Category NEW DESIGN CONCEPTS FOR HIGH SPEED AIR TRANSPORT

Before the airline makes a cost economic comparison it should first evaluate how the aircraft fits in the current fleet in terms of maintainability, passenger comfort, interior arrangement, handling and turnaround lime Reference (37) states that environmental factors such as domestic air poli­cies, expected traffic demand, capacity and frequency, financing and domestic traffic infrastruc­ture also play a role.

It is outside the scope of this section to go into all of these factors but a few are directly related to the aircraft design and should therefore be mentioned

• The cruise speed is an important criterion since it influences the number of flights per day on a given route. Airlines greatly favor even numbers of flights a day since this enables them to perform the aircraft’s (nightly) maintenance in its home port each day. Mach 2 is therefore a good cruise number for the transatlantic range.

• The operational flexibility. Performance in terms of turnaround time, maintenance and take­off field performance should be as good as the aircraft the new aircraft is compared with, to he able to use this method.

I%7 the Air Transport Association of America [31] published a standard method to es­timate tlie direct operating cost of aircraft This method is no longer used to obtain actual direct operating cost, but it is still used to make comparative and parametric studies. Since 1967. the airframe manufacturers have been updating this method to reflect the technological progress. These studies were generally not available to the public However, in 1978 American Airlines and NASA pul out a new study 128] and (33} that reflected the added experience in aircraft op­eration since 1967.

One of the suggestions in the new meilwd. the introduction of “aircraft related operating expenses” instead of the DOC. was not followed, as far as the author can tell. Since some of the items in the new definition have a large variance for different operations in different nations (e g. fuel servicing fees, training costs), the аиііюг has decided to keep the original definition for the direct operating cost, but to use updated relations from the American Airlines study for the actual equations. The factors in these equations were sealed to represent 1994 European conditions.

To enable us to compare configurations all costs and profits are expressed per scat km PROFIT * REV – (IOC ♦ DOC) (5)

The indirect operating cost IOC will primarily depend on the operator’s type of organ­ization and policy. This cost includes maintenance of buildings, servicing of flight operations and administration and sales. In 1985 the average IOC of the world scheduled airlines was 3.5 Set/ pass km. The direct operating cost DOC include the cost of flying, airplane maintenance and de­preciation. Ticket revenue REV depends on the load factor «averaging around 65*5- of the maxi­mum capacity) and the pricing policies of the airline. In 1985 the average revenue per passenger was 5.7Sct/km. today (1995) it is closer to 8$ct/km

In ihc following statistical formulae we have used daia published in the US Department of Transportation in reference (36].

Defining the block-speed:

The flight time is calculated during ihe mission integration. The loss time is taken to be 30 minutes. Figure 5 relates the cost of a flight crew, including training and employee benefits, per block hour can be expressed to the aircraft’s maximum takeoff weight.

There is an estimated additional 80 dollars for international operations and 190 dollars for supersonic flight per flight hour. The current cost of fuel can be used to calculate the fuel cost:

As mentioned in the previous section the block fuel includes non-revenue flying and maneuvers Figure 6 shows that there exists an empirical relation between the the annual utiliza­tion of 618 aircraft and the average route block time. In the literature sometimes higher values arc quoted, but in practice values over $000 h arc seldom achieved on average. A year has 8760 hours so that means the aircraft would have to be in the air 60 % of its life (including all inspec­tions).

иш Fa (tb) <9>

We can now calculate the cost of insurance based on an insurance rate і betw een 0 5 and

1.0%.

The capital cost can he calculated assuming a depreciation to 10% of the original value in period P of 14 years Four percent of the airframe value is needed for spares, while 30% of the engine value is needed for spares’:

In the published data of reference (36) a large discrepancy can be observed between the above depreciation and the value found in our estimate It must be noted that in reference (3b) the depreciation is made on the original purchase price and not the present purchase price We can synchronize both methods by correcting the depreciation of the airframe each >car with the airframe deflator.

