Wednesday, June 4, 2014

Success!

I finally received my last attempt at printing a car from Shapeways ... and good timing too as I’m out of money! It is true when they say all the costs are in the R & D.

But ... success!



I can finally say that I have printed an O scale rapid transit car that can pass casual inspection.

This time, as mentioned previously, I split the car sides each into five sections: the two end doors, the center doors, and the right and left window sections. I had to do this so I could make the parts small enough that they could be printed in Shapeway's Frosted Ultra Detail.

Below shows what the parts (the doors and the window sections) look like when first taken out of the box (fresh from the factory, so to speak):



These parts came with a very heavy coat of wax. So, they were soaked in Bestine for a few minutes then rinsed with water. I cut them from the sprues before soaking.

Once dry, all the parts (six doors total and four window sections) were prime with Rustoleum grey auto primer.

Some close ups of the parts once primed. There has been no sanding of these parts.




As shown above, the finish on the parts (roughness, lines from the printing process) is very good. Good enough that I haven't sanded. Also the fit of the parts (doors to window sections) is also very good. No sanding required. The sides look crooked as they don't lie absolutely flat. I had to "bump" the opposite side of the door recesses out a bit as the material got very thin and had a tendency to warp. But, I put the sides along a straight edge and they are straight.

Below shows the entire "kit" of parts: two ends, roof, six side doors and four window sections.


This is definitely a "shake the box" type of kit. Below shows the sides assembled. Again, no sanding to make the parts fit, its just a friction fit at this time:


To assemble the sides into one piece, I used bits of 0.040" styrene along with ACC glue to join the door sections to the window sections. I then glued a piece of brass square stock along the top of the side. The ends were then glued to the sides.





Some observations:

The whole process has been way too expensive. Granted, some of this has been my own fault (destroyed masters due to casting mishaps, design errors) but the cost of 3D printing is still pretty high for a one off model. A fleet of same cars might be better served by making molds.

Rivets - I just don't get it. Superman has his Lex Luthor, Batman his Joker, my arch enemy will be rivets, especially 3D printed O scale rivets. Granted, on this model, being printed in Shapeway's Frosted Ultra Detail, the rivet detail (where it printed) is very good. I say where it printed as there are spots where either the rivets just didn't print or were knocked off during processing and shipping. They are very fragile. 98% of the rivets printed without issue. But, there are a few that are just missing. This isn't a deal breaker as I'm just happy the car turned out. But if I were making a commercial mold, I'd be miffed.

The roof - That single piece was the most expensive part. So, I will make a mold of that and use that to make copies. Now, you purists might note that there are no rivets around the edge of the roof. This material wouldn't allow those to be printed so those will have to be added via rivet decals. 

Next up will be painting...stay tuned.

Sunday, June 1, 2014

First coat of paint on the 6200's is done

The first coat of paint is done on the MTH 6200's. So far. I am happy with how the cars are turning out.




My lighting isn't the best, so the colors aren't 100% accurately being reproduced in these pictures. But, I think the colors are close enough, in person, to pass. Once "Dulcoated" and weathered slightly, they should look pretty good.


Sunday, May 18, 2014

Updates for May 2014 ... painting and printing

As a diversion from the endless issues related to 3D printing, I've decided to repaint my MTH 6200 Bicentennial 6200 set to the Alpine White and Mint Green. I've always absolutely hated the Bicentennial scheme the MTH chose, so changing paint schemes has always been on the "to do" list.

After a day's soaking in 90% rubbing alcohol and a coat of primer, the first car waits for space in the paint shop at the Sheridan Road station.


The below car has just emerged from its alcohol bath is waiting for primer.


And the whole car that has been primed:


MTH actually did a very nice job on these cars. The details are good, its just a shame that they are slightly out of scale for 1:48, more like 1:45. Luckily, they were relatively easy to disassemble and the paint came off very easy.

As for the Alpine White and Mint Green (and without custom mixing) the closest match I can find are Vallejo Model Color Ivory 70.918 and Tru-Color CP Rail Action Green TCP-32. For my poor eyes and layout lighting. I feel that the colors are "close enough".

Below is a paint sample I painted on styrene, with grey primer underneath. The white is more ivory in person. To me, again, its close enough.


As for 3D printing and the CTR 4000 baldies, I got my 2nd set of car sides (with end doors as separate pieces) back from Shapeways. These were printed on Y axis (that is, vertical) and using the material frosted detail. I am not happy as to how these sides turned out. 

After priming and some light sanding:



The fit of the end doors (printed in frosted ultra detail) to the sides (printed in frosted detail) is good, so no issues there. However, the resolution and quality where the rivets are printed just still isn't acceptable for an O scale model. From a casual viewing distance (3 feet), the parts look good. But, up close, the roughness in areas that have a lot of rivets is very apparent and distracting.

