Wednesday, August 31, 2016

Thinking Out Loud - Part 9

I think it’s never a bad idea to revisit older concepts from time to time because new information is always unearthed in the most unpredictable ways. If you are like me, you have a dedicated folder on your computer hard drive with several subfolders dedicated to specific layout projects and prototypes. Over the year, most "newer" prototypes are rehashes and refinements of older concepts. A few recent examples are related to recently discussed ideas about a “future” layout of mine. Let's see how thing evolved over the last few weeks.


Last Friday, I met Jérôme to deliver the first batch of weathered boxcars to Erie’s Harlem Station. He showed me old CN Murray Bay subdivision timetables ranging from the 40s up to 1992. The information in it was very interesting because it puts light on the decade-long absorption of QRL&PCo into regular CNR operation from 1951 to 1959.

Another interesting aspect was that older timetable cared to identify private sidings by their customers’ name and car lenght. Strangely, only Dominion Textile never appears on timetables… my guess is their siding was considered to be not privately-owned. But the best part was about Montmorency Distillery. All timetables listing this industry mentioned two sidings for the company. One facing East and one facing West. According to CNR documentation, they were two distinct and unconnected sidings. I’m not the kind of guy to take information at face-value, but all other sidings listed in the timetables were strictly correct. The weird thing is that my copy of CN early 1980 movie over the subdivision clearly shows both Montmorency Distillery sidings to be connected: it’s a double-ended siding.

Unfortunately, later timetables dropped the concept of listing private sidings. So, the big question is if the sidings were merged together later in the late 60s or 70s or the listing of them as separate entities was an error. Surprisingly enough, Omer Lavallée’s 1959 track diagram of the line (otherwise very precise), don’t list any siding for Montmorency Distillery while they are attested by historic photographs and he travelled the road often. Well, that leaves me with something to search and only CNR engineering diagrams could put the issue at rest.

However, from a layout point of view, the two-ended siding is a blessing. When I designed the S scale Beaupré layout plan, I was concerned only eastbound trains could handle operation but not, we can clearly assumed the distillery was switched on both directions, making it much more interesting in the long run.



I used to think sound was a gimmick and now I find I’m no longer able to operate without it. I made a scale mockup of Lehigh Valley New Woodstock station during the week end to test the operation potential. A printed track diagram on a hollow core door, tracks and a few boxes were enough to get into action in a matter of minutes.

However, operation was DC and quickly, making precise moves at slow speed without sound lacked interest. It seems to me the sound is much more than an ambiance, it is an indication about what’s happening. Every small action is accompanied by a sound “consequence” that makes sense and bring a touch of reality. Remove the sound and there is no longer consequence to the performed act of switching. In the end, I found out New Woodstock was extremely boring but I know I can’t use this experiment to ditch such a generic and common station track plan. Because of the lack of sound and the fact the layout height was low I wasn’t feeling I was part of the action going on. To me, being immersed in a scene is the primordial goal in model railroading. When I was 4 years old, I used to watch my Bachmann trainset with my head put against the floor and with an eye closed to see the action from a realistic point of view.

But the lesson is learned: never underestimate how the way you interact with a model can dramatically change your perception. Next time I mock up a layout, I'll do it at a decent height and with my NCE Power Cab DCC. It won't take more time to set up and I'm guaranteed to get results closer to the condition I would operate a real layout.

More Dangerous Gimmicks

Continuing with the concept of immersion, I’ve always felt Ciment St-Laurent was a nice big industry but that something was desperately lacking when switching it. Nothing is more boring that shoving hoppers over an unloading pits. Sure, you can set a time limit per car before moving the next one, but it feels fake and a pure waste of time because it seems nothing happens.

There is only one way to bring interest and it’s to “physically” make something happen. Setting a car over the pit must trigger a consequence, an action that legitimates that you wait before resuming your move. It can be performed by sound or by real action (or both in an ideal world), but simply looking at your watch for a given amount of time isn't enough. At some point, it's so artificial you'll skip it out of boredom or simply because your brain don't see the point to wait when nothing really happens. In that respect, sound is the easiest and most convincing way. When a hopper is spotted, the sound of an unloading car is heard for a realistic amount of time. When the sound ends, you remove the fake load and spot the next hopper. This is the easiest way out there. Add the ongoing surrounding sound of the conveyor and you are in business!

