Friday, December 30, 2011

Modeling project: M&StL box car

At the meeting in Lisle, Illinois last October, I was intrigued by a clinic given by Clark Propst, entitled “M&StL Freight Cars You Should Build.” Though this railroad was not large, it did have a freight car fleet of some size; of greater importance, every model freight car fleet does need a few representatives of “unlikely” railroads (unlikely in the Gilbert-Nelson sense). Whenever samplings of prototype data are done, a few cars always show up from relatively small railroads.
     Although none of the car projects Clark presented directly inspired me, I did begin to think that some suitable Minneapolis & St. Louis car might be a good subject. With that thought in the back of my mind, I was browsing in Railroad Prototype Cyclopedia issue 18 awhile back, specifically the article on the ARA 1923 standard box cars. Turns out that the M&StL bought 500 of these cars, their 24000 series (even numbers only). The article includes four prototype photos.
     Mental wheels started to turn. I realized that the old Train Miniature single-sheathed box car is somewhat like the single-sheathed ARA car, though details are not up to current standards. Might I have one somewhere in my freight car stash?
     Yes, I did. Moreover, it turned out to have some of the features of the M&StL cars. (You always have to check the details of the particular model, because Train Miniature or TM mixed and matched car sides, roofs and ends, sometimes in ways that had no prototype.) Though lettered as a Wellsville, Addison & Galeton (WAG) car, it had wood ends, outside metal roof, and wood doors, all features of the M&StL cars--though the ends weren’t quite right--more on that in a moment. The sides are a Howe truss arrangement instead of the ARA Pratt truss, so the model can only be a stand-in, but would have the right general look. Here is the model out of the box:

The WAG lettering already present, though interesting, would only be correct for cars obtained by the WAG in 1958, well after the time I model, and in any case the WAG cars were plate-end cars purchased from Boston & Maine, and the execution of the TM lettering is poor.
     How about decals for M&StL? I always like to be sure I can letter a model before embarking on the construction work. I sent e-mail to Clark to ask where I might get appropriate decals, and he replied that Mark Vaughan offers a set which, though intended for steel cars, could mostly be used for the ARA cars, with different dimensional data. I promptly ordered a decal set, Mark’s set BB. If you’d like to take a look at Mark’s listing, here’s a link: .
     Meanwhile, I studied the TM kit further. The ends need to have the bracing somewhat like standard ARA bracing, instead of TM’s smooth end. Here’s a drawing of the prototype standard ARA end, an end which most buyers of these cars did not select, but apparently M&StL did:

although the M&StL end was not quite like this (more in a moment). The car body in the TM kit has this end:

Not only the incorrect placard board needs to be removed, but also the ladder, as the prototype ladder was outside the diagonal brace. The inset roof end shown in the drawing was apparently not used by M&StL, though it could be simulated with a crosswise brace. I decided to neglect the annoying fact that the ARA-design end had horizontal sheathing boards, and the TM end has vertical ones.
     For this kind of work, I like to use an Exacto No. 17 (chisel) blade, to carve off the parts to be removed. I keep the blade under control, so when it cuts through the part it doesn’t skip ahead. After carving off most material, I sand lightly to make sure the surface is level. Then the board divisions are re-scribed through the areas where material was removed. The end then looks like this:

     Next the end bracing needs to be added. I used styrene scale 4 x 4-inch strip, which is oversize for the prototype bracing but is a good match for the simulated Z-bar bracing of the TM kit. (Might as well have sides and ends look about the same.)
     The styrene strips were sanded to approximate the shape of ARA hat-section braces and cemented to the car ends. I also drilled holes for wire grab irons on the ends, and for two grabs at the left of the sides (the TM model has only one side grab, as is correct for mid-1920s freight cars, but after 1931 two were required, and the second ones were added to older cars when next shopped).
     The six-rung ladder for the ends was a minor problem, as none of the commercial ladders I had in my stock had a rung spacing matching the side ladders (normally ladder rungs on sides and ends are equally spaced and located in the same plane). The nearest ladder spacing I had was the InterMountain 1937 AAR box car ladders, which I obtained (along with many other useful parts) in their Body Details set, part no. 40700-59. I cut off the lowest rung and added styrene spacers to permit the ladder to stand off above the end braces (the prototype ladders were arranged the same way). The completed A end looked like this before painting:

Obviously roof detailing remains to be done in this photo, and I’ve omitted the flanges by which the car’s sheathing was bolted to these end braces.
     The door tracks on the TM kit are rather oversize, especially the immense ledge which acts as a lower door track. I removed most of this, as well as the large “claws” on the door, which guide the TM door along the bottom track if it is operated (mine are glued on). The TM doors are intended to represent a bottom-supported door, whereas the M&StL cars had top-hung doors, but I decided not to correct this detail.
     The TM running board is seriously over-thick, but can be used if it is filed and sanded down to a realistic thickness. I did that, and replaced the cast-on corner grab irons with Tichy wire parts. I also glued the TM kit weights to the inside of the floor, using canopy cement (R/C-56) and adding about one inch of steel strip at the door location to add a little weight. I cut off the coupler pocket box lids so they would be separate parts, and removable if necessary for coupler maintenance.
     Of value beyond this one car project is a trick I often use for addition of metal sill steps. Drilling into the car body itself can be problematic if the shell is thin. So I glue a short length of scale 4 x 4-inch styrene inside the side sill, thus thickening the area where drilling is to be done. Once the cement is well set, I usually drill approximately along the interface between the car shell and the styrene block. Here’s a photo to show how it looks when A-Line Style C steps are attached:

     Finally, after painting, applying the Vaughan decals, and weathering the model, here is the final result. I used Tahoe Model Works Buckeye trucks.

     At this point a prototype photo may be useful. This one is from the Bob’s Photo collection, and location, date and photographer are unknown.

