Sunday, April 12, 2020

Weathering Track: The Last Step

When weathering track, most people will understand the need for having the right ballast and using realistic colors for ties and rails. Others will had oil spills and other such dirty deposit. However, a simple trick to improve the look is adding a coat of weathering powders on the track.

A light earth brown from AIM or other manufacturer over a camouflage brown paint base does wonders. Time consuming to apply, but a difference maker if you ask me.

Front track not weathered, rear track done. See the difference!

The pigment lighten and dust the rails, creating a realistic rust powder appearance. Also, the light shade makes details such as rail joiners, spikes and tie plates pop up, improving the overall appearance.

Front and rear track now weathered.

Finally, the powdery nature of the medium means a little bit will spill over ties and ballast, creating a realistic blend between the track and surroundings. It brings the track into the decor nicely.

Saturday, April 11, 2020

Dufferin-Montmorency Highway Overpass - Mark II

Ten years ago, I built the first version of Dufferin-Montmorency Highway overpass. At that time, the club layout was focussing on Limoilou yard and it was used to create both a sense of place and hide the hole in the wall.

It was also my first modelling blog ever written here, so you can say this structure was part of Hedley-Junction DNA. For the new diorama/switching layout, I also need a model of the exact same structure. However, this time I decided to learn from my mistake and make a better job out of it.

Back then, I basically never used any kind of putty, never really thought about sealing wood and MDF components before painting even if I knew it would help a lot. I wasn't very keen on sanding too before applying primer and between coats of paint. But that structure did look cool and was a favourite of many people how visited the layout. Often, we considered reusing it, if only it could have fit the scene and space available.
Original overpass: blobs of glue and damaged material.

Before build the new overpass, I had to determine where to put it, how it would relate to the tracks and where pillars would be. Easy to say, but as you will find out later, I made a little mistake and had  to take strong measures to correct it.

Original overpass: paint had a different texture depending of material.

First step was to assess the structure. Back in 2010, I made several compromises in terms of height and pillar spacing. I also made the parapets far too high while making a few errors how the crossbeam supported the stringers (concrete beams). I thought I should follow the prototype closer, but in the end, I reused the same dimensions as my first scratchbuild. The reason was simple, the prototype span is larger than my layout width! Also, having only one span would look silly and wouldn't hide the fact there is nothing on the other side. Having a forest of pillars really helps to create the illusion of a real visual barrier. In that respect, artistic considerations has the upper hand.

The pillars create a new and interesting vantage point to railfan the trains.
Another decision was to have the overpass complete on both side. On the first version, it was facing a wall and I didn't bother. However, this time, the overpass is floating in the middle of nowhere and needs to look good on both side. Also, even if there is a wall, I still would do both side, because I thing it looks better when you see all the lanes. Cutting a lane in the middle wouldn't be visually interesting and creates that "cropped" image appearance I'm not too fond of. Think of it in the same way as flat structure against a backdrop. You reach a threshold when it no longer looks believable and the same apply with overpass hiding the end of a layout. Too many great layouts are visually ruined by these compromises that should have simply be avoided.

Original overpass: the drain pipes made of sprues are a detail to replicate

While workmanship on the original overpass wasn't top notch, it was still a well-thought piece of modelling, so I decided to keep the recipe and only improve it where I could. The original overpass was made out of styrene and MDF. The decking and beams are concrete and thus MDF was my first choice material. Pillars have neat flutings on their corners, thus they would be easier to scratchbuild out of styrene. The parapets have very intricate 1970s random grooved pattern which was extremely fashionable back in the days in civil engineering circles. They can only be convincingly modelled with styrene. As for the crash barrier in the middle, MDF was my first choice too.

As many knows, MDF and styrene glue together extremely well with thin CA glue. The bond is permanent and almost impossible to break. The MDF will delaminate before you can break the glue bond. Also, the bond is instantaneous. You don't have even half a second to wiggle the part in the right spot. It must be perfectly dry aligned, then glued in place. It may be, as you can guess, both a blessing and a curse. Learn to work with it, and it will open a lot of possibilities. Interestingly enough, CA glue on MDF can be easily sanded down to a glass-like finish. This is of extremely great importance because when painting materials with different porosity, you get variation in shines that look horrible and only get worst with weathering.

