I’ve been working on my pin and eyebar bridge for my Coquihalla canyon scene on and off for months.  Much of it has been guesswork – the actual place (now a park) was closed for a while due to a rockslide that damaged one of the original bridge footings – and with guesswork comes mistakes.  Details missed.  And some things you just can’t figure out from pictures, like the width of the deck.  Some things I deduced, but even then I didn’t trust my numbers, because you tend to remember things being much larger than they actually were.  I was doing my best though, not having the opportunity to get into the field, and I think it was coming pretty close.  However new photos, especially a few aerials from the talented photographer Andrea Coughlin, kept coming up with info that contradicted what I thought I knew.  The eyebars that rise diagonally from bottom left to top right of each ‘panel’ in each truss I thought were oriented one way the whole length.  Nope.  Midway through the bridge on the 5th panel they switch directions.  Also discovered that the east end of the bridge features lacework rather than eyebars, just like the west end.  I still didn’t see anything that answered my question about whether there was any sort of reinforcement under the bridge deck where the rails would be. Along the way I was learning a lot about processes for creating such tiny parts.  I discovered rather than installing all four bottom eyebars for each panel painstakingly with styrene (which loved to warp and get stuck to one another with glue), I could just cast them as one piece.  I started thinking, man, if I had known all of this before, I could have been more consistent.  In changing techniques on the fly, I ended up with bridge panels that were not even width, nor even height.  The variation is slight and not that noticeable, except to me.  And I am the one I am doing this for.  Darn!  I needed to get to the canyon

So I went back to Othello and I thought, okay, if I notice any more inconsistencies with this bridge, I’ll give a thought to doing a Version 2.0.  Why not?  Then I can apply all I learned.  I eagerly measured the deck width, which in pictures seems quite wide, but in reality is a skimpy 12.5 feet across. I also discovered that the plate girder the ends of the ties rest on has another beam of some sort running parallel to it that the ties were also resting on.  Oh, and that there is nothing beneath where the rails would be.  Just the ties themselves, and the odd crossbeam.  the ties are literally supported at their ends.  That’s it!  Wow.  Those ties must be strong.  The bridge is definitely kind of an odd design: most pin connected deck trusses I’ve seen have one truss per side, but this one has two.  One gentleman speculated that it might be that the bridge, in being recycled from another location due to WWI steel shortages, might in fact have been double tracked in its original configuration, or perhaps had been longer, or two separate bridges.  Interesting to contemplate, although I’m not sure why you’d need to combine two bridges in a space where you only needed one.  Perhaps the Coquihalla canyon is one of those places where you want to put off repair and upgrades as long as possible?

I learned enough there to now proceed with redoing my model of bridge, which had progressed almost to the point of painting and joining the two double trusses.  That’s okay, this will give me an opportunity to put into practice a lot of what I learned as I went the first time.  And I can use my first attempt as a stand-in for the ultimate bridge, so I can finally get the scenery done in there.

One thing about the canyon – it never ceases to amaze me how utterly beyond description it is.  Superlatives don’t cut it.  It blows my mind early 20th century railroad builders managed to put a railway through here.  With 19th century tools!  Also each time I return I find new visual cues and features I end up adding to the model.  I’ll probably be doing that forever, always working just another inch closer to perfection.  I’m very pleased with how it’s come out so far – especially considering I had zero knowledge of modelling going in.  I wanted the Coquihalla canyon in my office, and darn it, I think I may be close to having it!

Othello Revisited.

Way back about 6 years ago when I first started modelling, I chose to begin at what is known as Othello.  Othello was one of the ‘stations’ along the Kettle Valley Railway’s Coquihalla Subdivision.  And the word ‘station’ really doesn’t apply well – there was no actual train station – just a tiny passenger shelter at one end and the section foreman’s house at the other, along with a crew bunkhouse.  