So, as one last measure to see if it is possible to create a 3D printed O scale model of acceptable quality, I've decided to split the side into five separate pieces (two end doors, middle door and 2 window sections). This way, all the pieces can be printed in the highest resolution material, frosted ultra detail. Splitting the side into smaller pieces will also help eliminate warping and errors I've encountered with larger parts. I've noticed that quality along the length of a part is inconsistent as I think the part shifts or cools and a different rate during the printing process.

So, the idea is to "sprue" parts together so that they will fit within the limitations of the frosted ultra detail and have the highest resolution possible.



Then, when received, just cut apart and assemble. Again, the fit of the parts is very good, so there should be minimal gaps.

Obviously having a side made of separate pieces isn't the idea, but the quality just isn't there for the side being printed in its entirety in frosted detail. Additional bracing will need to be done on the inside of the model, but this is the best compromise that I could come up with knowing the limitations of the material. In addition, I've heard that master model builders such as Bill Clouser used to make sides out of separate pieces, so if it was good enough for him, I'll give it a try.

Monday, May 5, 2014

Skokie Swift's 50th Anniversary Party

I was fortunate enough to be able to attend the Skokie Swift's 50th Anniversary party, where the CTA ran the historic 4000 series cars 4271 and 4272.

Below are some photos and a video I took of the day's festivities.













And some video of the day:


Sunday, March 30, 2014

3D Modeling Update

I finally got my prints back from Shapeways, and overall, its still a work in progress.

Below are some of the baldie samples placed loosely together.








 Overall, the quality is pretty good, The end in the above photos has NOT been sanded. The side however has. I think as printed, with some sanding, the prints can make a decent model.

The side does have some warp-age to it, but that could be corrected by internal bracing.

My original plan was to make molds of the sides and then cast the sides out of resin.

Well, that didn't work out too well. I tried to make a mold out of one of the sides, and I didn't prime the side beforehand nor did I add mold release. BIG MISTAKE as seen by the below "after" pictures of what was left of the side:



As you can see, the silicone rubber stuck to the part and basically broke it when I tried to remove the side from the material.

I was able to successfully make a mold out of the side shown at the top of the page. However, I haven't really been too happy with the quality of the castings I've been getting. I think for a model, just working directly from the prints is preferable.

My Met L structure turned out very well. However, I realized that copying directly from prototype plans was a mistake. The center line of the track girders on the prototype is 5 feet. With O standard being 5 feet, I thought it wouldn't be an issue in replicating the measurement in the model. 

As can be seen in the below photo, the O standard gauge, while 5 feet, puts the center line of the rail on the outer edges of where the girder would be. The flat areas on the cross girder are where the long, horizontal girders would be.


The girder center lines should fall directly under the center line of the rail, NOT the gauge.

I will need to redraw the model and resend for printing. The difference isn't large, but if going through the trouble of making the lattice style, the girder / rail spacing needs to be correct.

The model I have is correct for proto48 track, but not standard O gauge (5' gauge track).

I've also made progress on my newer, more detailed structure. This L structure is a combination of cast girders, styrene shapes and 3D printed (then cast out of resin) corner braces.




I've tried to add as much rivet detail as I can to the various components. The column height is a little taller than I'd prefer, but that was dictated by the benchwork in this area of the layout. The track obviously needs to stay level with the rest of the layout, which dictates the column height.

Tuesday, March 11, 2014

Rejected, Part Two!

Ugh! Rejected again by Shapeways!

This is getting to be a bad habit!

Anyways, on the ends of the 4000's, I had a bracket to hold the destination sign. Apparently 1/2" thickness in O scale is too small due to "thin wires".

The pictures from Shapeways:



So, no built in brackets allowed. I've left a small, square pad where I can attach the cross piece to hold the destination sign.

This error is unlike my first error. That first one was just sloppiness on my part. I didn't properly size part of a model.

On this error, I guess I just pushed the boundaries of the material too far. Oh well, corrections have been made (again).

Saturday, March 8, 2014

Virtual Rapid Transit

All work and no play makes for a boring life ... so ...

When I'm not working on either the layout or making 3D models, some virtual modeling / rapid transit planning is in order.

First up, is a rapid transit game / simulator called Cities in Motion 2. In this game, you are the owner of a transit company and can create bus, trackless tram, streetcar and subway lines. You buy equipment, set schedules, set fares and attempt to run a transit empire.

Some screen shots of the city I've been working on. So far I've got street cars (light rail in the European style), buses and some subway. I play on the easy setting, so unlimited money means unlimited fun.








This game is available on Steam, for about $30. Steam often does sales, so watch for those and it can be had for a lot cheaper.

A new (and FREE) game is called Mini Metro:




In this web based, minimalist game, you create a subway system as new stations pop up. For a very simple, minimalist game, its quite fun and engrossing. 

It does require a plug in to be installed, but I've encountered no issues (it isn't Flash, so there is that!)/

The link to play is: Mini Metro