All surfaces sealed with CA. When primed, the wood grain disappear.

So here's another tip for working with MDF. Seal your component with thin CA glue. Apply it over all the faces and when dry, with fine sandpaper, polish the surface. Not only it will reducing warping in the long run, but the surface is perfect for painting. And it works also on regular wood. Since I didn't have enough MDF for the concrete beams, I used leftovers from a 3/4" thick pine plank. The CA glue helped me to completely hide the wood grain.

The central crash barrier was quite a challenge. The particular profile on such concrete barrier is hard to replicate. Back in 2010, Louis-Marie did it on a table saw, but as he explained me, it was the perfect recipe to lose a few fingers in the process. It was time to find a safer way to do it. The new process is simpler, I used a 1/4" plank of MDF that was about 3 inches high to have enough material to grasp in a safe way. On each side of the plank, I cut notches about 1/4" deep and 1/16" wide. Once the profile was right, I then cut it from the plank at the correct height, leaving an inverted T-shaped base. The barrier was then liberally covered in CA glue and when dry, I used a file to flare the base and top of the wall. I repeated the process several times, sanding down material until satisfaction. Interestingly enough, CA pooled into the "T" crevice which smoothed the profile, making a visually compelling slope. On my prototype, the crash barrier plinth is covered in galvanized sheet metal and I really wanted that detail to be carried on the model.

As for the order of assembly, I started by gluing the concrete beams (stringers) first on the deck. When dry, I cut the deck and beams on the table saw to length, making sure to induce an angle for visual interest. Then, the crossbeams made of 1/4" MDF were added. They do wrap at each ends around the beams, creating a "C" profile or a kind of clamp. All surfaces were sealed and sanded twice at this point.

It was then time to make the pillars. They are basically styrene boxes. To create the notches on the corners, I simply glued another sheet of styrene. All angles were sanded and round to look like cast concrete. Then, using a square and sandpaper, I made sure each pillar was perfectly perpendicular. As I said, CA glue, MDF and styrene aren't forgiving. you must dry test your parts before gluing. When everything was exactly as wished, pillars were located on the crossbeam in their exact position and thin CA glue was applied. It soaked right into the MDF, creating a very strong bond. Full disclosure: I did glue two pillars in the wrong place... I thought I would break the model when I had to take them apart. Not an experiment to replicate.

Oups! The pillars are too close to the track!

The last step was to build the parapets. They were crafted on the workbench and made longer than the overpass. I prefer to cut them to length once glued on it. They are made of several layers of styrene sheet and stripes to match the random prototype pattern. Later, a styrene railing with grace them.

A man can barely walk by the train without getting hit by the locomotive.

With the overpass almost completed and ready for paint, it was time to test it on the layout... and the biggest of my mistakes became instantly apparent. Maybe by sheer naïveté, I thought having the track quite close to the pillars wouldn't be too bad. Unfortunately, it looked both unrealistic and visually unbalanced. Basically, about 4 days of works for nothing.

Serious alignement issue
However, while looking at the track, it was evident they weren't straight. It was a little bit weird since I was certain the alignment with the pillars was correct when I planned the layout. Using a long straight edge, I found out the misalignment was severe, at least 3/8" which in HO scale is almost 3 feet. In another place, I wouldn't have minded such an insignificant detail, but given the trains were almost scratching the pillars, I had to to something.

Gaining 3/8" makes a great difference

Thanks to Lance Mindheim, I glued the track with white glue. After wetting the track again with water, it became soft again and I was able to move the track to the correct alignment. It certainly wasn't a walk in the park, but it did work. After soldering back the feeders, I tested the overpass again. Now it looked right! Yes, I lost half a day correcting my mistake, but I know it would have been detrimental to keep it as it was. In the process, I had to scrap the complex road I had painstakingly built to fit the track geometry... so no, it wasn't an easy decision to make.

