Sunday, February 5, 2017

Temiscouata #1 - A Project?

It will be no surprise for people following this blog that TMC #1 is going to be my first 1:64 locomotive. Many factors lie behind this decision, but mainly, it's the easiest one to build with my actual skills and available components.

To be honest, knowing where to start with such a crazy layout project isn't a piece of cake... A module? A piece of track? A freight car? A structure? or a locomotive? Well... I don't know but I'm looking at several possible project to officially start working on something.

At least, I was recently able to get old CDS Lettering dry transfer sets and that's a big problem behind me! In that regard, I'm now hunting old Ridgehill Models Fowler boxcars suitable for Intercolonial and Grand Trunk. I guess they had the original wood roofs and 5' doors back then.

C-5-b diagram, almost virtually similar to a C-3-c (credit:

But a far bigger problem is finding suitable motive power. Old time 4-4-0s are on my list, but it probably won't happen until much later... and certainly, a 4-6-0 would be sweet in the long run. However, I've got to start somewhere, taking into account my skills, resources and what's available.

I've been studying the sole Mogul ever owned by Temiscouata and it seems the little locomotive could be kitbashed in S scale. After looking at CNR locomotive diagrams (C-3-c and C-5-b) and a few pictures including #1 builder picture (which can be seen in Donald R. McQueen's Canadian National Steam! Volume 3, p. C-7). It must be noted I found some discrepancies between McQueen's and others histories about #1. McQueen states it was purchased in 1916 from a NTR contractor (Cavicchi & Pagano Contractors #5) and sold circa 1920 (other sources say 1918) while others indicate it was purchased new from MLW in 1910 and sold in 1911 when found unsastisfactory.

From what I've seen, and this is only my opinion, I find it weird TMC sold the original #1 in 1910, left the number not used for almost 6 years while acquiring many other locomotives and reassigning available numbers. It makes more sense to have scrapped the old #1, replaced it with a local mogul used by a contractor in the area and swapped quickly with Canadian Government Railway #4537 (ten-wheeler) when deemed unfit for its purpose. However, other sources indicate the original #1 left Temiscouata in 1916 for Davie Shipbuilding in Lévis, QC.

But that said, we know NTR opened the line between Moncton & Estcourt in 1913. It was operated on a limited scale by ICR then. Could McQueen be right? Maybe. He got the builder pictures and MLW order story to back up his claim. Given the road was still under construction up to 1913, it's unlikely the contractor sold the engine before that date. He also indicates a lot of misunderstanding did exist about this particular MLW order. One thing is sure, I'm seriously starting to believe the roster published along with Mr. Lemon's article on Old Time Trains musn't be taken at face value. Thus, the second #1, replacing the old Dübs 4-4-0 in 1916 is starting to make a lot of sense...

Well, we could speculate a lot and be completely wrong. At this point, I won't make a ruckus about that, it seems quite trivial to the grand scheme.

But leaving aside the historical inaccuracies, I have enough technical data to start to figure out how to build this particular locomotive. I’ve discussed the project with David Clubine and Simon Dunkley, which have parts that could help to build up a decent drive. We discovered the old Rex 2-6-0 frame has the correct driver wheelbase (approx.. 12’-6”) and that River Raisin Models 0-6-0 51” drivers are compatible. Add to this a NWSL gearbox and good motor and we can say we have a solid base to start.

On the other hand, almost everything else on the Rex model is inadequate and can’t be directly reused. At least, some detail parts and bits can be of use. However, it is a fact virtually everything else will have to be scratchbuilt including the boiler, cab and tender. It shouldn't be that much of a problem at this point using brass, styrene and other materials. Anyway, my HO Grand Trunk 2-8-0 kitbash thought me sometimes it’s better that start from scratch than pour countless hours modifying a part.

The next step will be to measure and draw the drive to scale in CAD then start designing the superstructure to fit it according to dimensional data from old steam locomotive diragrams. Then, I'll bash the model in SketchUp to see how everything fits together before building. I wouldn't be surprised if some parts are 3D printed and maybe some involve lasercut down the road. That aside, it will probably built the good old way.

Meanwhile, the drive will have to be assembled, drivers quartered and motorization fine-tuned. At this point, I’m glad I’m modelling an early 1910s locomotive that didn’t saw too much modifications since its original construction.

Monday, January 30, 2017

Fine Tuning the Era

Determining your era when modelling in Ho scale is a matter of preference. Whatever you choose, there is a chance you'll find out what you want in the long run.

