Schuchardt

[Schuchardt]

Solved.

Obviously there was a bug in the algorithm. Today I have installed the new update (now version 8.08.293), and everything works fine now.

[Schuchardt]

Hello,

one important notice:

When you intersect two layers. set the accuracy to ‘highest’. I seems, that DS uses the internal edges for the calculation of the intersection lines, and these depend on the accuracy.

You can easily see this effect, when you intersect the default model (new model) with a horizontal plane at z=1 and insert the new points into the plane. Before, leave the hull as it is, but subdivide the controlnet of the plan several times in order to get a higher accuracy. Then perform the intersect-operation with low and highest accuracy, respectiveley. You will get two different intersection lines.

Test file is attached.

Regards
Klaus

Attachments:

• Intersection.fbm

[Schuchardt]

Hi Dave,

its not easy to merge both these edges, but I have used your last model data base and tried the following way (see attachment):

I have moved the inner and outer hull by 0.1 in the negative z-direction to separate their edges from the transom. Then I have defined a new surface between opposed points of the hull layer and the transom layer. If you do this for all leak points, the space between the edges will be closed and the leak points will thereby be killed. Finally, the inner and outer hull will have to be moved back by 0.1 in the z-direction.

This solution isn’t smart at all and involves tedious work, but it is straightforward and i think it will work at least.

Interesting problem, interesting discussion.

Regards
Klaus

Attachments:

• getridoftheleakpoints.PNG

[Schuchardt]

Here is a small amendment. The effect is still present after a large move of the lower surface of the keel (test3a), but it disappears after a relatively smaller movement of the upper surface (test3b).

Attachments:

• test3a.PNG
• test3b.PNG

[Schuchardt]

Hi marven,

thank you for your comment, but I am not convinced.

I have reduced the hull to a small segment and have merged all unneeded internal edges in order to demonstrate clearly what happens. In the attachments you can find the results of the original ‘two-part-model’ (test1) and of the ‘one-surface-model’ (test2) that you have suggested. The disturbance is still there. The entire construction is very clear, very simple and very wrong in both cases.

The corresponding fbm-files are also attached.

Regards
Klaus

Attachments:

• test1.PNG
• test2.PNG
• test1.fbm
• test2.fbm

[Schuchardt]

Hi Dave,

congratulations, I think you’ve got it. And all leak points are gone, did they?

As far as I know (Mr. Admin will know better) it is not possible to import a DS file to FreeShip. I suggest, that you produce a body-, sheer- and lines-plan in your current DS implementation, then use these plans as backgrounds in your FS software and remodel your ship with much less points.

Regards
Klaus

[Schuchardt]

Hi Dave,

I had a look at your file. My suggestions are:

In your model, the inner part of the bottom and the transom are in one layer. Transfer the inner part of the bottom to a separate layer (lets name it ‘hull inner’).

Next, apply the function ‘subdivide controlnet’ two or three times. The crucial point is, that intersection points are inserted in edges only. If you have only few edges, no or very few intersection points are found, and these points ar too far away from the real hull surface (remember, that in subdivision modelling the edges do not lie on the surface). In this case you will get garbage. But the more edges you have, the tighter the control net will become, and it will represent the actual surface increasingly better. Therefore, you have to subdividethe net.

Finally, intersect ‘hull inner’ and ‘hull outer’ with ‘transom’ each and insert new points in ‘transom’.

Perhaps this could be the way to solve your problem, as I understand it.

Regards
Klaus

[Schuchardt]

Hi Dida,

you will have to define additional internal edges along lines above and below the deck line and then refine your model manually point by point. This is (as far as I know) the only way to reconstruct the primordial continuous shape of the frame at the deck line.

Regards
Klaus

[Schuchardt]

Hi Dida,

just rotate your model in the body view around the vertical axis by 180 deg.

Regards
Klaus

[Schuchardt]

Hi queqen,

it is a switch. When it is on (Same x and y scale), the Set-scale-command uses only the first value (for the horizontal axis), the second value is ignored and derived from the first value.

For a test, just modify only the second value in the Set-scale-command. If there is no change of the background image, the toggle is on. Otherwise it is off.

Regards
Klaus

[Schuchardt]

Hello Marven,

thank you for your quick response.

But what do you mean by ‘recreate’?

To be sure, I have performed the following reproducable process:

(1) deinstalled my current DS Free
(2) deleted all directories established by DS
(3) installed 4.38.130 from the DS site
(4) started a new model
(5) exported the default hull as a part
(6) tried to import the part – FAILED
(7) updated to 4.46.158 with the update function
(8) started DS and opended a window
(9) tried to import the part – FAILED

I have tested this procedure on two notebooks (Windows 7 64 bit) with the same result.Because I use DS for about three years, rookie mistakes should be excluded.

Exporting and importing parts with 4.46.158 is free from defects. Of course, in some respects this problem is a theoretical one, because there is no need running 4.38.130. But if the failure is not due to an operating error on my side, this could perhaps point to a bug.

Regards
Klaus

[Author]

Posts