The story of Canada Atlantic flat cars

The first step is to get the information organized, and so I pulled some old pages over to my Canada Atlantic Railway site, and started to look through resources that might help.   Most of the data comes from the Official Railway Equipment Register, which I photocopied at Stanford University Library almost twenty years ago.  As I recall, I showed up at the library with something like a dollar in change to drive the photocopier; I had to be selective with my choices!  Even so, a little bit of analysis tells a brief and not altogether uninteresting story - well not uninteresting to me and the other student of the Canada Atlantic.

First off, we can see that by Pembroke's 1905 era, 1/3 of the flat cars were 20-year old cars from the 333 series.  This is good, because we happen to have a photo of one!  These were followed by the 855 series and the 600 series, which made up 1/5 and 1/6th of the fleet, respectively.

A surprising note is that, even in 1905, half of the CA flat cars were 33 feet long.  The other half were evenly split between 34 footers and 40-foot mammoths.  Those 40-footers were the first to get pulled into the Grand Trunk when they assumed operations at the end of 1905.

Speaking of the GTR, they also had a weakness for cars with higher capacity.  They immediately snapped up those that the CA had upgraded to carry 40 tons, leaving the older lighter cars for lumber service on the what had been the Canada Atlantic.  Note the line downgraded the capacity of some older cars to 15 tons as they neared the end of their lives.

Photo credit: Bytown Museum.

Starting point

What is your starting point for modelling the flatcars?  Using the old Juneco kit or straight scratchbuilding?  I assume these would have been all wood construction for your era. 

Marc Simpson

Old Juneco kit

Hi Marc,

I think I've actually built one of those old Juneco kits.  There's not much difference between one of those and scratchbuilding, and so, I am planning to scratchbuild.  

I like a porous material to represent unpainted wood - either card or wood itself.  However, the sides will be painted and so I'm considering styrene for an outside layer; this will enable me to put Grandt Line stake pockets in without drilling holes.  

I've been looking into getting a tungsten core, but that's likely to be too pricy, and so, I'll probably just go with lead.  1/16" tungsten would be heavy enough to hit the NMRA RP for weight without a load.

Cheers,
Rene

Weight plan

The thing that makes flatcars challenging is surely how to make them heavy enough to stay on the track when they are empty.  If it makes you feel any better, it looks like real railroads had the same problem back in the wood car era. They appear to have blocked empty flats near the back of the train for this reason.

My friend Brian Pate made his many flat cars with lead decks, which he scribed to represent individual boards.  Some years ago, I went and purchased enough 1/32" lead sheet to outfit a large roster of flat cars in this way.  Maybe I should still take this approach.

My current preference is to hide the weight as a false floor above shallow stringers.  I think Jack Burgess suggested the use of printed paper to disguise the weight.  The question is, how much lead do we need?

The NMRA preaches one ounce plus half ounce per inch of length.  These 33-foot cars will be just over four inches long, and so, they should be at least three ounces.  To get a sense of the size of such a piece of lead, I drew up the flat car in OnShape and weighed it.  Two layers of my 1/32 sheet will come out a little light at 2.3 ounces.

I investigated tungsten as well, and this comes out at a comfortable 3.5 ounces, assuming 1/16" sheet.  There is an easy online source for tungsten at tungsten.com.  They will even cut the material to your specifications, which is well worth it because the material is very difficult to work with.  I found their service exemplary, and received a quote the day after uploading my drawing.  Unfortunately, it came out at over $30 per car; that's a little rich for me, but it may be okay for others.

I really like this facility to predict the mass of a model in the drawing stage.  It's yet another reason to use OnShape.

Casting?

Rene;

Have you considered making a master and having it completely cast? You could add stake pockets after I suppose if they would add to much complication. Perhaps Shapeways could quote a metal version for comparison since you have done the drawing already. 

I made a few HOn3 flats while at UW and used pieces of brass between the stringers. It was handy to screw in trucks and couplers as well as the weight they added. Lead sounds sketchy to be scribing although I may be to far gone to notice if it would affect me. No sense chance 'em though. 

Curious...

I haven't started down the 3D printed road yet but am slowly getting ready to think about it.  One can print with metals.  Wonder if this would kill a lot of birds with one stone?

A further upshot is that the files could be sold or left with a Shapeways to reproduce for a little cash on the side. 

Nick

Juneco Kits

This is one reason I like the Juneco kit, it includes the metal bolsters that add weight low down. They are too light, but some fishing weights jammed between the joists helps. The fishing weights are easy to hammer into a shape the will sit between the joists where it is hard to see.

Full metal frame options

The frame as designed with the recess for the weight would weigh only half an ounce if printed in stainless steel.  I could also cast it in lead, which would weigh 3/4 of an ounce.  In basswood, it is worth less than 1/20th of an ounce - basically nothing.  