The next option is more akin to a gimmick, but nevertheless could be interesting if correctly implemented. You use live load and the hopper is emptied for real into the pit. While attractive, this system is far to be fool proof. I’ve seen many videos online of realistic loading of hopper, but unloading them is far less practical. But that said, the old toyish Tyco hopper car, while not prototypically interesting, self-unload pretty well. The system being basically a kid toy is almost fool proof. It could definitely be adapted and improved for realistic coal unit train operations. Unfortunately, the system works better with cars in motion, which doesn’t fit the criteria of a car that must be spotted first.

Another older design for a self-unloading scale hopper was made by Ulrich back in the 50s and 60s. This time, the hopper realistically reproduce a common 3-bay car that can be used on many layouts. The unloading mechanism is hidden in the underframe and subtle enough to not be detracting. Unfortunately, information is scarce and I didn’t find anything about the efficiency of the system or car unloading time. That said, after studying Ulrich original instruction sheet, I firmly believe one could fit their metal underframe and mechanism under a modern and well-detailed plastic shell. In fact, having an all metal underframe would enhance the car weight which would be a good thing.

Such systems, while extremely attractive in concept, needs to be experimented. In my case, the unloading track is also used to sort out other cars (including hoppers) which mean the unloading device between the rails should be removable or not interfere with other actions. Another problem is that while the Ulrich system is compatible with older 3-bay hoppers, it isn’t with more modern or larger prototypes. It means I would have to reproduce that mechanism in different sizes and configurations to fit other specific cars. That could prove to be a foolish attempt plagued with frustration.

Finally, while self-unloading scale hoppers are a nice idea, they seem to be extremely fast to unload (at least, the Tyco version). To be honest, they seem to be faster than a real car and a realistic sound file could last (but some real-life hopper unloading time is quite fast according to videos on YouTube). So we are back to square one for this one. In the end, it seems a decent sound file and removable loads should do a better job with less efforts. That said, I’m still curious about the Ulrich car. If anybody has experience with this hopper, let me know.

Temiscouata As It Should Be?

If people could probe my mind in the last two months, they would see a labyrinth of thoughts, concepts and ideas. I think the “Thinking Out Loud” series gave a good instant picture about how I can connect diverting interests. Among all that mess, it seems my interest to model Temiscouata railway as it was in the early 20th century is winning over a later period rendition.

I went back and studied my motive power diagram compiled from the roster. What will follow is only speculation based on available information and cold hard facts. The 1948 Railroad Magazine article by Mike Runey was interesting, but mainly covered operations after the heydays.

When you look at the roster, it seems the new 4-6-0 started to be added to the roster circa 1910s, mainly to pull heavier freight trains. Funnily enough, this is exactly the moment when Temiscouata ceased to be a relatively isolated branchline and became somewhat a bridge line between various between New Brunswick and Quebec. The new National Transcontinental line and various other factors probably increased local traffic over Temiscouata, requiring more modern and powerful locomotives which incidently, were acquired second-hand from contractors building the NTR line. On the other hand, we could speculate NTR was a shortcut and prime competitor for long distance traffic between Quebec and New Brunswick. The Monk subdivision was built over an ingrate landscape but to exacting standards, making it a fast mainline served by state of the art locomotives. It was far to be the case with Temiscouata. However, we must take in account railways were still in their pre-WW1 expansion without any worthy competition.

It is interesting to note that even if 4-6-0 were in use in the 1910s, a large amount of 4-4-0 were kept in service. This fact can indicate the need for power was substantial and new locomotives weren’t bough to replace older engines but to provide additional steel horses to catch up with growing business. The diagram show that prior to the 1920s, about half the fleet were 4-4-0. Interestingly, the longest trains and steepest grades were all between Rivière-du-Loup and Edmunston because the line crossed the St. Lawrence and St. John rivers divide. On the other hand, the Connor Branch was a relatively flat water level road following the St. John River north shore. Knowing Temiscouata operated 4 daily trains (2 on Connors Branch ), we can imply a substantial amount of motive power was allocated to service Connors to pull freight and passenger trains.

That means we can surmise most old 4-4-0 were used over the Connors Branch while the 4-6-0 worked on the mainline to pull freight consist over the steep grade near Rivière-du-Loup. In fact, old pictures seem to support this hypothesis because most of the time the 4-4-0 are shot while working at Connors when 4-6-0 are often seen near Rivière-du-Loup or Edmunston. Sure, we know 4-6-0 ventured to Connors, but that information comes from the later period and I suspect, at least circa 1910, the very first 4-6-0 available worked on the mainline. This is a personal hypothesis and I could also be wrong.

