3D Modeling

[John Mytton]

Good work Jerry. Thanks for sharing!

+1

Jerry's 3D imaging prowess just gets better and better.

JohnM

[Jerry Organ]

+1

Jerry's 3D imaging prowess just gets better and better.

JohnM

Thanks, John and Charles.

I think we have a winner with the Hum3D model. Unless you have day-to-day measuring access to the limousine, scoring some 80% alignment and near-alignment is pretty good.

If one is competent in their photo-matching, then a few tweaks of the Hum3D model would settle things. In this case, the back seat area of the model would have be moved back a bit, some door seams and the parade bar nudged forward a bit, the front tire and wheel well shifted back a small amount, and so forth. Not a big deal except the facet-count of the Hum3D model is astronomical and I can't do any reconstruction on my set-up. I'll have to instead adjust my old SketchUp model to match the new information.



On the other hand, my photo-matching could be really off and someone better at it could match up everything between the model and the factory photo! Incidentally, I had to adjust the vertical-scale of the factory-photo that was from the Henry Ford Museum. That's another thing that could be throwing matches off; early scanners sometimes were off with aspect ratio.

[Jerry Organ]

Was applying some of James Hackerott's on-site measurements to the Underpass model.

And was looking at some pictures when I noticed something odd.

East Side

West Side

The two single pillars in the center of the bridge project outward from the bridge more than the walkway tunnel twinned pillars. Maybe half a foot? Am I wrong?

[Charles Collins]

Was applying some of James Hackerott's on-site measurements to the Underpass model.

And was looking at some pictures when I noticed something odd.

East Side

West Side

The two single pillars in the center of the bridge project outward from the bridge more than the walkway tunnel twinned pillars. Maybe half a foot? Am I wrong?

I think I see what you are referring to Jerry. I think that some if the effect may be due to the angle. Example, the pillar at the far right appears to me to project out further than the one in the middle. Also, the fancy facade of the section of the bridge over the walkway doesn’t exist in the sections of the bridge over the roadway.

[James Hackerott]

I was using these images to improve my 3D model of the overpass by overlay of the renderings. As typical for me, I put the project aside for another and have not finished that work.
The images were taken during my “Storm Drain Panoramas” project and include the overpass under sun conditions showing useful shadows. I find that method fairly sensitive. If you can view your model with these parameters you might find the time stamped images useful.







And from the west
20220927_125830 Alt 55.1, Azimuth 147

[Jerry Organ]

Hi, James,



I used one of your images to show the subtle curvature at the bottom of the three traffic spans. I think the "arch" effect is mainly due to the raised steps at each end of the spans. I would say the central portion of the bottom edge of the spans are level.

Thank you for pointing out the shadows. The twined-pillars of the pedestrian walkway cast a smaller shadow onto the superstructure facing of the bridge ...



That adds to my suspicion that the twined-pillars are a little closer to the superstructure facing than the two single pillars.

[James Hackerott]

Here are two more images. One shows the arching steps from the south tunnel and the second shows the south column profile from the north tunnel-showing the column is recessed from the roadway column. I think I have dimensions for the pedestal so we can estimate the offset of the two columns. I will look later today. Jerry, I’m curious, can your software create a file for 3D printing?
 
S Commerce   20191101_120859


N Tunnel   20220907_124406

[Jerry Organ]

Here are two more images. One shows the arching steps from the south tunnel and the second shows the south column profile from the north tunnel-showing the column is recessed from the roadway column. I think I have dimensions for the pedestal so we can estimate the offset of the two columns. I will look later today. Jerry, I’m curious, can your software create a file for 3D printing?

It can in an indirect way using the most basic file format for 3D printing. I use the SketchUp 2017 application, and can export .STL and .DAE (Collada) files, in addition to SketchUp. I haven't tried anything with 3D printing but, through Googling, have some information.

1) The SketchUp model must be a solid. So I will make a copy of the Underpass model and on that copy, I will collapse the groups so the model are one entity (or alternatively divided into several entities that each are one group) with all the faces oriented the same way and "leaks" corrected. Because my models are made from scratch and not from intensive on-site scanning that usually over-samples, the facet-count is low, so the file size will be small.

2) The best straight-forward format is .STL.

• "After you check that your model is ready to be 3D printed, you need
  to output your model to a format your 3D printer understands, also
  known as the STL format."  -- SketchUp ( Link )
  
• "For example, if you are looking for the “easiest to use” among the
  properties of the best performer in any software area, STL is the best
  3D print file type. ... Though it has limitations in resolution/file size,
  cannot contain texture or color information, and in many regards is
  considered primitive, it has the advantage of universality. ( Link )

I did a sample using the walkway arch. I don't know if it's the complexity or my inexperience, but it became a six-hour project. The "leaks" can be a minute amount apart and not readily visible (the Solid Inspector will highlight in red the general area affected). I kept at it until I succeeded.

Alternatively, I could create .STL files as is (with no "errors" corrected) so others could correct them using more-modern software. Those exports won't be for while, when the model is at a reasonably-accurate stage.

Now I'm using 2017 software (including the version of "Solid Inspector 2"). It may be that a modern 3D printer could itself have software that corrects the minute "leaks"?