Terrance Egolf

Forum Replies Created

Viewing 15 posts - 1 through 15 (of 81 total)
  • Author
    Posts
  • in reply to: Offset Table import #44942
    Terrance Egolf
    Participant

    Hello MBerlin.

    I apologize for not responding sooner. The past several weeks have been quite busy for our family.

    I have worked over both the text file and the model itself to see if there is a solution to your problems. I have to confess that I can’t sort it out. The import (and export) Offset feature has issues that defy explanation.

    First, I imported your original Offset-table-2.txt file and this is how it displays:

    Original-Model
    Notice the inward fold where the upper waterline data is zero.

    Then I attempted to clean up the file to approximate what I think you were looking for:

    Cleaned-Up-Model

    I eliminated the fold and smoothed up all the stations and waterlines.

    Then I exported an offset table based on the cleaned up model from the program. But when I imported THAT txt file, I got this result., which is worse than the original.

    Exported-Offset-Table-Model
    This model derived from the program’s exported offset table is full of errors, and again includes the fold where the upper waterline data is zero.

    So, I’m afraid I can’t help you. I’ve attached the revised model file for you to take a look at. I would consider relying on an offset table in this program to be a last resort.

    Terry

    Onemeter-Offset-Revised

    Attachments:
    in reply to: Autofair Function #44937
    Terrance Egolf
    Participant

    Charlie,

    When you get into a interesting discussion, you can quickly reach the minimum number of MSW posts needed to change your profile.

    Regarding your question: I have found that the best way to match hull surface shape to an existing plan is by the use of the program’s “Intersections” feature. The process should follow something like this:

    1. First, add your deadflat image to Reference Images, properly sized for your model, and so that the image baseline corresponds to the model’s. (I describe how to do this in one of my MSW posts. Contact me there to get my tutorial.)
    2. Next, in the Home ribbon, click on the Edit button in the Layers menu group. In the row corresponding to the model’s hull component under Description, check the box under the Intersections column. This makes any intersections you define visible in that component of the model. This step must be done first before any of the following, otherwise, you can’t designate any intersections in the model’s hull.
    3. Assuming you have already built a model and are tweaking its shape to the hull plans, go to the Intersections tool in the Project Tools menu group under the Tools tab. In the dialog box, select the Stations button to display all the active station intersections.
    4. If you have not done so, add the x-position of the station that represents the position of the Deadflat station. Normally, this will be the widest point of the ship’s hull. It is probably best if you delete all other stations in the vicinity of the Deadflat station to minimize interference with the view. (Note that this doesn’t change the hull’s shape, just the station intersections visible in the model.)
    5. Finally, back in the Home tab, click on the Control Net and Stations buttons in the Hull Display menu group to display both the Control Net and any Station intersections you have added.

    If you only want to manipulate the deadflat shape, the simplest approach would be clip the hull section where the deadflat is located using the Clip tool, located toward the right end of the toolbar under the ribbon. However, you will want to go over the hull first and bring it into something close to its final shape before you go any further. Realize that the Control Net influences not only the surface directly under a given control point but the connected surfaces adjacent to that control point.

    Once the hull surface is close to what you want, then activate the clipping tool at the deadflat, which should have control points either on the station line or very close to it. (See the image below.) Using an end viewport, carefully adjust the control points in the y-direction to move the associated station line (not the control point) so that it lies on the corresponding deadflat station in the diagram.

    Mesh-Arrangement

    Clipped section of the hull at the deadflat station.

    Realize that any major adjustments to the control points on the deadflat will change the shape of the hull surface adjacent to those control points, so this process should be done only when you are very close to finalizing the hull.

    In my brigantine’s case, the plans were not very good, so you can see that the deadflat station didn’t really follow the plan’s lines. I used the plans to get close to the desired hull shape, then used the shaping tools in DELFTship to smooth and fair the surface.

    Front-View

     

    Hope this information helps.