Temiscouata #8 as built in 1909 by MLW (Al Paterson Collection, "Canadian National Steam! Volume 4)

Now, when piecing together these circumstantial observations, I come to the conclusion Connors was mainly served by 4-4-0 for during the first two decades of Temiscouata history, maybe later. The grades and the traffic wouldn’t have made a good use of 4-6-0 when the rest of the line was far more challenging. With the acquisition of more 4-6-0, declining traffic and aging 4-4-0, the American Standard became a rarity rather than a norm over Temiscouata.

Temiscouata #4 built in 1888 by Manchester (credit: Musée du Bas-Saint-Laurent)

I'm pleased to say I already have all the required parts to build a good fleet (let’s say about 3 locomotives: 2 regulars and one spare). The Bachmann modern 4-4-0 is a good starting point to model the most recent locomotives serving Connors while the old IHC Genoa is an excellent starting point to model older locomotives from the 1870. A good picture of locomotive #4 exists and it’s a good match. Sure the IHC locomotive is OO scale, but given I’m not modelling an early 1860s 4-4-0 but something newer and bigger, it should fit the bill better. I have also an Athearn/Roundhouse 2-6-0 with high drivers. This locomotive could be converted into a 4-4-0 by removing the front driver and replacing the pilot wheel with a truck. Even the superstructure could be altered significantly by using a Bachmann 4-4-0 shell.

Also, it is interesting to note that Connors Branch, back in the early days, saw more traffic and was served by four daily trains (2 freight and 2 passenger trains). This makes for slightly longer trains and complex operation.

From a rolling stock perspective, old photographs also show more Temiscouata lettered cars in the first years of operation than during the late 40s, showing that the declining railway didn’t invest in its own fleet after a certain point in history (on the other hand, ORER are a very fine tool to figure out the fleet). As I said in a previous spot, I think Temiscouata is interesting in every period, but I must admit the sweet spot is still in the early 1910s when the line was connected with the National Transcontinental and Canadian Pacific in Edmundston. Add to this that the as-built Connors track plan is well-known while the later version can only be implied from partial photographic evidence until real track diagram or timetable can give a reliable overall portrait. 

As Trevor Marshall suggested, Connors could be operated both as an early 1900s railway and as a dwindling late 40s line. Each structures could be built twice; in its original colors and in an altered and weathered version. Since the amount of structure is very limited, that wouldn’t be a crushing endeavour, but rather a nice way to beef up the modelling potential. As much as I’m fascinated with early 1900s Canadian railways, I can’t brush off my interest for more modern eras. Trevor’s suggestion cover both bases without too much problem except for glossing over the fact Connors track plan did evolve over the time.

Finally, the biggest decision is to base Connors on a definite track plan. I drafted two versions showing the evolution of tracks. It sure ain't 100% accurate, but it gives a good idea of what can be achieved. As you can see, after a while, the turnouts were significantly rationalized to optimize operation. Personally, I think the oldest version is interesting for a particular reason: most moves are done in front of the station. Remember when I talked about immersion. This is particularly crucial on a small layout where you want to maximize the impression of distance. Concentrating the switching moves in front of the build structures is a good way to immerse yourself in Connors instead of using a non descript main line in the middle of meadows. To me, this is an important aspect. We pour efforts in reproducing scale version of fancy structures and I believe they should play a more important role than background decoration, particularly when they are directly related to the railway.

So, 1894 or circa 1948? I think I've made my choice! And you? 

By the way, if you are interested to learn more about Temiscouata Raiway, it's region, Rivière-du-Loup and Intercolonial Railway, it's nice to know Musée du Bas-Saint-Laurent put together a bilingual virtual museum named "À fond de train" (At Full Throttle) well worth a visit. Lots of very old photos of early Canadian railways including the original Grand Trunk line to Rivière-du-Loup.


  1. Good.
    (No need to justify your reasoning, by the way. I just wondered.)

    1. Ahaha! Temiscouata in HO. I have enough spare turnouts, tracks, steam locomotives, rolling stock and decals to make it possible + non need to compress the scene too much. I perfectly know it won't be my last layout, so when I'll be ready to do something with QRL&PCo it will probably be in S. Anyway, if I ever want to redo the Temiscouata in S, it won't be too late. As it is, I could build Connors as a module and later add the staging and mainline when the new room will be available.

    2. Rumbled.
      You keep on down this path: no matter what scale, it will be rewarding.