: 19 times
Was thanked: 35 time(s) in 24 post(s)
Hi.

Sorry for the delayed reply.

I am yet again rewriting my "sharable tutorial" in order to condense it for others to use it.

As for your comment:

"A very nice task would be converting the imported DXF meshes into a real surface. I've been playing with it a bit, but the current feeling is that it is VERY labor intensive. Unless of course, I'm not tasking VCP properly."

One way to get some VERY clean surfaces with minimal count in ViaCAD from Freeship is to set the precision to LOW. However, that causes several compromises:

-- setting to LOW precision may go against your model since the Low, Medium, High, and Highest settings change the way the Freeship surface is displayed and exported. For instance, simply export to DXF some sample boats, or a complex larger hull of one of the sample ships available at Hydronship.net. Now, compare the models at the various precision levels. You'll notice they do not match up.

-- setting to a low precision affects the final results of the hydrostatics reports obtained in Hydronship. This may or may not be acceptable to you depending on the percent of variation/deviation.

Using LOW means Freeship exports less complex surfaces to meshes. Once in VC/VCP/Shark, those simple, LOW res surfaces seem a dream to work with. However, they are not truly representative of the hull you want to model. If the hull has very complex or very compound curvature, and you export a Medium, High, or Highest resolution model, the number of vertices/ facets will go up dramatically, as will the file size.

Worse, as you've experiences, these meshes and vertices are an absolute hair-tearing nightmare to have to cope with.

What would be really awesome is if Tim and Team could come up with an algorithm to look at the governing curves that form boundaries for given meshes. It might require the user to export sections of a model, then let VC/VCP/Shark analyze a smaller number of hull or model panels with the Highest, High, and Medium and maybe Low resolution segments of a hull. Then VC/VCP/Shark could report back and after the user has had say 15 (in my case 15 or 16 compartments) compartments' compound curvatures analyzed, the user picks the one he or she wants to run with.

Then, VC/VCP/Shark could regenerate new governing boundaries suitable to Tim's brew of surface building. Then, the user would be spared of making replacements by hand. The new surfaces could probably be made in a way similar to what Marcus Bole does in PolyCAD, in which control nets of sorts allow the user to "magnetically" deal with snapping curves to other curves and using a type of gravity to manipulate surfaces into a desired curvature. Unfortunately, IIRC, those surfaces only stay in PolyCAD, or if you get them into VC/VCP/Shark, there may still be more work ahead.

But, if at least Tim could help us analyze the boundary curves, and then if VC/VCP/Shark could fire "rays" between the imported mesh and a VCP/Shark-proposed replacement, we could sleep better knowing that a new, faired, locally-better surface ----not a mesh--- is being generated. It would need to retain the colors, since my hull plates are color-coded per compartment.

Technically, I think this is feasible. It just might take Tim 10-20 hours of fiddling and being left undisturbed. I've sometimes punted to Marcus requests I needed help with in PolyCAD, and to my surprise and elation, he found fixes (retention of colors on import of Freeship dxf's; retention of layer names (IIRC... been a few months since last looked)...) even when he didn't believe his ancient code or the ancient compiler would permit what I expressed.

But, meshes definitely are an absolute, aggravating, worthless hell-hole and demoralizing waste of time when trying to use them coming from compound curvature models such as ships. Every time I have to cope with converting High Res meshes with bazillions of facets to a surface, I feel like telephone poles are being SWAT-slammed into my chest. Meshes need to be shot, LHC vaporized, and kicked sideways into an unreachable continuum.

In January, I spent a week or two in PolyCAD trying every which way to Sunday to find a way to get most of my Freeship work done in PolyCAD, but even if I could, PC and FS do not interoperate. I need FS's Hydronship reports that are not available in PC. PolyCAD has a lot going for it, though, once you become fluent in the interface. Marcus, when he gets a chance, wants to remove many legacy options that are more interesting to him but hardly touched by users, so the program is much friendlier to users.

