Technology Rediscovery

Funicular 1: Roof Garden Transport System (RGTS)

I grow plants on top of my garage roof, and moving plants and persons up and down from the garden seemed to be more fun to do by mechanical device rather than some fangled and complicated stairway system. Thus was born the funicular .

Jump to a section:

Origins

The need for convenient transport to and from the roof garden created the "creative catalyst" for constructing Funicular 1. Original plans called for 1 or more vertical stantions (with well-built footings) to prevent sagging of the track support structures (TSS). I also had envisioned the car running directly from the rear edge of my back patio to the lower right (facing back of garage) end of the rear garage gable. I imagined setting the saws to whatever custom angle was required to make the journey a straight line from the start and end points.

While talking through design ideas with my friend Brian, I decided to take his advice and scrap the plans for a perfect landing on the back patio in exchange for immense time and material saved by using a standard 45-degree angle of incline (AI). This obviated the need for more than 1 stantion, and perhaps both if I could create a sufficiently rigid track support structure (TSS). While mulling over this essential design decision, my regular reon missions to Construction Junction bore sweet fruit: 15-foot I-beam style manufactured floor joists were being essentially given away ($5 a pop).

Obviously not engineered to bear load at a 45-degree incline, my loads (in the hundreds of pounds) is well under typical manufacturer ratings for this size and type of floor truss. The metal tracks came The Junction also and needed only slight modifications from their original construction as shelving support brackets. Construction on the track support structures (TSS) began before a cart design was finished to facilitate adaptations to the layout during construction of the TSS.

Back to top

Construction process overview

Key construction steps:

1. Design: Use basic right triangle math to determine 2 hypotenuse length given the desired rise from the base of garage to bottom of gable end siding. (About 17 feet hypotenuse.)
2. Cut and join floor trusses with but end joints and gussets to create two track support beans of identical length from 15-foot stock lengths which form the backbone of the track support system (TSS)
3. Cut floor trusses at 45-degree angles with a circular saw (hand held)
4. Design and implement track support beam (TSB) "webbing" from 2x4 pine studs. (Think: Budget Funicular).
5. Prep and install four track rails (re-purposed steel shelving brackets)
6. Layout and pour bottom landing pad (concrete, 4- to 8-in thick with 3/8" rebar, photo below)
7. Install ATV winch and rig up charging box to the the backside of the garage footings
8. Built and install the upper hanger of the track support structure (TSS)
9. Fenegle a way to lift track and support structure weighing a few hundred pounds into position by myself--a pretty scrawny person.
10. Hoist the track with its support structure (the TSS) onto the upper hanger and lower launch pad with the winch. Secure to hanger and pad.
11. Build and install platform car. Rig upper hoisting hardware (UHH).
12. System tests and adjustments.
13. Paint and seal (TODO)
14. Christen (TODO)

Back to top

System component diagram

Back to top

Construction Notes

Track Support System (TSS)

Creating the TSS was the meat of this project. The distance between track support beams (TSB) was set to maximize the available wall support width between the edge of the garage and the rear southeast window on the garage. Leveling and squaring the beams remained the critical process throughout the construction process: I used many wood shims and scrap to ensure that each of the four points of support for the frame together held the beans parallel (square) to one another and flush on each end. A half hour or more of fidgeting was required to setup the beams correctly. With my handy dandy framing nailer, the rest was just a bunch of "thumps" of the gun.

I call the 2x4 boards that hold the track support beams parallel to one another the "TSS Bean Webbing". Note that in the right photo below, the TSS is rotated 180-degress along its major axis from its installation position. I chose to use use parallel space beams (2-feet apart--probably too much distance for some climbers) on what becomes the top of the TSS to create a built-in ladder for climbing the distance with one's legs. The 45-degree offset beams on the bottom provide angular rigidity for the TSBs.

Back to top

Concrete TSS Landing Pad

A sturdy base provides resistance to settling which is critical in this project because the upper TSS hanger assembly does not provide any significant angular retention of the TSS beams--meaning it only holds the TSS "up" and does not prevent it from sliding backwards towards the house. Since my hard is dirt and clay for several feet, locking the footing into stone of some kind was not possible. With dirt and clay my only substrate material available, the key became drainage to avoid the soil from washing out over time and sinking the landing pad. To promote drainage, I layered the pad's base with 4-6 inches of "paver base" from Lowe's and filled the pad's side's in with the same rock after pouring.

Concrete forms are essential for any work of this type and extremely handy for creating a sound block during the rough-and-tumble action of mixing and pouring concrete. I asked a guy pouring a driveway skirt how thick he makes his slabs and aimed to match or exceed his four-inch application. You can see in the diagram below that my form is 3.5-inches deep and rests on a bit of a dirt ledge all the way around, leaving the deepest parts of the hole about 7 or 8 inches. Since I'm generally cheap (and it was late and I was worn out) I used some hunks of concrete block lying around my yard to fill in some volume and save on sackrete bags.

To anchor the track support system to the landing pad, I embedded 1/4" threaded rod all the way through the pad and maintained their orientation with a 1x3" furring strip as their base. I created a tied web of 3/8" rebar to add lateral strength to the pad and prevent cracking over time. Remembering to tape up and seal the threads above the expected concrete level from splattering rock paste was a victory for my over-eager personality!