Final location with road

Nobody is immune from mistakes and addressing them when they appear is what makes you a good modeller. In fact, a good deal of being a better modeller implies simply to learn to not compromises with shoddy results and try to fix them. It's the only way to improve as rebuilding from scratch this overpass proved us.

Thursday, April 9, 2020

Modelling CN SW9 & SW1200 Switchers - Part 2

Work continues on the pair of SW.

The next step was to rework the cabs: I removed the side window details so I could plug later some Rapido's all-weather window. Some styrene stripes were used to narrow the window and to add new arm rests. Also, the molded on grabirons were replaced with phosphore-bronze wire. Miniature by Eric sturdy Sinclair antennas added and correct horns installed. New rain gutter/sunshade rail were crafted from modified Plastruct C-beam. Basically, I cut one wing to get a L-shape angle and sanded it down to size. I've done it on several other models with success and they provide a sturdy base to apply sunshade to.

At this point, I stopped working on the cabs. I knew I would have more work to do later, but not knowing how to do it, I moved to the hoods. Since I didn't have correct switcher bells, Athearn's were kept but modified. The stock parts look like a steam era bell, so I trimmed them to fit better a diesel locomotive. Once trimmed, I cleaned the flashes and puttied the casting holes.

On the hood itself, I started to remove door louvers that didn't fit the prototypes and the handrail stanchions. The stacks were cut and filed down. Some holes had to be plugged with styrene to take into account the bell location was different on 7007. New square holes were also drilled so I could fit extra Rapido CN spark arrestor from my SW1200RS kit that I won't use. Luckily enough, out of three locomotives, only one was equipped with spark arrestors on CFC. These are beautifull casted parts, but they will require some cleanup to remove the mold seam lines.

The next step was to modify the locomotive sills. Indeed, while the SW7 have steep angled metal sheet linking the staircases to the frame, CN SW9 and SW1200 have 45 degrees metal sheet there. My idea was to add some 1mm thick styrene there and, making a jig, to cut the shell and styrene in the right angle. It did work better than I thought, thus recreating a spotting feature on the prototype.

I figured out the molded overhang on Athearn sills didn't exist on the prototype, so I filed it down to get smooth-sided sills. Unfortunately, when I put on the cabs again, I found out they extended over the frame in a very unrealistic fashion.

Thus, I had to cement a new layer of 0.25mm styrene sheet over the sills and sand down the ends. I had to be extremely careful while cemented this with solvent cement because it could warp the styrene surface badly. The styrene sheets were slightly larger than the sill, providing enough material to sand it down flush with the running board. It is often easier to get better result filing down material when in place than trying to figure out before hand the exact size. It also minimize the use of putty while providing a seamless surface between the old shell and the new styrene.

From there, I started to add the new hood grabirons, but I didn't foresee their location was absolutely wrong in comparison with the prototype. That would come to bite me later, but that, is another story for another time!

Monday, April 6, 2020

Modelling CN SW9 & SW1200 Switchers - Part 1

As stated, I've been in self isolation since early March and it will continue up until early May according to Quebec provincial government. It provided me with more free time even if I still work from home. I recently discussed about modelling in difficult time and this new project is a study case of what can be done with less than perfect models, parts, supplies and paint. I better times, many parts would have been sourced from hobby shops and online stores. Unfortunately, shipping times from Ontario to Quebec are now about 2 weeks and stocks are depleting at a fast pace. This can no longer be considered a viable option which is why I've decided to do as much as I can myself in the old fashion way.

CN 7034 as handpainted about 20 years ago and damaged by storage.