Both lead and steel seem like they don't yield enough benefit for the additional complexity and cost.  

Thanks for the suggestions, though

Rene

Juneco kit weight

My assembled Juneco flat car is 1 3/4 oz, including metal wheels.  The white metal bblsters and brake cylinder definitely help, but they're not enough.  

Juneco sold the parts separately.  I may indeed have some.  I wonder if they're still available.  

Thanks for the idea, Brent

Rebe

Getting organized

It's funny how jobs stack up. I needed to design the bolsters for the flat cars, which meant I needed to measure the truck bolsters, which meant that I had to find some trucks. These, along with most of my detail castings, live in one of the cabinets above the layout, and they're a bit of a pain to get at. So, while I was up there ferreting out the trucks, I also dug out some packs of nut-bolt-washer, queen post, brake gear, and other sundry castings.

These all went into one of my three project boxes - the one vacated by the roundhouse. So, I will be able to make progress on this project, but also put it away in favor of the locomotive as that project gets rolling.

Three is enough

Ellen Hood played flute in high school band, across from me and my clarinet.  Apparently, I was poorer than most kids at hiding my hobby as she found out about it, and introduced me to her dad, Tom.  For several years afterward, I joined Tom's work sessions on Tuesday evenings, I guess until I moved to Vancouver.

Through all those years, a board with a half-dozen half-finished box cars collected dust in a corner of the benchwork.  When I asked Tom about them, he said that six was just too many.   You have to drill every hole six times, and that feels more like a job than a hobby.  Three was his limit, and so, his boxcar project languished.

Now, hats off to folks like Chester Mackniewski with his mega-projects, or JD over at 78 Miles to Yosemite, who is working on 38 freight cars at once!  Chapeau to my old friend (and Tom's), Bill Scobie, who once assembled 16 P-B-L high side gondolas in a week.

But for me, as I was realigning twelve brake beams on swing-motion double archbar trucks from Shapeways, I thought back to Tom.  He was right: three models at once is enough.

Three

Nice progress. I've wondered if 3D sideframes and bolsters would be strong enough or brittle 

My limit seems to be two. Three gets to be a work out.

Size limits and brittle FUD

I've taken on building boxcars and would have to agree that size limits are a good thing.  I've got about 15 or so ready to paint and it did become a bit much.  3 at a time might be more manageable.  Economies of scale are one thing, doing the same thing over and over and over is another...

I have 3D printed trucks using frosted ultra detail plastic from Shapeways, and they are brittle, you have to be very careful when drilling or inserting wheelsets.  Thus far none have failed in service but I'm not running operating sessions yet.  My 3D printed trucks will be used in a smelter setting where the cars run from the mine to unloader, a distance of about 10 feet or so.  Hopefully they will hold up in this short distance service.  

Marc Simpson

Brittle FUD

These particular castings have been sitting in a box almost since Shapeways introduced FUD.  I was surprised to find how flexible they still are.  However, I also cleaned them in isopropyl alcohol, which is supposed to increase flexibility for FUD somehow. 

Marc, your smelter run is not much shorter than my whole layout!

cheers,

Rene

Towards swappable loads

It is a strange fact that I have never scratchbuilt a flat car before.  I clearly remember the Kalmbach book, Easy to Build Model Railroad Freight Cars, recommending that you should start out with a flat and work your way up to house cars; of course, I also clearly remember ignoring that advice and leaping in on a dustbin-fated hopper.  However, what makes it stranger is that I've had the project rolling around in my head for years.

Over the course of many bike rides, and much laying awake at night, I had come up with a plan for making flat cars so that a load built for one car could slip into the stake pockets of another of the same class.  The theory revolves around setting the stake pockets precisely using an alignment jig.

To avoid gluing pockets to jig, I made the jig from brass.  It turns out that 3/64" square brass almost fits into the Grandt Line single-bolt pockets that are a match for my cars.  A little filing on the pegs' faces makes them slip into the pockets nicely.

I only need a jig long enough for half the car as I can flip it to do the other half.  The spacing between the middle two pockets is different on the 333-series cars anyway.

Canonical flat car and load

In order to get swappable loads, I need to not only get the stake pockets all aligned, but also the side sills the same distance apart and aligned relative to one another.  I should count my blessings that these cars do not appear to have end pockets!  

I had thought that simply building all cars with identical side and end sills would achieve this alignment, but then I realized there could be minor errors in width or squareness that would throw everything out. What I really need to do is to build the flat cars so they all fit the same load. Likewise, all the loads should be built around a single flatcar.

So, I put together two jigs that I am pretentiously calling the canonical flat car and load (proper engineers probably have a name for such things). These jigs fix the locations of the stakes in the case of loads, and of the stake pockets in the case of cars. The "flat car" is, of course, not an actual flat car, but merely a piece of styrene with the pockets in the right locations. Similarly, the "load" will never travel the rails.