If I win a lottery of around USD$50M, I will send Tim at least $3M to deal with the stuff that I want fixed, added, or changed. During 2010 and 2011 made a list of maybe 80 original ideas or problems, but to submit then would just overwhelm them or demand a headcount increase by maybe 20 people. That's not tenable. So, i decided to just sit on my list.
blowlamp  
#3 Posted : Saturday, August 4, 2012 9:18:53 AM(UTC)
blowlamp

Rank: Senior Member

Joined: 6/28/2008(UTC)
Posts: 648

Have you seen this video of Tim's? http://www.csi-concepts.com/Dem.../connectedTorusCheck.mp4 from this thread http://forum.punchcad.com/showt...ad.php?p=21727#post21727

Would it be of any help?

Martin.
brandgw  
#4 Posted : Saturday, August 4, 2012 8:33:42 PM(UTC)
brandgw

Rank: Junior Member

Joined: 2/15/2012(UTC)
Posts: 7

Thanks ZLC. Individually manipulating mesh facets is not going to be the answer. Curiously, I converted a few to surfaces and their common edges are nowhere near each other, relatively speaking.

Blowlamp, that thread looks promising though I am unable to view the most important piece of it. I can't get the video to load. Is it still loading for you?

The mesh conversion would be ideal, especially if it could be manipulated.
ZeroLengthCurve  
#5 Posted : Sunday, June 2, 2013 6:01:04 PM(UTC)
ZeroLengthCurve

Rank: Senior Member

Joined: 5/15/2008(UTC)
Posts: 987

Thanks: 19 times
Was thanked: 35 time(s) in 24 post(s)
Hi Brandgw,

The tapering thing bothers me after i do a thicken operation on the surface of the hull in Slt or VCP. Although Freeship accounts for thickening, i end up thickeneng the vcp or slt surface and flipping it or.throwing it inboard so that i can use a plane to lop off undesired beveling going on. I wish the program would just ask "Is the top or skywad edge supposed to terminate flat and parallel to baselind so as to be welded upon?", and then just do it.

An alternative is to create more hull than needed in fs and then lop off by planar cutting that material which will not compromise your design.

Due to years of trying to force my way with meshes, including recently using control point splines to recreate a hopelessly large facet or vertice count and then use the Guide Skin Surface tool, i gave up when i suddenly realized the Conic tool would result in almost zero vertices, vertice bulging, and other things that i spent years trying to creatively circumvent all because i cannot just use the meshes like surfaces... No cutting, trimming, thickening...



Originally Posted by: brandgw Go to Quoted Post
Here is a task for ZLC or anyone else that delves into boat design with the help of Freeship.

ZLC,
I was reading this thread...

http://forum.punchcad.com/showt...7&highlight=mesh+surface

I haven't covered it all, yet, but you are doing things that elude me. A very nice task would be converting the imported DXF meshes into a real surface. I've been playing with it a bit, but the current feeling is that it is VERY labor intensive. Unless of course, I'm not tasking VCP properly.

First question:
How are you converting Freeship meshes?

I've been using a kayak model for my explorations into VC. It's a hardchined with a V-bottom and a curved upper deck. Very similar to this one.

http://www.boatdesign.net/forum...ak-loggerhead-42099.html

The garboard panel were easy to model. A simple Cover-Surface command along the chine and keel were sufficient. The sheer strake was an easy model also. A Net-Surface command worked nice using the sheer and chine for one direction and the bow and stern edges and intervening stations in the other.

The trouble comes in with the deck. It tapers to a point and has contour. I have tried to work the Net-Surface approach with little success. The problem is that both M and N components converge at the bow and stern and VC doen't like this. I may have to settle for a minutely small M or N component at those places, but this messes with my internal sense of order.

Question 2:

How you deal with this situation?
ZeroLengthCurve  
#6 Posted : Friday, July 19, 2013 4:04:34 PM(UTC)
ZeroLengthCurve

Rank: Senior Member

Joined: 5/15/2008(UTC)
Posts: 987

Thanks: 19 times
Was thanked: 35 time(s) in 24 post(s)
Hi,

Recently, I decided to

-- import a hull as a mesh, but not use the mesh except as a template

-- create interpolated splines to track the curvature or the bow, midships, and afterbody of a hull

-- add surfaces for the bow section, the midbody section, then the after section

-- adjust the splines as necessary to bring the surface as close as possible with the least amount of crossing over

-- add splines to the edge of a surface in highly-compressed areas, and snap the spline to it's opposite edge, then adjust the spine, snapping it to the surface, so it is part of the surface

-- adjust the splines as necessary to further refine the hull curvature

Had I thought of that I would have saved years of time.