Based on the original ATA’67 publication and the data published by the U. S Depart­ment of Transportation, a new model for estimating maintenance cost was proposed. The model is correct for aircraft five years after their introduction and four years after purchase The [1]

original formulae in references (28) and (33) have been simplified by reasonable assumptions and variable estimates as presented by the methods in the section on weight prediction1 .

Equation (13) predicts the number of labor hours per airframe per flight cycle. These cost arc associated with the number of passengers (for example: cleaning the scats) and the size of the aircraft

И, РГ, ш <2.14 ♦ 0.0000079 m, + 0.0046 s) ‘Af (12)

А/С to

Equation (13) predicts the number of labor hours per airframe per flight hour. This in­cludes repairs to the flight structure, and the passenger facilities:

//r -(3.08 + 0 000032 m,+ 0.0041 а) Л/ (13)

AN. of

For high subsonic operations die Mach number M is set to 1. According to Air France[2] [3] [4] “twice as much effort" is spent servicing the Concorde for every hour of flight as it does for planes that cannot fly at the speed of sound. However, this is primarily due the small size of the Concorde fleet and the square root relations as proposed by ATA will probably be more accurate for larger supersonic fleets. Airframe labor cost are expressed in equation (14) as a function of the labor hours per flight hour and the labor hours per airframe.

The number of labor hours pci engine per flight hour can now he expressed as:

1452 +0.530 * m

HFL = ———— ———— 4- 0.143 (16)

Д/

r

We can now express the engine labor cost in equation (17)

Airframe parts costs is again divided into the costs per flight cycle and the costs per flight hour The materials costs per airframe per flight cycle is.

The materials costs per airframe per flight hour

The engine parts costs is expressed in equation (21):

In equation (22) the maintenance burden, indirect maintenance costs such as supervi­sion. inventory management are related to the total labor cost:

Cn = HC„.5 + CJ (22)

Since 1967 the overhead burden has gone up twice as fast as labour inflation. As is clear in Figure 7. a large variation in maintenance соки exist*. New aircraft like the B757 and the B767 have much lower maintenance costs than predicted, while some more unusual and older aircraft like the BAEI11 have higher maintenance cost. On average the model will arrive at a 24T than actual maintenance costs because so many new aircraft have much lower costs than aircraft that arc 5 years or older.

Figure 7 Predicted and Actual Aircraft Maintenance Costs

An estimate of «he direct operating cost per seat kilometer can now he obtained:

Tabic 1 shows some published duect operating cost and the costs predicted by this method. A very good correlation is achieved overall

The indirect operating cost can be subdivided into two groups: aircraft related and non – aircraft related operating costs.

The aircraft related operating cost per km are:

Ground handling, equipment and landing fees are roughly proportion») to the maximum lakcoff weight, end inversely proportional to the aircraft range

С «Д0023тм / R (24)

o. lO I to

The navigation fees and the aircrafts administration per kin are proportional to the max­imum takeoff weight:

С,, – Д 10.58 ♦ 6.70-7 m ) (25)

0.11 / to

Cabin cress cost per km is proportional to the number of scats and inversely proportional to the cruise speed:

C „•D.42B/V. (26)

0,1it D

The non aircraft related operating cost arc currently around 1.5 cts/s. km assuming a loadfactor I of 100 %. They arc typically one lime expenses per passengers and therefore inverse­ly proportional to range. These costs include: passenger food, passenger service, baggage han­dling. reservation advertising, commissions and airline non-aircraft administration.

C D. 13 s l /R (27)

o,13 <

IOC = y" ^ 128)

^-r = 10 s

The production cost of for each aircraft in a scries will greatly depend on the following consid­erations:

• Aircraft Size – Assuming that all aircraft arc designed with a comparable degree of sophisti­cation. irrespective of size, then и is clear that the aircraft cost of dcsclopmcnt increases lin­early with size. Even though increased research can decrease empty weight, if all designers have a common set of weight to economy tradeoffs they will all stop developing at a compa­rable degree of sophistication. Figure I shows relative inflation corrected airframe prices as a function of size. Except for very small aircraft a constant cost per pound can be assumed.