Construction was halted for a few moments to allow the Official Project Monitor (OPM) team to investigate some concerning aromas emerging from the recently dug pit filled with gravel. No report was filed and construction promptly resumed, but I'll tell you, we were all on edge for a few moments before the assessment results were released. I'd rather be safe than cross the OPMs.

Junior official project monitor (OPM) on a routine site safety check. These occur at random times during the construction process and, while somewhat inconvenient, form the backbone of the site safety process (SSP)

The Junior OPM detected concerning scents coming from the work site and called in the Senior Official Project Monitor (SOPM) for consultation. Work halts entirely during this process per the Uniform Funicular Construction Code (UFCC).

On the left below: the cured pad and track support system (TSS) mounted with a pressure-treated 4x4 beam crossing between the two sides of the floor joists. 1/4" hex bolts with washers connects the 4x4 to the pad. A 7" x 5/16" lag screw and about a dozen 3" outdoor screws connect the 4x4 to the TSS frame. On the right: side view of the TSS mounting assembly which amounted to a 2x4 stud cut at 45-degree angles to fit in between the layered floor truss end caps. This ought to prevent any lateral slippage from a lumber standpoint. The friction between the TSS beam ends and the concrete was enough to raise and lower the car without a single piece of hardware secured (only backup braces that were not loaded). Given this test data, the real concern for lateral slippage remains migration of the concrete pad through the soil and clay as a consequence of natural settling of the ground and water-induced erosion.

Back to top

Upper track support system (TSS) hanger

Force the "slip out" tendency of the track support system (TSS) is handled entirely by the landing pad and not the upper track support system hanger (images X and Y). Imagine leaning a ladder against a wall--the top of the ladder is just "resting" against the wall and isn't actually "holding the ladder up." Thus, the functional role of this component is maintaining the location of the TSS against the garage. Weight is transfered to the concrete by simple normal force compression of the hanger against the wall. One could very well get away with only mounting vertical slide prevention braces on either side of the TSS.

Designing and implementing this assembly was a whole afternoon's project:

• As you can see the concrete block used to raise the garage roof has silly (decorative) pyramidal notches extending beyond the vertical plumb line. As a result, the rear of the hanger required chiseling to lie flat against the wall.
• Forming the horizontal, downward angled support cross piece and bolting it to the rear hanger plate took longer than expected due to my mistaken placement of several bolt holes. Each hole had to be bored out to fit the washers and nuts.
• The upper hanger is called a "hanger" since it is only supported by two 3/8" sleeve anchors installed directly into the garage wall brick block--with no mounting hardware preventing the whole assembly from lifting up. If a monster came through the backyard and lifted up the TSS from its base, the upper TSSS hanger would function as a large hinge as the strap plates bent against the block wall. Installing the sleeve anchors in the block was a major undertaking since don't have a suitable extension ladder (yet) for reaching this high up on the rear of the garage. Maintaining force on the hammer drill to drill the rough holes required lent itself to setting up a rappelling rig such that I could hang off the top of the roof of the garage to maintain enough pressure on the bit to bore 3 inches into the concrete block.

View of the hanger from the outside of the assembly. Note the single hex head of the sleeve anchor. (All bolts into concrete must be installed with appropriately rated anchor hardware.) The hanger back plate and angled cross bar are both constructed from pressure-treated timber. The grey hose running under the angled cross bar is a water supply line for the roof-top garden.

View of the TSS hanger assembly from the in between the TSS's floor joists. Note how the TSS beams are simply leaning against the angled cross piece (with a few 3" screws to prevent wandering over time). Think of this as a ladder that's leaning against a wall. The 45-degree cut on the support cross piece is essential such that the force is transfered across a rectangular intersection and not just on a single point which would be the case if the TSS beans were simply leaned against the garage wall (the angle may not be a perfect 45 degree).

Back to top

Track & Platform Car

The car design emerged once the track support system (TSS) and its track rails were completely installed. I created this component with my father whom I call "D". The first iteration of what will likely be many versions, this simple platform car was designed to be as narrow as feasible such that the car itself could be lowered as near to the ground as possible. Many sizes and shapes were considered, including a multi-level car with a lower and upper platform such that the lower platform would be almost flat against the ground when the car is lowered and the upper platform almost parallel with the lower corner of the garage roof gable end when raised. This simple design was implemented to avoid too much complication for a first go.

This is a first-pass, ultra functional but still somewhat rickety structure. My dad and I were slated to build the roof garden boxes the next day and needed a machine to help hoist framed boxes onto the roof safely. The use of a scrap piece of the floor truss was handy since, when placed on its side, the overhang of the I-Beam top and bottom plates latched nicely onto the leading edge of the 4x4 central support beam. The "other side" of the I-beam ends forms a 3/4" lip on the long edges of the car's platform. This is handy in preventing skydiving by potted passengers during transit.