Originally Posted by: ZeroLengthCurve Go to Quoted Post
Hi Brandgw,

The tapering thing bothers me after i do a thicken operation on the surface of the hull in Slt or VCP. Although Freeship accounts for thickening, i end up thickeneng the vcp or slt surface and flipping it or.throwing it inboard so that i can use a plane to lop off undesired beveling going on. I wish the program would just ask "Is the top or skywad edge supposed to terminate flat and parallel to baselind so as to be welded upon?", and then just do it.

An alternative is to create more hull than needed in fs and then lop off by planar cutting that material which will not compromise your design.

Due to years of trying to force my way with meshes, including recently using control point splines to recreate a hopelessly large facet or vertice count and then use the Guide Skin Surface tool, i gave up when i suddenly realized the Conic tool would result in almost zero vertices, vertice bulging, and other things that i spent years trying to creatively circumvent all because i cannot just use the meshes like surfaces... No cutting, trimming, thickening...
ZeroLengthCurve  
#7 Posted : Sunday, August 18, 2013 11:21:58 AM(UTC)
ZeroLengthCurve

Rank: Senior Member

Joined: 5/15/2008(UTC)
Posts: 987

Thanks: 19 times
Was thanked: 35 time(s) in 24 post(s)
Hmmm:

"The text that you have entered is too long (11565 characters). Please shorten it to 10000 characters long." Hence, two-part posting for this one...

First,

Apologies for any frustration or stress I may have caused along the way with my drawing techniques and by pushing VCP and Shark to do things, which sometimes failed because I imported wonky curves and tried to build surfaces and thicken them when wonky curves were the underlying problem.

In 2011, while playing around with PolyCA, I learned a much easier way to create a very faired hull, but as I was jumping around between Freeship/Hydronship and PolyCAD and VCP/Shark, I set PolyCAD aside for a while, then failed to apply to VCP/Shark what I'd learned in PolyCAD because I kept thinking too hard/deeply about different features of each program.

The method I will explain below initially only requires control point splines and the Skin Surface tool, along with the translate tool


Obviously, the fewer number of control point splines involved in the whole of the bow and sonar dome area, the better. Doing this in VCP/Shark will be vastly easier than doing it in Freeship/Hydronship with one caveat: in Freeship/Hydronship, you explicitly control the number of control points involved in the entire model. Importing to FS/HS the VRML model from VCP's/Shark's skin will introduce THOUSANDS of control points, none of which you'll want to manually manipulate relative to fairing. At the very end, to compare between your model and a parent model the differences in efficiency, propulsion, acceleration, and other, you'll need to do the following:

-- orient your model bow-to-right so that Freeship/Hydronship (3.41 by now) can properly do the calcs

-- close off the transom by using the FS/HS Extrude tool

-- close/patch any "leaks" in the FS/HS model by using the "Face" ("Create a new control face from selected control points") tool

-- set the file's project parameters (length, beam, draft)

-- create, in the Intersections interface, stations and waterlines, and optionally, buttocks and diagonals

NOW....

For those wanting to design ships in Freeship/Hydronship (3.41 by now), but do so most expeditiously and with less stress, then, what I recently learned and which is not new information, and is easy to do would be:


1. Find a parent model that has reasonable hydrostatic properties as reported by Freeship/Hydronship

2. Turn on the interior edges and export that model as dxf 3D mesh (alternatively, export the 3D-polylines, too, but you may not need them depending on your near-term goals) as a half-breadth model, in Low resolution (to save on file space and loading time)

3. Into VCP or Shark, import the 3D mesh

4. Just as do real naval architects, modify the parent model sufficiently enough to not infringe on the copyright of the hull design (modify the length, beam, draft, freeboard, bow rake, hull flare, transom shape and rake, fins positions, keel/skeg angles, and other atttributes) by doing:

(((((((-- Interjection here: TURN ON THE CONTROL POINTS so you can see them. Save often, as in my case, there are times when, after the hull surface was created, random crashes would occur if I grabbed an errant/"troubled" control point -- the result being the Shark interface flashing once, then the dialogs going "outline" black and white, then restoring, then the entire crash of windows -- unless I managed to exit the "aborting" file before Shark internally moved into a collapse. Also, if this crash happens to start on you, avoid saving the file manuall as it might not open later, and if it does open, you'll be luck because Shark may have had a chance to create its own backup image with an odd suffix on the file's name. Save, and save often, as the mantra applies to ANY software. end interjection )))))))

IN PROFILE VIEW:

Use longitudinally-spaced, keel-up-and-outboard-to-the-top-deck direction control point splines to "trace" significant areas of hull curvature. Unless you plan to use the Guide Skin Surface Tool and unless you plan to early on create keel/skeg features, do not bother with creating the keel or the transom as these will be automatically "generated or inferred" in VCP/Shark, and later you can "link mirror" the hull and then just create a surface between them and not fiddle around with centerline-based cutting/trimming/creating)

(Note: be consistent about whether you draw the control point splines from keel up from from top deck down to keel -- it is handy latter when troubleshooting crossing curves in Bodyplan view; use color codes to avoid hair-pulling "de-crossing" of hull breadth stations)

4a. Create control point splines for the stem, but not the transom

4b. Create a second control point spline a few centimeters parallelt to and abaft the stem rake line

4c. Create between the .10 and .20 distance from the bow two to 4 forward-raked control point splines to control the bow flare and, later, the beam transition, spaced as desired,

4d. Create along the .25 to .50 distance two to 5 vertical, mid-body control point splines to control the top deck and keel "levelness" (since if not done with precision, VCP/Shark will introduce an almost imperceptible bulge below the baseline of the hull where you begin the keel's "cut-up" or "upturn")


4e. Create along the .60-.70 range two aft-raked control point splines

4f. Create along the remaining or .80 to .90 distance three control point splines running vertically, being careful to not introduct tight or overly-spaced curvature, to avoid unwanted gondolas longitudinally or convexes transversely

4g. Create at the desired location the transom, raked or vertical as desired, to terminate the hull's length overall (LOA)

4h. Save for later any skeg treatement unless you actually want to start trying to crease the skeg curves or mix-match control point splines, interpolated splines, and curves

[edit: corrected para 5]

5. Create three "waterlines"-based control point splines that run longitudinally from the stem spline to the transom spline just above the model's waterline and about 3/4 the way up to the top deck, and one for the top deck so you can control "fairness" along the top deck terminating control points. (If your hull has any hard "stepping" aft or amidships, this may complicate your modeling since there probably will be stress points in the splines in the transition from horizontal to vertical.)

6. As much as possible, hand-fair (but do not yet use Sharks auto fair tool on) the splines in the profile view, introducing typical nautical hull curving/dipping around 1/4 to 3/4 the way, or as desired

[Added note here: Later, if desired, use the Shark "Smooth Curve" tool to fair the curve, but note that it will detach the connection of any control points you paired up for hand-fariing/shape-making purposes.]

Remember that you are using the control point splines to mimic the body shape of the hull, the translate tool to make points overlap, and selecting relevant pairs of control points (only two points at a time) to affect the shape of the hull.

IN PLAN VIEW:


7. Adjust the beam position of the outboard control points to influence the shape of the hull. If a parallel midbody is needed or desired, then accordingly set the points along the relevant length of the parallel midbody. Even if one is not required for the waterline area, it may make modeling much easier to initially make the top deck control point splines relatively parallel to the centerline and then make minute adjustments in fairing later

8. Create control point splines that intersect with the shell-making splines and the two waterline-shaping splines, influencing the beam of the hull for each curve's elevation

9. Zoom in and select each sideshell/frame control point spline's control points and translate them to the control point of the waterline spline, taking care to not have too wide a stem nor too fat or narrow a transom, by translating or moving the corresponding waterline control point "station" control point as a pair

