Back ...

By September 1998 I was educated that Newfoundland was a faraway island and that it was going to take some money to put this project into serious motion. At Les Hamilton’s and John Chirtea’s urging, the Society for Technical Aeromodel Research (STAR) was formed. A pro tem board of mostly DCRC members began twisting friends’ arms, asking for contributions to the effort in exchange for membership, which promised only an occasional newsletter. We were off and running.

Paul Howey and Ted Rollins designed and built the front end of an autopilot; i.e., gyro, wing leveler, roll stabilizer, and servo controller coupled to a Futaba receiver. Joe Foster came on board in the summer of 1998 and started on the massive job of software for GPS navigation and altitude hold.

During the winter and spring of 1999-2000, we flight-tested in a horse pasture on a farm that Beecher Butts owned. He was 88 years old and a legend in the area. He earned his pilot’s license in 1946, and at 88 he was still flying his ultralight and riding horses. When riding to the hounds, he jumped fences with people one-third his age.

If the grass got too tall for us, he’d knock out a runway with his farm tractor.

If the grass got too tall for us, he’d knock out a runway with his farm tractor. The combination of our admiration for Beecher and the need to rise above the technical struggles we were having in these early days (crashes!) led me to suggest naming our transatlantic model “The Spirit of Butts’ Farm.”

In April 2001 we were fairly confident that the few remaining problems could be solved by August, so we bought some cheap (therefore, nonrefundable) airline tickets—still a great expense for our treasury. We asked Andy Gutow to make us some crates in which to transport the models.

All sorts of problems cropped up in May and June 2001. Facing the situation square on, we had to postpone our attempt until August 2002. August is the best, and nearly only, month of suitable north-Atlantic weather.

Making use of those cheap tickets, John Patton, Roy Day, Joe Foster, and I flew up to Newfoundland to look at the terrain and meet Carl Layden: the Atlantic province director for the Model Aeronautic Association of Canada (MAAC). Carl had volunteered to be the Canadian FAI observer for our record attempt. He had told Saint John’s newspaper, The Telegraph, about our expected visit and its purpose, and a pleasant story was published on the day of our arrival.

Bingo! Nelson Sherren, a former RAF Lancaster-bomber pilot, read the article and called Carl to ask us to meet with him because he wanted to help. Nelson is a Newfoundland/Labrador aviation-history buff. He has retrieved parts of busted or sunken WW II failures and owns a spare wing rib of the Vickers Vimy biplane that Captain John Alcock and Lieutenant Arthur Whitten Brown flew for the first nonstop crossing of the Atlantic in 1919.

In 2001 Nelson was president of the 150th wing of the North Atlantic Royal Canadian Air Force Association. He told us he would arrange for low-cost housing on a military base when we came back in 2002 and that he could also provide a large workspace with telephone service at the association’s clubhouse on the base.

His word was good; these things were available to us in 2002 and 2003. What a blessing! The arrangements were top-notch, especially since we had envisioned ourselves working out of motel rooms. We spread out on eight workbenches in a 30 x 40-foot room with three computers, a drill press, an oscilloscope, and plenty of folding chairs.

During the winter and spring of 1999-2000 I built five airplanes with 6-inch-diameter fuel tanks. When we got around to measuring speeds, I decided that these models were too fat. They flew only roughly 38 mph when pulled by the allowable horsepower and fuel rate.

During the winter of 2000-2001 I built six airplanes with skinnier fuselages, with 3⁄4 of the fuel in the fuselage tank and 1⁄4 of the fuel in a wing tank. The plumbing and fuel-pressure system was complicated. Putting precise amounts of fuel onboard was a nasty challenge. These models flew approximately 43 mph at full weight, but three were crashed during tests in the spring of 2001 and two more were totaled after our 2001 postponement.

During the winter and spring of 2001-2002 I built four more airframes. By this time I had constructed 21 fuselages and 12 wings. Hundreds of hours of engine testing filled in the spots between construction.

July 26, 2002, my wife Gay and I started a six-day journey to Newfoundland in a rented Dodge Caravan. I don’t have a driver’s license because of my degenerated vision, so Gay, who loves to drive, took the wheel while I had a much-needed six-day rest. Preparations had approached a frenzy during the final weeks of testing. We had four all-up models packed in the van along with boxes and bundles of auxiliary stuff.

