This was an expedition truck design and build project that I completed over an 18 month span with my wife Yvonne. The goal of the project was to build and travel in a custom four-wheel drive expedition truck.
We built the vehicle from a 1987 Mercedes-Benz 435 military truck. In addition to refurbishing and modifying the truck chassis and drive-train ourselves, we personally designed and built the habitat living module and systems specifically for this vehicle.
The chassis is large enough to accommodate a 75 square foot living module with a total vehicle weight of up to 16,500 lb. The habitat module has a bathroom, shower, kitchen, eating area, permanent bed, refrigerator, heating, full solar charging system, and a pass-through into the refurbished truck cab. The vehicle is capable of remote operation for several weeks without re-supply.
Trip at a Glance
Number of Days – 133 (19 nights above Arctic Circle)
Total number of Canada/US border crossings – 10
Number of ferries/barges – 13
Miles on Pavement – 9,478.3
Miles on Dirt/Mud/Gravel – 2,521
Miles on Ferries – 618.7
Total Miles – 12,618
Up to date information on the project can be found on our blog and Instagram:
This project was particularly fun because I did it jointly with my wife Yvonne. The goal of the project was to refurbish a M1078 LMTV for civilian use as a chassis for an expedition truck. The chassis is large enough to accommodate a 110 square foot living module with a total vehicle weight of up to 26,000 lb. A completed expedition truck based on a M1078 can be capable of remote operation for several weeks without supplies, can be US registered as an RV, and drive-able with a normal Class C driver’s license.
The M1078 is a member of the basic LMTV utility truck family in the US military. The family includes 4×4 and 6×6 chassis for a wide variety of uses. We acquired the base 1996 4×4 M1078 truck from government surplus.
This project took us a little over a year and required refurbishment, painting, power coating, and replacement of many vehicle parts as well as the fabrication of many new custom items.
I became interested in long distance sailing after reading the experiences and life philosophies of Bernard Moitessier and Joshua Slocum. The simple efficiency and economy of traveling the world by wind is appealing. Proas are a type of traditional South Pacific two hulled outrigger sailboat. From a design vantage, I’m interest in modernizing ideas while respecting heritage. Proa are asymmetric with one large hull (aka) on one side and a small hull (ama) on the other. On a Pacific Proa the smaller ama is always kept to the windward side of the boat. The ama acts as a counterweight and provides righting moment against sail load subsequently negating the need for a heavy keel. Proa are therefore not tacked like a normal sailboat; they are “shunted” keeping the small ama to the wind. This means that the bow and stern of the boat are identical, and the direction that the boat is traveling changes with each shunt. The absence of a keel and having the small ama lifted high in the water makes Proa have unique sailing performance and potentially very high long distance cruising speed.
I sized this Proa for a couple to comfortably live full time while long distance worldwide bluewater cruising. The aka accommodates a full sized galley, head with shower, and large storage areas. There is ample room for a normal queen sized bed in the overhanging lee pod.
Before I started this project, I received flight training in several different gyroplanes to better understand the differences in handling and performance. Of all the aircraft that I have ever flown, gyro flying has by far been the most enjoyable flying experience. The intent of this project was to build a one-off, stable/docile, open, two place, gyro with good visibility for fun local flying and short cross country flights. I started my configuration study by benchmarking my basic layout against 15 other gyros for comparison.
During the design process, I checked 82 different weight and balance scenarios and iteratively refined the configuration to establish as little CG movement as reasonably possible. The structure was monolithic carbon fiber/epoxy/Divinycell with localized fiberglass and Kevlar buildups. There were also localized carbon unidirectional tow reinforcements. The horizontal stabilizer was epoxy/fiberlass/polystyrene with unidirectional pultruded carbon fiber spar caps. The fuselage tooling method was a seamless composite shell molded over a removable CNC milled foam male tool.
The vertical and horizontal tail volumes compared well against other benchmarked tractor gyros. The vertical stabilizer and rudder were positioned in the cleanest air available on the underside of the fuselage. The intent was to maximize yaw stability with the nose raised at high power and minimize spiraling-slipstream turn/roll tendencies.
I compared 44 different engines for this application in the 85-150 hp range, and I decided to go with the Rotec Radial R2800 swinging a 76×57 Culver wood prop. The overall gyro design was based around the R2800 from the beginning. The rotor was intended to be a RFD 28 ft aluminum rotor with RFD double-bearing rotor head.
The gyro is now being finished by a gyro enthusiast in Oregon.