BODY PLAN VIEW

10. As above, move control points in pairs

11. Using color codes you should have set up earlier, visualize your fairing process in three groups:

A. Forward area of highly-raked/flared curves
B. Midbody area of relatively similarly shaped curves
C. Aft body of currves converging toward, but not meeting the centerline (unless your model has a sharp, canoe-like stern, or is double-ended as such)

Within each group, colors still should be distinct so you can flip around between forward and aft views of the model and maintain your orientation as you zoom and pan and move spline control points

12. Move the waterline/station control point splines as necessary to effect a well-shaped sonar dome or bulbous bow. No need to separately model and then attach or join multiple surfaces

12a. DO NOT join the bow stem and the near-bow-stem splines. If. You do, you might get a long "sword" shape flying off 300 meters into model space, excruciatingly slowing down zooming, panning, or other modeling activities

12b. Do not terminate the near-bow-stem curve too high above the termination point of the bow stem/rake spline, otherwise you'll have an upward-curving surface that might initially confuse your troubleshooting efforts

12c. Do not terminate the near-bow-stem spline too far AFT of the bow spline as it may introduce unwanted shaping of the bulb/dome
ZeroLengthCurve  
#8 Posted : Sunday, August 18, 2013 11:23:26 AM(UTC)
ZeroLengthCurve

Rank: Senior Member

Joined: 5/15/2008(UTC)
Posts: 987

Thanks: 19 times
Was thanked: 35 time(s) in 24 post(s)
CREATING THE HULL SKIN SURFACE

13. At some point, and this may be a good point, create a skin surface.

13a. Use the Skin Surface tool, starting with the stem, working your way aft in succession (skipping around can create bizarr, non-maritime geometry that could take seconds to minutes just to create, so, be careful), press and hold the Shift key as you select each station, then at select the final station and release the Shift key

You now should have a half-breadth hull of a single piece. Link mirror the half-breadth surface, then close off the transom.

Alternatively, rather than have a half-breadth model, link morror the half-breadth control point splines, except for the centerline.

Using step 13a, create the hull this way, then close off the stern/transom. This approach is useful for creating full-breadth, from-plan-view-downward transverse watertight bulkheads and also for creating plan-view-down stations at which you can then create sideshell stiffeners by using the One-Rail-Sweep Solid tool with "Allow Self Intersections" option enabled. Then use infininite planes to split the stiffeners into Starboard/Port pieces and color code them as desired, and to help declutter the view (unless you prefer to use the Arbitrary Section tool).

EXPORTING THE HALF-BREADTH MODEL BACK TO FREESHIP/HYDRONSHIP

14. Select the half of hull will use in Freeship/Hydronship

14a. Export it as VRML, using the "Selected objects only" option

14b. Intro FS/HS, import the VRML model. No need to initially create a model. Just open FS/HS, and click on File, Import, VRML, and select the relevant file you're bringinng into FS/HS.

Refer back above to the initial few paras telling you what you can expect about density of control points.

14c. This process will be iterative, and you'll want to practice it 5 or 10 times to make it second nature. If you're only using FS/HS to model efficient hulls, this process may be very efficient for most users. But, for those who want to model weights and structural proxy geometry in FS/HS, keep those items in another file saved as a part file, and import the part file. Such items may be of two types: Appendages and Internals. Rapid-prototyping hydro-related files may be best off if you're importing only the Appendages parts, to save time and to reduce risk of turning on or off any Internals parts that might cause troubleshooting to be needed.

Enjoy!
Tim Olson  
#9 Posted : Sunday, August 18, 2013 12:19:05 PM(UTC)
Tim Olson

Rank: Senior Member

Joined: 2/2/2007(UTC)
Posts: 5,447
United States

Was thanked: 499 time(s) in 353 post(s)
>>How are you converting Freeship meshes?[/B]

I've been doing some work on converting meshes into NURBs. In my "test" folder I have the Freeship examples (~9 files). If you have some interesting test cases, plz zip and send to me:)

Unfortunately I did not save the NURB results. I need to wrap up some ViaCAD stuff and then I'll be back onto testing the mesh to NURB technology again. It's pretty exciting but there are some limits/guidelines to getting good NURBs from meshes.