Being new to the hobby and new to the KVR at the same time, I had to make do with the best info available.  Using Joe Smuin’s excellent Mileboards book, I figured out the general lay of the land for Othello and set about to plotting where it was using Google Earth.  At that time I hadn’t ‘invented’ my strategy of printing out aerials in N scale for overlays, so I relied on measurements and guesswork to lay it out on my foam base and get carving.

But there was one flaw.

As I’ve mentioned before, a lot changes over the course of 50 years.  Especially with rivers.  Rivers seem so permanent, don’t they?  But they’re not.  They are dynamic.  Especially mountain rivers.  They are always working like busy beavers, eating away at embankments and rock, cutting a path, ultimately, to the sea.  And as they do their work, they change.  They create and remove landforms at will.  They change course a little.  Or a lot.

In the case of Othello, it was the latter.  When I first visited Othello in person in 2009, I was *lost*.  The old photos showed an idyllic, wide, lazy emerald river lumbering alongside the tracks before making a bend south.  The passenger shelter appeared to sit right on its banks.  But when I went to Othello, I saw no river.  Instead I saw forest.  TONS of forest.  The roadbed was still there of course (it is now Othello/Tunnels Rd. and leads to the provincial Quintette Tunnels parking lot).  But the river?  Couldn’t see it.  And in fact, I came to doubt Mileboards and wondered if maybe the site was further east, because there was a site back there that looked pretty much identical in many respects to the one in the old photos.

In emails with Joe however, he patiently explained how much could change over a few decades.  And now, 6 years later, I finally have my hands on a late 1940s aerial of the area via the National Air Photo Library offered by Natural Resources Canada.  If you’re a modeller and haven’t used a resource like this, I highly recommend it.  I would assume most countries have something similar.  Since the invention of the airplane, governments have been surveying their territories from the air, and Othello was no different.  Only from the air are we able to appreciate how dramatically Othello has changed.

The first thing that jumps out is the islands.  What appears in the Google Earth photo (and on the ground) as a mostly solid mass of land and trees was once active riverbed, with a series of islands along the north side.  In the older aerial, there’s even some bridges to a couple of them!  Wow!  That area is *unrecognizeable* now.  The islands are now just part of the landscape.  And the river is long gone to the south.  And it appears to have taken that chunk of mountain jutting into it in the middle of the old photo completely out!  I don’t know if maybe later highway building crews dug it out for fill or if the river was purposely rerouted for flood control, but the whole main body of the watercourse has shifted significantly south.  Any wonder I couldn’t get my bearings!  It’s clear the section house is, as Joe said, where it always was.. just a few feet further back on the property post-abandonment.  And in fact, the shelter survives also, albeit as a funky bed and breakfast type deal at the Kwakwehala Resort.  There’s some pics of it here, now billed as the ‘Othello Cottage’: http://www.eco-retreat.com/sleep.htm

The thing that I am confronted with now though is whether I’m prepared to live with my Othello scene as it is today, or take the benefit of this new information and bring it that much closer to its historical appearance.  It is tempting.  Although I’m proud of the scene as it is and loath to get into a redo, there is some appeal to the idea of taking everything I’ve learned in the 6 years since I first did it and reapplying it to today.  I’d love to capture a few bits of those islands too if I can.  Wow.  But yes, Tip #1 to modellers: invest in the aerials.  The aerials will tell all.

Just a few more pics of the pin-connected deck truss at MP 49.7 of the former Coquihalla sub.  As posted below, this is the bridge standing between me and being able to complete the first level of my Coquihalla scene and run some trains.  And of course, it is a serious piece of engineering.  For bridge lovers, it’s a visual feast: box girders, laced box girders, stringers, eyebars, pins, you name it it has it.  

Above I did a drawing on computer, trying to rationalize it and understand how its components went together before attempting a model.  I wanted to figure out precisely what supported what, how things interconnected, partly for my own interest but also of course to dictate how assembly would work.

