MoonBoard Build: Our Outdoor, Variable-Angle, Cable-Suspended Home Wall
This is a mostly-complete (and rather long-winded) log of our home-wall project. We found lots of amazing resources across the internet, but very little that covered a project from beginning to end. Neither of us are structural engineers and nothing here is advice, but should help other potential DIYers learn what goes into such a project.
The Project’s Goals
As with any project, it was important for us to understand our goals early on. This wall was designed and build with the following in mind:
Safety
By far the most important feature here was safety. A climbing wall of this size will weigh many, many hundreds of pounds and can easily kill a person if it falls. The wall is intended not just for the two of us, but also for my younger cousins, so in addition to being strong, the design needed to avoid systems that were finicky or prone to misuse.
We designed everything with a minimum safety factor of 5, meaning that every component has a breaking strength of at least five times its maximum anticipated working load. This isn’t some magic number and each project is going to have to make its own trade-offs, but this was a comfortable and practical number for us to use. Every component or connection was assessed for its strength either using manufacture ratings or general design tables found on the internet: from the wood beams and studs to the connectors, the soil to the cables and hoist rings.
We also built redundancy into every aspect of the design, with one exception: we assumed that the large Oak tree will not unroot or split at its base. Given the size and health of the tree, and the impracticality of adding a second factor, this was a call that we were comfortable making.
Durability
It was important to us that all our effort wouldn’t wash away with the first rain. This goal is partly about respecting our time and effort, and partly about preserving safety: when components corrode or rot, their strength is greatly compromised. This wall was designed to last at least 10 years with periodic inspection (mostly of cable clips) and minimal maintenance.
Flexibility
This wall is intended for users with a broad range climbing ability. To make it as useful and fun as possible, we wanted to follow a specification with an app and community for easy route sharing. We needed the same “set” to function with the app at different angles, and we wanted the option to re-set the board in the future. MoonBoard was the obvious choice to satisfy all these criteria, with a variety of different “sets” reusing many of the same holds, a large community, and standard 25° and 40° overhangs.
We also realized that the wall’s angle needed to be easily and safely adjustable by a single person, or else in practice it would be left at one angle.
Settling On A Design
Like most things I design (from a simple bracket to a computer algorithm) I started out sketching different ideas on paper. I calculated some “ballpark” loads and we walked the property looking for an ideal spot to put a climbing wall. Once we picked our favorite ideas, we refined them with real measurements.
With a solid direction, we broke out AutoDesk’s Fusion360 — which is FREE not only for students, but also for hobbyists like myself, and even companies with gross receipts under $100k! It definitely has a learning curve, but its huge feature set and an incredibly powerful workflow easily justify the time investment. A number of the components that we purchased from McMaster-Carr, and some from Grainger, had compatible CAD models that were trivial to import, and obviated the need for us to model them ourselves.
We’ve made our final model public here. Feel free to use it however you wish. Our design involved a ground-anchored kicker supporting the main wall through a continuous-style horizontal hinge. The wall was to be suspended from a large Oak tree, and adjustable via a hand winch. Our chosen location put the wall on slight grade, and required us to vary the length of the ground-set posts. While none of us are professionals, we were confident in our ability to build precisely, and this is reflected in our design. Certain choices would be inappropriate for less experienced builders, like the prevalence of complex cuts or our reliance on a perfectly positioned kicker to avoid twist as the wall is raised and lowered.
The MoonBoard Spec
We have mixed feelings about MoonBoard. Generally we love the idea and are incredibly grateful to Ben Moon and team for making it a reality. However, I do feel compelled to note that it is quite evidently a “first generation” specification. The specification itself has some notable quirks which should nonetheless be followed, such as an additional 0.75in between rows 6/7 and 12/13 (gaps that span plywood sheets).
It also has some quirks that don’t seem to matter, such as the distribution of additional space on the sides. While the specification puts additional space on the right side, the pre-drilled panels sold directly by MoonBoard have the additional space on the left side. We instead chose to center the holes on the board, preserving the specified distances between all holds.