• Technology risk. As Mach number increases, more expensive materials must be u. sed and ex­pensive systems make up a larger fraction of the empty weight which increases the airframe cost per unit weight. Figure 2 shows the relative cost per kilogram airframe as a function of maximum flight speed corrected for inflation. Thought the primary source of this data is mil­itary it can still be used for civil transports because the cost of avioncs and weapons systems was not included. In addition, the Concorde data point was predicted well, as well as the av­erage for subsonic commercial jets.

• Production Size. Based on market rescurch the company has to decide on where we Пх its breakeven point. Reference (27) provides sales figures for some successful aircraft programs For big and medium sized passenger transports a breakeven number of 400 seems reasonable. To obtain the sale pnee per airframe we now use Wright’s 809r learning curve as described in reference [39] and (32) and a representative cod per unit weight and thrust for commercial aircraft. Figure 4 shows actual recorded learning curves.

Based on current pricing policies the follow ing equation can be used to predict the cur­rent (and) future price of aircraft.

Where Fi is shown in Figure 3.

Obviously it is not true that an aircraft can be sold for more money if it is heav ier or has bigger engines. The market price of the aircraft is solely a function range, payload and speed. Such definition of price would however not be useful at this time to tradeoff engineering features of a design. The assumption here is made, that aircraft manufacturers can only stay in business if they produce an aircraft with the same technical refinement as their competitors and charge a price that is closely iclaied to their production cost. So wc assume that aircraft are a commodity, which would not be true for an innovative design that would not have competitors.

This simple relation should be corrected when non standard materials or technologies arc used. Usually the price per pound goes up more than proportional with the introduction of weight saving techniques This relation will have to he modeled if such weight saving techniques are considered. For composites every percent reduction in structural weight will at least cause a percent increase in the cost per pound

The economic model presented in this paper uses f«>ur deflators to include the effect of inflation

on cost:

1. The SIC 3721 deflator О ^ for the airframes industry.

2. The SIC 3724 deflator Df is used for engines and engine parts

3. The Consumer Price index will be used to deflate labor costs. This is by no means an ac­tuate deflator for all labor involved Maintenance workers, pilots and the “average" Ameri­can all have different deflators, but the CPI is a reasonable deflator for all these professions.

4. The fuel pnee deflator Df

The definition of a deflator is:

^ _ Price for year=X

D><«’**- Price in 1994 Ш

These deflators arc often revised and reliable data is often five years old. they can be found in "Aerospace Facts and Figures" [27) Note: All deflators are defined as 1 for 1994.

A. Van der Vclden
Synaps Inc., Atlanta. GA, USA

2.1 Abstract

Before the go-ahead is given on the further development of a new transport aircraft design a number of questions need to be answered The airframe manufacturer needs to know whether it can breakeven on its initial investment easily. The airline will only order this new product if it can expand its market, reduce its cost and increase its revenues. The traveller wants a low ticket price and high comfort The society as a whole wants this new technology to improve the econ­omy while safeguarding the environment.

Since all of these views arc in conflict we can only evaluate a new design by comparing it with existing transports, keeping in mind that the life of a new design can span a quarter of a century or more. The content of this section follows is loosely based on my report [ 38J “An Eco­nomic Model for Evaluating High-Speed Aircraft Designs” of 1989. This report has been updat­ed and reevaluated after my experiences at Airbus.

The present model has been developed to give realistic results to tradeoff engineering features of a design to improve aircraft economy. It is not intended to present an accurate picture of pricing policies of airframe manufacturers and airlines

The first sections of the paper deal with a market in equilibrium The economic view ­points of the manufacturer, airline and pussenger arc based on the commodity product jet travel is today. These equilibrium market conditions can also be used to make design trade offs for a supersonic transport but we would have to be very careful to infer more. The question of whether such a supersonic aircraft can be sold and at at which price is very much a different

issue. Daimler Benz Aerospace and other companies use very refined forecasting models to assess the marketability of a new aircraft and such models are highly proprietary In the last see – lion of the paper I will present a highly simplified qualitative overview of such a non equilib­rium market model

