Authors: John Harding, John K. Pollard, Leonore Katz-Rhoads, Peter Mengert, Robert Disarlo, E. Donald Sussman

Date of Publication:  March 1999

Sponsoring Agency:  U.S. Department of Transportation - Federal Railroad Administration

Performing Organization:  USDOT Volpe National Transportation Systems Center High Speed Guided Ground Transportation Safety Task Force

Report No:

Abstract:

Proposed high-speed ground transportation systems, such as Maglev, may have motion
characteristics affecting passenger comfort that set them apart from anything previously
experienced. Operating at aircraft speeds along rights-of-way established for conventional ground
vehicles, these systems may subject passengers to significantly larger vertical accelerations and roll
rates than they have ever felt on existing common-carrier modes. If the design limits for guideway
curvature are set too high in the interest of achieving the shortest travel times and/or maximum
utilization of existing, short-radius right-of-way, substantial numbers of passengers may find the
ride quality unacceptable because of excessive vertical acceleration and roll rates. In that case,
speed would be reduced, resulting in moderately longer trip times. In areas where new right-of-way
is unavailable, the question becomes how can a Maglev or other high-speed-system guideway be
optimally fitted to it and what speeds should be used.

Previous research carried out by the Volpe National Transportation Systems Center for the National
Maglev Initiative demonstrated that more than 95% of the public would accept isolated Maglev
maneuvers involving bank angles up to 37 degrees and roll rates up to 7 degrees/sec. Since these
limits were higher than those contemplated in most Maglev-system-design proposals, passenger
acceptance did not appear to impose any significant constraints. However, further reflection on
motion sickness as experienced in other modes suggests that the frequency of occurrence of
motions, as well as their power spectra, are as important as their magnitude and that the view out
the window may strongly influence the passenger’s likelihood of becoming ill. Hence this study
was undertaken to explore comfort and motion-sickness effects of Maglev travel in corridors
characterized by frequent curves.
Four segments of the New York State Thruway, totaling 277 km (172 miles), were chosen as the
hypothetical route for evaluating passenger acceptance for the following reasons:

• These segments are representative of a great deal of the hilly terrain found in the United States.
• Their length of 277 km (172 miles) is typical of the distance between several major city pairs which would be good candidates for Maglev service.
• The State of New York was willing to supply detailed maps containing the required data to construct the hypothetical route.
• The State of New York provided significant financial support to the construction of the simulator used for part of this study.

To facilitate both the experimental design process and subsequent data analysis, a procedure was
developed for estimating the propensity of a given set of ride motions to induce motion sickness.
This procedure is based upon the work of M. J. Griffin and British Standard 6841:1987 for ride
quality. It generates a number called the Motion Sickness Dosage Value (MSDV), from which the
proportion of passengers who will experience nausea can be estimated. The model predicts the
incidence of kinetosis from the magnitude and duration of exposure to low-frequency (0.1 - 0.5 Hz)
vertical accelerations. For the hypothetical route, 27 alternative sets of design limits for bank angle,
roll rate and longitudinal acceleration and deceleration were initially considered, which had MSDV
scores ranging from less than 2 to 13. British Standard 6841 provides an approximate method for
convenient interpretation of these figures. In a “mixed population of unadapted male and female
adults” BS 6841 gives the estimate:

Percentage of persons who may vomit = 1/3 * MSDV.

Also, the scores may be used for comparative purposes; motions leading to high MSDV scores may
be expected to produce more motion sickness than motions leading to low scores.
The only means of simulating trips with realistic accelerations at reasonable cost is through the use
of an airplane. In turning, aircraft naturally bank at just the right angle to eliminate lateral forces on
the passenger, just as a Maglev would. Conventional ground vehicles would produce unpleasant
and unrealistic lateral accelerations in rounding turns at high speeds, since they are restricted to low
amounts of super-elevation and generally lack tilt-body suspensions. The principal disadvantage of
using an airplane as a simulator is that it cannot provide a realistic out-the-window view a future
Maglev passenger would see. Only a laboratory simulator can safely expose passengers to the
visual effects of scenery rushing by at 400 kilometers per hour (about 250 miles per hour) at ground
level. The laboratory simulator can also add realistic amounts of vibration.
To provide facilities for testing subjects in both the airliner and laboratory simulations, a contract
was awarded to Grumman Aerospace Inc. (now Northrop Grumman Corp.). This contract
supported the development of computer-generated-imagery of the New York State Thruway right-of-
way, use of the simulator and staff for testing subjects and use of a 21-seat Gulfstream I and
crew for flight experiments. Due to the merger with Northrop and the ensuing downsizing of the
corporate fleet, a Beechcraft 1900C replaced this aircraft.