The new project started when rearranging my ever growing collection of motive power and rolling stock in new storage cabinets in the hobby room. I discovered an old pair of Athearn Blue Box SW7 once bought in the late 1990s with the goal of kitbashing Chemin de fer Charlevoix iconic SW1200RS. However, as you can surmise, I was a teenager and when I compared the prototype pictures (which were extremely scarce) to the models, I quicky concluded it was far beyond my skills. Also, at that time I already had a Proto 2000 locomotive and I was absolutely dejected by the atrocious running qualities of my new Athearn locomotives. Later, I thought about using the drive and trucks to create QRL&PCo electric steeple cabs. The idea was simple: find Roundhouse boxcab shells and bash them. I had a hard time finding the shell (pre-Ebay era) and couldn't order MDC's 3-in-1 kits featuring the shell. Thus, the project died.

CN 7007 was never completed and lettering wasn't applied.

Later, I simply repainted the switchers with in a CN livery, assigning them number 7007 and 7034. I recall my choice was based on a Microscale decal sheet... I simply used numbers that were close together on the sheet, limiting the number of parts to piece together! You see I'm lazy!

Shoddy craftmanship: I simply glued the lift bar on the pilot, without eyelets.

However for this project, I'll keep these numbers because they have been part of my roster for so long and both offer different challenges. CN 7007 was a SW9 while CN 7034 was a SW1200. One will be painted in the early all black CN noodle scheme and the other one will have the later all-orange cab scheme. Interestingly enough, both locomotives had different sets of handrails, their air horn and bell location differed too. Better, they both had the classic CN spark arrestors.

CN 7007 in classic "wet noodle" scheme (credit: Gord  Hilderman,

I'm well aware these blue box kits have serious limitations, but given I have them and most parts required to bashed them into decent models, I'll do it. It is both practice and amusement, but also a way to improve my skills, including soldering brass parts which I never did before. Also, I'll upgrade various parts, including the stanchions.My goal here isn't to make a funny half-baked project, but really to see, with my actual skills, how far I can go to improve outdated tooling to the best level I can using limited resources. This is the old fashion way: if it doesn't exist, do it.

CN 7034 with modified paint scheme (credit: Jack Smith,

The Athearn shell isn't half bad, but it has big issues due both to tooling and the fact it is based on a SW7: the front radiator grill isn't right, the door arrangement on the hood doesn't match, the headlight are far too small, the cab windows are out of scale, the running board skirts are correct, etc. Name it, you have a endless amount of work on your plate. However, with some exception, these can be addressed. At this point, I will only point out, I'll accept the cab windows are wrong, that doors on the hood aren't correct and that headlights are somewhat dubious. Everything else will be improved though.

CN 7007 shell after a dip in 90% alcohol and a bath of SuperClean.

The first step was to stripe the paint. Alcohol and SuperClean did a great job on 7007, but for some reason, a layer of Floquil CN orange acted as a primer on 7034. Don't ask me why I brushed painted a thick coat of orange under a black model back then, I can't recall, but it was a weird choice to make. Floquil don't stripe easily, so I used my air eraser and baking soda. After one hour, a lot was gone, but orange paint was still there around the details. I knew it would look atrocious when repainted. Then, I decided to applied some Testor Universal Enamel Thinner with a stiff brush and scrub the surface. The paint lifted in a matter of a few seconds! Armed with this new knowledge, striping the paint only took a few minutes and I ended up with two clean undecorated shells.

CN 7034 before scrubbing with enamel thinner.

It was time to think about a kitbashing and detailing strategy, which will be covered in future parts to be published here.

Sunday, April 5, 2020

Negative Foundations For Structures

Generally, it is known among hobbyist that to get realistic structure foundations, you must embed them into the scenery. SOme with use sealed MDF, styrene, wood stripes or any other suitable material to lay a footprint and build scenery around it. Basically, it is the most efficient way to get buildings that look to be really part of the layout. Fellow Ontarian modeller Stephen Gardiner is recently experimenting extensively with this method with his excellent Liberty Village layout. I had the chance to visit Stephen's layout back in early February this year and you can now follow his progresses in the recommended website list. More on his compact yet fascinating layout in a future post.

Basically, this method is excellent when working with structures with a foundation and a superstructure. However, what do you do when your human element doesn't have a dichotomy of materials? Such as a concrete highway overpass. Depending on your prototype, you may not be able to create a separate foundation. Worst, you don't want any obvious gap between the model and the scenery. And as much as possible, your structure should stay separate and movable for maintenance and scenery work.