In S scale, you got to tailor your goals according to what is available or doable. In my case, finding freight cars isn't exactly a walk in the park, but nice stuff like Fowler boxcars, early steel twin hopper cars and good old wooden gondolas and truss rod boxcars and reefer can be found or scratchbuilt.

However, finding decals is another matter and as you can't do everything, I decided to see what was available. I was lucky to find a bunch of CDS Lettering dry transfers for classic Canadian roads of the day: Canadian Pacific, Canadian Northern, Grand Trunk/GTP and Intercolonial. These cars were often seen in the area back then. However, there is a catch. Most decals represent cars in service between 1906 and 1914. While I set my era in 1911, it means I should push the date forward to the eve of the Great War. It seems 1913 would be my safest bet. Does it change anything on the layout? Nothing at all but it's good to know that at that time, the National Transcontinental Railway was now serving Edmundston and interchanging with Temiscouata. Also, it wouldn't be out of place to see a motor car by the station at this date and newer railway designs such as the Fowler car can run on the layout without creating discrepancies.

For Temiscouata, the decals available are all for cars repainted in the mid 1930s. I think the reporting marks and data can be salvaged.

And as crazy as it sound, I love dry transfer. I used to hate them until I made a bunch of cars last summer for a friend's layout. When you know how to work with them, they are much pleasant than decals which always leaves some weird ghosting whatever you do. Dry tranfers also works better on real wood cars than waterslide decals.

As for locomotives, given S Scale is a rich man's world, I decided it was better to start of with one excellent model rather than waste my resources on many of them. I'm convinced a good old time 4-4-0 will do the job better than anything else. It will have to be built from scratch so I'm in no hurry. In that regard, I've examined many old Temiscouata locomotives and it is evident that in the 1910s, their boiler jacket sheeting had got a serious amount of abuse over the time. They were no longer the new kids on the block and I feel it would be neat to model boilet jacket full of scratch and dent. I've never seen that kind of weathering and think it would be very neat to represent an old lady doing her last waltz on the dance floor.

To be noted, I don't expect to work on S Scale projects in earnest before next fall. My schedule is loaded beyond control until then and I'll be happy if I can buy and build a freight car. I must admit I'm lured by the idea of building a Ridgehill Intercolonial Fowler boxcar.

Until then, I'll concentrate my efforts on Hedley-Junction and building my new 1:1 workshop. I reworked my design from last year to look more like a 1890s-1910s building. I'm seriously thinking about applying wainscotting and moldings to the second floor future train room to get a nice turn of the century railway look. The main floor will be the real workshop and I plan to have a comfortable space for modelling. I think most people would be facepalming if they knew I'm modelling and airbrushing on my computer desk which generally have no more clear space than a 1 square foot area!!!

Sunday, January 22, 2017

First Ballasting Test

Yesterday, while scenery was drying on the club layout, I made a quick ballasting test using my homemade ballast. This material is sampled and sifted from an abandoned limestone and clay pit located nearby.

To be blunt, I'm far to be satisfied by the results for many reason I'll take a few minutes to explain. This isn't what I'm looking for at all and it doesn't fit Temiscouata pictures.

Let's talk about the prototype. On all picture, there is virtually no roadbed. The track is at ground level except when topography required a fill, an embankment or cut. The very notion of ditch seem to have been highly ignored by Temiscouata Railway. The ballast isn't ballast as I suspected. It is nothing more than earth and dirt. Very fine dirt in fact and exactly the same that can be seen on adjacent lots. When you look at this picture of a construction train back then, you can clearly understand anything goes on.

Temiscouata Railway construction crew (credit: UMCE)

Let's take a look at the prototype ballast before analyzing my test. The first picture was shot at Rivière-du-Loup with the glamourous Madawaska business car (which I'd like to reproduce). The ballast is plainly dirt.

Rivière-du-Loup station (credit: Musée du Bas-Saint-Laurent)
On this other picture, taken at Cabano, the ballast as some bigger chunk in it, more akind to a dirty gravel. But still, is not what you would call real ballast.

Cabano station (credit: Musée du Bas-Saint-Laurent)
Now let's look what I did and why I think I failed...

First, my ballast is too coarse compared to the prototype. Given the ties are roughly 9 inches wide, it means some rocks are about 3 inches in diameter, which is clearly too big. It doesn't look like earth to me.

