We finally threw in the towel! Using wood stringers as a roadbed is just not working out. It’s been one thing after another using wood. The first setback was rot. We finally found the answer to that, but the next thing was keeping the track fastened to the stringer – with a couple puppies constantly pounding on it. Even adding screws every 8″ didn’t help! We now have track with much larger holes in the ties where the screw heads pulled through them.
Fast forward. We’ve added yet another puppy to the team, Jasper. Alright, three full sized German Shepherds! None of them are little puppies anymore. We were barely keeping ahead of things with two puppies pounding the track into the ground. But now we’re fighting a losing battle with three, we’ve been removing track, not adding it. And that’s moving in the wrong direction!
Concrete roadbed should certainly hold up to the constant pounding, if sidewalks and driveways are any indication that is… Beyond that, the idea is to shape the concrete with a center depression, a “well” of sorts, that will “cradle” the track. The only question left is whether molding roadbed bricks will work better than forming concrete then screeding the profile.
After all the trouble we went to on the forms for the trial version of casting concrete roadbed, only to be thoroughly disappointed with the results, we’re hoping the roadbed brick approach will be more successful. This time around we’re counting on the forms to produce a much more defined profile and a better defined well for the track to sit in.
The Design
The molds are all 3D printed a “standard” 8″ in length, with a ballast profile. A picture is worth a thousand words. At the top of the diagram are the individual parts on the right and how they are assembled into a casting mold, or form, on the left. At the bottom are the “dividers” for the various diameter curved segments.
The molds are designed with the deep well profile, the center part of three. The other two are meant to emulate a standard ballast profile. The three parts are glued together, held in place with three of the “assembly jigs”, seen beneath the assembled 8″ brick mold. This assembled mold consists of a 2×6, with sloped 1×4 sides, and the assembled 3D printed parts set inside the “U” shape formed by the lumber pieces.
The dividers are designed for the three different curved track segments we use, 10′, 14′, and 20′ diameters. Each has a different angle and number of segments for the associated diameters and there are left and right version for each diameter. You can see the text slightly raised above the divider body, and mirrored so as to emboss the brick casting correctly. Can’t wait to see if it works.
The relationship between the three designs for 10′, 14′, and 20′ diameter curves are shown in comparison below. A quick explanation will help to understand the information it contains. Obviously the arrangement of the bricks into the three different curves demonstrates the 8″ segmentation, but not so obvious is the arrangement with respect to the track segments.
The Explanation
Note there are two angled guidelines. One is 22.5° with respect to horizontal and the other 30°. These angles are related to how many track segments required to complete a circle of track. For 14′ and 20′ diameter circles, 16 pieces are required, 22.5° per piece. The 10′ diameter circle requires only 12, 30° per piece.
Let’s look at the 10′ curve to begin with. Counting the number of bricks required to complete the curve to the 30° mark, we count four. Observe that while the nominal size is 10′, it’s really 59.055″ because Aristocraft decided to try to stay compatible with European track makers. Doing the math, one of those track segments is ~30.92″ long, just shy of 32″… That’s roughly four 8″ bricks.
If we round off to ~31″, we need to lose roughly ¼” per brick, making them 7¾” long. Note that these dimensions are along the centerline of the circle. Adding the extra 1¾” to get to the outside edge of the ties is ~31.84″, which rounds nicely to 32″.
Now consider the 14′ diameter curve. This time we’re using the 22.5° guideline to count bricks. Again four are required for a single segment of track. And again, it’s not really 14′, but 82.677″. That yields a length of ~32.467″, again roughly four 8″ bricks, only this time we add about 1⁄8″ to each brick.
The 20′ (118.11″) diameter curve is similar, using the 22.5° guideline except now we count six of those 8″ bricks to match a segment of track, ~46.382″ in length. Again we subtract roughly ¼” per brick, making them 7¾” long. By now it should be abundantly clear why the 8″ length was chosen.
The Decision
If all that math hasn’t turned your brain to mush, the key point is all the bricks are roughly the same length, with roughly the same angle on their ends. The angles range from 3.75° to 7.5°, but they must be halved to obtain how much the end angle diverts from perpendicular with respect to the centerline. It’s 1.75° to 3.75°, which is large enough to be discerned with the naked eye.
So let’s start with the 20′ bricks, the ones with the smallest end angle of 1.75°, and see just how noticeable it is. While the picture shows an independent 8″ mold for each brick, a more practical approach starts with an eight foot or ten foot long 2×6, and matching length 1x4s. The ten foot length is preferred since an 80# bag of concrete mix will just fill it, give or take.
Unfortunately all we have on hand is eight footers, but that just means a bit of wasted material, roughly three bricks worth. The thought is to attach the 1×4 sides in a sloping fashion to provide enough draft that the casting will slide out of the form. The 3D printed ballast profile has a built in draft angle spacer effect, this is it’s slightly wider than the 2×6, which should cause the 1x4s to splay out by just enough.
Sounds good anyway. In practice, putting this together demonstrates how imprecisely everything actually fits. Every single piece of wood is warped. Trying to precisely fit the dividers at the precise angle is an exercise in futility. At this point, those 3D printed parts are held in place with hot melt glue. The main 8″ piece gets a dab in all four corners and the dividers get a bead along both edges. That rechargeable hot melt glue gun sure did the trick though!