My suggestion is simply to create a negative foundation; slots in the scenery in which the structure will nest itself into, laying below the ground surface and hiding any possible gap between both elements. I decided to test the idea with the Glassine Canada diorama I'm actually building in my hobby room.

Using illustration board, I created a series of plank with a hole fitting each overpass pillar. They are snug enough to fit without leaving obvious gap, but they should also allow the structure to slide without effort into them.

Then, I laid them on the layout and installed the overpass. When everything was aligned as wished, I glued the cardboard parts into place. When dry, I removed the overpass and was left with a series of perfectly located holes. I must admit I feared the illustration board would swell with all that water and applied a bead of carpenter's yellow glue on the perimeter of each cardboard element to seal it.

Sculptamold was used to blend the new negative foundations with surrounding ground. While still wet, a first layer of gravel, limestone powder and fine sand was applied and glued down with alcohol and diluted white glue.

Once dry, it was time to test if the overpass would fit again in the hole. Answer? Yes! And it's realistic as can be. I've been convinced this is a viable technique and will use it again for the main building featured on that layout.

Thursday, April 2, 2020

Modelling In Difficult Times

Over the last few days, I've been redetailing a pair of junker Athearn SW7 I once painted back in 2000 if my memory serves me well. These locomotives were purchased in 1998 with the goal of replicating CFC SW1200RS. After closer inspection, it was evident the conversion would be too complicated to handle for me as a teenager. Thus, I dropped the project. Circa 1999, I thought I could use the drives and slap a kitbashed Roundhouse boxcab shell to create a QRL&PCo steeple cab. Not a bad idea if you ask me. But at that time, finding boxcab shells was hard and I didn't have access to Ebay or other flea markets/train shows to find the part. Thus, I whimsically painted them in CN colors, handbrushing Floquil paint and decalling them with Microscale CN switcher lettering. The drives were horrible and at the end of the day, I simply boxed them with no intent to ever use them. One was CN 7007, a SW9 and the other CN 7034, a SW1200. Back then, I had very little knowledge about variations among SW-type locomotives.

But now, with self-isolation being the norm, businesses shutting down and postal service running at a snail pace, these botched locomotives gain a new value: they are excellent to provide hours of meaningful occupation. The goal itself isn't to build winning-contest models, but try my best to capture the essence of a SW9 and SW1200. Starting with Blue Box Athearn models mean a lot of compromises have to be dealt with. The question isn't about creating a perfect replica but rather what elements must absolutely be addressed and which can be accepted as is. At this point, we are entering artistic license territory. We are dealing with a completely different beast.

For the sake of entertainment, I've posted my process on Facebook so it could provide a meaningful diversion to many people while enticing others to start doing actual modelling. Many well-meaning posters pointed out I should purchase better detail parts, replace the cab or buy other stuff so my model would be perfect. While I perfectly understand their point of view, I feel they completely miss was is my goal by projecting their own standards on my work. Honestly, I don't care that much. It can provide an interesting exchange and I take it for that.

However, I feel many are completely oblivious about what's going on with the world nowadays. From an economical standpoint, investing so much in these locomotives make no sense. I feel it is an utter waste of money to buy so many detail parts for a cheap and inaccurate shell. The total amount would far offset the cost of a correct model at this point. Buying a Walthers SW1200 with sound and DCC would be a better choice at this point. Should I be out there purchasing new models when my weekly revenues are cut in half and our dollar is taking a serious plunge back to what it was back in the late 1990s?

For this reason, and given the actual world conditions, I don't feel it is a logical choice to start purchasing various detail parts at premium price, from dozen of sources and wait for weeks, if not months to get them. My goal is to have fun and make sure I keep myself occupied during confinement. Until a week ago, I had absolutely no interest in modelling a pair of CN SWs. I do with what I have on hand, recycling extra parts, salvaging broken details and parts myself. Crafting details by myself is a good way to get better with my tools, lean skills and become a better modeller. Anybody can slap dozen of commercial parts on a model with glue. Shaping doors, handles, exhaust, fuel filler cap and such other things is much more rewarding, even if the result won't be as slick as a thin wall Cannon & Co cab (discontinued), a Rapido shell (pricy) or any other option available on the market. 