     This project was chosen as another example to show the kind of upgrades and modifications one can perform to get closer to a particular prototype, without any particular goal of creating an “excellent” model. And remember, this model is a stand-in. For me, this will be what I call a “main line” model, normally only used in passing trains, so correcting every detail just isn’t essential. But now I’m glad to have an M&StL freight car in my fleet.
Tony Thompson

Monday, December 26, 2011

Ballard, Part 3: first train

I have posted before about the town of Ballard on my layout. There was a general description of the town and its industries, under the title “Layout design: Ballard,” which can be viewed at the link: , and I also have described the role of Ballard in my layout scheme, in the post entitled “Layout design: locale” (that post can be found at: ).
     Placing the town on the layout required installing the track board, which was a landmark event in reassembling the existing parts of the layout, since Ballard is one of two major components in my layout design. I related this installation process in the first two posts with this title, “Ballard-1” (at: ) and “Ballard-2” (at: ).
     With the track board installed, the backdrop repaired and retouched, and some industrial buildings repaired and reinstalled, it was time to address operation. Track and wiring connections have now been completed, and I was able to operate a train into Ballard for the first time. Here it is on arrival:

This train is small, just SP Consolidation 2836, C&NW box car 65114, and PFE reefer 73617, with caboose 783, but it was still fun to operate. In this photo, the staging transfer table can be seen as a level below Ballard. At the moment, Ballard is only operated with DCC, and I may continue to make that the case.
     More needs to be done, but now that I can switch Ballard, waybill operation will become more relevant to the layout, and I can expand the things I can do. I have always enjoyed switching, so having more of it handy is fun.
Tony Thompson

Wednesday, December 21, 2011

New tank car article in RMC

I’ve finally received in the mail my copy of the January 2012 issue of Railroad Model Craftsman (RMC), containing my article on tank car modeling (pages 61-66). The article describes two tank car models, a Stauffer Chemical Company acid car, and a Buffalo Electro Chemical Company (Becco) peroxide car. These were first mentioned almost a year ago in this blog, in a description of the modeling projects I was taking to display at the annual Prototype Rails meeting in Cocoa Beach, Florida (see: ). Following that meeting, I posted photos of the models which had been displayed, including the ones in the RMC article, and those are at this link: .
     I was tempted to summarize in a blog post the model-building process for these two cars, but Bill Schaumburg, editor of Railroad Model Craftsman, was interested in an article for the magazine, so I wrote it up from that perspective. Whether an Internet description of the same models would in any way compete with the printed magazine, I don’t know, but it seemed to me that a single submission and presentation was a more ethical approach.
     I also want to acknowledge again the inspiration of Mark Feddersen’s tank car articles, fully 25 years ago. There were two of them, and I’ve now done models derived from both of his projects, although with somewhat different modeling approaches. Here are citations to Mark’s magazine articles:

Mark Feddersen, “ICC-104 Insulated Tank Car,” Mainline Modeler, October 1985, pp. 63-69.
Mark Feddersen, “The Becco peroxide tank car,” Prototype Modeler, January 1986, pp. 16-19. 

It’s the second of these articles which inspired my models in the January 2012 RMC, but the first article was part of the inspiration for my previous RMC article on tank cars, in July 2011, pages 65-71. Some comments and further details about that article were posted earlier in this blog; you can read them at: .
     Additional tank car posts in this blog include my scratchbuilding project of an asymmetrical tank car with two compartments, which is at: , and a description of my method of modeling Southern Pacific tank cars from Athearn models (revised from an SP Society Trainline article), which can be found at: .
     Anyway, it’s nice to be in print in RMC again. I continue to work on tank car modeling projects, and I plan to describe them in future posts.
Tony Thompson

Monday, December 19, 2011

Route cards — 6

In prior posts on this topic, I have given dimensions of placard boards and route card boards as I have found them on drawings, in the Car Builders Cyclopedia and other sources, but as I mentioned, these parts are often not dimensioned, suggesting that there was a standard size. I inquired of Guy Wilber, who has access to documents like the AAR Manual (a loose-leaf book issued by the AAR Mechanical Division, containing current standard and recommended practices), whether there was in fact a standard for these boards.
      Guy was kind enough to look this up and send me a series of extracts from documents of various ages. For the situation before 1937, in the early days of all-steel cars, an AAR Manual comment states that “Apparently about half the roads are using the board shown on Plate 250 and others are using boards with clear space smaller than 5" x 8", but which probably serve the purpose for most cards.” Plate 250 initially had shown a route card board of 5" x 8" but in the 1937 AAR boxcar design, this was increased to 5-1/2" x 9" for the boards on boxcar doors.
     A 1948 Car Construction Committee report addressed the continuing irregularities of the routing card boards (as is visible in many photographs). The 1948 revisions to the Manual clarified size (5-1/2" x 9") and changed thickness from 1-1/4 inches to, “not less than 25/32 inch.” Wording of Interchange Rule 49(c) was also changed at this time. It is important to realize this was all “Recommended Practice” and was neither mandatory nor universally followed.
     In 1952 came further revisions for routing card boards to, “Card board to be located on house cars on one door on each side directly under placard board and as near the bottom of door as lifting handle will permit.” This followed a circular letter in 1950 to car owners pointing out difficulties utilizing card boards on steel house cars which were located higher than 72" above ground level. Only card boards were lowered. The Car Construction Committee attempted lowering the placard boards as well, but needed approval from the Bureau of Explosives, which would not be granted until 1953. This was also agreed by union carmen and others affected.
     In 1953 agreement was reached to lower placard boards also, to go into effect in 1954. Here is an extract from the 1954 edition of the AAR Manual:

“Recommended Practice, adopted in 1914; Revised, 1948, 1952 and 1953.
     “The space available for placards should not be less than 16 by 24 inches on each end and each side of car. Steel house cars should be provided with placard boards made of soft wood not less than 16 by 24 inches and not less than 25/32 inches thick. The vertical edges should be reinforced with metal protection pieces fastened to the board with not less than six bolts, three through each, and at least four of these bolts should be used to fasten the placard board to the door, car end, or placard bracket when used. The boards may be made of more than one piece, and should be tongued and grooved, shiplap, or notched edge. The distance from the floor line of car to the bottom of board should not exceed 2 ft. 6 in. for end board and 22 in. for side boards.  These will be located on each end and on one door on each side.
      “On steel house cars routing boards with space not less than 5-1/2 by 9 in. available for tacking cards will be placed on one door on each side to left of placard board with lower edge in line with lower edge of placard board and so as not to interfere with lifting handle of door.  On steel refrigerator cars, routing card boards shall be located on each side directly under placard boards.
     “On other than house cars, card board must be located on each side of car near bottom at left end, facing side of car, or on outer end of end sill.  On cars equipped with center sills only, card boards must be located near center of car attached to outer end of running board support, or attached to outer end of body bolster.”

