Author Topic: Pennsy Centipedes  (Read 10229 times)

0 Members and 1 Guest are viewing this topic.

peteski

  • Crew
  • *
  • Posts: 32839
  • Gender: Male
  • Honorary Resident Curmudgeon
  • Respect: +5276
    • Coming (not so) soon...
Re: Pennsy Centipedes
« Reply #45 on: October 10, 2012, 11:15:22 PM »
0
I came up with the scale by dividing 1 by 160. I got 0.00625% which is obviously total;y wrong. 0.625% may be correct. I'll try it.

Frank
Right, but then I moved the decimal point 2 places to the right (as percents are expressed in 1/100s).
. . . 42 . . .

GimpLizard

  • Crew
  • *
  • Posts: 526
  • Respect: +51
Re: Pennsy Centipedes
« Reply #46 on: October 12, 2012, 07:47:34 AM »
0
The question is... what scale was your original TSM model? If, like Peteski alludes, it was full (1:1) scale then you do need to scale it (multiply) by .00625 (1/160) to get N scale. If you modeled it in HO (1/87) scale, you would then need to scale it down by .54275 (87/160).  Moving the decimal point two places to the right, gets you 0.625% & 54.275% respectfully.

Now, once you get all this worked out, promise me you'll do up a Great Northern NW-5 next*. Alrighty?  :D

*Seeing as I haven't had any luck getting Atlas to do it. (Not to mention that 1964 Rambler Ambassador 9 passenger station wagon.)

kelticsylk

  • Crew
  • *
  • Posts: 781
  • Respect: 0
    • Milepost 15
Re: Pennsy Centipedes
« Reply #47 on: October 12, 2012, 12:46:45 PM »
0
Not to get your hopes up, but I think I do have drawings of the NW-5 in my library. The only Great Northern unit I modeled was the W-1 heavy electric.

Don't have any '64 Ramblers, but I did do a bunch of trucks from the 1940's, including White Mack and Studebaker COE's, semi-trailers and a Diamond Reo pickup.

I'm also revising the Centipede to have better detail and higher resolution than I could use in Train Simulator. There was a limit on how many polygons the sim could handle so rounded shapes like cylinders had to have 16 sides or less (we did a couple with 32).

Frank

kelticsylk

  • Crew
  • *
  • Posts: 781
  • Respect: 0
    • Milepost 15
Re: Pennsy Centipedes
« Reply #48 on: October 12, 2012, 02:59:30 PM »
0
I tried scaling this thing again and this is the size TSM reported. I'm think this is about 6" or so. .560547 feet x 160 is about 89 feet.


The dimensions are approximate because you have to eyeball the "measurer"  :)

If the techs are getting different dimensions, maybe 3D Canvas changes it somehow. Gonna check.

Frank

wcfn100

  • Crew
  • *
  • Posts: 8837
  • Respect: +1218
    • Chicago Great Western Modeler
Re: Pennsy Centipedes
« Reply #49 on: October 12, 2012, 04:19:33 PM »
0
I recommend viewing your models with MiniMagics .stl viewer. It will show you any problems and has some tools for fixing them.  I think you can measure your models too.

http://software.materialise.com/minimagics

I used it to help me build better models with less errors.

Jason

kelticsylk

  • Crew
  • *
  • Posts: 781
  • Respect: 0
    • Milepost 15
Re: Pennsy Centipedes
« Reply #50 on: October 12, 2012, 05:43:42 PM »
0
I checked 3D Canvas (I use it to convert the TSM files to .stl) and the scale still seems OK...

The dimensions at the bottom of the page are about right 9.4 feet x 10.6 feet x 89.4 feet. I'll double check the dimensions with the info I have on the real deal.

I downloaded MiniMagicks and looked at the .stl file.

I'm not sure what I'm looking at as far as the part information goes.  The -1.9 does not look promising. I also viewed it with 3D Exploration and got a different set of numbers but they didn't make sense to me either.

I requested a download of trial version Magick RP so I could run the "fix my part" function. It has to be approved.

I created three files from the three different programs. I'm sending them on to Metro.