Another problem is the ballast "powder. that gets into the distressed wood grain, killing a good deal of the effect by making the cracks look whitish.

I tried to remedy to that problem by cleaning some ties and applying another light coat of alcohol and India ink. The improvement isn't that much great.

Finally, I put too much ballast, even if I tried to not bury the ties. The big problem is that ballast get caugh under the rails which kills the effect of "floating" rails so typical on Temiscouata.

Also, I'm not too sure about the ballast coloration. It is far too gray to my taste. I believe it should be a little bit browner, maybe slightly tan. I'm not sure. Should there be more contrast between the ties and ballast? I'm not sure either. Maybe the soil should be slightly darker to make the greyish and decaying ties pop out? Or maybe the ties should be a little bit more brown/tan. I say that because I've just looked at some old color slides from QRL&PCo - namely an old stub switch - and the ties are squared logs and seems untreated and they have a nice golden coloration... very similar to what Bernard Kempinski does. I think he was right!

At this point, I'm alsolutely open to any suggestion to improve the look. Meanwhile, I've ordered some supplies including rails, ties and spikes. As you can see, my glued rails popped at the first occasion, but it's because ofo that I was able to discover ballast gathered under the rail.

This brings another set of question. Should the ties and ballast be placed before laying rails? Or should the final distressing should be done AFTER ballasting and before laying track, making sure the ballast wouldn't fill the cracks? The answer will probably be got by laying a few other test tracks and trying each techniques. My biggest concern would be for turnout...

Also, I think I'll have to make the tie edges rounder to better represent half-sawn logs.

Saturday, January 21, 2017

Early Temiscouata Track Standards (Tentative)

After looking at many pictures and trying to determine various dimensions from known components here are a few rough observations that will probably need to be refined when more information will be available. One thing is sure, the trackwork was so full of discrepancies many jigs available on the market just fail to convey that look. I'll probably have make a few of mine.


-60 lbs
-28 feet long (to be confirmed by standard 1890s practices, TMC built the line with 32 feet flat cars)
-4-bolt joint bars


-Height: 7 inches
-Width: highly variable, from 8 to 10 inches
-Length: average 8 feet 6 inches(it varies wildly in fact)
-Spacing: 24" center/center


-Type: stub switch
-Switch machine: wooden with large target

3-way stub switch in Ste. Anne de Beaupré. TMC had longer central post (O. Lavallée)


-Mainly dirt with some bigger chunk
-Doesn't reach the top of the ties (about 3 to 4 inches of clearance under the rail)

Telegraph Poles:

-Single pole, no crossarm, single insulator on wood corbel

Friday, January 20, 2017

First Time Handlaying Track

Today I decided to handlaid a 18 inches long piece of track to see what I could do with available materials and decide on ties looks and colors. While I didn't have spikes nor track gauge, I decided to go forward. I what to use this piece of track as a scenery design study case on color, ballast, details and vegetation.

Having watched again TrainMasters TV's Roadshow track laying episode, I custom-built a tool to distress wood. I used an old ruined paintbrush, discarded the hairs and kept the socket.

Four dulled old X-Acto blades were glued together with 5-minute epoxy then glued onto the paintbrush. Before gluing, I made sure the socket was shaped in such a way the new blade assembly would fit tightly. This makes for a sturdy yet almost professional tool. As you will see in the following pictures, I think it yields great results without effort.

19th Century Trackwork

As much as people like to complain commercial track lack details, when you model early railways, you quickly find out there is far too much details including tie plates.

All the pictures in Connors should Temiscouata never really improved the track up there and things were quite spartan. What I wanted to achieve was similar to this picture shot in Connors.

Yesterday I browsed Bernard Kempinski Aquia Line layout blog to seek information and find how he did model his tracks. Unfortunately, as much as I have great admiration for his work, I thought something was missing and it's not until this morning I found out what it was. Every picture of old railways I found, particularly the ones when they build track, ties are nice dimensional treated lumber but rather squared logs with rounded corners. In the case of Bernard, he does a great job at coloring and he even goes as far has laser cutting extremely irregular ties for a incredible effect. However, I've not seen him rounding the edges. I believe this is seldom modelled and I decided to try my hand at it.

I think the best way to replicate an element is doing it the same way it was done back then. Using rough 3.5mm pieces of wood I once made on my table saw for a wooden bridge project, I started to round the edges with a large rasp. Unfortunately, the result was far too perfect and smooth compared to prototype pictures.

