Pullman-Standard Mini-Hy Cube boxcar in 1:29

[Burl Rice]

I don't know how many of you here will be familiar with the name Brian Briggs, but several years ago, he built a Pullman-Standard Mini-Hy Cube boxcar in 1:29 scale. As far as I know, he only built one. I tried to get him to offer it as a kit, but he never would give me a definite answer.

Unfortunately, he passed away about a year ago. To my surprise, his wife, Margaret offered me patterns, molds & extra castings he had made for various projects. Among them were the patterns for his PS Mini-Hy Cube boxcar. He made it as a flat kit, with no patterns for the underframe. It was my intention when I acquired it (with Margaret's blessing) to produce a kit from it. Lately, I spent a little time working out the design of the underframe. I think I can make an operating cushioned underframe, with sliding center sill.

I am working primarily in Sketchup:

Since Brian cut up USAT boxcar ends & used them as his pattern, I felt it best to re-make this part. I am learning Autodesk Fusion 360 to draw these, since it has a filet function that I like better than Sketchup.

 

[Paul M]

Looks good so far, best wishes for the build

 

[daveyb]

what year did these enter service

 

[Burl Rice]

1976-ish. Will check for sure when I get home.

 

[Burl Rice]

I had my dates mixed up.

 

[Burl Rice]

A little closer to having all the parts worked out for the underframe:

 

[Burl Rice]

My inexperience in F360, coming to light, I ended up redrawing the end from scratch. Watching their tutorials, they put an emphasis on making everything components.

When I started cutting the mounting holes for the ladders, I realized making each rib a component was not going to work. Making a hole in one, resulted in a hole in every rib.

When it was time to scale it down to 1:29, I also encountered odd behavior in how it handled what they call “patterns”. Where I had distributed rivets every 2 inches, it maintained the 2 inch spacing after the scale operation.

Redrawing it also allowed me to correct something I missed the first time. The ribs are not completely straight, but slope gently from the middle to the end.

 

[Burl Rice]

Got everything broken down & arranged for a Shapeways order.

I'm taking a different approach on the ladders this time.

In the past, I have had them 3d printed in SLS nylon. The texture was a little rough that way, but I could live with it. However, Shapeways has a habit of revising their printing guidelines. I guess they’re still going through growing pains. I have sometimes found things I printed 6 months ago, they say are no longer printable. Occasionally, things have also arrived broken (I assume they were handled too rough during cleaning).

This time around, I’m going to try making the ladders in brass. I’ll use the acrylic prints (below) to make a mold for waxes. These waxes will be sent off for investment casting. The rungs (with NBW) will be one piece, and the rails will be a separate piece. I made them with locating pins to make assembly easier. I will find out in a few weeks if this is a better approach.

These parts will be used as patterns for resin casting:

 

[Burl Rice]

Laser cut parts came in today. I had these made at http://ponoko.com/
Kind of like Shapeways, but for laser cutting. I have not taken the paper off yet.

 

[Burl Rice]

My prints came in last week:

 

[daveyb]

looking good!

 

[Burl Rice]

Progress on the underframe pattern:

The printed parts are only there for reference right now. I will be making resin copies of them later.

 

[Burl Rice]

Poured the molds for the underframe this weekend, and I immediately saw some things I wish I had done different.

Since this car has a cushioned underframe, it has four beams for center sills, instead of the usual two. I had planned on making the outside center sill as two pieces. I could see right away they were going to be too floppy to glue in place and get a proper alignment. My goal was to make the sliding underframe operational, so this would be critical.

I also had all the bearings aligned to take vertical load only. I can see now, it will function better if I added some to maintain the clearance between the inner & outer center sills.

The other problem I see is that one beam has to be cut to make clearance for the trainline piping. I’m afraid this will be easily broken if its made out of resin. Brass would be much better.

So, I think I’m going to revise it to make the outer center sills part of the floor casting. Just wish I had figured this out before I used four pounds of RTV, but you live & learn.

 

[Burl Rice]

After stewing on it a couple days, I have found a design I like better. Funny how simpler designs seem like they are harder to come up with.

The outer center sill (green) will now be part of the main floor casting. It will have a channel that the bearings will ride in. The inner center sill (pink) will have slots for the bearing shafts. This should lock the whole thing in & keep it from sagging vertically. It also means I only have to reprint the center sill, and I can use the rest of the floor pattern with no further modifications.

This also allowed me to change the orientation of the pattern in the (eventual) RTV mold by 180 degrees. Now I can also model the spring for the cushioning unit as part of the casting.

 

[Burl Rice]

While I’m waiting on my revised underframe parts to come in, I decided to work on the side patterns. Started with the vertical panels. Looks odd right now, because the glue shows through in places where the castings are very thin. It should all even out when its painted though.

 

[Burl Rice]

My revised underframe prints came in last week. When I started to glue them in, I was having trouble keeping them even. After some double checking, I realized I had some variance in the I-beams. So I stripped out the old glue, and made up this sanding stick out of aluminum C-channel & 220 grit sand paper:

With that, I was able to fine tune the fit & get a fairly straight line. When I got ready to try gluing them again, I made up a spacer to represent the sliding part of the centersill (cut the same width), and added .015” styrene for clearance:

With that in place, I could tack it from the outside, remove the spacer & finish gluing from the inside.

Now everything looks like it fits well. I don’t want to get in a hurry on this step, so I’ll look at it for a while. If I can’t find anything wrong with it, I’ll pour another mold tomorrow (I have found sleeping on it helps me avoid mistakes due to haste):

 

[Burl Rice]

Poured my molds for the revised lost wax parts yesterday, only to find the two I really needed had refused to completely cure. I was about to write them off as a total loss, but then I realized I could still get patterns out of them sufficient to finish the underframe casting.

I think the last thing I need on it is the divots for the new mounting holes:

I will remake the lost wax molds later. Still trying to figure out what inhibited the RTV, but they say platinum-based silicones are very touchy. When this has happened before, I cleaned the patterns, applied a fresh coat of paint, remolded, and they were fine the next time.

 

[Burl Rice]

Finished the centersill pattern & molded everything. Made up some castings to check the fit:

 

[Burl Rice]

Sort of got derailed on this project. Shapeways took their sweet time on my last order (3-4 weeks). When it did come in, I got covered up at work.

Anyway, I got back around to working on the sides again. I’ve re-done some more work here. What I had before was OK, but I felt like I could get it closer. I had previously used my old trick of using Aluminum tape to imitate oil-canned sheet metal. Sometimes it worked well, sometimes it was hard to control the effect. And there was always un-evenness between the panels that made it hard to apply the weld bead decals later.

What I did different this time was to model the oil-canning in Fusion 360. I started with an image like this:

I run this through a script that interpolates the z-coordinate based pixel lightness/darkness. Sort of imitates the result a 3d scan, without all the expensive hardware. I do some mesh reduction & smoothing on the raw image, and scale it to get the print to the thickness I want. Which looks something like this:

To avoid obvious repetition, I had three prints made: one regular, one flipped horizontal, and one flipped vertical. The prints came in with obvious scan lines, which I expected. I sanded them with 220 grit sandpaper until they no longer looked like topographical maps. Then I made an RTV mold, and cast copies in resin. Using a combination of castings, some rotated 180 degrees, I built up the side pattern:

 

[royale]

Craftsmanship at its best - beautifully modelled.