     Illustration plates in the Manual were changed, as was wording of Interchange Rule 49(c). Again, remember that this was recommended practice, not mandatory.
     It is also worthwhile 
to note that receptacles for defect cards were required for all cars built new or rebuilt on or after January 1, 1937.  This was to separate defect cards from route cards, which had apparently caused some confusion. I will illustrate this with a repeat photo, PFE class R-40-14 built in 1941, shown before (see:

The route card board here is on the side sill tab located at the bolster. The small, slanted-top appliance directly above it on the car side is the defect card holder. This car does not have car fans.
     Thanks to Guy Wilber for the background on this. The main point here is that there was a recommended minimum size for placard boards and for route card boards; the former appears to have almost always been applied in that size, but the route card boards continued to vary somewhat in size and certainly in placement. And here again, as in so many modeling issues, a prototype photo is essential for accurate modeling of these details.
Tony Thompson

Saturday, December 17, 2011

Operations: car flow and car storage

An aspect of car flow on any layout is the handling of cars which are not on the layout at all times. This can of course arise from just having too many cars, but a good reason to rotate cars on and off the layout is to increase the apparent fleet size encountered in trains and on sidings. In addition, unusual cars like 65-foot mill gondolas or depressed-center flat cars usually need to spend more time off-layout than on.
     Given that many and maybe most cars in a fleet will cycle on and off the layout, a couple of practical issues arise. An important one is, when cars are off the layout, how are they handled and stored? I want to describe in this post an experiment I’m currently trying. But first, I will outline my existing system.
     I discovered some time back that the bottom sections of cartons for canned beverages (cut about two inches high for supermarket display), usually available in supermarket discards, are mostly a slip fit into a standard 10-ream paper box, which conveniently has a lid, too (and is readily available empty at any local copy shop). Since these can-carton bottoms are deeper than an HO freight car lying on its side, they can serve as storage trays, and I have stored “extra” cars in these trays for some years. Below is a photo of one such tray (foreground), inside a paper box, with lid behind, and on the second paper box in the background is a second tray (actually the top tray from the foreground box). I have used soft rags and facial tissue to provide a cushioned setting for the cars.

Usually five trays fit in a paper box, the bottom of each one resting on the top of the walls of the tray below, and around ten to a dozen cars fit per tray. This isn’t bad for storage or transportation, but falls down badly when it comes to quick and easy access. I’ve made up a location tally sheet or inventory, listed by box and tray number, for individual cars, but access remains slow even if I can go right to the correct tray for retrieval.
     My new experiment is to try using open shelves for at least the most-used cars. Here is a photo of a recent status of these shelves:

The intermediate shelves are quarter-inch plywood, with a strip of molding glued to the edge to prevent rolling outward. In this photo, some of the car sleeves and waybills are standing alongside the cars, but I’ve since moved this paperwork out of the shelves, because it tends to get in the way of car retrieval.
     This amount of shelving can only accommodate my most-needed cars. As can be seen, about a dozen cars is the practical limit for these 26-inch wide shelves, if 1:1 scale fingers are to be able to reach between cars to pick one up. This shelving is thus only the equivalent of about one paper box. The shelf spacing is about 5 inches, also mostly to accommodate 1:1 hands and fingers reaching in to place or retrieve cars.
     This system seems to work well, as long as I make sure that only the most active cars are here, and I will continue to experiment with this arrangement and see if it keeps meeting my needs.
Tony Thompson

Tuesday, December 13, 2011

Small modeling project: NYC drop-center flat

Some time back, Walthers had one of their 90-ton depressed-center flat cars on sale, lettered for New York Central. Since I remembered that this model of a cast-steel car was quite close to an actual NYC car, I sprang for it. Now modelers have always loved unusual cars like drop-center flats, but they were actually pretty rare on the prototype. In fact, the only one I owned before this was an ancient white-metal Devore car, still backshopped in need of serious upgrade. Probably this NYC car will represent the only FD (AAR code for depressed-center flats) in my fleet.
     But I was also aware that the Walthers car, with its roller-bearing trucks, needed a truck change to represent the steam-era NYC car. As it happens, Eastern Car Works catalogs a truck which is very close, their “Commonwealth High Capacity” 90-ton truck, stock number 9063. As it also happened, my local hobby shop had these in stock, so with that purchase, I was on the road to making this model closer to “correct.” These trucks can also be purchased readily on line.
     This is the stock model, with the stock shiny wheelsets obviously uncorrected.

The car number and NYC Lot Number are all right, so trucks are the main concern. I know some modelers don’t like Eastern Car Works (ECW) trucks, because you have to glue the sideframes to the bolsters, but with generous amounts of styrene cement, the parts really weld together and make a solid truck. The parts have well-designed alignment bosses, so gluing should result in a square truck. I always glue one sideframe to the bolster and let the joint attain full strength overnight, then do the second sideframe the next day.
     The best-fitting Reboxx 0.088-inch wheelsets for the ECW trucks were the 1.010-inch axle length. Comparing the original Walthers truck with this ECW 90-ton truck shows how different they are, with the Walthers truck (left) having not only a longer wheelbase but also 36-inch wheels. Those characteristics may well be correct for the trucks on later-era drop-center flat cars, but they are not at all right for this 1941-built NYC flat car. The photo below compares the two model trucks.

