A Little Bit of Boeing History: or How to Do a Barrel Roll in November 707

By Andrew Herd (15 July 2005)

August 7th 1955 was the day of the Gold Cup race for the unlimited hydroplanes on Lake Washington, Seattle, a big event in the city, which at the time could claim to be the boating capital of the US. To celebrate this final event in the Seafair, there was going to be a display by the Navy's Blue Angels, but everyone knew that something else was going down - Boeing had been trailing their coat all week about how they were going to demonstrate their new jet airliner, the four engined 367-80. The demonstration was to take place between the second and third heats of the Gold Cup and given that the plane was a make or break design for the company, word on the street was the show was going to be an impressive one.

As the time neared, there was a general drift of people to good viewpoints along the lakeside. The more air-minded members of the public had already chosen the best spots on rooftops with a good view of the race course and the floating bridge, landmarks which were likely to mark the flight path of the new jet liner. There was an almost vacation-like atmosphere, this being before the days when a new plane could only debut at a major airshow, and Boeing had been known for showing its appreciation for the support of the people of Seattle in the past.

As the second heat of the race drew to a close, even the powerboat fans were looking up. It was a Sunday, so Boeing's workers were out in the sun, watching the Angels' usual great display and waiting to see what Tex Johnston, their chief test pilot was going to do with the new baby. The minutes seemed to drag and the question on everyone's lips was, “Where's Tex?” Then they saw him. Racing over the floating bridge, apparently at mast height, was a big yellow and brown jet airliner doing over four hundred miles an hour and pointing right at the race course. Then the left wing began do dip and everyone held their breath, because there had been accidents before and every Boeing worker on the lake knew this could be another one. “The Dash 80's going down!” But the right wing kept on rising and the plane kept on rolling, over and over, past inverted, until it was back on the straight and level, still going flat out and close enough to the water that some of the people on the rooftops were looking down on it.

Then, just in case anyone had missed it, Tex went round and did it again.

Boeing's president, Bill Allen, was on his yacht in a marina near the race course along with a clutch of potential customers and suppliers he wanted to see the new product and the story is that he had no idea that Tex was going to do anything other than an ordinary flypast. It is also said that Allen had seen one of his clients take a heart pill earlier in the morning and said, 'Larry, give me one of those pills, I need it more than you!' when he saw Tex fly past upside down.

Next morning Tex was invited to an emergency board meeting of the Boeing Company; as usual, he was late, and as usual, Allen got straight to the point, "Why did you do that?" he asked, trying to stop his voice rising a notch. “Selling the airplane” replied Johnston, as laid back as he usually was. “The airplane does not recognize attitude, providing a maneuver is conducted at one G.”

There was a long pause, and then the corners of Allen's mouth twitched very slightly before he spat out the most famous line ever delivered by a Boeing president, “You know that. Now we know that. Don't do it any more." But Allen can't have been that pissed at Tex, because he invited him out to dinner that evening, bringing along Eddie Rickenbacker as a guest. The 367-80's registration was November 70700 and they say that is where Boeing got the idea for the model number series. I guess the board didn't find it an easy number to forget.

The Captain Sim 707 is one of the most outstanding airliner simulations ever released for Flight Simulator and has always been a firm favorite of mine, partly because it exudes character in a way that modern glass cockpit jets don't, but also because it is a damned fine piece of code, with a wonderful panel, great flight dynamics and a selection of classic liveries. I hadn't visualised it as an aerobatic steed until Allan Kallohan released a Dash 80 livery for it (B367-80.ZIP), but once I had installed the repaint, the idea of doing a barrel roll in it became irresistible, particularly because we are lucky enough to have a very fine photographic scenery of the Seattle area, in the form of PC Aviator's Pacific Northwest MegaScenery. It was while I was driving home from work that it suddenly occurred to me that August the 7th '55 was fifty years ago, so what better way of celebrating the anniversary than by recreating Tex Johnston's famous exploit?

I weighed up the pros and cons. On the negative side, the Captain Sim addon simulates late variants of the 707, but against that, these had more thrust, no bad thing. Then again, the 707 was never known for having a high rate of roll, but against that, you don't need a fast aileron response to fly a barrel roll if you get it right. A third potential nail in the coffin is that Tex flew his maneuver at a height of less than a thousand feet, which leaves no margin for error; and finally, he was a really good pilot and I am not.

