Mill Creek Central
Background: Mill Creek Central is a 7.5” gauge
model railroad. The 7.5” gauge is roughly one eight scale of the full size
prototype. Both live steam and gasoline driven models of diesel
locomotives are operated. The trestle described here will span a ravine to
enable the expansion of the railroad to the western end of the property.
It will be about 150 foot long with maximum height a little over
21 feet. A single track will run across the trestle.
This document will be updated from time to time to reflect additional changes.
Trestle Site: Photo above shows the site of the trestle. The white line is the approximate height. The dirt on the right end of the line is where the north end abutment will be. The structure will be about 20 feet behind the Ford Escape and about 20 feet high at that point. It will be behind the trees to the left of the Escape.
Photo above shows a prototype steel trestle located beside US23 10 to 20 miles north of the Virginia border in far eastern Kentucky. Our design is similar in concept.
The beams we plan to use range from 7" high structural tubes to 12" high I beams. Note in the photo above that this prototype trestle has two different deck or beam heights. Our different height beams will appear to be prototypical.
Trestle Load: A major design parameter is the load the trestle is expected to support --- the design load. There are two components of the load:
These loads are assumed to be uniform and expressed in lbs/ft so that the
loads on individual spans can be easily calculated.
A pair of identical height beams will be used on each span The type of beam that will be used on each span and the length of each span is shown on the previous drawing. The design load of 300 lb/ft is estimated to be less than 20% of the normally allowable load on the pair of beams and maximum deflections of less than 1/4" are anticipated.
Towers: The beams will be supported by towers constructed of 5 foot high panels. Each panel has two parallel vertical tubes and three horizontal cross braces These panels were salvaged from an application where the panels were stacked 150 high along the side of an industrial chimney. The panels were the track and support for a small elevator that ran up the side of the chimney.
The drawing on the above left shows one of the panel units. The photo at the right shows one of the panels after wire brushing and priming.
Securing the Beams (updated again on 10/2/2006): The project progressed to the point where we had to make decisions about how the beams will be attached to each other and to the towers. Our initial thought was to just weld everything together. After more thought we decided to bolt everything together so that we can make minor adjustments. Also, bolts permit part of the trestle to be disassembled for repair if for example, a tree falls on it. The next decision was how to deal with expansion. Steel expands 0.000076 inches per foot of length per degree F. The maximum- minimum recorded temperatures for central Ohio are about +110 F and – 30 F for a range of 140 degrees. The trestle is 155 feet long so the maximum change in length is 1.65 inches. Our next plan was to clamp the beams to the towers and let the ends float on the abutments.
We were a little concerned about the stresses on the end towers that would be caused by the expansion/contraction if we clamped everything together. We decided to measure the flexibility of tower 1. We tied a rope and tape measure to the top of tower 1 and strung both to the north abutment. We put a linear scale in series with the rope and couple guys stood on the abutment and pulled on the rope while a third guy used the tape to determine the movement of the top of the tower relative to a mark on the abutment. The tower flexed about 0.15 inches when a 300 lb force was applied. (This was much stiffer than we expected ---- guess that means we have strong towers.) Tower 1 is 11’ 1.5” (133.5”) tall and the base in the direction of the track is 25.5”. The 300 lb force will produce a down force of 1571 lbs on the north legs and an equal up force on the south legs. A taller tower will require less force to move the same distance ----- this should be inversely proportional to the tower height. However, the same force on the top a taller tower would produce a greater force on the base --- proportional to the tower height. These two effects should cancel so the torque on the base for a given movement at the top should be roughly the same independent on the tower height. Hence, from the above data we can say the added force on one side of the foundation for movement of the top of the tower in the direction of the track is 10473 lbs per inch of movement. We only took this one crude measurement so considerable tolerance must be applied to it.
After some more thinking we decided that connecting all the towers together would unnecessarily stress the the tower foundations, especially those at the ends where the greatest expansion-contraction would occur. The next option was to spilt the trestle into three groups of towers, towers 1 & 2, towers 3, 4 & 5, and towers 6 & 7. The beams within each group would be tied to the towers in the group. The beams would float on the abutments and there would be expansion joints between the groups. The expansion joint positions are shown on the sketch above. The beams between towers 2 & 3 would be tied down on tower 2 and be permitted to slide on tower 3. Similarly, the beams between towers 5 & 6 would be tied down on tower 6 and be permitted to slide on tower 5. The deck would have expansion joints at the same points. With this configuration the maximum additional load on the foundations due to expansion and contraction would be less than the live load of the structure. This is the current plan.
Expansion/contraction causes forces which tend to pull towers together or push them apart. When a train accelerates or brakes or if a moving train gets caught in the deck the resulting forces are all in one direction. These forces, especially braking or the force resulting from something caught in the deck might be much greater than the force due to expansion/contraction.
The sketch above shows how the I beams will bolted together and attached to the towers. The channel at the top of the tower will be captured between the clamp angle and the beam thus securing the beam to the tower. The tie angles will bolt to adjacent beam webs keeping the beams aligned. The separation angle will bolt to the tie angles to maintain the correct separation between the parallel beams. The separation angle is near the top of the beams to help keep the beams upright. The bolts on the clamp angles will be left slightly loose at the expansion joints. Nyloc nuts will be used on these bolts. Similarly, the bolts in the tie angles will be left loose at the expansion joint. Also, the holes in the beams for the tie angles at the expansion joints will be drilled such that there is a gap between the beams to allow for expansion and the holes in the tie angles will be slotted to allow for beam movement.
The closed structural tubes used on the south end of the trestle require a different arrangement as shown above. Nuts will be welded to the bottom of the tubes for the bolts through the clamping angle. Small tie angles will be welded to the sides of the tubes. Tie plates will bolt to the tie angles to connect adjacent tubes. The separation angle works the same as for the I beams.
The transition from I beams to structural tubes on tower 5 requires a combination of the two approaches described above.
The deck post units will be constructed off site. Five foot long sections of the wood ties will be fabricated in the work area at the shop and transported to the trestle site. The deck will be assembled by simply clamping a post U to the beams and then laying down a five foot long section of ties and bolting it to the post U and then repeating the process. The railing angles will be added after each 20 foot of deck has been assembled. The cables and track rails will be installed last.
The first part of the project is to construct the towers --- check out Constructing Towers I.