There are those who view the “correct truck syndrome” in freight car modeling as overly picky, but in this case it should be evident that the truck on the right is far closer to the prototype I want to model. Here is a portion of a prototype photo of NYC 499056, from the Joe Collias collection, showing the trucks clearly:

The ECW truck can be seen to be very similar to this prototype.
     The only further work required was the use of spacers to raise the car body above the trucks, since the ECW and Walthers trucks do not have the same bolster height. I added the height by gluing squares of 0.020-inch styrene sheet, with corners cut off to make crude octagons, to the top of the truck bolster (thus making the “spacer” part of the truck). I then drilled out the center hole and enlarged it with a round jeweler’s file to match the original bolster hole. With that done, here is the truck:

     Then reattaching the trucks to the car body completes this project. Of course the car needs weathering and a load, but those are different subjects. The assembled model looked like this:

The appearance of the trucks can be compared to the prototype photo.
     There might be an alternative approach, to raise the coupler boxes in the car body, so that the body (like the prototype) sits as low on the trucks as possible. But the cast metal body of the Walthers model makes this challenging, and the coupler box is already practically flush with the top surface of the car. Another possible approach would be to use Kadee offset-shank couplers, but adjusting bolster height seemed to me the easiest procedure.
     I chose to post this project not because of its great importance but because it shows one kind of adjustment sometimes needed in changing trucks on freight cars. This can be a small but essential part of creating a more nearly correct model.
Tony Thompson

Monday, December 12, 2011

Route cards — 5

In my various route card postings, I began with some prototype photos and information, in the post entitled “Chalk marks and route cards,” available through this link: . I posted additional information on locations and size of route card boards in my third post, “Route cards-3,” which can be viewed here: . This new post is to show the modeling process of adding route card boards, a very small project to be sure, in response to a question received in email.
     As I mentioned in “Route cards-3,” route card boards were visibly not all the same size, but a size seen often, and shown in some drawings, was either 6 or 8 inches high and 12 inches long. To model them, I just cut one-foot slices of scale 1 x 6-inch or 1 x 8-inch styrene strip. These boards were often located on the side sill at the left bolster, but particularly on open-top cars, prototype photos are essential to determine actual locations.
     To choose an example, I’ll use the prototypes in Richard Hendrickson’s by-now-classic articles on modeling various mill gondolas by kitbashing the Athearn gondola. This was a two-part series in Prototype Modeler, in the September-October and November-December issues of 1982, pages 31 and 12, respectively. In the articles it can be noted that route card board locations are all over the various cars. Likely some were even moved at shoppings of older cars in later years.
     I needed cars of this type at the time the articles came out, and followed some of the suggested prototypes in those articles to build my own models. Here is one of them, an Elgin, Joliet & Eastern car.

The car has been modified from the Athearn original per Richard’s article, including addition of Dreadnaught drop ends and wood floor, and of course cutting sides down to a straight side sill. I also added a styrene overlay to represent the rivets on the interior of the car sides, since Athearn provided none of them. A retainer and lever handbrake were also applied. But there was no route card board.
     For an intriguing look at the variation in route card board locations on gondolas of a single railroad, I recommend Dorin’s book on the EJ&E (Signature Press, 2009), Chapter 10. I made use of the EJ&E prototype photo in the second Hendrickson article. This particular board is 1 x 6-inch styrene, located as on many (though not all!) EJ&E gondolas. Here I am applying it:

After application, of course, the route card board will be painted black and then touched up with weathering colors, next time I weather a batch of cars, before adding a route card.
     This kind of detail addition is quick and very easy, and it enables you to have a location for those interesting route cards on your freight car models.
Tony Thompson

Friday, December 9, 2011

“Shake ’n’ Take” modeling project: Kahn’s reefer

This modeling project is one of the “Shake ’n’ Take” models from the Cocoa Beach meeting that’s called “Prototype Rails,” held every January in Cocoa Beach, Florida under the direction of Mike Brock. The  idea came from Greg Martin, with assists from Jim Singer of 5th Avenue Car Shops, Richard Hendrickson, and Bob Walker and Dennis Storzek of Accurail. The Accurail refrigerator car kits which were the basis for this particular project, presented at the 2007 meeting, were decorated and donated by 5th Avenue Car Shops.
     The prototype of the model refrigerator car is a former Fruit Growers Express (FGE) reefer, leased by E. Kahn’s Sons Co. of Cincinnati, Ohio, a meat packer, from National Car Company, a Fruit Growers subsidiary. To approximate one of these cars, the Accurail reefer kit was the starting point. That kit has a USRA fishbelly underframe, whereas the FGE car had a straight center sill (and several other differences in underframe construction). Correcting the stock kit underframe was one part of the project.
     A second part was upgrading details, primarily removing cast-on grab irons and replacing with wire. This is normal detailing work and went easily. Lastly, end and roof details are modified. The entire directions for the project were provided in PDF form by Greg Martin, and with his permission are posted on Google Docs. Here is a link:
     I might mention that this handout states the paint scheme on this specific 5th Ave. Car Shops kit was introduced by Kahn’s in 1953. But there is a fine photo of a Kahn’s car with almost the exact same paint scheme, a car painted in November 1951, on page 197 of the book Billboard Refrigerator Cars (Richard H. Hendrickson and Edward S. Kaminski, Signature Press, 2008). So the scheme is earlier than stated by Greg, and poses no problem for my modeling year, 1953.
     I decided to minimize the underframe modifications, since they are not very visible on a finished model. I did cut down the kit center sill to a straight configuration, and added a cover strip of 0.015 x 0.125-inch styrene. Changing the number of cross-bearers from five to four did not seem essential to me. Greg points out that the Accurail-provided locations of the brake gear are not exactly right for FGE cars, but they are in fact pretty close, so I chose to ignore that discrepancy also. Prior to adding brake rigging, here is my underframe appearance:

Let me emphasize that I am not suggesting my approach is “better” than the more complete and more accurate modeling outlined by Greg. I just applied my personal standards to this particular car and its intended use.
     Carrying on to the body modifications, I used a chisel blade in my Exacto knife to remove molded-on side and end grab irons, and roof corner grabs. Using a scriber, I then made sure the grooves simulating wood siding boards were continuous through the places the molded grab irons had been. I then drilled holes to add wire grabs to all these locations. I cut away the poling pockets at car corners and filed the bottom of the car ends smooth, then added styrene channel (Evergreen 2 mm) at the bottom of the end to represent the end sill on these cars. With some grab irons added, at this point the body looked like this:

The end-sill grab irons were approximated with styrene rod. After these additions, the side grab irons were touched up with Floquil Reefer Yellow.
     Upon completion of the underframe detailing and painting Grimy Black, I installed Kadee #58 couplers and the Accurail trucks with Reboxx wheelsets, which are my current standard couplers and wheels. Then turning over the underframe, I glued on two large steel nuts (5/8-inch size) using airplane canopy cement (I use an older Wilhold product, R/C-56, but its equivalents are readily available at any hobby shop selling plastic airplane models). Here’s the underframe at that point:

You can just see a bit of the brake rodding underneath the car.
     On the car body, I added the “Modeler’s Choice” part 602 wood running board which was supplied in the “Shake ’n’ Take” kit, again using canopy cement. I added supports for the brake platform with styrene strip, scale 1 x 2-inch size, and 0.019-inch brass wire as the brake staff, and applied the kit brake wheel. Finally, short pieces of styrene strip, 0.020 x 0.060 inches, were added at the side sill corners (where poling pockets were removed), to represent the side sill extensions under the end sill channels. All these added parts were touched up with Floquil Boxcar Red, except the side sill, which I painted black.
     Here is a photo of the A end of the car at this point, prior to weathering.

The extended side sill is just visible under the end-sill channel. There is a prototype photo of this same area, included in the directions linked above to Google Docs, which clearly shows the prototype end sill arrangement.
     I added a reweigh date and a repacking stencil to my car prior to weathering, since I want to present the car as recently painted, rather than with the usual paint patches on older paint. I used both reweigh dates and repack data from the Speedwitch Fruit Growers decal set D119.1, which contains a variety of FGE (or National Car) reweigh locations like Plattesmouth (PLT), Hillyard (HILL), Alexandria (AX), Indiana Harbor (IH), and Jacksonville (JAX). If you’re not familiar with reweigh dates and all that they imply, you might like to read my article on the topic, in a corrected version from what appeared in Railroad Model Craftsman, available at: .
     My weathering method employs acrylic washes, as summarized in the joint clinic handout by Richard Hendrickson and me (if interested, it can be downloaded at: ). Here is a photo of the lightly-weathered car, on the house track in my layout’s town of Ballard, being picked up as an empty after delivering a load of packing house products to my wholesale grocer, Peerless Foods.

     That completes a pleasant and interesting model project. Thanks again to Greg Martin and his “Shake ’n’ Take” series at the Cocoa Beach meetings. As you can see, I have already put the car to work on the layout!
Tony Thompson

Thursday, December 8, 2011

One year and counting

Today is the anniversary of my beginning this blog. I started without any great confidence that I would have much to say for very long, and certainly without any sense that very many other people would be interested. In fact, if I’d made a prediction that day last year, I would have guessed that interest from readers would be rather slight, and that either that low response, or exhaustion of my own motivation, would have wound down the blog in a matter of months.
     To my considerable surprise, there has been a lot of interest. As of today, there have been well over 51,000 page views of this blog since December 8, 2010 (excluding my own), and many posts have attracted multiple comments. And I have received a number of emails privately, from folks who I suppose didn’t think their question or comment needed to be public as a blog comment. And some of those private communications have stimulated further posts to expand upon or explore the subjects raised. These extensions have been an interesting experience for me, and hopefully informative for readers.
     The total number of posts so far is 126, to me a surprising number, and though I thought these posts would be mostly brief, a few have been pretty long. I hadn’t realized I would use as many photos and other graphics as I have, and didn’t really expect to touch on so wide a range of topics. So this has been a more complex and gratifying experience than I imagined. I feel like I’m still learning how to use this medium effectively, which adds to the interest. Summing up, I have to say I’m more than pleased with it so far, and certainly still enjoying it.
     If you’re a regular reader, thanks for your participation. Both for you and for those just occasionally visiting, I hope I can continue to provide material of interest.
Tony Thompson

Monday, December 5, 2011

My column in MRH

The current issue of MRH (Model Railroad Hobbyist) magazine contains a column by me, for those who might be interested. The magazine is entirely electronic, no paper copies unless you print one out for yourself, and can be downloaded as a PDF (portable document format) to your computer. The magazine is free to anyone. To obtain a copy, go to .
     The column is part of a series called “Getting Real,” intended as views of prototype modeling opportunities and challenges from a variety of prototype-oriented modelers. To date these include Joe Fugate (editor of MRH), Mike Rose, Jack Burgess, Marty McGuirk, and me. Each of us is writing a column in rotation, so we submit one every five months.
     My piece this month is entitled “Choosing a Model Freight Car Fleet” (abbreviated in the MRH Table of Contents as “Freight Car Fleet”), and essentially follows a couple of my blog posts on the subject, though not repeating any of them. Focus is on both the Gilbert-Nelson idea of foreign-car presence on any railroad, and on the appropriate percentage of home-road cars. The former topic was first discussed in this blog at the post entitled “Choosing a model car fleet-2,” available at: , and including the bar graph of railroad fleet sizes. The latter topic was discussed first in my post called “Modeling freight traffic: Coast Line, 1953-Part 7,” which can be seen at: , a post which described finding an SP home-road percentage in a conductor’s time book.
     I did include in the column the bar graph of SP car types in 1950, a graph included in each of my volumes on SP freight cars. Beyond that, the column included an SP yard photo, just for flavor, a photo which was also included as a double-page spread in my Volume 5 on Southern Pacific Freight Cars (Signature Press, 2008), pages 104 and 105. Here is a smaller version to show the overall impression:

     I feel that Joe Fugate is doing something worthwhile with this series on prototype-oriented modeling, and hope I can keep up with the distinguished company of my fellow columnists.
Tony Thompson

Thursday, December 1, 2011

Modeling the Glasshopper II

In the mid-1980s, when I lived in Pittsburgh, I participated in a project of the Pittsburgh Model Railroad Club (as it was then named), to build a small demonstration layout for the Lord Corporation of Erie, PA. Lord produced a number of railroad-related products, from rubberized grade crossings to a wide variety of coatings and linings for freight cars. They wanted a small HO scale layout which would feature their products, for publicity purposes.
     The club set to work on this project, and produced a nice portable layout which I understand Lord did use, taking it to various conventions and so forth. One part of the project I worked on was to produce a model of the fiberglass covered hopper called the “Glasshopper.” Lord not only supplied an adhesive used in the car’s construction but also the polyurethane coating used on the outside. So what the heck was a Glasshopper?
     The Glasshopper story had started back in 1973, when a severe car shortage caused James Springrose, then Vice President-Transportation for Cargill, to think about ways to increase payloads and thus need fewer cars. From then until 1978 Cargill worked on its own toward development of a lighter car body for covered hoppers, with fiberglass uppermost on the list of materials.  
     In 1978 a joint venture between Cargill, Southern Pacific, and ACF 
Industries was created to build a prototype car, named the “Glasshopper,” 
based on Cargill’s ideas. The novelty in construction was that long fiberglass filaments were wound onto a large mandrel to make a one-piece car body of otherwise conventional polyester-fiberglass material. This is in contrast to most of the familiar fiberglass products one sees, which use chopped (short) glass fibers in a polyester matrix. The goal of the Glasshopper project was a tare (light) weight of 52,000 pounds, well below that of other 100-ton covered hoppers. 
     The production process was dramatically different from other freight cars in that innocuous-sounding phrase, “one-piece car body.” The mandrel for the filament-winding process was the size of the complete car body. Here is an ACF photo of the process, which was carried out at ACF. Fiberglass pressure vessels can be fabricated this way too.

Once the basic body was completed, final assembly, including steel side sills and bolsters (the Center Flow design has no center sill), was accomplished at ACF. There was a detailed story about the first car in Railway Age for 30 March 1981 (pages 56 and 57), when it was put into service. 
     The car was considered very successful, holding up well in service with a variety of loadssome of them corrosive. An additional unplanned test occurred in August 1982, when the car was involved in a derailment which included several other cars. As a spokesman for the joint venture put it, “the impact created by the derailment was such that it broke the Glasshopper’s coupler shank and caused the car to skid several hundred feet, leave the track, and slide down an embankment on its side. The car was righted, re-trucked, and returned to ACF’s Technical Center in St. Charles, Mo. for evaluation and minor repairs. These repairs primarily involved steel components such as ladders and outlets. The FRP (fiberglass-reinforced polyester) body was virtually unharmed.” (from Railway Age, January 1984, page 28)
     Here is a photo of the car being retrieved from that derailment (photo courtesy Lord Corporation); note the upper left logo of Cargill, SP and ACF.

The Glasshopper was part of an imagined “new age” of freight car design, judging from the slightly breathless article by Railway Age senior editor Gus Welty, entitled “The new breed of freight cars: A material difference.”  (Railway Age, May 25, 1981, pages 26-29)
     The experience with the Glasshopper led to modified specifications for Glasshopper II.  The original car was “somewhat conservative” in design, according to ACF, and still lighter weight was targeted.  In addition, a cleaner interior design without bulkheads was considered feasible. The new car was announced as entering service in January 1984, although its “built date” was April 1983 (testing was conducted in the interim). Like the original Glasshopper, it superficially  looked much like a conventional ACF Center Flow covered hopper, although its 15' 6" height put it into the “Plate C” extra-height category, and was again a one-piece filament-wound FRP body.
     Here is a three-quarter view of the Glasshopper II (courtesy Lord Corporation), with the same paint scheme of colored stripes as the original car, but numbered RNDX 166 instead of RNDX 163. Some of the bolt heads near the ends of the car sides reflect the assembly process. You can click on this image to enlarge it and see some of the end sheet configuration.

The Glasshopper experiments seem to have worked all right technically, and showed good durability and corrosion resistance. But they were more expensive to construct and harder to repair than conventional cars, with only about a 6 percent advantage in total weight. As far as I know, they were not duplicated. You can read more in a couple of places: D.C. Ruhmann’s paper in the journal Composite Structures, Vol. 27, 1994, pages 207-213; and in the book Reinforced Plastics Handbook, 3rd edition, D.V. Rosato, D.V. Rosato, and J. Murphy, Elsevier, 2004, pages 530-535.
     Modeling either of the Glasshoppers presents some challenges if the car is to be duplicated accurately. The car body had numerous, though minor, differences from the conventional ACF body, and the stripes and lettering of the paint scheme are not available in any commercial decal or dry transfer. I decided to minimize the modeling problems and essentially convert the Athearn ACF car to an approximate representation of an HO Glasshopper II.  Once this decision is made, most effort focuses on the car ends, since they are the most distinctive major difference. The Giasshopper II was also a 3-bay hopper, unlike the original Glasshopper and the Athearn model, which are 4-bay cars.
     As a starting point, I chose the Athearn “Co-op Fertilizer” car, kit No. 1909, because it is white. Ordinarily I prefer to work on an undecorated body, but since the final base color of the Glasshopper is white, some of the paint coverage difficulties of white paint over black plastic can be avoided by using Athearn’s paint as a base coat Unfortunately, little of the Athearn lettering can be salvaged. I used only the “Caution” notice at the center of the side sill and the dimensional data (these were later covered with tape when overspraying the completed model with white).
     The underframe modification is simple. The end hoppers on the Athearn car are in very nearly the correct location, so the modification is to replace the two center hoppers with a single one. The Athearn model has flat (horizontal) “floor” segments between each hopper. I sawed through the center of each of these, then used styrene cement to attach the two frame ends and one of the hoppers to a piece of 0.040 styrene sheet, about 30 scale feet long, for stiffness. The Athearn weight attaches to pins on the underframe, and can be used as a spacer to maintain the correct final length of the modified frame. The outlet gates on the Glasshopper were horizontal sliders, more like the ones on the Athearn Pullman-Standard (ribbed side) car than the ACF model, but I chose to ignore this detail. If you choose the model the original Glasshopper, you need make no underframe changes (again, except outlet gates).
     The ends, as mentioned, are distinctive, and were essentially the same on both Glasshoppers. Here is how I modified the Athearn slope sheet/end part to resemble the Glasshopper end. The Glasshopper had a smooth slope sheet except for a horizontal angle, and the end below the slope sheet had deep, formed ribs running vertically. The large diagonal stiffeners on the Athearn end accordingly were removed. There is also a small horizontal rib on the Athearn slope sheet; this was retained, and a piece of 0.020-inch styrene, 6 scale inches wide and as long as the end width, was glued on edge along that rib to represent the Glasshopper’s horizontal angle. 
     The photo below shows the stock Athearn end (lettered for Athearn’s version of the Glasshopper I, discussed below). You can see the large diagonal stiffeners which had to be removed.