Unfortunately this is one of those situations where HO scale is king and the limitations of N scale become painfully apparent.  First, they do not sell any laced box girders of the type seen on this bridge in N -scale.  They do in HO, of course, but not N.  Initially I was going to fabricate something myself, and had been, using a styrene ‘box girder’ (minus the lace) as my model, I had been casting the legs one at a time using Alumilite.  The plan was to affix lace to them afterwards.  Unfortunately, that plan fell victim to the reality of trying to manipulate lace that is so tiny and fragile it’s almost impossible to manoeuvre.  I tried some etched brass lacing (X shaped rather than W, the plan being to cut the Xs in half to arrive at a W shape) from Gold Medal Models, but even it was the wrong angle, too big, and too unwieldy.  Cutting it to make the W wasted half the lace and warped the other half.   So that died.   I even tried making lace myself using the contraption pictured above.  I pictured sliding up my bridge legs in a measured way and then affixing a dab of glue, sliding in some thin styrene strip, cutting it, then on to next, creating a series of / / / / / down the bridge leg one way, before flipping it and going the other way.  Nope.. nothing doing.  N-scale Fail again – it is impossible to glue with sufficient strength and accuracy when you’re dealing with pieces that small.  Too often glue went where it shouldn’t, and the lace got stuck to the jig.  Scratch that.

So now it looks like we’re going for the ‘best we can do for now’.  N-scale is unlikely to ever have the array of options HO does, but there is a truss bridge kit made by Central Valley with some laced girders I can cut to the right size.  The lace pattern and style is not correct, but, what other option is there?

Hope in 1961.  Am getting some great help in FB forums on identifying buildings.  Apparently the building perched right against the tracks in the first photo had sliding doors and coal on the floor.  Not sure what it was for, but in the second photo there is another building way up at the top left, just in the shadow of the highway embankment that looks similar in size and dimensions.  Wondering if they’re both the same thing and what they were actually for?

Morning arrives and the Realistic Water has cured to a point that now if I touch it, it holds its form.  I grabbed an old, small paintbrush that had dried solid with something or other on it, and began ‘stippling’ the top of the water to create tiny mini waves.  Hard to see in photos, but noticeable up close.  I’m very pleased with how this has turned out.  By layering paint and RW, I was able to achieve depth and a little bit of the mottled look rivers have up close. I’m very pleased with it and am going to call it essentially done. 

Union Bar Rd. will be the project tonight.  I have on decent authority it was paved.  My ‘paving’ technique involves using very wetted down spackling, applied very gently with any flat bladed tool of the right width.  My choice this time was a paint mixing tool.  After it’s dry, a light sanding, wipe down and application of top coat should seal the deal.  The trick will be figuring out the crossing at that spot.  Was it wood?  Pavement?  Can’t tell and nobody seems to know for sure.  Guess it’ll be more guessing. 🙂

A few beauty shots with the scene up on the shelf.  The bridge is off-kilter because the base is being tilted to keep the water from pooling at one end of the river.  Looking east, we see the abbreviated run to the tunnels.  West, towards the long-demolished signal tower at CNR crossing.  The green is looking a bit too green for me but I’ll get that with some adjustments when the water dries.

More modelling, more learning.  I’m quite pleased with how my Coquihalla river crossing scene is taking shape.  It’s always kind of neat recreating things that no longer exist – not just the railroad itself and its associated hardware, but also the land and water.  Seeing the river come to life in a form and following a course it hasn’t been in for decades is really neat.  And disorienting:  I’ve been to the site a few times and the changes are significant enough that in my mind’s eye the model at times just can’t look right.  

My first run at the river, as usual, ended in stalemate.  I tried the usual green paint, but it looked wrong here.  Photos gave the water a more brownish colour.  So I tried filling the bed with brown talus, and then pouring Realistic Water over it.  As it set, I’d dip a fine paintbrush with white paint on the tip in and create water ripples/foam.   This produced something closer, but a bit too clear.  I made a mistake too – using modelling clay to try to seal off both ends of the river so the realistic water wouldn’t pour right off the edges.  The Realistic Water very realistically find every nook and cranny and pretty much emptied itself all over the table.  On the plus side, the talus that began to float when I first poured it settled and the whole river bed became hard and fixed in place.  I then began applying lighter dashes of green, and then when that dried, another dollop of Realistic Water, this time with the ends of the river sealed off by DAP white silicone caulk.  I also worked on Union Bar Rd., which I was pleased to see existed in the 1950s and provided one more crossing to do on a subdivision that only had a few of them.  Have to figure out how to handle the crossing: there are many questions.  Was the road gravel?  Or dirt?  Or pavement?  Was the crossing wood?  Pavement?  Black and white aerial photos unfortunately have their limitations.