Now, I don't have any issue with loads having to always be aligned in one direction - I can mark their orientation on the bottom. However, I was tickled when I was able to put the "load" on the "flat car" backwards, and it still fit.

Sides and ends

The evening of detailing did nothing to shake my conviction that three models is enough.

Once I'd trapped the little wandering blighters with some tape, the stake pocket jig and its low-tech companion, the end sill template, worked beautifully.  Okay, the "end sill template" is just a piece of paper with ticks on it; to use it, I simply lined each bolt head with the tick.  It sounds easy until you have to do it 60 times!  By alternating side sills and end sills, I was able to not only stay sane, but also to let the stake pockets set before peeling the sills away from the jig.

Hehe

Imagine building a steel bridge and adding all the plates and rivets. Three will seem like making pancakes. What happened to doing this via OnShape and 3-d print them? Seems a natural. 

3D printing

With Shapeways tends not to come out as clean as building the parts up by hand.  Besides, I'd still be waiting for them, and then the shipping would kill me!  Finally, it's simply more satisfying to make them by hand.

Cheers,
Rene

A Lean Lesson

The Toyota Production System -- Lean Manufacturing here in North America -- will tell you that batch-building is bonkers.  Lean suggests that we should limit work in progress, and build things one at a time.  By building in batches, we build batches of blunders.

Maybe they're right.  As I started assembling the flat car sides and ends tonight, I found that the outermost nut-bolt-washer casting is almost beyond the edges of the side sills.  I will probably have to shave all twelve of them off and reapply them.  That's twelve NBW castings that I could have saved myself.

I only got as far as assembling a single car before deciding that I need a better way.  Predictably, the wood layer on the inside of the sills wicks solvent away from the bond, and makes it difficult to get a strong styrene bond. Adding more solvent yields a car that is permanently fixed to the canonical load - unless you panic quickly.

I think I will assemble the sides around the weights instead. Offsetting the hard lessons of the evening, the weights came out pretty close to bang on the measurement predicted by OnShape. Being 1/32" lead, they're very soft, and I had to roll them with a rolling pin (not the food one) to get them flat.

Integral weights

Okay, so the approach of welding the sides to the ends didn't work so well.  The weights themselves pose a much larger gluing surface, however, and so, I made them integral to the cars.

To do this, I first coloured up some paper to look like the bottoms of the deck boards, and spray-glued it to the bottoms of the weights.  I actually glued this to the wooden parts of the side sills with white glue, before adding epoxy between the plastic part of the side sills and the weights.  To ensure the two 1/32" thick weights fit beneath the decks, I clamped them to set.

To tell the truth, the plastic side and end sills seemed like a good idea at the time, but they're turning out to be more trouble than they're worth.  Being only .020", they're too thin to hold a straight line without the wood backs.  The result is that holding these cars square and true is more challenging that it should be.

Ends and stringers

Measure once cut twice.  These are words to live by.

Reasoning that there are only two truss rods on these 20-ton, 33-foot cars, and given the number of bolt heads on the ends, I felt six stringers were in order.  That makes 18 in total, which would be a disagreeable amount of fiddling to get them all to be straight.

So, I spent some time making a gauge that fits between the side sills, and positions the stringers for gluing.  When I've done this with styrene, I positioned all stringers first, and then fed solvent in from the side.  For the wood to paper bond here, it turned out to be easier to position and glue the stringers one at a time.

A cannier modeller would have left one end off and trimmed the stringers to length in place.  Sadly, at the time, I was not so cunning, and wound up performing a disagreeable amount of cutting to fit.  I'll aim to improve next time.

Ends finished

Please excuse the dreadful iPhone photo.  It's been one of those nights.  One of those nights when I should have just stayed upstairs.  

As expected, I wound up shaving off the outer-most nut bolt washer (NBW) castings so I could centre them relative to the side sills.  This also made it easier to add the ends of the end sills, which are actually lengths of 2x6 styrene added to the side sills.

Then it was time to add the NBWs back on, and I don't know what is going on with my motor skills tonight.  First I couldn't pick up an NBW without having it spring across the workbench to be swallowed up by....what?  There's nothing there to swallow the castings, but still they disappeared all the same.  Then I would drop them just as they were getting to the end sills, and they would evaporate.  Finally, those few that made it all the way to the end sills would be the only ones that would stick tenaciously to the tweezers like virgins at the volcano rim.  

Twelve castings applied.  Countless lost.

Sometimes you just have to laugh at yourself.

NBW’s

I think I must have a few hundred nbw’s floating around under my workbench. If I get 1 in 3 that actually make it to the model I think I am doing good! I keep thinking that there must be someway to modify a pair of tweezers so that they would pick these things up and not fling them across the room. Maybe some shrink wrap tubing on the ends of the tweezers? I will have to try that next time...