And let's not fool ourselves. With RTR becoming the norm in the hobby, hunting detail down parts is getting harder and harder. Detail Associates is a semi-inexistant entity which most useful parts are never in production at the right time, Cannon & Co. too, Detail West is getting scarce too and others aren't better. Basic commodities such as eyelets are selfom produced... Try to find them (not in plastic or photoetched), good luck! Canadian hobby shops (real location and online) have long depleted their stock and purchasing detail parts from the US is a sad joke. As I posted on Facebook, buying a set of KV Models SW grilles would cost be $51! At that price, I can find a neat Proto 2000 SW900 and repaint it! Worst, these grilles aren't correct for a SW1200 with horizontal louver, so I'm back to scratchbuilding! Who in his right mind would take such a decision.

Worst, Canada Post is paralyzed by the large influx of mail and parcel. I've ordered PSC stanchions from Ontario for this project 10 days ago. Didn't receive anything yet and generally, it take 2 days. Should I rely on such delay when my goal is to keep me occupied? No. So back to the basic, do it yourself!

So let's face it, you change only a few parameters in how the world is run, and what was the best option 2 or 3 weeks ago has now become impossible. As much as possible, I try to source parts from other canadian modellers via the various market pages on Facebook. Not perfect, but together, we can help each other improving our modelling without going broke. Shipping delays are still there though. On a positive note, I've found modelling older CN SW9 and SW1200 is quite interesting. A lot of variation within the fleet. We often only think about the cool SW1200RS with all their Canadian-style details, but the rest of the fleet isn't boring at all!

Wednesday, April 1, 2020

Improving Sidewalks

Sidewalks are often modelled with less regard than other elements of our layout. While some modellers will indeed pay attention to the details, scribing joints between concrete slabs or adding finely detailed curbs and manholes, I often feel the painting and weathering is quite plain.

Some of the top European modellers however will carefully weather their sidewalks until they get a very realistic look. I'm not at that level, but I thought it could be easy to level up my game in that regard.

Typical sidewalks in Canada (Quebec City)
After looking at various sidewalks in my area, it was evident snow plowing takes a serious toll on concrete structure, which is amplified by freeze and thaw cycle. A particular effect is the degradation of slab sides which start to delaminate. After a while, sizeable part of the surface is completely removed, exposing the aggregate. Also, the plow scraping creates whitish marks on the concrete sidewalks and granite curbs where the steel enter in contact with the surface. To be honest, I've never seen anybody modelling that commong pattern and decided I should.

Many media could be used to render this particular weathering effect, but I felt color pencils were the right tool since they create granular lines similar to exposed aggregate. A brownish dark gray pencil was used to create the exfoliated concrete while a light gray pencil provided the whitish scraping marks.

Pencils give you a lot of control over your work and that's their greatest quality as a dry medium compared to wet medium such as paint. I'm well aware someone could carve and sculpt the crumbling effects directly in the sidewalk material, particularly when dealing with larger scale. However, the pencils give quite a convincing effect without loosing your mind. Also, since my sidewalks are made of illustration board, carving them could create a fibrous surface which wouldn't look very realistic. If made of styrene or plaster, it would be another story.

I also decided to use pencils to try to create the crumbling asphalt effect common on grade crossing where moving trains induce a stress into the pavement. This is relevant when modelling older grade crossing that weren't equipped with rubber plates. I'd say it is less convincing upon closer inspection than the sidewalks, but from a normal viewing distance, it works wonder. Once again, it is all about suggesting weathering pattern, as film makers and theater staffs would do it. We want an impression.

If you are interested in these techniques, I can only recommend Gordon Gravett's books on scenery published by White Swam in the UK. His "Modelling Grassland and Landscape Detailing" is worth every penny invested.