But what the hell. How much can you hurt yourself in a simulator?

So let's examine the problem.

We have here one simulated airliner, powered by four 18000 pound static thrust P&W JT3D-3 turbofans. It has no lack of speed and a comforting initial rate of climb of over 2000 feet per minute. It does not have any airbags fitted and the cabin crew threw up the last time we attempted a hammerhead, but on the other hand, the plane has low mileage and only one (reasonably careful) owner since new. We also have a seriously nice scenery of the Seattle and the surrounding area, so it shouldn't be a problem aligning ourselves for the run over Lake Washington, and if it all goes pear-shaped, there is always the reset button.

I took time out to read what eyewitness accounts of the event I could find and came to the conclusion that people's memories were unreliable after the event, which as any police officer will wearily affirm, is more or less what you would expect. There are some frustrating hints that a video of the flight is available on the net, but if it is I haven't been able to track the thing down - if someone does come across it, the Cineplex is ready and waiting! However, Tex Johnston does give a pretty good account of what happened in his autiobiography 'Jet Age Test Pilot'.

Tex says he began the maneuver at 490 mph at a height of 200 feet, before pulling up into a 35 degree climb, releasing the back pressure and pushing the stick all the way over to the left. Tex might well have been a showman, but he was also a very smart guy and there is no way that he would have conducted a maneuver at a speed that might have exceeded any of the plane's limits, particularly at an altitude which left no room to recover. The Vno of the plane was around 375 knots, or 430 mph; Vno being the maximum structural cruising, or maximum normal operating speed, a limit that normally should be observed. But the Lake Washington flight wasn't a normal event and by the time he flew the Dash 80 on that famous pass over Lake Washington, Tex would have known everything about that plane and the curves of the engineers' V-n diagram would have been etched in his memory, as the more stressful parts of his test flight program would have been dedicated to proving that their calculations were right.

Just for interest, I have included a sample V-n diagram to show the sort of calculations that would have entered into Tex's mind. He would have wanted the Dash 80 as light as possible for the flight, so I have put in some ballpark airspeeds based on this assumption. If you click on the graph and load the larger version in a second window, it will make it easier to see how the thing works - the vertical axis is G, the horizontal airspeed in knots. Go to the zero G point where the two axes meet - the curve going up and away to the right represents the point at which the aircraft will stall at a given G-loading: so for example, if you draw a line from 1G on the vertical axis to this curve, then drop a vertical, it meets the airspeed axis at 115 knots. Given that the stalling speed of any wing rises with the square root of the G-force applied, at 2.75 G, the stalling speed has risen to 191 knots with the wings level - any bank increasing the stalling speed quite considerably, a good reason for not flying any more than a 1 G maneuver. In the absence of any more detailed information, when I did the diagram, I put in two structural failure lines at 2.75 G positive and 1 G negative - the other boundary of the flight envelope being the never exceed speed, or Vne, which given a Vno of 375 knots, was probably around 450 knots, or around 520 mph at sea level. 490 mph gave Tex a margin for error.

In our made-up graph, the 367-80 will keep flying as long as the airspeed and the amount of G pulled keep the plot within the white area. On the limits I drew in, if 2.75 or -1.0 G is exceeded, or the plane exceeds the Vne in level flight, there is a very high chance of something breaking. In other words, it would be possible to pull 2.0 G at 380 knots and get away with it, but the plane would stall if the pilot pulls 2.0 G at 120 knots - easy as that. In practice, the Dash 80's upper G limit was at least 3.5 G, but the graph is there to give a general idea of how these things work. A fully aerobatic plane would have much wider G limits and the stall curves would also be steeper, giving more freedom to maneuver, but we are talking an airliner here and sustained positive or negative G capacity would have been outside the designers' brief - but even given a Vn diagram like the one above, it is possible to fly a 1 G maneuver as long as the airspeed limits are observed - which is why Tex chose a barrel roll. An aileron roll would have been tricky given the slow rate of roll of the 707 and a loop would have put him too close to the positive and negative load factors for comfort; but expertly done, a barrel roll pulls neutral G all the way around and a good pilot can balance a glass of water on the glareshield and not even see the meniscus tilt. So a barrel roll it had to be.