Frank

wcfn100

  • Crew
  • *
  • Posts: 8837
  • Respect: +1218
    • Chicago Great Western Modeler
Re: Pennsy Centipedes
« Reply #51 on: October 12, 2012, 08:58:13 PM »
0
  The -1.9 does not look promising.


I think the 0 Volume is your biggest problem right now.

Jason

kelticsylk

  • Crew
  • *
  • Posts: 781
  • Respect: 0
    • Milepost 15
Re: Pennsy Centipedes
« Reply #52 on: October 13, 2012, 09:16:39 AM »
0
I think the 0 Volume is your biggest problem right now.

Jason
Didn't see that Thanks!
Frank

kelticsylk

  • Crew
  • *
  • Posts: 781
  • Respect: 0
    • Milepost 15
Re: Pennsy Centipedes
« Reply #53 on: October 21, 2012, 11:35:13 PM »
0
I was able to get some volume in the last version I sent to Metro, but the scale is still wrong and I have no idea what is causing it. The resolution of the model needs to be better, tech reports that the surface is faceted.

I went back into TSM and used the highest resolution it allows. You can't create an actual curve only a polygon with 64 sides. Since the model is drawn full size (1:1) and then "shrunk" to 1:160, the facets may not be as noticeable. The program has a function to "smooth" by subdividing the polys but this does weird things to the shape.

Short story...I thought it might be possible to turn these 3D shapes into real models. I have a bunch of different 3D locomotives but they require too much work to modify for 3D printing and there is no guarantee they would make good masters. I think I'll have to switch to another design program or just build a master if I want a centipede for the Allegheny Eastern.

Regards,
Frank Musick

wcfn100

  • Crew
  • *
  • Posts: 8837
  • Respect: +1218
    • Chicago Great Western Modeler
Re: Pennsy Centipedes
« Reply #54 on: October 22, 2012, 12:21:52 AM »
0
I think I'll have to switch to another design program or just build a master if I want a centipede for the Allegheny Eastern.


Probably your best bet.

While maybe not the best geometric modeler, Rhino is a program that's pretty easy to get a hold of and not too hard to learn.  I'm not sure if all versions export to .stl or if it's just the more recent ones.

Jason

bbussey

  • Crew
  • *
  • Posts: 8881
  • Gender: Male
  • Respect: +4709
    • www.bbussey.net
Re: Pennsy Centipedes
« Reply #55 on: October 22, 2012, 09:59:32 AM »
0
The question is... what scale was your original TSM model? If, like Peteski alludes, it was full (1:1) scale then you do need to scale it (multiply) by .00625 (1/160) to get N scale. If you modeled it in HO (1/87) scale, you would then need to scale it down by .54275 (87/160).  Moving the decimal point two places to the right, gets you 0.625% & 54.275% respectfully.

Now, once you get all this worked out, promise me you'll do up a Great Northern NW-5 next*. Alrighty?  :D

*Seeing as I haven't had any luck getting Atlas to do it. (Not to mention that 1964 Rambler Ambassador 9 passenger station wagon.)

87.1:1 is the accepted HO ratio.

You should build most of the digital model in full 1:1 scale, then reduce to desired scale when done and adjust any details that are too fine to render.

I checked 3D Canvas (I use it to convert the TSM files to .stl) and the scale still seems OK...

The dimensions at the bottom of the page are about right 9.4 feet x 10.6 feet x 89.4 feet. I'll double check the dimensions with the info I have on the real deal.

I downloaded MiniMagicks and looked at the .stl file.

I'm not sure what I'm looking at as far as the part information goes.  The -1.9 does not look promising. I also viewed it with 3D Exploration and got a different set of numbers but they didn't make sense to me either.

I requested a download of trial version Magick RP so I could run the "fix my part" function. It has to be approved.

I created three files from the three different programs. I'm sending them on to Metro.

STL files are solid files converted to a series of triangles.  If the file contains too many triangles, it can cause an error.  The number of triangles is determined by how fine the model is rendered to.  You should reduce the model to scale before converting to STL format.

Rhino accepts a number of 3D formats for importing.  I would convert to a format that utilizes actual curved vectors to import there.  STL should be reserved for the output files sent to the 3D contractors for solid model rendering.
Bryan Busséy
NHRHTA #2246
NSE #1117
www.bbussey.net