    Terry

    Attachments:
    in reply to: Autofair Function #44935
    Terrance Egolf
    Participant

    Charlie,

    Yes, the Galilee has been my main nautical historical research project for almost 23 years (off and on). My maternal grandfather sailed in her for nearly two years (1906–1907) as a scientific observer while the vessel was under charter to the Carnegie Institution as a testbed for conducting geomagnetic data collection at sea. I have been corresponding with the recently-retired librarian of what is now the Carnegie Science Library since 2001. He was most helpful providing needed research materials.

    After frustrating myself trying to construct a set of workable plans using 2D software (CorelDraw), I finally discovered DELFTship Free back in 2014. I used the program to reconstruct the vessel from the only valid extant set of plans of the ship, which consisted of simply the three standard views of the hull, and the major deck and inboard bulwark structures. (No masting, no rigging or sails, and, of course, none of the unique magnetic observation structures and modifications.) The undated plans were developed by a member of the Pacific Marine Research Society named G.C. Berger. I have no information regarding the provenance of the plans. It seems that the plans were created sometime late in the ship’s history, well after she was converted from a brigantine packet ship to a three-masted fishing schooner—and then a retired houseboat—so I don’t know where he obtained the deck arrangement details as a brigantine.

    Thankfully, I was able to obtain many contemporary photos of the ship during the timeframe of interest, and was able to make numerous, if mostly minor, adjustments to the hull and other details. I am currently using Blender to construct the masts, spars, and rigging, since that program is more convenient for modeling those kinds of details.

    One significant issue I ran into is that the registered length of the ship couldn’t be correlated to the plans. On top of that, Berger used an unconventional method for designating stations. Consequently, I had to adjust the ship’s length nearly three feet to make it conform to the US merchant documentation practices in place in 1891 when the vessel was first built.

    The DELFTship program helped me realize that the three views of the hull were not consistent. For example, if one tries to digitally construct a hull from just the profile and deadflat views of the plans, the resulting waterlines don’t match those shown in the plan’s half-breadth view. If you try to construct a hull using just the deadflat station lines, the resulting hull isn’t fair. That is why I encourage you to use the program to construct the best compromise of the data you have available.

    From correspondence with other MSW members, this kind of problem appears to be quite common when researching historic vessels with limited documentation. Very old plans tend to be distorted and faded with age. Photocopied plans are often distorted by the scanning process. I found this to be the case with plans of Galilee obtained from the Smithsonian HAMMS project.

    If I might ask, what is your MSW handle? The MSW forum’s website is convenient and private for members to correspond.

    Take care

    Terry

    in reply to: Autofair Function #44931
    Terrance Egolf
    Participant

    Charlie,

    This program is probably still the best for modeling ship hulls of any digital software available. The need to add more patches and multiply polygons in regions of sharp and irregular curvatures is the nature of the way surfaces are modeled in DELFTship. Sailing merchant and naval ships prior to the last century had very curvy sterns and tumblehome. For example, this is the control net for my 1891 model’s stern:

    Galilee-Aft
    If I were to do this hull over again, I would have run the horizontal edges all the way from bow to stern, and the vertical edges from keel to rail. This avoids the breaks where edges stop in the middle of a surface. However, increasing the number of patches in a surface makes it harder to create a fair and smooth surface. Also, try to avoid triangles and polygons with more that four sides, if possible.

    You can use the built-in Grid view to adjust the control points to create a fair surface. However, I have actually found that building your own grid of station line, waterline, and buttock line intersections, in conjunction with the Gaussian surface display, works better for shaping the hull in most cases. There is even a Diagonal line option, which was often used back in the day.

    Depending on what plans you are referring to, for example, if they are copies of originals from a national maritime museum, you will likely find that they won’t produce a smooth hull in this program—that’s a problem with the old plans. You just have to do the best you can.