The spec also puts the top row of kicker foot chips very close to the seem. While this isn’t inherently a flaw, it makes it extremely difficult to accommodate adequate support structures while providing clearance for a T-Nut and bolt. We were forced to deviate from the spec here, and moved the kicker bolt holes down by exactly 1 inch. This does change the nature of the routes, and is something we very reluctant to do… but ultimately chose structural soundness over perfect compliance.
Finally, the build guide published by MoonBoard is flawed, calling for a stud directly behind a column of bolt holes. For anyone considering building a wall of any sort, it’s essential to double check bolt clearance before making cuts. By modeling the entire wall in software first, this was fairly trivial for us.
- MoonBoard Build Instructions (HTML)
- MoonBoard Build Guide (PDF)
- MoonBoard Specification: Imperial (PDF), Metric (PDF)
Building The Wall
This is our first time building a climbing wall, and we learned a lot in the process. Our design put the build process somewhere between typical “construction” and more precise “furniture” style work, and we ended up requiring exceptionally straight, true pieces of dimensional lumber. Despite our efforts to pick the best, driest lumber from Home Depot, we ended up with reasonably substantial warping in one fairly visible piece between cut and assembly days. We might have been able to avoid this by paying closer attention to grain structure.
The Frame & Exterior
The frame is constructed primarily of standard Douglas Fir, with pressure-treated pieces used for below-ground and potentially ground-contacting parts of the kicker. The untreated fir has an excellent strength-to-weight ratio and allows us to use normal fasteners (unlike redwood or cedar); the PT (pressure treated) wood resists mold and is essential for pieces that have to handle sustained exposure to moisture.
Our design used complex cuts to provide the desired shape and properties, including rabbets to recess the plywood, dados to make room for the load beam, and 45° rip cuts to provide clearance for T-nuts/bolts.
Our giant 6-foot-long, stainless steel, continuous hinge sits between two symmetrical 4x4s. The main wall is supported by direct compression of the studs between this bottom hinge-attached beam and the suspended 4x4 load beam (which is actually amplified by the orientation of the pulleys to the winch). The load component directed perpendicular to this and directly towards the climber is supported by various Simpson Strong Tie hangers. All framing pieces are attached with some sort of SST connector (mostly A35’s which are incredibly versatile), the appropriate StrongDrive screws (which substantially increase the strength of the connection compared to using nails) and construction adhesive.
We built the face out of 23/32in ACX plywood (meaning one face has an A-grade veneer, the other a C-grade veneer, and the sheet is constructed with eXterior-grade adhesive) and the backing with 1/4in ACX. The climbing surface was primed then and painted with several light coats of BEHR pro outdoor-rated, matte acrylic paint; all other exposed wood was treated with several coats of BEHR premium solid color acrylic-based stain/sealer, which is designed for decks.
Every panel was cut to allow a 1/8in gap for humidity- and temperature-induced expansion and contraction. We installed the panels into the rabbets cut into the exterior framing members, recessing and protecting the plywood edges which are particularly susceptible to water damage. All gaps were then filled with DynaFlex 230 to prevent ingress of water while still allowing movement.
We secured the climbing face to the frame with 2.5in decking screws, every 6 inches around the perimeter and every 12 inches to interior studs. We followed a similar pattern on the back panels with 1.5in decking screws.
Ground Attachment
The kicker is embedded into the ground via 4 posts of varying length, due to the grade of the hill. The ground at our chosen location happened to be primarily sandstone beneath the topsoil. While this made for a lot of work with the rock bar, it provided excellent bearing capacity and prevented us from having to contend with or damage any of the tree’s major roots. With holes at least two feet into solid earth/rock (ignoring loose top soil), we added and aggressively packed a small amount of gravel to the base, assured the kicker was level in both directions, and locked it in place with Quikrete. We chose to keep the holes quite narrow through the rock, and only used 9 50lb bags in total. Note that we live in an area that never freezes and didn’t have to account for a frost line when deciding on the post depth.
We actually installed the wall before attaching the backing, our rationale being that adjustments due to any deviations from our intended measured wall position would be much easier.