     The Athearn  lower end sheet, the vertical part, was turned into the five ribs of the Glasshopper by dividing the end horizontally into eleven parts, marking parallel vertical lines, and razor-sawing along each of the lines up to the slope sheet. Then every other sawed strip was removed, starting at either edge, so that five ribs remained. The center one lies across the hole in the original end sheet, and to make it full length, one of the removed strips was fitted. Now a piece of 0.020-inch styrene was glued behind the ribs, so that when the slope sheet/end piece is in place, the face of the ribs will be flush with the edge of the cut-out in the car side, as on the prototype. At this point, I installed (but didn’t glue) the slope sheet/end part into the body. The Glasshopper end also has lighter diagonal angles flush with the plane of the end ladders. I modeled these with 1/16-inch Plastruct angles, cut to fit, and added small gussets from 0.010-inch styrene sheet.
     Here is an angled end view of the completed model, to illustrate the work on the ends, though it would have been better to photograph the modified end piece before inserting it into the car body.

     The car side can be made to resemble the Glasshopper II by removing the narrow rib of the Athearn car side, about two-thirds of the way up the side, and by filing the “overhang” at the top of the car side so it is vertical rather than overhanging. If you wish to model the original Glasshopper, the sides curved smoothly onto the top, with about a 2-foot (scale) radius; the Athearn body has enough material to permit filing to this contour. 
     Lastly, the “trough” hatches were added. These hatches do not duplicate the fiberglass hatches on the Glasshopper, which had ribs running across the hatch rather than the lengthwise ribs on the Athearn hatches, but the general appearance was acceptable to me. 
     The model car resulting from this work is only an approximation to the Glasshopper, not only because there are several detail differences remaining, but also because the Athearn car is not as tall as the Glasshopper. To remedy the height would be a larger project than I was interested in.
     Turning now to finishing, I assembled the underframe to the body, carefully locating the slope sheet/ends with the underframe tabs, and used modeler’s putty to fill the gap where the top of the slope sheet meets the body. This would also be the time to fill any other gaps or gouges from construction work. I also added a Walthers white metal Ajax brake wheel. Several light coats of Floquil Reefer White were sprayed on until the color was uniform. This was given several days to dry thoroughly before starting work on the stripes. 
     Using the prototype photos as a guide, drafting tape (not masking tape) was cut into strips 3 scale inches wide to define the gaps between the stripes. These were then applied with 9 inch separations for the stripe width. These were pressed down firmly, along with tape to cover the rest of the body. Additional pieces of tape were prepared with low stickiness or “tack” by applying and removing them several times to the sheet of glass I use for cutting. These covered the stripes not being painted at any one time.
     The stripe colors require some mixing. Choosing the exact color is difficult because the photos of the prototype Glasshoppers differ considerably in color rendition with the lighting conditions. Accordingly, I felt that an exact match was neither necessary nor very practical. The upper stripe was mixed from Boxcar Red with a few drops of Roof Brown; the second stripe was painted with straight Pullman Green; the third stripe was mixed from Reefer Orange with several drops of Black; and the bottom stripe was mixed from Foundation with several drops each of Roof Brown and Black. 
     I painted all of these Floquil colors with light spray coats, in one painting session, by allowing about 15 minutes between colors for drying (while I was mixing the next color) and occasional using of a hair dryer to speed the process. I had no trouble with covering the prior stripes with the low-tack tape, but if it sounds risky to you, you could spread the process over two or more sessions.
     The “Plate C” label and the small lettering on sides and ends was taken from a Walthers modern covered hopper decal set, No. D561. Some of the capacity data was pieced together, since I decided to reproduce all the correct numbers in that group. These are:

CAPY 200 000 LO
LD LMT 208 800
LT WT 54 200 NEW 4-83

On the side sill under this lettering is the legend, “1 1/2 IN HF COMP SHOE” (this can be pieced from the Walthers set; it refers to the 1.5-inch-high friction composition brake shoes). This also appears at the other end of the car, beyond the “lube plates.” My lube plates were taken from a Herald King set remnant I had on hand; they are available from several decal makers.
     The remaining lettering comprises the reporting marks and number, the Cargill-SP-ACF logo, and the large Glasshopper II name. Since no commercial products are close to these items, all were made as dry transfers. Artwork was made for these by tracing from an enlargement of a prototype photograph, then having the transfers made commercially by the 3M “Image-N-Transfer” process (this process was later sold to Letraset; I understand there are still graphic arts suppliers who do custom transfers this way). I have the lettering artwork available and would be happy to make it available to anyone who wants to make decals or dry transfers.
     After applying these transfers, the car was oversprayed with Testor’s Dullcote. The prototype is actually rather glossy, as a polyurethane coating was applied to the exterior, but HO scale models with high gloss simply do not look realistic. I finished the car with a bit of Grimy Black on trucks, since the car is modeled in its nearly-new condition. Here is how it looked (the photo is a little washed out):  

