Tuesday 26 January 2016


Our post today makes this the fourth in a series on CPR Pile Trestles and specifically, the Bridge at Mileage 102.7 of the Princeton Subdivision of the Kettle Valley railway.  They are unexpectedly long posts and there are still at least two to go.  This is a surprise to us as we only envisioned two or three maximum.  Here follow some interesting details which could apply to many railroad bridges and trestles.

Barrel Platforms: In each of two places on the 
deck of our trestle, 3 ties were omitted to allow for construction of the water barrel platforms.  The three empty tie spaces were fitted with 8" x 8" x 17 foot long ties to support the platforms.  Upon the tops of these ties were placed five or more planks for the platform itself. Our observation is that the platforms were placed on alternating sides of the bridge and in the case of Thalia trestle, they were located such that they were roughly equidistant from the ends of the bridge and from each other.
We provide here a CPR drawing which is taken from the set of plans dated September 3, 1957 but there are some minor differences from earlier construction practices to note.  One factor in considering these dimensions, is that by the date of this plan, 1957, specifications for bridge ties on trestles had reduced the lengths of bridge ties to 10 feet in length from the earlier standard of 12 feet.  But the size of the platform was (thankfully) increased so that the platform ties were still 17 feet as specified in the drawing.  This later version is seen in this photo by Dave Love.  The barrel is long gone as the Carmi Sub had been inactive for some time by the date of the photo.

First a few words of explanation of the abbreviations.  The "S4S" note means that the railing boards were surfaced (i e planed smooth) on all four sides.  The abbreviation "Rgh." for the platform floor boards stands for "Rough" which means that these boards were not surfaced thereby affording some measure of grip for work boots.  These platforms are called refuges in other places.

One feature which this drawing from the 50's and the photo above reveals is that the platform had safety railings.  This was not the case for bridges from earlier times where the specifications made no allowance for them.  (One surmises that old timers were a hardy bunch with an acrobat's sense of balance.)  The arrangement in this case can be seen in one of the detail photos, two 2' x 4" boards being placed in a "V" to prevent the barrel from falling off the platform.  Nothing else.  Old timers must have gripped the barrel (with white knuckles?) if caught out on the bridge as a train approached.   But the structural support of the platform is the same in both eras.  Here are two views of the model.
The barrels seem to have been the standard steel drums of 55 gallons capacity with a cover, although some photos show them uncovered.  In earlier years the barrels were made of wood.
We provide here some interesting excerpts about these platforms from an engineering book of the Pacific Great Eastern Railway Co. issued in 1939.

The book is entitled: MAINTENANCE-OF-WAY Rules and Instructions.
397. Fire protection water barrels must be placed at all wooden bridges and trestles and at all steel bridges with wooden decks.  At each such bridge or trestle 30 feet or less in length one barrel must be provided.  At each such bridge or trestle over 30 feet in length a barrel must be provided at each end... and at intervals of 150 feet and on steel bridges with wooden decks at intervals of 200 feet along the deck.
398. ...Those [barrels] at the ends of bridges and trestles must be placed 12 feet from the structure, and those along the deck must be secured to platforms outside of the outer guard rail.  A four gallon pail, in the bottom of which two small holes have been punched, must be placed inside of each barrel, and each barrel must be provided with a cover and kept covered.
 As to barrels situated at the ends of CPR bridges, there seems to be some correlation with the PGE practice, but only in some cases. In the example of the 100 foot long girder bridge at Brodie there is only one barrel and that is located at the west end as can be seen in this shot from 1989.  In photos from earlier eras, no barrels are evident.  Early photos of other bridges show barrels buried in the ground with only a foot or so showing above the grade.  As to trestles of the 1940's through the 70's, the only barrels we have observed in photographs are those placed on the platforms.