    If you aren’t a member already, you should consider joining (for free) Model Ship World. This is an international organization of maritime researchers and ship model builders (with 45,000+ registered members) who would be happy to provide information on ship plans and construction in the timeframe you are interested in. We also have a number of DELFTship modelers who can assist in your work as well. (My member handle is CDR_Ret.)

    Terry

    Attachments:
    in reply to: Autofair Function #44929
    Terrance Egolf
    Participant

    Hello Charlie,

    If you are using the program mainly for documenting or reconstructing old sailing ship plans, I believe everything you want to accomplish can be done using the Free version.

    Please check out my reconstruction of a 19th century US brigantine at https://forum.delftship.net/Public/topic/modeling-the-1907-u-s-brigantine-galilee/

    If you have any questions, I would be happy to assist.

    Terry

    in reply to: Offset Table import #44921
    Terrance Egolf
    Participant

    Hi MBerlin,

    Thank you for these files. Sadly, my computer is in the shop for at least a week due to a potential malware intrusion from an email correspondent, so I won’t be able to check these out for an indefinite time.

    Will get back to you eventually.

    Terry

    in reply to: Offset Table import #44914
    Terrance Egolf
    Participant

    MBerlin,

    Do you have a photo, diagram, or plan of the vessel you are trying to document as an offsets table?

    It would help tremendously if I could visualize what you want to end up with. Also, I assume the coordinate system is in metric units, correct?

    Thanks.

    Terry

    in reply to: Offset Table import #44908
    Terrance Egolf
    Participant

    Hi MBerlin.

    If you could attach your text file to a post here, I might be able to check out what it is doing. I don’t want to try and transcribe what you have shown in previous posts, which would likely introduce errors.

    Terry

    in reply to: Offset Table import #44903
    Terrance Egolf
    Participant

    Hi MBerlin,

    I am no expert on the importing of offset tables in this program, but I have used them before with mixed success.

    First, the “0.000” entries under the column 0.132 (m?) waterline must be nonzero in order to make sense. These entries represent the lateral (y-) coordinate at each waterline/station combination at the surface of the hull. Entering 0.000 in these fields tells the program that the hull beam is zero (at the centerline) at these locations. I think you should try leaving these fields blank to omit waterline date where there is no hull surface at these stations along that particular waterline. This will probably require, after importing the table, manually inserting some control points in the outboard sheer line (rail) to help smoothly terminate the 0.132 waterline net edges at these locations.

    To define the deck, I think you want to use the third and second columns from the right-hand side of the table. These columns are discussed on page 33 of the v. 15 Manual. I’m afraid that the function of these columns is not described very well in the manual. For example, the “left column” establishes the deckline height, but it doesn’t define whether that is the outboard height or the centerline height. I suspect it is the former, but without testing the coordinates myself, I can’t be sure. The “right column” is a bit more clear, and evidently provides the outboard lateral dimension (y-coordinate) of the deck edge at each  station. So taken together, you might assume that the “left column” probably provides the outboard deck height (z-coordinate) at the hull surface.

    [To DELFTship] It would be very helpful if all the terms used in this Table of Offsets discussion were formally defined in layman’s terms and illustrated on a model hull. For example, what does “contour” mean in the context of this table?

    Going further, if the above discussion defines the outboard edge of the deck surface, it would be helpful to program users if the Manual would describe somewhere how add a cambered deck surface so that it accurately intersects this line just defined! I have done this myself using a deck camber template, but it is very tedious.

    Hope this answers your questions, MBerlin.

    Terry

    in reply to: what is that, can i prevent it ? #44880
    Terrance Egolf
    Participant

    Flo, terminating an additional edge at an internal edge that guided the shape of a fair surface beforehand will create a discontinuity at that point. That’s simply the way the program works. You would need to continue the new edge beyond that point all the way to the bow, then tweak the resulting intersection points at each vertical edge to restore the fair surface (smooth station lines).

    Not sure if that clearly explains what is going on there.

