FreeShip Question

[ancient kayaker]

In FreeShip, when I insert a plane to generate new points at the intersection the points create a flat spot on the hull. Can I insert new points in a plane without changing the shape of the hull?

- I am trying to divide a hull into fore and aft halves by cutting it across at midships so the two halves have separate developments.

[lewisboats]

You would have to try and do it with the mesh on...the interior edges. When you divide the surface you have to ensure that all your "faces" have 4 points so the mesh isn't stretched or skewed any more than necessary. I expect that you will have to add more points to get to this. Each point "pulls" the line it is on in a different way and trying to divide a line and get the same curvature back that it had without the point is very difficult. I haven't found a work around for it yet. You may have to generate two different models...one with the front half and one with the back half.

[daiquiri]

AK,
I think it has been debated here in the past... And, if I remember well, the conclusion was - nearly impossible to do with Freeship.
When I need to add bulkheads, decks etc, I export the hull into a 3D software like Rhino and do the work there.
Cheers

[ancient kayaker]

lewisboats & daiquiri: you are correct about the difficulty; the plane intersection merely adds points along the existing control net lines which changes the hull shape.

It seems a limited tool; I could correct the positions of the new control points if there was a way to leave the former hull shape on the screen to act as a pattern. I don’t have another application to import the hull into, but I can probably figure a work-around during construction.

Thanks!

[lewisboats]

Markers will allow you to see the adjustments I think.

[ancient kayaker]

Maybe it's because I have the freebie but the program will only let me import markers not create them.

However, I exported the image from the plan view and then imported it back as a background image, so I would have a visible record of the hull shape before the plane intersection operation. For some reason the image was offset in Y so I used the Move tool to superimpose the lines over the background image. That allowed me to see what actually changed following a plane intersection operation. Then I performed the plane intersection operation.

Unfortunately all the control points - the old and the new ones - need to be adjusted. On a mature and detailed design distinguishing which control points affect which part of the hull is virtually impossible so I gave up.

This is a trick that might work ealier on in the design cycle when just the basic hull shape is completed, before adding appendages, decks, coamings, rigging etc. FreeShip is a great development tool for the casual amateur designer like myself but it can get a bit clunky - it's probably quicker to solve some issues during construction . . .

[mick_allen]

export the lines and re-import them as markers.
It's best to export with hi resolution as there's a noticable difference between the 4 levels.

It's still not quite what you want as the control pt locations are not noted, but maybe a combo of the 2 techniques (background image w/ grid turned on to align w/ dwg grid + markers ) will be of assistance. (don't use the freeship generated image, but take a screenshot of the control grid image and use that)

Another tech maybe that in the frontview, you can align the control pts with each other wrt ht and width, so maybe if that is done and then maybe a 90 vertical axis rotation may set the length locations. Then rotate back.

I can't remember, but I think the dxf panel output with sections placed may be noted on the panel - would that help? - it can at least show up on the image.

anyway, a few workarounds.

[ancient kayaker]

I haven't been able to export any file type that will re-import in any form, except for the image file. Turning some of the layers off simplified the view but did not help get needed results.

The only way I have found that works is as follows:

1) reduce number of stations (too many is confusing) so they are only near the existing and proposed control points
2) turn off all layers except the hull - which has no creased edges so it's all one part
3) export the image of the body plan (front) view of the hull and stations, without control points
4) re-import the image and center it, adjust scale so it is superimposed on the body plan *
5) juggle with the control points so the hull matches its former shape

* both the center and scale of the image change during the export/import process, but I have figured out how to correct them so the original and current images are superimposed

I'm not sure whether it's worth the effort though . . .

The hull shape I have is a close approximation to what I want, but without accurate markers from another program such as Carene it is impractical to get a FreeShip hull that is a precise match for a desired hull shape. FreeShip has provided pictorial images and hydrostatic and hydrodynamic data that are close enough for my purposes. For this particular hull shape - which is based on geometrical curves - it is possible - and probably much simpler - to derive plank developments and frame dimensions mathematically using a spreadsheet.

Many thanks for your valiant attempts to solve the problem, but I doubt it can be done, basically I'm fighting the fundamental NURBs-based nature of FreeShips.