The next component for the wood deck is the outer guard rail which by the 1950's was called a "Tie Spacer".  This was made of 5" x 8" material.  We substituted 4" x 8" boards which was the later practice and the closest commercial size available.  It is also advantageous to have this plank a little lower than the railheads to allow cleaning of track without scraping it.  Been there; done that.

N-B-W's: A few nut-bolt-washer castings could be added to the ties that are situated over the bent caps.  They are barely visible in the overhead shot of the previous post.  In that same post, the close-up shot shows the n-b-w's applied to the Main Stringers.  More on these items in the next post on the construction of the bents. 

Rails: We use Floquil solvent base paint to give the rail a rusty look.  Rail Brown will do or in our case we mix equal amounts of Rail Brown and Rust.  Fortunately, we acquired a goodly amount of these paints shortly after the announcement that Floquil paints was ceasing production.  There are alternatives available.  The tops of the rail are cleaned off in preparation for installing the rail to the bridge deck.  The main rails are code 70 and the inner guard rails are code 55.
Four or five squares of masking tape are placed on the bridge deck and the position of the main and inner guard rails marked out.  Slots are cut in the masking tape the width of the rail base to guide in the positioning of all four rails as they are glued down.
To affix the rail to the ties, we use Pliobond glue which is a solvent-based contact cement.  Walther's Goo or standard construction contact cement will work as well.  A generous coating on the bottom of the rail is applied and allowed to dry.  This can be a messy job in that the glue will easily smear on the sides of the rails.  We did manage one or two smears even though there was no need to hurry the job.  Normally contact cement is applied to both surfaces but a coat on one surface will do here because by applying heat from a soldering iron to the top of the rail, and working slowly from one end to the other, the rail bonds with pressure to the ties quite well.  It can be reheated and adjustments made soon after or even months later.  In our case over the years, we have had to make a few adjustments due to movement from seasonal expansion.  Once or twice we have had to work a little fresh glue into the joint.
The spacing of the main rails was done with standard rail gauges.  Our preference is Precision Scale Co. # 4958 but the rail spacing must be critically checked with NMRA gauge after placement.  The inner guard rails are spaced at 2' - 5 1/2" on centres according to CPR specs.  For these rails we cut a short block of wood as a guide at .310" which is close enough for the distance between the code 55 railheads in HO.  These rails are centered between the main rails.

Here is a photo with measurements.
And part of the CPR drawing.

Mileboard: One final detail for the bridge deck is the mileboard which can be seen in the crop from a photo in our collection which also shows the barrel platform.  This is the bridge near Portia at Mileage 33.9.

To the right is a sketch with some field measurements of a sign long forgotten; perhaps from the Myra Canyon.  The prototype is cut from a 2" x 10" plank for outside measurements of about 1 1/2" thick X 9 1/4" wide X 39" long.
We made our tiny model on a Word Document along with many other signs and mileboards.  It can be seen in a photo above.  The paper signs are cut out and affixed to a piece of styrene for strength; one for each side of the styrene.  These mileboard signs seem to be placed on the  engineer's side of the track when facing East.  This is not necessarily a CPR specification but simply the consistent observations we have made on various bridges in the field and from photos.  On the other hand, it could be that the signs were placed on the same side as the telegraph poles where the regular mileboards were posted.

Whew.  Glad that is done.  Questions to clarify things are welcome.  More to come on the bent construction in a week or two.

Coquihalla Man

Tuesday 19 January 2016


We continue our study of the curved trestle at Mileage 102.7 on the Princeton Subdivision with a description of some of the techniques used in constructing a model of it for the benefit of those readers who may be considering building a curved trestle of their own.  As is often the case in construction, curved structures are ten times more challenging than straight ones.  But in the case of this bridge, the roadway span introduces a further complication.  Trying to describe the process is a challenge of its own.

Here is a bird's eye view of the bridge deck showing its eight degree curve and the oversize roadway span.  You can enlarge it by clicking on the photo or by using your computer ctrl and + keys.