Tim
Tim Olson
IMSI Design/Encore
ZeroLengthCurve  
#10 Posted : Sunday, August 18, 2013 3:04:56 PM(UTC)
ZeroLengthCurve

Rank: Senior Member

Joined: 5/15/2008(UTC)
Posts: 987

Thanks: 19 times
Was thanked: 35 time(s) in 24 post(s)
Hi Tim, is that question for me?

If so, then...

I import the meshes, but I

- use them as visual cues and references

- for internal geometry (engines, uptakes, etc., stuff that are proxies but not product makers's 3d models of their wares

For geometry purposes, in Freeship, the are justt so I know the compartment sizes are sufficient or big enough. Once imported to VCP/Shark, I might occasionally convert only the planar or flat meshes to VCP/Shark surfaces. Those that are cirles, engine exhaust plenums and such I just rebuild by hand.

As for meshes related to the ship's hull/sideshell, there are too many instances of vertices not being aligned to their neighbors, and it's too much work to try to repair them. Sometimes on conversion, the surfaces would have holes in them or not thicken or would make weird things.

So, in short, for now, I cannot use the Freeship/Hydronship meshes for conversion in VCP/Shark. However, that could be because the underlying curves in FS/HS are not faird due to my having too many control points, added to force certain contours I need. If I build the model in Shark, export the VRML to FS/HS, the model's fairness (thanks to Shark's control point splines) results in a faster ship (negligible to any navy, but measurable in any event), more energy efficient hull.

If this might make your coding effforts more "energy efficient", would it be a worthwhile suggestion to have VCP/Shark:

-- "Scan" the imported mesh

-- Rebuild it based on control point splines

-- Highlight difficult/out of tolerance regions

-- Interactively allow the user to refine those regions

To try to improve the imported model?

In the case of a ship, it might mean mapping the hull, assigning control point splines, then mathematically generating a faired local area of plating, or a monohull that is faired from bow to stern. Manually that is more or less what I'm doing. I like this much bettter than what I was doing 3 years or so ago, which was based on my own ignorance.

Originally Posted by: Tim Olson Go to Quoted Post
>>How are you converting Freeship meshes?[/B]

I've been doing some work on converting meshes into NURBs. In my "test" folder I have the Freeship examples (~9 files). If you have some interesting test cases, plz zip and send to me:)

Unfortunately I did not save the NURB results. I need to wrap up some ViaCAD stuff and then I'll be back onto testing the mesh to NURB technology again. It's pretty exciting but there are some limits/guidelines to getting good NURBs from meshes.

Tim
ZeroLengthCurve  
#11 Posted : Thursday, December 12, 2013 3:41:35 PM(UTC)
ZeroLengthCurve

Rank: Senior Member

Joined: 5/15/2008(UTC)
Posts: 987

Thanks: 19 times
Was thanked: 35 time(s) in 24 post(s)
Hi all, again,

Lately, I've taken cues from PolyCAD and other hydrodynamics/ship design-related apps and videos.

I now just freehand the splines in VCP or Shark, then adjust them to be as close as possible to previous models I created. Since I design ships with bow bulbs/domes, it's a bit complicated, moreso than just creating a hull with a full baseline at z=0 .

I then use the surface command (no ruled/m-n surfaces) from bow to stern.

I then edit the splines to further smoothen the hull.

I rotate the model bow right (for Freeship purposes but not for CAD purposes), move the hull back to y=0, and slide it a few meters back to get the AP (After Peak) and/or rudder stock at x=0.

I then select the hull surface and export (only the surface) it as VRML.

In Freeship, I import the VRML file, set the project parameters, add stations and waterlines, then repair (manually) any damaged surface faces (usually 1 or 2 in the entire model), then look at the hydros.

Strangely, I now in Freeship have ZERO entry angle issues, whereas if I in Freeship/Hydronship created the model by hand and pushing control points around (hundreds of them for the shaping I need), I would get crazy entry angles of 80 degrees.