The Pour
The directions call for 3 quarts of water for the entire 80# bag. Despite using one of the measuring buckets to dispense three quarts exactly, the mix is too dry. Adding another quart results in a mix that’s still too dry. Not quite another quart and now the mix seems too wet. This time I used the correct (read larger) mixing tub and still manage to spill some of the mix over the edge.
Oh well, I’m more worried about getting it mixed and tamped into the form, which lays across the driveway just outside the gate. About the only place I can put it that’s pretty much level and leave it sit for a few days. Tamping the mix into the form then screeding it off with the same square shovel floats quite a bit of moisture from the mix.
I should mention I liberally sprayed WD-40 over the entire form as a mold release agent before mixing the concrete. Continuing to shovel the mixture into the form until it’s full, the extra goes in the spot along the drive where the old well used to be.
There’s quite the story about having the trees removed from the Barkyard and the crane’s front wheel “falling” into it up to the axle. We filled it in with chunks of the trees and old scrap concrete and whatever was available at the time, but over time it’s slowly settled, leaving a void beneath the edge of the drive.
The Cure
I think my phone is messing with me. I’m pretty sure I took pictures of the pour right after and the next day. Guess what? No pictures! GRRR!!! Looks like I need to get in the habit of checking afterward. Seems like this isn’t the first time it’s happened either. Guess I’ll have to describe what I saw…
Just hours after the pour, the edges had shrunken back from the sides of the form already! That seems odd. It’s definitely unexpected. The next morning, I noticed thin cracks had formed above the divider locations. At first I was worried they didn’t stand quite as tall as the 1×4 sides, but figured the stress on the small amount of concrete would make it easy to split the bricks apart anyway.
Removing one of the sides of the form later in the day, I noticed the concrete itself was still quite crumbly, more unexpected oddness. I did expect where the concrete seeped through the cracks would be weak, but not the entire brick. When the brick on the end started to crack and split into pieces trying to remove it from the form, it was decided to use a different approach.
Perhaps just rolling the entire form over with the other 1×4 side still attached would give them more support and aid in removing the entire form all at once. Considering how some of the dividers stayed with the wood and tried to pull the bricks alongside with them, that was probably a mistake too.
All in all, nine of the twelve bricks survived the rough handling. Two of them pretty much crumbled when exposed to any sort of stress. Another split along the entire length at the edge of the well. The “crumbles” got thrown in the spot along the drive where the old well used to be on top of the extra concrete.
The Result
While 100% perfection was not the expectation the first time through, a 75% yield of lesser quality, low strength bricks is less than pleasing. Worries that too much water was used soon turned into questioning whether not enough water was added, leaving the mix dry and crumbly.
If you’ve ever seen one of those videos where they compare the strength of “dry cast” concrete, tamped in the forms dry, then wet afterward, to casting already mixed wet concrete, the bricks had the strength of the dry cast method. Hopefully it’s as simple as that, mix it “soupy” next time and compare the results.
Perhaps a slump test would prove more useful, but it’s easier to just measure out the same amount of water per 80# batch and go with that. Once the magical amount is known that is. Nick says mix it soupy, so soupy it is for the next round.
The dividers sticking is another issue to be dealt with. A redesign may be in order both to help better align them without the need for tools and to eliminate the need to hot glue them in place at all. Aligning the divider angle within the accuracy of even a few degrees is difficult without some sort of alignment tool.
Hot melt glue is the quick and dirty method of “loosely” assembling the mold parts into the forms. There were a few that had too much hot melt and refused to let go. Clean up is a pain too. Normally the blobs of hot melt will pop right off with a bit of persuasion. Too much persuasion for wet fingernails!
The Comparison
As far as the amount of work involved, the most work is mixing up the concrete by hand with a shovel, but that much is required using either method. Let’s compare the two methods, which takes more work, and the pitfalls of each, starting with the trial version first.
By far the method involving the most work is setting the forms in the ground and screeding the profile into the wet concrete. It takes days just to get everything formed proper and level for pouring. If the mix is too dry, we know using a screed tool won’t work, it needs to be a trowel. And that trowel’s a guess since it hasn’t been tried. If the mix is too wet, the profile will slump regardless of what method is used to create it.
Using the casting molds approach is much simpler to prepare for pouring, using a dab of hot melt glue on each of the four corners to hold the mold to the 2×6 form base. A couple more dabs of hot melt to hold the dividers in place and we’re off to the races. The best part is the only screeding required is to level the top of concrete with the sides of the form.
As far as pitfalls, the formed in the ground method is almost as much of a pain to disassemble and it was to assemble. The form slats are held to the stakes by one small brass screw, removed one at a time, the same way they went in. Granted, disassembly can be accomplished in hours compared to days to assemble. At three days for every ten feet of right of way, it will take forever to finish.
For the brick molds, it takes a few hours to clean them up too, washing away the remaining concrete residue and removing all those hot melt globs. Those raised letters are nearly impossible to read. Vibrating the form may have helped alleviate some of the small voids and filled in around the letters more, but just working the mix into the form with the shovel did a pretty good job otherwise.
Next Steps
The advantage goes to the brick mold. Those parts can be turned around and assembled into a form ready to reuse in about an hour. The first thing to try is a wetter mix, perhaps even vibrating the form too. If that fixes the voids and the strength issues, then maybe a few tweaks on the dividers. It would be nice if they were thicker and snapped in place instead of needing hot glued.