While the water set, I worked on laying down the roadbed, carefully aligning it with the bridge, and doing some scenicking all around, including the dry, rocky floodplain area.  From the older aerial I could see it had vegetation in a wide swath in the middle, so it probably wasn’t flooded often.  It’s pretty amazing to see that that whole area is wiped out now, the river having broadened and shifted to cover it.  Had the KV stuck around, I guess CP would have needed to figure out another option to the pile trestle at the west end.

Speaking of pile trestle, that’s the next project on this scene.  I’m gathering details (and materials), but it looks like I need it to be about 15’ high, about 180’ long (shortened because I went with a 150’ truss bridge rather than the 130 that was there originally).  Onwards and upwards!

Work continues on the Coquihalla. Over the weekend I mostly finished up a Central Valley truss bridge kit I’d had lying around for ages. This bridge kit was very close to the prototype bridge that existed at Mile 53.4, near the Kawkawa Lake Rd bridge to the south. That steel bridge replaced what was originally a wood truss. As originally constructed, there was a wood pile trestle at each end with the truss in the middle. After the steel truss replacement was in, the pile trestle on the east side was filled in.

From the aerial photo of the site attached, it is clear the flow of the Coquihalla could expand considerably in heavy flow conditions and these flows periodically took out the pile trestle at the west end. I also note via the attached satellite photo from Google Earth how significantly different the area and river bed is. This is important when constructing historical railways: rivers do change course over time, and/or riverbank is sometimes filled in or diked for development.

I say ‘mostly finished’ with this kit because there are some differences I need to remedy. The bridge is 20 feet longer than the prototype, but rather than cut it to pieces are and manage the tricky business of piecing it back together, I decided to fudge it a bit and go with 150 ft. I am modifying the x bracing at the top to more closely resemble the original. I also acquiring some more box girders as the type of braces supplied in the kit are not correct to my prototype. I’ve never built a pile trestle but I’m looking forward to trying.

To accomplish this scene I have used the historical aerial photo and printed it on a Designjet 450c large format printer. I then overlayed it on my foam base, placed my bridge as per that photo, and am using long nails to mark off where details like bridge footing and river banks should be. Next I will carve out the banks, river, hillsides etc. Then probably to pile trestle.. then water.

Apart from the kit, I created the bridge footings from notes and measurements I took of the original onsite (pictured).  Tough to work with with all the plant life around it, but it was cool seeing that still extant footing, with the 1914 date stamp on it.  Where it is now it is almost impossible to believe that trains once passed over it.

The New Year has given me renewed vigor and desire to see at least one level of my Coquihalla model operational.

On the first level of my model, I have three bridges I must complete in order for trackage to connect.  Of the three, I have one completed (the HDPG between the second and third Quintette tunnels), I have another in kit form  to make the last Coquihalla crossing just east of CNR crossing that will be kitbashed into the right size and look, and then there’s the pin and eyebar bridge between the third and fourth Quintette tunnels.

I have to be honest.  This bridge scares the hell out of me.  Look at it.  The HDPG that took me a month to build is positively child’s play by comparison.  That one had a couple of big girders, some stringers and hangers.  Whoopdy doo.  This eyebar bridge – well, let’s just say I can see why they stopped building them this way.  Lace and lattice, eyebars and rods everywhere.  The thing looks like a spiderweb on some angles.