We made the first record attempt with the serial-number 19 model, identified as TAM (Transatlantic Model) 1 on our Web site (http://tam.plannet21.com). The launch was at 8 p.m. local time on August 8. The evening launch was made so that the model would arrive in Ireland during daylight hours. Minimum crossing time would be as short as 28-30 hours if there was a brisk tailwind. That would put it on the Irish coast in the dark. The maximum time had to be less than 40 hours. That was the maximum expected fuel duration.

Joe Foster manually flew TAM 1 to roughly 1,000 feet using a Futaba PCM transmitter. When he toggled the landing-gear switch to put the model on autonomous flight control, it started making lazy circles approximately 300 feet in diameter. The wind blew the circling model toward Ireland, but it soon fell into the ocean.

The cause may have been that the model was too far out of trim for the gyro and steering software to grab hold. We tried for a test-and-trim flight early in the afternoon, but high winds led to a rough forced landing. I think a gear was broken when the aileron snagged in tall grass. We’ll never know.

TAM 2 was launched two days later after a check flight. This time all looked good. The model made the half-mile leg to the north that Joe had programmed, to avoid flying over a Cape Spear visitors’ parking lot. We were on a gravel road roughly 1⁄4 mile west of the lot.

After passing the 1⁄2-mile point, TAM 2 took off on a beeline, straight as an arrow, but on a heading toward the Azores islands—not toward Ireland. Telemetry indicated that the engine stopped and the model dropped into the sea 171⁄2 minutes after launch.

We were doing poorly! I had some ideas about why the engine had stopped, but there was no way to check my ideas with the engine on the ocean floor. The heading error was something else; there was a flaw in the software that was undetected during our two years of flight-testing the steering functions. The reason it escaped detection was that we had never turned the model loose to fly long distances. We flew small ovals and short hops up and down a road using a convertible to keep the model in the pilot’s sight.

We were not irresponsible nor were we reckless.

Any modeler who intentionally programs an autopiloted model to fly out of sight over populated land is reckless and irresponsible. The potential harm done to the hobby would greatly outweigh any inflated ego. We were not irresponsible nor were we reckless. We aimed to fly over the ocean at approximately 1,000 feet of altitude. There would be no swimmers underneath, no airplanes flying that low, and no ships with masts that tall. “Safe” is the word that describes what we were doing.

Weather over the north Atlantic had not been good up to this point, but on August 13, 2002, it was predicted to be bad for the next four to five days. This was partly okay because Joe Foster and Les Hamilton had to recalculate, simulate, and insert new pieces of code from top to bottom; that was 100 pages and 10,000 lines! They were at it for three long days.

I took advantage of this break to retreat to a table in a remote corner of an adjacent room. In spurts I wrote a eulogy which I hoped would be read at a memorial service for Walt Good; he had died a month earlier, and his service was to be in Florida. I yearned to be there, but I was trapped in Newfoundland. Tears of grief dropped on my writing pad. Walt was a good friend and a fun scientist. He surely would have enjoyed participating in this technical miracle we were seeking.

Nelson Sherren had meteorologist friends in Gander and Goose Bay who faxed, on a daily basis, current conditions and 24-hour forecasts of wind and frontal systems over the Atlantic. Roy Day pulled predictions from the US Navy Norfolk Web site. Bob Yount in Maryland sent E-mails with recommendations. The situation was no-go August 15-17.

Paul Howey, who was to be the landing pilot in Ireland, was telling us via E-mail and cell phone that it was raining buckets most of the time. What’s worse, the rain was blown horizontal by winds that, at midnight on one of those days, tore down the tent in which he and his son Rusty were trying to sleep. They retreated to a bed-and-breakfast.

Nelson’s 9 a.m. weather briefing on August 18 informed us that things were far from ideal but not totally negative. We decided to give it a try. Time was running out; cheap tickets home were for August 22. We drove to Cape Spear under clear skies, only to see a dense cloud of fog over the launch site as we rounded the last bend in the road. We stood in that fog until 8 p.m. and saw no hope for clearing.

We returned at roughly 5 p.m. the next day, set up TAM 3, and launched it at 6 p.m. to beat that fog. This time TAM 3 took off on the 62° track it was supposed to follow and flew on course for the next eight hours.

The last report from the satellites used for tracking stated that it had gone 479 miles before dropping out of sight. The most likely cause of failure this time was that the model hit a rainstorm and severe turbulence. We joked that we could talk like military public-relations people who would call this “a successful test that showed the system to be working properly.”

We still had TAM 4 in the shed, but we had run out of time. We packed our goods and started home on August 22, 2002.

Continue ...