What is a barrel roll? I guess the first thing to say is that it is an aerobatic figure which is easier to describe than to fly well, because you need coordination and good spatial awareness to fly one really smoothly. The easiest way of describing the maneuver is to imagine you have a very large old-fashioned wooden barrel ahead of you in the sky and a mission to fly a spiral path around it from one end to the other with the cockpit always pointing towards the center line of the barrel. Another way of imagining what a barrel roll looks like is to think of a single coil of a spring - stretch it out a little and that is that path you are aiming to fly. For what it is worth, an aileron roll is done by pushing the stick over left or right, so the plane rolls around its own axis (an initial pull up is usually required to prevent the plane exiting the figure with the nose below the horizon); and a flick, or snap roll, which uses the secondary effects of yaw, can be executed in level flight in some aerobatic planes simply by treading hard on one of the rudder pedals. Call me Mr. Cautious, but I doubt that doing a flick even entered Tex's wildest imaginings.

So the next question is: what kind of barrel roll did Tex do? Barrel rolls can be short and fat, or long and thin, or any flavor in between. Get out the spring again and stretch it out a very small amount - a single coil represents a short fat barrel roll. Really pull the ends of the spring apart and you can see what a long thin barrel roll looks like. Fat barrels can be thought of as stretched out loops; they need you to pull more G to get the plane around, whereas thin barrels are better for planes which lack pizazz in the pitch department, but have plenty of speed in hand, which is what we have here. Tex undoubtedly went for something nearer to the long thin end, which is why so many people who saw him think he did an ordinary aileron roll - particularly because with his experience, he could have flown the maneuver with hardly any altitude loss at all, which is just as well, because he didn't have any to lose.

So let's do it (-:

Getting the 707 set up right is a vital part of completing the maneuver. The flight took place from Boeing's own airfield in Seattle (KBFI), which has Lake Washington on the downwind leg from 31L, assuming you extend your upwind leg down towards the northwestern end of the lake. This allows you the luxury of flying with a minimum of fuel, I loaded the fuselage tanks only - and I also used Captain Sim's 707TCE applet to set the passenger load to zero, on the basis that the fewer witneses I had to my first attempts, the better. Once I the was plane set up, I dropped a click of flap and advanced the throttles, enjoying the short run the 707 needed before it lifted off; this being a plane which normally eats pavement as if it is going out of fashion. I climbed to 2500, reduced power to cruise setting and enjoyed the flight out over Lake Union, taking occasional glances out of the right hand window until I could see the northern arm of Lake Washington where it bends around the knuckle of Kirkland Magnuson Park (all of this is clearly visible in the excellent PNW MegaScenery - if you don't have this, just fly three quarters of the way down the lake which runs northwards on the right of Boeing's field and position yourself to fly back down the length of it). Then I commenced a descending rate one turn, adjusting the throttles so the airspeed slowly built up, until I rolled level over the water with 300 knots or so on the clock. As the first bridge came into sight, I advanced the throttles all the way to the gate, so that as the bridge vanished under the nose, the airspeed indicator needle was red lining... and then I pulled the nose up and progressively fed in full aileron so she went round.

You may need to practice this a few times to get it right - if you are new to aerobatics in FS, try reading this. Do note that if you don't get the roll balanced exactly right, you will need to push the stick forward some to avoid the nose dropping as you pass inverted; if this happens, the 707's rate of roll is so slow that there is no way to avoid diving into the lake... The video was taken on my second attempt to do a roll over the water. The first time, I began the maneuver too high, didn't pull up enough to begin with and rolled level at what would have been a hundred feet had I started at a historically accurate altitude. Sweaty stuff, if had it had been real. With a bit of practice, I found I could do an almost perfect barrel roll, except that the 707 ended it several hundred feet higher than it began; which is moderately realistic as Tex actually climbed the plane during the roll, finishing the maneuver at 1500 feet, but then he had more than his pride at stake!

Now you watch the video and can see if you can do it any better (-:

Andrew Herd
andy@flightsim.com