Image result for trammel pointsLayout: The actual Thalia trestle was built on an eight degree curve which in HO is about a 98 inch radius.  No, that is not a mistake - slightly more than eight feet.  In O scale the radius would be 14' - 10" and in  N scale about 54".  For our model, we made an accurate plan starting with a centre-line and marking out the position of the bent caps and laying out the stringers.  To obtain the centre-line, an arc was drawn on a large sheet of paper to this 98 inch radius using trammel points and a long stick.  The photo to the right was downloaded from the internet to illustrate what we mean.  To repeat, this arc would be the centre-line of the bridge when laying out the deck and locating the bents.  The paper was then permanently mounted on a flat softwood board as a working surface on which to build the bridge deck.  A wood template of the 98" radius was also cut and trued up to assist in aligning the bridge ties later on.
Image result for drafting dividers
Fig. 150.   To Erect a Perpendicular to an Arc of a Circle.Once the arc was drawn, the main tool for accurately laying out the placement of the bent caps and the stringers is a compass or divider set. In order to locate the bent caps, we simply draw perpendiculars to the arc at about a scale 14' - 10" on centre.  Of course, the span over the road is larger requiring centres at 19' - 10" apart.  Perpendiculars to an arc are erected in the same way as perpendiculars to a straight line if you remember your high school geometry.  Here is an edited description and drawing from an old trades publication.

To Erect a Perpendicular to an Arc of a Circle, without having Recourse to the Center. – In the figure, let A D B be the arc of a circle to which it is required to erect a perpendicular. With A as center, and with any radius greater than half the length of the given arc, describe the arc x x.  With B as center, and with the same radius, describe the arc y y, intersecting the arc first struck, as shown. Through the points of intersection draw the line F E. Then F E will be perpendicular to the arc, and if sufficiently produced will reach the center from which the arc A B is drawn.
Read more: http://chestofbooks.com/crafts/metal/Metal-Pattern/Geometrical-Problems-Part-4.html#ixzz3wGpZNdMp
Stain: Before we go too far we should interject here that all scale lumber was soaked in a pot of stain, specifically Hunterline's "Creosote".  We wish that the stain was a little more concentrated so that occasional darker results were possible without having to soak them several times.  It is easy enough to dilute a stain but strengthening it is not so simple a process.  Most components had been cut to exact or rough size beforehand and after soaking, allowed to dry thoroughly.  We should also add that no progress photos were taken as solving the unique challenges of this particular project required a fair amount of concentration and experimentation.  With the arc and perpendiculars drawn in, we turn to the actual construction outline as much as we can remember it. 

Layout continued: These perpendiculars are the centre-lines of the standard bent caps which are specified as being 12" x 16" x 16 feet long.  There are two larger caps for the bents flanking the road which are 14" x 18" x 18 feet long.  Using dividers set to a scale 8 feet, we set one point on the intersection of the arc and the perpendiculars with the other point/pencil marking out the two ends of the bent caps (9 feet for the larger caps).  The scale lumber for these bent caps was cut and would be placed in position on the drawing.

The next paragraphs and photo deal with the allowances that must be made for this particular trestle with its non-standard roadway span, so the reader may want to pass over them for now and jump to "Superelevation" for a plainer outline of a normal curved trestle.
At this point, some full length temporary shims needed to be inserted under the standard caps to account for the oversize timbers of the roadway span.  There are two factors here.  One is the 6 inch difference in the stringer depths. This difference can be seen in the photo that follows.  It is noted in the elevation drawing in the previous post. There is also a 2 inch difference in the thickness of the bent caps; for a total height difference of 8 inches.  These shims were tack-glued to the underside of the bent caps.  The purpose of these shims is to bring the tops of the stringers level for placement of the bridge ties.  Keep in mind that when the bents are built, those flanking the roadway will have to be this same 8 inches shorter than the standard ones.  Finally, a permanent shim needed to be placed on the larger bent caps in order to bring the standard stringers even with the heavy ones spanning the roadway.  This shim is 6" high x 6" wide x 11 feet or so in length and was positioned over half of the bent cap that flanks the roadway.  See the photo below. 