Even without the backing and holds, the wall was very heavy and was most easily positioned using jacks and wooden blocks. Once in position and perfectly level, we attached the continuous hinge with 2.5in, #10 stainless steel screws.
Tree Suspension
Our wall is suspended from a massive Oak tree with 1/4in vinyl coated wire rope (containing a 3/16in galvanized steel cable). Two independent anchors distribute the load over three 2x4 blocks into a high saddle, pulling in the direction of the trunk. We spent considerable effort researching best practices for tree attachments, leaning heavily on publications intended for arborist, treehouse, and zipline communities. While the state of the art involves massive, purpose-built tree bolts that allow for adjustment as the tree grows, this strategy adds serious cost and complexity. The available literature also supported the effectiveness of blocking for preventing strangulation due to light (sub-2000lb) and variable loads. This is the strategy we chose, and we are incredibly happy with its results.
Rated triangular maillons and turnbuckles attach the anchors to the wire ropes, which run to pulleys on the upper corners of the wall. Because each line is independent and position-equalized with a triangular geometry, the system inherently resists both twisting of the wall and lateral (side-to-side) displacement. This makes for an impressively solid feeling wall for being suspended.
Each cable and fitting is rated to support the entire load of both the wall and climber should the other one fail, and the continuous hinge is rated to support the kind of torsion such a scenario would produce. However, the system does have a single point of failure in the axle of our split-reel winch. To address this scenario and any other catastrophic failure scenarios of the primary suspension system, we also installed a “backup” system, consisting of a third anchor in a lower saddle, attached to a half-inch stainless U-bolt in the middle of the load beam through an OSHA-compliant shock absorbing lanyard. This is designed to protect roofers and industrial workers from falls, and self-destructively expands (thereby absorbing dangerous shock loads) at 900lbf for up to 4 feet where it limits out with 5000lbf tubular webbing. To hide the bright “safety orange” we sewed a sleeve out of outdoor UV-resistant brown fabric acquired from JoAnne’s Fabric.
The cable is loaded by all-stainless-steel pulley brackets we designed and built for this purpose. Sadly, my drill press, welder, and most other metal working tools were at my brother’s place in a distant part of the state, and were impractical to retrieve due to the Covid pandemic. Accordingly, we designed brackets that required minimal fabrication, and improvised a drill press when needed. The brackets consist of square U-bolts that clamp a 3/8 stainless plate to the load beam. The plate is drilled and tapped in the center to accommodate a rated stainless steel swivel hoist ring made by Crosby, with additional thread provided by a stainless jam nut, recessed about 1/4in into the beam.
I want to call special attention here to the complexity of working with loads that change angles. It is easy to look at ratings for things like eye-bolts and ignore the consequences of off-axis loads. Even shouldered eye bolts lose 70% of their strength when correctly loaded at 45°, and lose far more when the load is across the eye. For our brackets, we chose to use swivel hoist rings, which are designed for angular loads. While these can be extremely expensive, we were fortunate enough to find the exact ones we wanted on ebay at an excellent price.
Also, while rope climbers should already be familiar with this concept, it’s also important to remember that the load on a pulley is double the corresponding load component of each line.
If free body diagrams and force calculations are unfamiliar to you or feel intimidating, Khan Academy has a great introduction. Doing the math may feel tedious, but it’s not hard and can prevent a lethal mistake.
Both lines are actually part of one continuous cable, which is fed through a hole in the divider of our split-reel, worm-drive winch by Dutton-Lainson. The winch is bolted through a custom steel bracket and 4x8 block, which is secured to the frame with a hanger, brackets, and additional blocking.
The cables are routed through side-mounted pulleys from Home Depot, which experience minimal load but direct the cables to the winch for consistent, symmetrical coiling. These were the only components we had to adjust after their initial installation.
Thanks to obsessive measurement throughout the build, we were able to achieve a path of travel that is perfectly level at any angle. Had the kicker been built even just a couple degrees off, this would have been impossible without substantially modifying the suspension design. We did build in some room for fine-tuning by allowing the pulley brackets to be moved along the beam, adjusting the system’s “center”. Fortunately, though, the wall lined up perfectly without any tweaking.