Then I decided to use my hobby knife to trim the edges into the right shape just like it would have been done back in the days (OK, they squared the log and I rounded a piece of lumber!). When I was satisfied with the slightly irregular shape, I use my new tool to distress the wood. The next step was to use a metal awl to split the wood according to what we usually see in real life.

My ties were then given a bath of India ink and 70% ink to give them a light greyish patina. Some other ties had already got a silvery finish by applying a custom mix of peroxyde and steel wool. That old trick age the wood in a matter of a few minutes and have the advantage to completely kill of the natural tan color of new wood. The recipe is simple, in a closed container filled with hydrogen peroxyde you place chunk of steel wool. A quick reaction with rust and disintegrate the steel, turning the liquid a deep oxyde red color. The next day, you only have to apply this liquid after filtering the mix. If you wish a darker color, just repeat the process. This way, I got wood to turn a deep black when I built my Erie Harlem station pontoon.

The rest was easy but quite tedious since I glued every single ties individually to the plywood plank. I know there are much more efficient methods, but that not to bad.

Rails were ripped from an old PECO code 70 piece of flextrack, pre-painted and glued on the ties. Detail West plastic fishplates were added at every 33 feet and I made sure the joints were staggered. A good coat of dark earth weathering powder was added to the rails.

Gluing track with beads of CA glue is the most unreliable way to lay rails and I will surely spike them later when I'll have the material on hand. However, I love the results and it's definitely something you can't get with commercial track.

The color is quite close to what real untreated wood kept in the soil would looks like. The added grain is also very realistic. When laying track on the layout, I will probably use exactly the same process except I will glue and sand the ties on the roadbed before rounding the edges and distressing the wood.

I think going back to HO and PECO will be hard!

In the end, I can affirm track laying is fun as long as you see it as building a structure or a car. No hurry and attention to detail should bring good results. And certainly doing so in S scale really helps to easily add details compared to HO. I would certainly not attempt building a 19th century layout in HO scale!

Thursday, January 19, 2017

Gathering Information

The process of gathering information on that fascinating prototype sped up thanks to many fine folks. I'm actually quite surprised by their enthusiasm will certainly acknowledge their efforts in the future.

As I'm writing, many new facts surfaced and started to bring a new light on Temiscouata Railway. In that regard, the work of Carl Riff is absolutely pivotal in providing first hand account and evidences.

While I won't publish parts of their collection by respect toward the ones that put hard work for years, I can assure you the line is much more fascinating than I ever thought. The small doubts that lingered in mind yesterday have all been swept away this morning we I saw the old pictures.

As a matter of fact, an extensive collection of good quality photographs of rolling stock at various epochs exist, showing the variation in paint scheme. Among that, many cabooses, roofless boxcars for pulpwood, strangely braced and probably custom-built boxcars and flat cars with their load of squared timbers. Add to that a generous amount of passenger equipment, including the very neat president's car "Madawaska".

Also, while not in great quantity, at least three old original 4-4-0 are now documented, including locomotives #1, #3, #4 and #5. The more modern 4-4-0 and 4-6-0 are also very well documented which will be useful in recreating these proud iron horses.

I've also got the photographic confirmation stub switches survived at least until the 1930s at Connors.

A friend also gave me recently a CNR reprint of old Intercolonial and Grand Trunk Railway plans of various trackside structures and elements. It includes stations, fences, turnouts, culvert, telegraph poles, signs, mile post, etc. All that is relevant to my era and the good thing is we know Temiscouata Railway took inspiration from Grand Trunk practices even if it was quite low scale.

Motive Power

Investigation on motive power continued over the last days. After figuring out Bachmann's On30 locomotives critical dimensions I can affirm none of them are close enough to be bashed. In fact, they are quite far from Temiscouata locomotives specs. Their 4-6-0 were extremely small!

That means all locomotives for the layout will have to be built from scratch. While it will cost a lot, it means the results will be closer to the prototype, which is a big incentive for me. The next step will be to find technical drawings for these locomotives.

The 4-6-0 being MLW products, a search in Exporail could yield results. We already have the builder photo and the CNR diagram, which is a good start.

The newer 4-4-0 used at Connors were Portland-built and originally owned by Quebec Central. I have good access to their archives so maybe I can find something there.