2.2 List of Principal Symbols

engine price operating cost airframe block hourly material cost airframe flight cycle material cost Maximum takeoff lift coefficient aircraft and paris deflator relative to 1994 engine and parts deflator relative to 1994 fuel deflator relative to 1994 labor deflator relative to 1994 engine labour hours per flight hour airframe flight cycle labour cost (h) airframe block hourly labour cost (h) insurance rate loudfactor maximum takeoff mass ikg) airframe mass (kg) block fuel (kg) engine mass (kg) number of engines depreciation period to 10 of value range (km) number of scats

AFM

ACM

Ljmaxjo

D

D < IK

D.

m

m

P R s

wing planform reference area maximum uninstalled sea level static thrust

time (h) block time (h)

maximum turbine entry temperature

yearly uuliiation aircraft speed Mock speed (km / h)

aircraft weight Greek Letters mean time between repairs

Subscripts

aircraft

airframe

block

engine

fuel

high subsonic

labour

takeoff

The development of a supcrvmic transport that can be operated at a profit by the airlines, and sold in sufficicni numbers for the airframe and engine manufactures to eventually realize a profit as well, remains a challenge. The U. S and European supersonic research programs now have very focused, and somewhat different, goals. These programs involve the companies that profit from the sale of their subsonic jets. It would take some bold competitive vision, not unlike that which led to the Concorde, for a supersonic transport production program to emerge from these studies. Such an aircraft faces the real possibility that it. too. w ill be a technical success, but not ал economic one This book, therefore, focuses much of its attention on the underlying tools for the study of such aircraft, as well as on unconventional configurations.

For unconventional configurations the technical and nsk harriers arc very high. It appears that an oblique flying wing (see Chapters 19 and 20) could provide a Mach 1.4. or higher, transport that operates with a minimum surcharge over future subsonic transports and that competes with them over land as well. If it is large enough it becomes the "New Large Air­craft" and. in this sire, such an aircraft may compete in fare with its subsonic counterparts. But without further research, considerable experimentation, and flight tests, this remains a conjec­ture Such an aircraft would also require rethinking of selected aviation regulations and perhaps even some minor reconfiguration of airports. Both were required with the introduction of the Boeing 747.

A conventional configuration, operating at a higher Mach number, benefits from high productivity and substantially reduced travel times. Because of pasi and current government research programs, including that which led to the Concorde, the needed research is largely done and the technology mature. Consequently, the development costs of such an aircraft appear to be reasonable. Because of its limited subsonic and transonic performance, and its restriction to intercontinental routes, this aircraft’s market is relatively small. As a fleet, its con­tribution to the acoustic environment in and around selected airports may be small enough to deserve continued regulatory relief.

A small, corporate, supersonic transport appears to have a significant market and. if small enough, might well be certified for supersonic operation over land Military technology and excess production capacity provide the basis for making such an aircraft affordable

At a meeting on sonic boom research in 1967. Adolf Busemann. having comprehended the concept of banglcss sonic booms, concluded this meant we would have to fly in the tropo­sphere to make the sonic boom acceptable. He stood up. placed his arm over his eyes, and said: ‘This is terrible; we will have to fly through the wind, the sleet, the rain, and the snow.” Further research showed even this would not be enough Large transports will not be able to fly at super­sonic speeds over populated areas.

It may be a long time before most of us can fly twice current speeds at affordable fares And we may have to fly obliquely to do so. Before this happens, some will have travelled at Concorde speeds in corporate supersonic transports such as the proposed Sukhoi S-21.

Remarkable advances have been made in propulsion since jet engines were introduced. Over the past 25 years there has been about a 204 reduction in the amount of fuel required to produce a unit of thrust (261 Because much of this gain has come from higher bypass ratios, take-off noise levels have fallen in some cases below those required by current noise regulations. Current SST engine concepts, without augmented suppression systems, are probably 15-20 decibels (equiva­lent perceived noise decibels) above these standards. Further noise suppression adds weight and reduces thrust. Low lift-to-drag ratios at takeoff demand considerable thrust, and this, in turn, leads to larger exhaust velocities and more noise At the moment there are sound ideas, but no tried techniques, on how to accomplish this noise reduction with acceptable weight increases. Unlike the some boom, however, we arc not up against a fundamental momentum balance A breakthrough is possible. Given that subsonic transport noise levels continue to fall, and the near certainty that conventional supersonic transports will operate only from selected coastal cities, current noise regulations need to be examined to see what airport noise levels might be accepta­ble from a small fleet of supersonic aircraft.