The suspension system uses both galvanized steel and stainless steel components, so we were careful to electrically isolate the different metals. In the presence of an electrolyte (ie wet outdoor conditions), metals of dissimilar electropotential will create a galvanic cell (essentially a battery-style chemical reaction), which will cause rapid corrosion and loss of strength. Avoiding this is as simple as ensuring the presence of a nonconductive material between the metals. In our case, this took the form of a vinyl coating around the wire rope.
Installing Holds
MoonBoard has published several different standard layouts. We chose the 2017 version for a variety of reasons. While the 2016 is generally regarded as superior, it is only specified at a 40° incline. Also, the MoonBoard app no longer accepts new submissions of routes for the 2016 layout. (EDIT: this is NOT actually true, despite their 3-year-old post.)
While the 2019 layout is available at both angles, it is heavily reliant on their plywood holds. Unlike their plastic holds or other manufactures’ solid-wood holds, these are extremely susceptible to deterioration from weather and totally unsuitable for an outdoor board.
We started with the original school holds, school holds set A, and school holds set B, and are planning to add school holds set C shortly.
We had some issues with MoonBoard’s shipment of bolts and T-nuts, and ended up sourcing steel stainless bolts from Bolt Depot and screw-on T nuts from Rock Candy. This style of T-nut is far superior to the hammer-in style at resisting pop-out, and when access to the back is difficult, these are essential. Do note that this combination is the only case of mixed metals in our system, and was a reluctant decision mainly due to the difficulty of finding reasonably-priced stainless steel T-nuts in the screw-on style. However, given that the T-nuts do not extend all the way to the front of the plywood, which is always facing down, the opportunity for them to become wet is very limited, and so the risk of corrosion is acceptable.
EDIT: After several months climbing the 2017 version, we began exhausting the catalog of climbs that didn’t require the wood set or red holds set C. Instead of purchasing the red set, we took the recommendation of others and just re-set to the 2016 layout using all the holds we already had. This proved to be an excellent call, as the layout is far superior, and despite the post by MoonBoard, the 2016 layout is still fully functional in the app.
Padding
Our padding currently consists of two Metolius crash pads and one Tumbl Trak we found on Craigslist for an excellent price. The latter is very low-density, and far too soft to use by itself. However, when underneath the high-density Metolius pad, it makes for an exceptionally well-balanced system. It’s stable when walking around yet remarkably supple when absorbing a fall; the combination is completely confidence inspiring when making moves up high.
All that said, our current padding still provides insufficient total coverage especially with the wall at 40° (where there is more area to protect). We’re keeping out eyes out for another good deal on additional padding.
Notes
Time & Cost
Building this wall was neither trivial nor cheap; while we’re still tabulating the exact costs (our Home Depot runs included purchases for other projects) it’s roughly on the order of three thousand dollars including holds, materials, padding, components, fasteners, chemicals, etc. We were fortunate enough to already have most of the big tools, but did have to purchase some additional smaller ones which are not factored into that number.
Despite the quarantine, my fully-committed work life prevented any real progress during the weeks, and so this project was spread out over a total of 9 partially-available weekends. COVID-related supply chain delays and long lines at hardware stores also added to the total required time.
Final Thoughts
While certainly frustrating at points, this was overall an incredibly fun and gratifying project. It required us to learn some new skills and allowed us to exercise some of our existing ones. This was the most ambitious design and build project we’ve done since the Boltline van, and the result has been everything we hoped it would be. The last-minute addition of bistro-style lights has enabled delightful sessions in the cooler evening temps, and a simple laser pointer has made it trivial to following routes without the LED system.
With COVID numbers making a resurgence in the US, we are especially grateful to have this in our yard. Emily and I have been actively sport and trad climbing for about a decade, largely outdoors. Even during seasons of gym climbing we generally stuck to the lead walls, and actively avoided any sort of “training.” We’re just now discovering a real appreciation for this kind of climbing, and are excited for all the fun in the months ahead, maintaining and even improving our strength and muscle memory!