An experimental apparatus was constructed to facilitate flying an airliner through a series of several
dozen roll maneuvers which would subject passengers to the same vertical accelerations and roll
rates they would experience in a Maglev built to a given set of design standards. This apparatus
was based upon two notebook computers linked to a roll-rate gyro and a three-axes accelerometer.
It generated a cockpit display showing what the aircraft’s bank angle was supposed to be at any
given time, what its actual bank angle was, and the direction of the next maneuver. The pilot’s job
was simply to keep the two bars on the display parallel. The apparatus also recorded the outputs of
the accelerometers and rate gyro at 0.1-second intervals, thus allowing MSDV and other measures
of ride quality to be calculated.

After training the crew to fly the experimental procedures and securing use of restricted airspace,
two preliminary tests were conducted using government and contractor personnel as subjects.
These tests exposed subjects to two intervals of flying with relatively high bank angle limits,
consistent with making the 277-km (172-mile) trip in about 38 minutes. More than half the
subjects began feeling queasy at these higher limits. As a result, a decision was reached to restrict
the exposure of subjects drawn from the general public to bank angles of less than 30 degrees and
roll rates of less than 9 degrees/sec.

The final experimental design specified nine flights with 14 subjects each. Each flight simulated a
277-km trip made with one of the nine possible combinations of limits for bank angle and roll rate.
The limits for bank angle were 14, 21 and 28 degrees while those for roll rate were 4, 6 and 8
degrees per second. Since the laboratory simulator seated only four subjects, two sessions were
conducted with each combination of limits, allowing more than half of the persons who had flown
to take the simulator trip as well. Subjects were required to rate ride comfort and their own
tendency to motion sickness (both on seven-point scales) five times during both trips and to read
magazine articles and answer questions about them.

Analysis of the data from the subject rating sheets and the instrumentation lead to the following
conclusions:
 

• Based on the results of this study there is no evidence that more than a small percentage of Maglev passengers would experience kinetosis on routes confined to the boundaries of existing highway rights-of-way. This study simulated a Maglev system traveling through representative portions of the proposed New York State route at average speeds that ranged from 320 to 400 kph (200 to 250 mph). While the vertical accelerations experienced by the subjects in the aircraft simulation were generally greater than those that would be experienced by Maglev passengers, only 2 of the 127 subjects vomited.
• Within the bank-angle and roll-rate limits tested, the majority found the plane ride comfortable and felt no motion sickness. These limits were greater than those specified for the Maglev Systems Concepts Developers.  However, a significant minority, 23%, felt slightly queasy at some time during the flight, while 8% felt intermittently nauseous or worse during some portion of the flight and two subjects vomited. The reported differences between subjects in their perceptions of ride quality and propensity for motion sickness were greater than the physical differences in bank-angle and roll-rate limits for different flights. Ratings of ride comfort and motion sickness were not significantly correlated with bank-angle or roll-rate limits.  The percentage of passengers showing signs of motion sickness in the flight experiments are probably greater than the percentages who would do so aboard an actual Maglev, because the flights subjected them to somewhat larger doses of vertical acceleration than they would have received aboard a Maglev with the same nominal bank and roll limits. Furthermore, the limited views through the small airplane windows and/or anxiety about the flight may have contributed to the onset of nausea in some subjects. Hence the foregoing conclusions are conservative.
• Cumulative dosage and duration of exposure showed significant correlation with motion-sickness ratings. The implication of this finding is that average values for bank angle and roll rate should be lower on longer routes than on short ones.
• In the laboratory simulation, no subjects vomited and only one of 71 reported even intermittent nausea. Thus, the visual effects of scenery rushing by at 400 kph (250 mph) do not appear to present a problem when that view is limited to a side window, even as large as the 89 cm (35”) video monitors used in the experiment.

No. of Pages:  100