Whenever wc bum hydrocarbon fuels using air. wc impact the atmosphere and. in some cases, the local air quality. Whatever fuel we burn using air will produce oxides of nitrogen. A concern during the late 1960s was the effect of water vapor from SST engine exhausts on stratospheric ozone levels. It was soon realized, however, that the oxides of nitrogen were much more impor­tant (22). This led the Department of Transportation, in 1972. to launch the Gimatic Impact As­sessment Program This monumental and highly regarded 7200 page study, comprising the work of over 500 individuals, concluded that a limited fleet of supersonic transports, such as the 30 Concordes and TU-144s then envisioned, posed an insignificant threat to the atmosphere. This study also aided the extraordinary discovery of the reduction of atmospheric ozone by CFC re­frigerants (Freon 11 and 12), culminating in ihc Montreal Protocol (1987) which will lead to the eventual elimination of these refrigerants

The oxides of nitrogen catalyticaHy destroy ozone above about 13 kilometers in mid- latitudes; they cataiyiically create ozone below this altitude Aircraft emissions arc the major unnatural source of these oxides in the stratosphere. They are also an important source of them in the upper tn>posphcre. at least of mid-latitudes in the northern hemisphere [23]. Thus it appears that SSTs in the stratosphere may reduce our protection from ultra-violet radiauon by ozone on the one hand At altitudes of 12-14 kilometers (13 kilometers = 42.650 feet), the effect of these oxides on ozone is minor. The calculated ozone column change due to the injection at 20 kilometers of the amount of NO* expected from a full fleet of SSTs was about *12** in 1975. New knowledge changed this to +34 in 1979. Since that lime, increasing knowledge provided a result of -104 in 1988, about double the -54 predicted ozone depletion if CFC releases remained at their 1974 rate (24). Recent results show NO, to be less significant than was once thought, but raise the issue of the effects of engine emissions on stratospheric aerosol surface area. This could also play a role in depleting stratospheric ozone [25].

Just as wave drag due to lift is inescapable, so is the sonic bang. Adolf Busemann liked to illus­trate this by depicting the conical shock wave system and its reflection from the ground as the crow – bar that supported the weight of the aircraft (17) Ironically, w hile the weight of the aircraf t is to be found m the integral of the pressure signature over the ground, it is not to be found in the first-order pressure field there 118). In the U S. we call the sonic "hang" the sonic "Ычил " The "bang" in the some boom derives from the abrupt pressure increases through the two. and some­times more, shock waves emanating from a supersonic aircraft Wc call the integral of the posi­tive phase of the pressure w ith respect to time the "impulse" The bang is directly related to the outdoor annoyance of animals and humans; the impulse is related to structural damage and. to some degree, to indoor annoyance.

The increasing acoustic impedance (i. e., the product of the density and (he sound speed) below the aircraft in a real atmosphere freezes the shape of the pressure signature before it reaches the ground. In the approximation of an isothermal atmosphere this occurs in ЖҐІ atmospheric scale heights, or about 40.000 feet. This knowledge set me and my colleague A1 George to tackle the minimization of various parameters of the sonic boom signature, including its bang and its boom, or any weighted average you might use of the parameters. Indeed, for the cruise characteristics of the Mach 27 Boeing 2707 at 60.000 feet lifting 600.000 pounds, an air­craft 527 feet long need not have a sonic bang at all. i. e.. the pressure field below the aircraft need not steepen into shock waves (19J. But as we noted then, reducing or eliminating the "bang” in the sonic boom increases the impulse, or total pressure loading, for obvious reasons: the bang part of the boom, that is the shock waves, dissipates the energy in the signature. Conse­quently. reducing or eliminating the shock waves makes the impulse worse.

