Fixing a transom leak in Freeship 2.6

[drshaddock]

I'm working on my first design in Freeship and so far have worked out most things. But my model indicates a leak point (schooner displacement hull, by the way) aft where the topsides, deck, and transom all come together. Oddly, if I turn off all layers but one of them and check the corner point in question, it's identical for all three layers (hull, deck, transom).

Anybody have advice on how to fix this? My hydrodynamics seem to be working anyway--although I'm not really ready because I want to add a ballasted area to the keel and re-determine the center of gravity. I have other questions about that; will ask on a different thread. Thanks!

[marshmat]

Provided the three corner point have exactly the same co-ordinates, and by this I mean that you have manually entered the exact same 3-dimensional position for each one (not just dragged or nudged them until the numbers matched), the "Check Model" tool should automatically fix the duplicate point. Every time I have encountered this problem, re-entering the co-ordinates of each point by hand has allowed the model check tool to correct the error.

[drshaddock]

Matt, I really appreciate the help--but that's exactly what I've been doing. When Check Model reports a leak (it started out reporting a dozen or so--and I realized they were all back at the transom, so I had to extrude the aft edge transversely and then bring the control points back to the centerline) it gives, um, four significant digits. But Freeship is using 5 significant digits internally (I think five--I did this last week and haven't had time to ask the questions until now). When I saw that several of the leak points looked like identical coordinates, I went out and found them, and discovered that where one was, say, -3.91565, the other might be -3.91570. (I'm only giving the X coord for simplicity's sake; of course there are three). So I made a note of the first one I saw and corrected all the others to match. I got rid of all my leaks this way, except for this three-surface intersection.

Am I creating the deck improperly? Maybe it's the triple corner that's throwing the program off. I extruded a transom, cleaned that up and curved it to make it prettier, and then extruded a deck from the sheer line because I couldn't get the program to let me create a surface at what I WANTED to be the deck line and let my sheer line be the caprail on my bulwarks. (But that's another thread).

[Fanie]

It seems the entered resolution may not quite accurate enough. Entereing a value 0.0000 doesn't get the lines exactly there. I had to zoom in quite a bit and drag the point to the ie centre line to get it precisely right. Bit time consuming.

[Rick Willoughby]

For hydrostatics I just use a single layer. I just continue extruding the deck until I close along the centre line. Any layers added after this can be nominated as non-hydrostatic layers.

If you are getting hydrostatic information then the leak really does not matter. It will only come into play when you do cross curves. The angle at which the leak goes below the surface will be the last data point.

Rick W.

[drshaddock]

Thanks, Rick. I had, indeed, discovered that I could get hydrostatics once leaks were minimal--but didn't know if I could trust the results because the manual states hydrostatics don't work unless you have no leaks. (but I could watch the balance curve disappear or generate as I moved points around on my transom, either causing or curing leaks). I turned my single layer into multiple layers just so I could turn off one at a time and click on the appropriate layer's corner point--that's the only way I could keep from just getting the last point entered at that corner.

By the way, what's the purpose of checking the 'Corner' checkbox when you're examining a point? It doesn't seem to make much difference what I do with that--and sometimes after I uncheck the box and look again in a few seconds, it's checked again.

[Rick Willoughby]

The corner checkbox will give a sharp corner when ticked or a curve if not. It is best to observe if you switch to the rendered view rather than just lines.

Rick W.

[drshaddock]

That's very helpful, Rick. I think I'll hang in there with Freeship (when I have time to get back to the design); it seems to have a lot more going for it than the (limited!) number of other pieces of commercial design software I've perused. I wish there were more to the manual; it leaves a lot of questions in my mind.
I like the way you can enter densities for materials when you build layers--but I can't use that very well until I can work out methods to get all my bulkheads, cabinhouse structure, bulwarks, and mechanicals into place. I've addressed that in a different thread and was offered the idea of building them as separate parts, then importing them, which makes sense. I think, however, I could do bulkheads by placing a new vertical surface inside the hull, running in a transverse direction.
Anyway, thanks again for the help!

[LP]

It sounds like you are trying to get multiple "pieces" to join up at a single point. Just having co-located controlpoints is not going to produce a leak free model. The pieces actually have to be connected to form a continuous shell. Extruding will create a joined surface as will "Create a new controlface from selected control points" (upper righthandish function icon).

Also, you will get a "checked" box at any control point where 3 creased edges come together.

====================================================

I went ahead and added a screenshot and the hydrostatics results of my latest project. I've developed a full interior with independent layers for each of the components with density data applied. There are different ways to build different conponents. Importing a part is my least prefered. To build within the model, you need to create some "seed" material. Do this by extruding from a clean edge a small surface. Then extrude again from the new surface and "delete" the initial extrusion. Now you have a surface that you can manipulate through control points and extrusions et. al.

Hope this helps.

[drshaddock]

LP, those are outstanding suggestions.

You know, I had originally extruded the sheerline to form my deck, and extruded the aft hull edge to form the transom. As I manipulated these new faces to give them a nice curve, my leaks probably were created.

I think your suggestion that just having control points on the same spot is not enough is exactly right. I'm used to working in AutoCAD where I can extend an entity to the point(s) where it intersects another entity. I had expected at first that the Intersect command in Freeship would do something similar. So I tried to pick my two layers of deck and transom, for example, and Intersect them. But it seemed to have no effect. I had also tried making my deck extend further aft so that it 'passed' the transom and went through it, thereby (I foolishly assumed) sealing the junction. This didn't work, and I ended up with some odd tucked curves at the aft end of the deck. Maybe I stopped too early--maybe I was on the right track, and needed to run the Intersect command again at this point...? I don't have time today to go back and work with this, but thought I'd write about the idea and help me lock it into memory.

But if I'm on a wild goose chase, would somebody please let me know?

By the way, that's a pretty boat. Didn't see any hydrostatic data, but I like her lines.

[LP]

Freeship is very limited in some ways while being very veratile in many others. Don't expect it to work as a CAD system would. Overlapping work surfaces will not create a "seal". The intersection command will not work to those ends either.

Without looking at what you've done, I'd say your best bet is to nibble your deck away until you have a "V" shaped notch between the deck and transom. You could do that by taking the aft most centerline deck control point and shifting it forward until it is forward of the transom. If you have more than one control point on the aft deck edge (not counting the outboard corner) it/they will need to be adjusted too. Once you have your gap, select all of the control points of the gap in a sequential manner around the perimeter of the gap (hold the control key during the selection process) and then hit the create a control face icon. That will bind all of the points together with a new face and you will need to recrease the transom edge.

Thatshould do the trick.

The hydrostatics are in the .txt file at the bottom of my last post. Just click it and open it.

====================================================================================================================

Another way to create your deck would be to start from your completed hull and transom portion. Extrude both hull and transom edges towards centerline, collapse the resulting spikes and move the rest of the extruded control points to y=0.

[drshaddock]

Nice trick--I hadn't seen the 'Create Control Face' icon. I'll look for that tomorrow when I get back in. Thanks again!