In addition, one can reference Dave Love's photo of the prototype (published on the June 10th post) which clearly shows the difference in elevation of the caps of bents 8 & 9 compared to the others.  The shim in the above photo shows what is only a best guess as there are no specifications on hand for this part of the bridge and the original was destroyed some time after abandonment. 
Superelevation: For any trestle built on a curve, a tilt is specified in the engineering drawings.  To produce the tilt called for in the drawing, all caps were elevated on the outside edge with temporary scale blocks of 8".  This is close enough for a 2 1/2" superelevation in 5 feet.  It was important to build the tilt into the deck as the components were being assembled and glued so that the deck would sit naturally on the bent posts without twisting or distorting.  The caps were held in place on a flat board by pins and/or spot glue.

As mentioned earlier, the varying lengths of the stringers were marked with an X-acto knife after placing them in position directly on the plan.  The stringers were all glued to the caps and to each other and to the spacers in accordance with the colour-coded drawing in the previous post.  Except for the larger span over the roadway in which case the six stringers were laid parallel over the span.  Common white glue was used for all wood joints. 

Bridge ties:   As mentioned earlier, bridge ties had been previously cut to length and then treated with "creosote" and left to dry thoroughly.  Mount Albert Scale Lumber was the choice for the 8" x 8" x 12 foot long ties.  Note that one tie is placed directly over each of the bents.  These critical ties were placed with care and allowed to set hard before the rest of the ties fitted in between at about 12 inches on centre.  To aid in alignment of the edges of the ties, we employed the aforementioned template of 98" radius.

Now that the deck has had the ties placed we will cover in the next post some interesting details that give the bridge deck a lot of character: water barrel platforms, rail and guard rails, mileboard and a few nbw's.  We welcome comments and questions.

Coquihalla Man

Tuesday 5 January 2016


Well, it has been a while since our last post but here we are with a follow up to our study of the CPR pile trestle near Thalia, BC at Mileage 102.7.  This bridge was rebuilt several times over the years and we are modeling it with reasonable accuracy for the 1949 era.  We provide here some drawings with a few details and comments on how we built this miniature, focusing today on the bridge deck.  In preparing this post it must be said that it is not an easy thing to describe the construction of the model or the prototype and this is partly why we have taken so long to get to it.  As it is, this article will have to be broken up into several posts as it has already passed 2200 words.  Another reason for the long delay in posting is that we have been going through a major building boom on the layout with developments by the B&B gang, and the scenic and Rolling Stock departments.  And we assisted in making a motion picture of our operations which we will present eventually.
Much of the information presented here can be applied to other trestles of course, but this one has several features that are not usually found in the model railroad press: approach fills; a prototypical curve; a non-standard span to cross a roadway.  For this reason we will give some detail on its construction from actual CPR engineering documents and our model-building techniques.  To start with here is a photograph of our interpretation in HO scale of the Bridge at Mileage 102.7 on the Princeton Subdivision.

In considering a location for this bridge on the layout, we decided to make a small compromise in the overall length and number of bents.  There were14 bents in the photos and plans that we have presented earlier.  For the model, we felt that reducing that number to 12 would not seriously affect our reproduction in miniature.  All other dimensions and details are accurate.  Bridge Mi. 102.7 was one of numerous crossings of the Otter Creek between Tulameen and Brookmere which the VV&E and the KVR bridged with pile trestles for the most part though sometimes a frame trestle was constructed.  We count fifteen of them listed in Joe Smuin's book, KETTLE VALLEY RAILWAY MILEBOARDS.  They varied in size, some being a mere 30 feet long to the large one at Tulameen which is 334 feet long.  The subject of our study is some 200 feet in length but on the Kettle Valley Model Railway it has been shortened to 170 feet by eliminating bent Nos. 4 and 11. 