The older 4-4-0 were of various origin. #1 was a 1872 ex-Intercolonial Dübs-built engine. We have a picture of it and there is a good deal of information available online about CNR #40 (Museum Train) and its original appearance in 1872. While not from the same builder, general dimensions were quite standard between companies. Don McQueen's Canadian National Steam! books have a few pictures of Dübs product. It could be doable.

Other 4-4-0 built in 1887 were from CLC. I have little info on them and would like to know where in the world CLC archives and technical drawings ended up.

Finally, the last pair of 4-4-0 was built by Manchester, which was merged into ALCO. Yes, even when doing steam, it seems I always have a soft spot for MLW and ALCO. As of now, I have no idea where Manchester archives are stored. I'm well aware the ALCO legacy is well alive and some historical associations exist. Finding technical drawings would be great and probably a builder picture did exist.

Yes, it's the beginning of the search, but at least it's promising. Over the last 10 years I've been doing research on New France colonial monastic architecture and I can tell you looking for documents made 400 or 300 hundreds year ago isn't exactly a walk in the park. Their archives often burned once or twice in their long history, leaving large gaps in the collective memories. At this point, the Temiscouata has already shared a few secrets.

Track plans

Temiscouata and Quebec Central shared the same management board during my era. A recent visit at Groupe TRAQ in Charny shown QCR kept extensive records of their customers and stations. Every little upgrade and modification was drawn on paper or survived as blue print. I'm pretty sure Temiscouata did exactly the same thing, both routes had too much in common.

Now, the big question is finding where the Temiscouata Railway archives ended up. It's a given CNR inherited the papers, but right now, I have no idea where they are located. Maybe Exporail, maybe the Canadiam Museum of Science and Technology, maybe somewhere else. One thing is sure, before driving any spike on the layout, it would be truly useful to find such documents. Connors being a terminal, it's almost certain several plans do or did exist.

Wednesday, January 18, 2017

Updated Track Plan

I finally updated the original HO scale track plan after a lenghty process of scaling and adjusting each element. I generally use xTrackCAD to draw my plan, but I couldn't locate a Fast Track S scale Library, so I draw it by hand using Tim Warris PDF templates as a guide.

Redrawing the plan allowed me to find out new details I overlooked in the past. Among them are fences and each of them tells us a peculiar story related to its function.

The first fence is the ubiquitous one separating the right of way from the fields. I think the old picture showing a cow grazing on the siding at Connors is enough to make this point clear! While we, modern people disconnected from large animals, can hardly imagine how a cow could be a match against a steam locomotive, we ought to remember 19th century locomotives, particularly 4-4-0, were extremely lightweight by today standards and could be easily derailed at the most insignificant obstacle on the track. Forget your YouTube train videos plowing through trailer trucks as if they were paper toys. Back in the days, any large farm or wild animal posed a serious threat that could results in serious injuries, casulties and material loss.

This became quite clear in my mind when I recently read the book The Old Somerset Railroad: A Lifeline For Northern Mainers written by Walther Mcdougall. The author had the pleasure to meet several former employees of the Somerset that worked in the early 20th century under the old management. Many of them recalled tales of 4-4-0 badly derailed by cows taking advantage of neglected fences. The most fearful event occured when a train met a bear crossing the large Gulf Stream steel trestle up Bingham. The train crew knew perfectly they had no chance. The locomotive would hit the bear, derail and crash in a 125 feet deep ravine. Chances for survival were inexistant. Fortunately, when the train was about to strike the dangerous animal, the afraid bear simply forgot it's instinct and jumped off the bridge, meeting its deadly demise on a rock by the stream.

Less impressive and dramatic are other fences found near the engine facilities. Near the coaling shed, a snow fence protect the turntable pit from getting filled up with snow. To complete this defensive wall, a cedar hedge was planted between the coaling shed and the engine house. This is the first time I see a railroad using deliberately vegetation as a mean for snow fighting. It was a common knowledge back in the days in rural areas, which is implemented again by farmers since the 2000s. I think it would make a very neat detail on a layout, one seldom seen, even on diorama.

The last fence is a large and opaque one located south of the storage/MoW siding near the engine house. It separate the railway property from the public road. At first, I was surprised they cared to build such an elaborate fence in open country, but it makes sense since the most dangerous activities were perform in that area and it needed to be secure. Also, just on the other side of the road are employees houses which could explain with they bothered putting up a fence there.

These fences are neat little details that can be easily modelled, particularly in 1:64, and that gives a deeper sense of the place. Incorporating them on the layout add interest but also helps to visually frame the scene by creating hints at the railway property limits.