This project was the design of the spring steel landing gear for the tandem tractor gyroplane that I was concurrently developing at the time. The main landing gear were modified Cessna 140 landing gear legs.
I used an iterative Finite Element Analysis (FEA) method to tailor the leg spring stiffness to produce the desired deflection at max landing weight and expected landing descent rates. I milled the width of the Cessna 140 gear legs down to the specific analyzed dimensions. Also, I canted the legs out for a wider stance while still producing the same root bending moment of the heavier C140. This wider stance required special axle mounting blocks to set the proper axle angle. I analytically designed the axle block angle to produce level axles with the aircraft sitting static at a nominal weight. The gear was sized per configuration applicable FAR23 load cases. The rolling stock was FAA/TSO Parker Hannifin 500-5.00 with single puck differential hydraulic brakes.
This project was a complete restoration that I did on a 1943 Ford GPW WWII army jeep. This particular GPW was originally used by the US Army Corps of Engineers in Los Alamos, New Mexico during WWII on the Manhattan Project. This chassis was delivered to the government on June 9th 1943.
The military vehicle designation for the WWII jeep is technically “G503”. The G503 is the original ancestor of the modern-day Jeep. The design started its history in 1940 under development by American Bantam Car Company. For mass production reasons, the final production version of the G503 was produced by both Willys-Overland and the Ford Motor Company during the war. There were slight variations between the Willys and Ford chassis, but the vehicles are for the most part identical. Willys retained the rights to the design after the war, so most people know these vehicles as “Willys jeeps”, but there were several companies involved in the wartime development and production.
I took every part down to clean metal and repainted with period correct paints. One of the most fun parts of the project was researching all the historical details. When I got the vehicle, it was in relatively good shape, but it was 70 years old and needed complete restoration. I repaired all structure and body damage using original materials. No body filler was used. I rebuilt or replaced every system component and returned the electrical system to its original 6 volts. The project took almost a year and a half of diligent work, but overall it was very satisfying to make every part as clean as new and historically correct. I owned and drove the GPW for a year after I completed the restoration and later sold it to a collector in Texas.
I wouldn’t say that I’m really into astronomy, but I do like the idea of the infinite vastness of space. I have wanted to try astrophotography for some time, but my main problem with astrophotography was that the cost of a telescope capable of doing what I wanted was outside my budget for a short term photography project.
My solution was to design and building a very simple all-mechanical camera mount that would passively rotate at the same angular speed that the earth rotates. I knew I couldn’t make anything that could track accurately for hours, but I thought that I could probably make this tracking mount accurate enough to take 30-60 second exposures without noticeable motion blur. The tracker runs on a set of nice ball bearings and is completely mechanical. It uses a simple homemade gearbox, a curved threaded lead screw, and a normal clock motor to move the camera at the same rate that the earth rotates. When the tracker rotational axis is properly aligned with the North Star, the tracking error is unnoticeable for exposures up to about 60 seconds.
My homemade sky tracking gizmo.
This is the Andromeda Galaxy. At 2.5 million light-years away from the earth, Andromeda is one of the closest galaxies to the Earth. Andromeda is actually fairly large in the night sky, but it is very dim. I took about 840 images with a Nikon 300mm, f2.8 manual focus lens for a combined exposure time of about 7 hours. There are actually three separate galaxies in this image. The small light smudge below Andromeda is a small galaxy called M110. M110 is 2.9 million light-years away from earth. The fuzzy looking “star” on the upper edge of Andromeda is another small galaxy called M32. M32 is 2.65 million light-years away from earth. All the smaller stars in the image are stars in the Milky Way between Earth and Andromeda. These stars are much closer to us and are probably only 50-70 thousand light-years away from Earth.
This is the Orion Nebula. It is about 1,340 light years away from earth. You can faintly see the Orion Nebula as the center “star” in Orion’s “sword”. This image is about a 7 hour exposure made by stacking 30 second sub-exposures taken with a Nikon 600mm, f4 manual focus lens.
I solved the long exposure image noise problem by simply avoiding the problem altogether. I take 30-60 second exposures one-after-another for 7-8 hours. This effectively gives me ~6 hours of gathered light information in something like 800+ separate image files. I can take all the “sub-exposures” and “stack” them to get a final image with an equivalent exposure time of many hours. By dividing up the exposure this way, it is fairly easy to line up all the stars in the image from one frame to the next, effectively eliminating any tracker error. This technique also subsequently improves the image signal to image noise ratio of the final image. This “image stacking” method is by no means new or original. This is the same basic technique that NASA uses for processing Hubble images.