[liki]

Wouldn't it be enough to just insert a station in the middle? It will be visible in the developements and they are quite a lot easier to work with in a cad program after developing plates.

[mick_allen]

I haven't been able to export any file type that will re-import in any form,

Assuming you're using FS2.6, export the offsets. The offsets are now in a simple text file that has points for stations and chines. Go into that text file and delete all the station stuff and any other info that does not match the file info necessary for importing markers. (It's really easy, basically delete everything except the points at the front and put 'EOF' at end of the file). Then import that file as markers. (Read the manual). Remember to export at highest resolution otherwise the marker input is much more jagged.

except for the image file.

As you are trying to match control point locations, don't just use the Linesplan bitmap export, but also do a screen capture of the control net so you can more accurately place those. Leave a similar grid turned on for the image so that you can resize it to the correct size and location. Screen capture at the highest size you can so that readjusting is more accurate. This may be a little more helpful than the linesplan.

anyway, good luck.

[ancient kayaker]

Originally Posted by mick_allen

Assuming you're using FS2.6, export the offsets. The offsets are now in a simple text file that has points for stations and chines. Go into that text file and delete all the station stuff and any other info that does not match the file info necessary for importing markers. (It's really easy, basically delete everything except the points at the front and put 'EOF' at end of the file). Then import that file as markers. (Read the manual). Remember to export at highest resolution otherwise the marker input is much more jagged , , ,

I exported the offsets, deleted the lines containg “Station” and also all buttock and waterline data, added EOF at end and re-imported the files as markers. I have now have something that looks like it may be useful with a few changes; I will work on it.

Thanks!

[mick_allen]

Hi Terry,
As you'll quickly find, one of the issues with this type of import is that you cannot exactly locate the control points for normal curves. With creased curves you can get a good idea of those locations, but the normal (interior) curves - not so easily.

So if you have a bunch of normal curves and wish to locate their control points:
As you now can edit these files fairly easily, a little trick is to export your markers as before with fine resolution and save it for getting creased edge locations and without saving the next steps - 'select all', crease ALL the normal edges and select coarse resolution. Then export this as another offset file. When you import this , you can fairly easily locate the normal edge points as they are every 2nd marker point (in coarse exports). Then get out of the model without saving - and you haven't screwed anything up.

And I presume you've played with exporting and reimporting the model as a 'part'? If not, this might be another option to try out.

anyway, other ideas.

[LP]

Terry,

You may have moved past this problem by now, but I see that it is only a month or two old. I was doing a search for other info and found this thread.

An option that you might try:

Place a station at midships or the location of your split. Add a plane at the location and add control points as needed to define the plane that matches your hull form precisely. This is an independent element. Essentially, a bulkhead that you create to manually match the station line of your choosing. You can the delete the control edges at midships and stitch the hull section to the "bulkhead". A second, matching, bulkhead will have to be made for the other hull section.

There will be some hull deformation at the control points adjacent to the bulkheads, but that could be corrected by similarly placing reference bulkheads, beforehand, along those points and readjusting the control mesh afterwards to match the ref. bulkheads. Any further deformation should be limited to non-existent. I think.....

[ancient kayaker]

From what I read on another thread adding new control points using a plane inserts a crease. I haven't tried changing the crease back to a normal edge yet but I will. I will also try your idea but I don't know if I can insert bulkheads with FreeShip as it doesn't handle structures only shapes. I don't think FreeShip can handle the extra calculations needed to insert a new point and keep the old shape as other points would have to be relocated.

[LP]

Inserting a plane was probably wrong terminology as it references a function in Freeship.

The point(s) that I want to emphasize are placing section lines at the point where you want to a) cut the hull and b) at the control points adjacent to the hull cut. c) The bulkheads inserted at those locations are independent/separate of/from the hull.

There is a way to make a bulkhead that is connected to the hull, but that does cause a crease and it screws up the hydrostatics. The bulkheads are matched to the hull stations and then become place holders for the hullform as you start to manipulate the hull itself.

You may have already understood of this, but here it is again, hopefully explained differently.