We did retain the obvious features of the road and creek that pass under the bridge which we found to be very desirable in that they suggest there is some civilization in the area.  For the most part, the Kettle Valley Railway traversed a landscape that was fairly remote and unsettled.

Unfortunately, our model plan of the Thalia bridge deck was rendered illegible during its construction but we can offer some prototype drawings and descriptions of the process.  Here is part of the drawing which was presented in a previous post on June 10, 2015.  The construction of the bridge deck was typical for our era for both frame and pile trestles.  This drawing shows a straight bridge deck but most of the detail applies to the curved trestles as well. 
Next is a detail from an actual CPR drawing issued on September 3, 1957 which shows how the stringers are laid out for a curved trestle.  The plan for the "E-50 Loading" seems to match the drawings for other KV Trestles in our collection.  For the sake of clarity, the stringers have been coloured.  The pink stringer (A) is the longest one at the standard length of 30' and it is beveled a little at the ends as are all the stringers.  On the full plan which follows, a table is provided which gives the lengths of the other stringers according to the curvature of the track.  For a curve of 8 degrees, Stringer "B" is 3/4" shorter than "A" at 29' - 11 1/4". Stringer "C" is shorter again.  However, in building our model we did not use these measurements.  We simply placed the model timber directly on the plan and marked the cut line with an X-acto blade and then sawed it to length.  The slight bevel on the ends of the stringers was made with a file.  On the drawing below, note how the joints are staggered and spacers placed as illustrated.  These spacers were made of wood for the model version.  Also note that stringers are butted to each other on the centre of the 12" bent cap so that they each have 6" of bearing.

Here is the full page of the CPR trestle plan issued on September 3. 1957.

As an added challenge to our modeling efforts, the engineering drawing in our previous post (and see below) calls for a superelevation of 2 1/2 inches for the eight degrees of curvature in the track as it crosses the bridge.  This, we are pleased to say, was accomplished in the model.  It is evident in the photo to the right that the caboose is listing somewhat and the cars ahead of it are progressively straightening to the upright position as the rails are easing out of the superelevation and of the curve.  Standard railroad practice for superelevation is accomplished by lifting the outer rail while maintaining the inner rail "at the established grade line".  This practice can be followed in our miniatures as well but this really is a small detail that some modelers might determine to be not worth the added complication, especially if the viewing angles render it not so noticeable. 

In a future post we will treat of superelevation in more detail.  For now we will point out the basic principle that as the curvature of track increases, the amount of superelevation increases and this applies to all track whether on a grade or on a bridge.  One other factor in determining the amount of elevation is sometimes called the "design speed".  This is simply the average speed of the different trains which use that track whether passenger or freight.
For our trestle, the plan calls for a curve elevation of 2 1/2" in 5 feet (see red arrow) which according to a table in one of our engineering books suggests that the design speed is 22mph.  The Kettle Valley was not renowned for its land-speed records.
Taking another example from the same table, we note that a sharp prototypical curve of 12 degrees (in HO about 65" radius) and a design speed of 22 mph, the outer rail will be 3 7/8" higher than the inner.  This would be quite noticeable.  However it is likely that a curve of 12 degrees would have a permanent slow order of about 15 mph, so the superelevation would be reduced somewhat.
One further point to add is that a transition from level rails where the tack is tangent, to the full superelevation is accomplished gradually through the curve easement (aka spiral easement; transition curve; or tangent lead-in).
To the right is a blow-up of the "NOTES" in the upper right hand corner of the general plan given above.  In the figure is the letter "a" which is the amount of tilt or superelevation.  The word "sprung" suggests to us that it was preferable to bend the timbers or use their natural bow on broad curves rather than use a bunch of spacers.  This would not apply to stringers in sharp curves as in our case study.

We will continue soon with another post on the construction methods we used in building our model of Bridge Mileage 102.7 on the Princeton Subdivision.

Good to be back;

Coquihalla Man