Unfortunately, I no longer have this model. It was part of the Lord layout project, and went with the layout on its travels. I’ve never learned what was the final disposition of that display layout, or of its structures and rolling stock. 
     Athearn has done a commercial lettering version of the original Glasshopper by simply putting the lettering on their standard ACF Center-Flow car. The ends are unmodified, but the appearance in a side view is good, and the car does have the correct four bays of the first Glasshopper. Here is a side view:

The stripe colors here seem deeper than what I modeled, and deeper than the prototype photos I have, but do represent the general appearance well.
     Although this prototype is well beyond my own modeling era, it was an interesting challenge to research and build, and after all, this car does have a Southern Pacific connection!
Tony Thompson

Sunday, November 27, 2011

Freight fleet rosters

I’ve discussed in several prior posts the approach I have taken and continue to take to developing a freight car roster. In particular, there are two parts to the problem: the mix of foreign cars, and the proportion of home-road cars. I follow the Gilbert-Nelson idea for my fleet of foreign cars (see my post at: and have settled on about one-third of all freight cars to be home-road cars, based on conductor’s time books for the Coast Line (using data presented, analyzed and described previously at: ).
     But beyond the railroads which are chosen, what cars and car types should that fleet of foreign cars contain? In this post, I discuss my approach to analyzing my various groups of foreign cars.
     Gilbert-Nelson tells me what proportions of foreign cars to have, but does not tell me which cars. My choices, then, are based on several factors. First, numerically important car groups of various foreign railroads. Just as Gilbert-Nelson suggests that the total number of cars of a foreign road should be in proportion to their proportion in the total national car fleet, it seems reasonable to assume that the more numerous groups of cars would similarly be most visible. Second, available models: in most cases, I am not sufficiently fascinated by foreign roads to want to scratchbuild freight cars for them, though kitbashing is within the bounds of feasible work, to me.
     A third factor is car types. Gilbert-Nelson, as often observed, ought only to work for free-running car types, meaning box cars, gondolas and flat cars, provided of course that the cars have no special equipment. Other car types have to be selected based on what is known from photographs and time books, along with expected industrial needs on my particular layout, leavened with some sense of what is a rarity and what is merely unusual.
     For example, I have a photo of San Luis Obispo yard in the 1950s which includes a MKT stock car; similarly, there is a Morris Abowitz photo at Los Angeles in the late 1950s of an empty Reading hopper car. I conclude that both are genuine rarities, and I have decided to model neither car.
     On the other hand, despite the absence of specific photographic evidence, I know from interviews and other documents that domestic coal (widely used on the SP in the 1950s for everything from section houses and cabooses, to depots, foremen’s houses, and sandhouse sand driers) came largely from eastern Utah and western Colorado. The mines were served by D&RGW, UP, and Utah Coal Route (UCR), and most of this traffic at this time was carried in drop-bottom (GS) gondolas. I am developing a number of such cars to handle my layout coal shipments.
     (I made this point about coal traffic and GS gondolas in my post about my freight car fleet of gondolas; see the post at: That post also contains a photo of a UCR gondola.)
     My basic roster analysis is to break down my fleet by railroad, and look at what I have (and plan to have) for each individual road to see if it makes sense. I will use D&RGW as an example. Though a Western road, it had a tight relationship with Western Pacific and Missouri Pacific for long-distance traffic, and thus minimal interchange with SP. But of course its cars still carried traffic into SP territory, and California photos do confirm the presence of D&RGW box and stock cars, as well as coal cars, in the early 1950s.
     My roster, existing and planned, for D&RGW looks like this. I have a Westerfield single-sheathed box car, representing the 1916-built cars which were a mainstay of the Rio Grande fleet, and a Sunshine kit for the 1941 steel box cars, which may be enough for that car type. My stock car is an illustration of the pitfalls of trying to achieve accurate models. Combining reading about Rio Grande stock cars, and looking at a variety of photos, I concluded that I could model one of the cars D&RGW modernized with steel ends and roofs.
     Here’s what I did. The sides of the modernized Rio Grande cars look a lot like the Athearn stock car, but as is well known, the Athearn stock car’s diagonal-panel roof has a reversed panel arrangement, and in any case is too new for the D&RGW cars, as are the Improved Dreadnaught ends of the Athearn model. I simply cut the sides from an Athearn stock car and from an Athearn box car, and put the stock car sides into the box car ends and roof. Why is this wrong? What I didn’t realize from the photographs I had examined is that the Rio Grande cars had corrugated ends, not Dreadnaught ones like the Athearn car. Sigh. Here’s a photo of my model:

I had to carve off the Athearn molded-on lettering boards, and decided to model a single broken board just left of the door. (Though I’m not sure a car inspector would have passed such a car.) The car continues in service on my layout, but is definitely a stand-in.
     I have three cars for Utah coal, a pair of MDC triple hoppers (not quite correct but adequate stand-ins) and a 47-foot GS gondola, a brass W&R model. The old Ulrich D&RGW GS gondola I used to employ as a stand-in is only 40 feet long and thus not at all correct for the D&RGW cars, and is in my “sale” pile.
     This is a comparison of the 40-foot and 47-foot cars. The proportions are quite different.

I’ve added an InterMountain early covered hopper and a Proto2000 53-foot flat car to round out my D&RGW fleet. So to me it looks like finishing that Sunshine boxcar kit will complete an adequate Rio Grande car fleet. I don’t have anything really surplus (except that Ulrich car) and don’t need any additions.
     This kind of analysis is slowly being extended to other railroads, in each case looking to see whether I have too few or too many cars. The ultimate goal is a fleet of foreign freight cars which is at the same time representative of each foreign railroad, of an appropriate size, and of appropriate car types.
Tony Thompson