And then there’s the small problem of lacking plans or engineering details.  My HDPG came about thanks to an article in a modelling magazine.  The kit bridge already mostly exists.  But the pin and eyebar?  Nada.  Zilch.  No plans.  I can’t even find anyone with good field measurements.  I went out there myself, and apart from length, the rest is tricky to divine.  Tricky because the bridge is perched between two cliff sides with a river canyon underneath.  It’s impossible to measure certain parts, or even see some of them.  So there’s a lot of guesswork.  This is why I pays to take time.  As I’ve learned more about modelling, I’ve learned lots of terminology, but more important, I’ve learned what to look for.  What questions to ask.  What parts to measure.

The eyebar bridge looks intimidating, and it is, but in reality it’s pretty straightforward.  You have four trusses that run the length of the bridge and are its backbone.  Components of the trusses include posts, long horizontal stringers, and of course the eyebars.

Sandwiched in between the trusses is what appears to be plate girders.  This is what the ties (12 inches thick) appear to sit on at their ends.  From my examiniation of various photos, I cannot divine if there are any girders directly under the track.  There should be, but it’s not evident in my photos.  I suppose it’s possible these massive ties are thick enough to carry all that weight on their own.  If so, wow.  That means directly below the train is nothing but rails, the ties, some thin floorbeams and a few (torsion?) rods.  Impressive.

Now to the build.  You’d think a bridge this large would positively dominate the model.  But in scale, it’s actually only about 13 inches long and a little over an inch and a half wide.  Which makes all those rods, eyebars, channels, etc *very* finicky.  Another problem is that 6" plate steel scaled in styrene results in something that has the structural integrity of paper.  The little lips on the stringers are so small they’d almost require a magnifying glass to see.  Rivets?  Yeah, I think we’ll be skipping those.

Anyway, I’ve spent 3 years trying to figure this out.  My battle plan is to keep it as simple as possible.  First, I printed out a rectangle, in scale, representing the footprint of my bridge.  My plan is to build it upside down, starting with stringers and girders.  I’ll assemble four of those (to create four trusses).  Next, I’ll create posts.  The posts are laced on two sides, so that’s going to be tricky.  Nobody makes lace that matches the size present on the prototype.  Thus, compromise.  I found some etched brass in an X pattern rather than the > pattern used on the prototype.  I chose that because it’s a bit smaller than the > type offered was, and I can just cut the middle of the X’s to get my > pattern.  Probably what I will do is create each post in components – the sidewall, a bit of lip for the lattice to attach to.  I’ll get that done and then make a mold of it, and cast it in Alumilite over and over again.  I need 40 of them so I’m not eager to scratchbuild each one.  The bottoms will have the holes to secure the eyebars.  Between each post is four of those, plus a couple that go diagonally from bottom of one post to top of the next one down.  I’ll make a couple of each and use alumilite to replicate them also.  Once I have the trusses done, I can then install the girders, and then the floor beams and start tying it all together.  Detail bits like the rods will be done later.  Once it’s steady and strong enough, I’ll flip it over and work on the top deck.

Last night I began working on the stringers.  The stringers run the length of the bridge and each of the 10 posts per truss are bolted into them.  They have lace on the bottom and plate on top.  Immediately this presented me with a challenge – the available styrene C channel pieces were either too big, too wide, or the lips of the C shape were too long.  After finding channels that had narrow enough lips, I assembled them on some thicker-than-scale styrene strips.  I was pleased to see how strong these work, but dismayed to realize what I had was twice the size of scale.  Crap.  So I removed the c-channel and found some L shaped styrene and glued along one edge to create the c-channel shape.  Still a bit over spec, but okay.  It’s not going to be exact.  Exact would be so small and weak.  The toughest part is keeping this extremely flexible styrene straight and even.  It likes to twist and bend.  Getting one L- shaped piece in position after pulling the c-channel was about all I could do, but I think it’s looking about right.  I positioned two experimental posts I made using the brass lacing in between to check.  Those posts are more or less to scale (about 12 inches in real life), so if they fit I should be good. 

This is going to be tricky!!!