    Terry

    Terrance Egolf
    Participant

    Hi Flo, did you get a response yet? If not, I think I can suggest a solution if I understand your issue correctly.

    Terry

    in reply to: Printing stations separate #44864
    Terrance Egolf
    Participant

    Hi flo,

    I apologize for missing your earlier responses on this topic.

    I never actually used the Plot or Printer settings for Linesplan.

    When I did a test run for printing the hull frames of my 19th-century brigantine, I built in a reference template to register the positions and sizes of the frames. This template included a horizontal line at the model’s baseline, a vertical centerline, and a horizontal line marking the waterline—all lines that are visible in the Plans view. I also included waterlines and buttock lines to assist in resizing the image after importing to a 2D graphics program.

    I placed this template object perpendicular to the centerline of the model in the y-z plane at the aft end of the model and on its own layer. Make sure to check the properties of this object/layer as Visible and Linesplan in the Layer Settings.

    When you save the DXF file for a given station, then import it into your graphics program, you first should delete all the extraneous content except for the Body plan. The DXF includes layers for all the program’s reference lines and other junk that you don’t want. You can then resize the object to the proper size against the program guidelines already set to the proper dimensions for the baseline, centerline, waterline, etc., according to the scale of your output print job. At least, this is the way I was able to accomplish what you are trying to do. As I mentioned before, the DXF model lines came through the process as vector line segments, at least in CorelDraw X7, which will require some cleanup.

    In my model, I created patterns for  non-overlapping frame faces and the keel, rather than station lines. I also included all the waterlines and buttock lines for reference in the final printout:

    Lines-Export-Example

    Please let me know if this helps at all.

    Terry

    Attachments:
    in reply to: Printing stations separate #44857
    Terrance Egolf
    Participant

    Hello, Flo.

    One thing you might try is to print station lines from the Lines Plan, one at a time, without making the hull layer visible.

    In the Edit tool, and in the Intersections column, check the hull layer that you want to display station lines in (and ensure that all other layers are unchecked). Also uncheck the Visible setting to avoid displaying the layer surface itself. (If your model’s keel is on a separate layer, you should perform these steps for that layer as well to obtain the full keel-to-rail line.)

    Next, go to the Tools tab, and open the Intersections tool. Clear all the station locations then type in the  x-value corresponding to the station you want to print.

    Back at the Home tab, look over the display of the station line you want to print by selecting either the Front or Aft view. If the line is not visible, be sure that Stations is selected in the Hull Display menu group and no other lines are selected. Select Both Sides, also in the Hull Display menu group, to show the full-width station template, if desired.

    When you are satisfied with the station form, select the Tools tab, and then Lines Plan in the Project Tools menu group. You should be able to see and print the selected station as part of the overall plan, along with the Baseline, Waterline, and Midline as set in the Project Settings. These reference lines will help align the station template for your project.

    Repeat the above steps for each station in your project, displaying only one station at at time in the Intersections setting. This could get tedious, but works around the lack of a 2D vector graphics program. I found that importing DXF images into CorelDraw from DELFTship generates lots of artifacts and fragmented curves. Post-import processing to clean up the curves can be a pain with that program.

    You probably should also do a test print from DELFTship to verify the output is full sized for the scale of your project.

    Terry

     

    Terrance Egolf
    Participant

    I don’t think this forum is monitored on a regular basis. I can contact one of the DS people to see if they will respond on this thread.

    Terrance Egolf
    Participant

    Hi RKG,

    A review of the manual indicates these terms are not defined but probably should be.

    Based on context, PS seems to mean “Port Side” and SB is the opposite, StarBoard side(?). The default coordinate system makes y-dimensions from the centerline to port positive, while to the starboard side negative. Don’t ask me why those abbreviations were used.

    The settings you referred to can be switched if desired.

    Terry

Viewing 15 posts - 1 through 15 (of 81 total)