I resisted this process years ago, and it cost me years, but I am OK with that error and the learning process.
ZeroLengthCurve  
#12 Posted : Thursday, December 12, 2013 4:43:32 PM(UTC)
ZeroLengthCurve

Rank: Senior Member

Joined: 5/15/2008(UTC)
Posts: 987

Thanks: 19 times
Was thanked: 35 time(s) in 24 post(s)
Something I got around to experimenting with a few weeks ago pleased me greatly.

I found that Shark neatly "tracks" or follows the surface shape of my hull. So, when I start out with splines and create a surface, I can lengthen or shorten my ships if I thoughfully select a collection of splines.

(Normally, when "plugging" or "shortening" ship or yacht or boat, it would be done were least disruptive to the flow of the ship. If possible, this would be done in the parallel midbody area. Anywhere else means being bogged down in redrawing or re-fairing hull curve lines and means a more complicated plug to insert. Other considerations may affect things, such as how much structure, piping, wiring, etc would be cut or not cut based on the cutting location.)

So, when I moved a spline aftward, and it appeared to respect my hull's original shaping, i was excited. As a sanity checker, I imported or copied the hull and worked the curves of one to verify I was not imagining things.

To a good extend, this will be really useful for me when I want to lengthen (plug) and shorten hulls, saving an IMMENSE amount of time otherwise spend on breaking up the model and resizing it (the way I used to in Freeship/Hydronship before settling on prototyping in Shark, exporting to FS/HS, tweaking or verifying the hydros, then tweaking my splines to suit) to get what I need/want.

Your "mileage may vary" depending on where the splines you move are. I will revisit it, but it's definitely a good thing WRT (with respect to) the parallel mid body. Of course, of shortening the hull body means you end up taking space from adjacent compartments to make the shortened compartment worth having, then moving the related, adjacent splines in the hull may or may not affect your model.

For compartments, I set station locations and set them apart for thickness of the bulkhead plating, then create a surface betwen each pair, then project/rail sweep them "to surface", terminating at the hull surface. If your sweeps are not further complicated by additional stuff such as deck sweeps, compounded by transverse girders and beams, and by sideshell stiffeners and stringers, then you should be able to increase or narrow local beam dimension. My model, however is dense and so I cannot benefit from that. I tried, and trying to re-fair one curve that is in too far inboard took 20+ minutes, maybe 30, before a soft-abort, with an equally long back-out process.

However, if your model is vastly simpler than mine, having deck plates and bhd plates, you can use other VCP/Shark tools to associatively re-project/extrude those surfaces in a better-faired/improved hull.

Beats redrawing the entire model from scratch.

Cheers
memphisjed  
#13 Posted : Thursday, December 19, 2013 1:26:32 AM(UTC)
memphisjed

Rank: Senior Member

Joined: 12/19/2010(UTC)
Posts: 110

Thanks: 15 times
Was thanked: 6 time(s) in 6 post(s)
Are you a Web graduate? The only people I ever hear use 'plug' to mean lengthen.
ZeroLengthCurve  
#14 Posted : Saturday, January 4, 2014 1:10:10 PM(UTC)
ZeroLengthCurve

Rank: Senior Member

Joined: 5/15/2008(UTC)
Posts: 987

Thanks: 19 times
Was thanked: 35 time(s) in 24 post(s)
Nope.

I've used the term since maybe 1980s or early 90s. I'm around 48 years old and have since the 4th grade been studying ships, but never really liked higher math. Geometry was awesome in that he helped me account for volume and area in my drawings of ships, subs, and space ships. Later, when went on a tear for information about the DDG-51 class ships back in the 90s, I eventually (and recently) learned that there had been considerations to "plug" the ship as a number of SLEP/FRAM ships may have seen or experienced in the 70s'/80's. And, I learned that cargo ships, even MSC ships and USNS ships received plugs during "jumboization" programs, to permit longer loitering (making holes in the ocean, circling, waiting to tank-up ships transiting the large seas) and topping off more customers.

But, I'm just a lowly civvy designing ships (naval) with the aim to make stuff someone'll want to license in gaming platforms. Hopefully, reinvent myself out of debt.
Users browsing this topic
Guest
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.