A critical adjustment this car needs is to shift the side track wheel fixtures to pinch the rails more firmly. My initial thinking was to build a noticeable amount of play into the car-to-track calibration so as to prevent jamming as the structure settled and shifted over time. After a few months of near daily use, I'm convinced that I'd rather have the car grip the rails more snugly than await shifting later. I can always tweak as needed. This is not a big job, but the list is long.

The wheel assembly consists of a horizontal arm, vertical wheel mounting post, and the wheels themselves. The vertical wheels bear the load and the side-mounted wheels prevent the car from jumping out of the rails.

Hitch hardware close-up: the welded eye-bolt is attached to the load-bearing beam on the car with a nut and washers. The eye hook is rated to about 700 lbs, which is well outside the tolerance for normal use and could take quite a shock load. The standard winch hook is used to attach the winch cable to the car. Note the spring-loaded gate on the hook to prevent accidental sliding out of the hook itself.

And to wrap up the car, a view of its underside:

Back to top

Hoisting system: winch, pulley, and cable

Funicular 1's car is hoisted up and lowered down the track with an ATV winch rated for a 2500-lb load. Many childhood dreams were realized when I purchased a winch to fit this bill from Harbor Freight tools for about $70. Link to product page . It is designed for a sturdy vehicle and as such as a 12VDC input and cranks at up to 100-amps. The wiring for this device seems sturdy enough and the weatherproof remote relay box is handy to mount near the winch itself.

My design- and mechanistically savvy friend saw Funicular 1 and was complementary. He told me later that afternoon that the standard mechanism for hoisting professionally designed funicular cars is via a counterbalance system and motor--the same principle used in building elevators. This is clearly a much safer approach since the current design relies on the internal winch hardware to hold the load during retraction and extension and pauses along the track. Single failure risk is extremely high in the current setup and is not ideal. If the automatic breaking mechanism on the winch fails, the car will likely zoom and crash in horror.

Expediency (recklessness?) lead us to skip another safety mechanism I was considering under the guidance of my first friend/consultant who suggested 45-degree angles: a single direction tooth and grabber setup as used on roller coaster cars during their ascent. We were thinking a retractable grabber would catch on screws drilled into the inside of each track support beam (TSB) unless it was manually raised by the operator. A solid idea for future iterations.

Now some close-ups of the winch motor and gearing along with a view down to the winch from in between the track support system (TSS) beams.

A view down to the winch from in between the tracks support system beams. You can see the upper hoist hardware which also consists of a central load-bearing 4x4 with another 3/8" welded eye bolt. A rusty--but still very sturdy--3/8" screw gate carbine clip connects the main block (pulley) to the fixed load bearing beam.

The ATV winch mounting bracket, motor, gears, remote control receiving module with relay. The battery and its charger have been living inside that plastic bin. Not a great idea--the battery gets baked by the sun each sunny afternoon.

Back to top

Revision and update priorities for Funicular 1

• Transition hoist system from the current winch-bearing model to a counterweight setup. I'll mount the winch at the top of the track support system (TSS) so it sits where the eye bolt and block (pulley) currently are mounted. A second car--the counterweight traveler--will get its own track on the inside of the track support system. I plan to run the cable around the drum of the winch about 1 or 2 times such that the winch will basically feed the line back and forth between the "car" side and the "counterweight" side. Safety through redundancy will be achieved by using two separately mounted hoist cables to connect the car and the counterweight.
• Move battery and charging hardware into the garage and send a dedicated cable through the wall or window down to the winch. I'm sure the stress on that system is unduly high given the sun exposure of the current system.
• Reduce play in the car's wheel and rail system by moving each side-mounted wheel toward the center of the car. Less jostling will occur when the car is pulled up and lowered down the track.
• Stabilize the upper end of the track support system (TSS) by adding diagonal cross braces between the two TSS beams.
• Build a suitable stair set and landing pad staging platform. The current pile of concrete block is not befitting of a wooden machine like Funicular 1.
• Paint and seal all surfaces.

Back to top

Project Process Reflections

• The design-as-you-build construction model was certainly fun since I have never built a funicular before and each step required a good handful of problem-solving moments. Building from a mental plan only, however, precludes more advanced structures that interlock and work together as a system. Lacking a master plan also makes gathering feedback and ideas about a design trickier.
• Building with revision and updates as a stated reality is psychologically handy for somebody like me who can border on "perfectionistic" during these kinds of projects. "A little too much play in the car-and-track system? For next revision! Keep Moving!" Just make sure it's reasonably safe for the time being.
• Designing a method for hoisting the entire tracks support system (TSS) onto the upper hanger using only the winch itself and creative pulley placement amounted to the most satisfying step. While I have no photos to document this process, it was quite a site to behold. I actually used a different placement of the main block (pulley) to allow for vertical raising of the TSS against the wall. The process of hoisting involved dragging the upper edge against higher onto the wall while guiding the lower end of the TSS toward the landing pad. Hoisting the TSS onto the upper hanger with only the winch worked out perfectly and brought a small tear to my eye. Using the winch to hoist the TSS itself reminded me of database systems which use its own data structure for controlling and monitoring databases. When in doubt, use the native system's structures for building and implementing the desired functionality.

Back to top