Very considerable studies by the NASA over the past decade have explored whether or not such shaping of the sonic boom signature would lead to an acceptable sonic boom The NASA’s conclusion reinforces ours of two decades ago. Unless a supersonic aircraft is very – light. but long, its sonic boom cannot be reshaped to be acceptable [201. Very small supersonic aircraft, such as a corporate supersonic transport, may have an acceptable, indeed nearly inaudi­ble, sonic boom. This stems, in рал. from a thickening of the shock waves as their strength is reduced.

SSTs will be constrained to subsonic operation over populated areas, and perhaps to supersonic operation over the oceans alone The penetration of the pressure field of sonic booms into water, versus their reflection from it, is now well understood (21). For aircraft traveling less than the speed of sound in sea water, this is simply a travelling source of acoustic radiation. Commercial transport at supersonic speeds over the oceans, and perhaps over unpopulated areas, is likely to continue to be acceptable. Rights over land areas with significant popula­tions of wildlife may not be allowed. Through constraints on aircraft routes wc can avoid the problems caused by sonic booms, but in doing so wc reduce the market for a second generation SST

As I’ve noted earlier, the U S. SST program was canceled in part because of environmental con­cents. The Concorde'(economics have been greatly affected by being prohibited from superson­ic High! over most land areas, and by the cost of fuel Ihc environmental, and thereby political, barriers to a successful SST arc: energy consumption, sonic bang, atmospheric impact, and air­port noise.

1.6.1 Energy Consumption

The fuel consumed by SSTs per passenger mile is several times that of subsonic transports Su­personic flight entails a new penalty, that of wave drag. Lift has to equal, and sometimes exceed, weight if there is to be air travel. Wave drag due to lilt is inescapable except for an infinitely long swept wing, best approximated by the way. by an oblique wing Volume can be moved through the air supersonically with no wave drag, but at considerable expense in skin friction drag from extra surfaces.

Sixty countries have ratified a treaty that commits them to better manage their genera­tion of greenhouse gases J16). Developed countries arc to provide plans by the end of this cen­tury that show how they will return to 1990 levels of greenhouse gas generation Does this argue against an SST? As Secretary of Transportation Coleman said in his decision to let the Con­corde operate: "It would border on hypocrisy to choose the Concorde as the place to set an example… (for energy efficiency) while ignoring the inefficiency of private jets, cabin cruisers, or an assortment of energy profligates of American manufacture’* |5J.

The Concorde achieves 17 scat miles per gallon and. at 679fr load factor, is equivalent to a car w ith only the driver, achieving 12 miles per gallon But the Concorde’s passengers arc going more than twenty times as fast and follow ing nearly a straight line to their destination. A future SST should not be rejected because of energy considerations. However, its economics and thereby its market, are more sensitive to fuel costs than its subsonic counterparts and these are not only variable, but jet fuels may eventually be taxed for their carbon content

Within a few months of the first flight of the French and the British Concorde prototypes (March

2. and April 9, 1969). the US SST finalist, the Boeing 2707. had booked 122 options from 26 airlines to purchase aircraft: the Concorde had booked 74 options from 16 airlines. Thus, nearly 200 SSTs were "on order." A year later, in 1970. the FAA predicted 500-800 SSTs would be in operation by 1990. It is now 1996.

Tvelvc Concordes operate today with a limited schedule and at load factors below 50<* These aircraft need only pay their operating costs exclusive of the amortization of their purchase, they were essentially free to the two airlines flying them |I0J. What happened^ The fares required to pay for their operation deter their use. Maintenance costs are said to be seven times those of a 747 and fuel costs per passenger mile at least three times that of the 747.

Studies by Boeing and by McDonnell Douglas predict a market for 600 to 1500 SSTs (11]. (12]. Міомо of Japan Aircraft Development predicted a market for 600 Mach 2.5 SSTs with a 5500 nautical mile range, and estimated perhaps a 50^ increase in this market derived from its stimulation by the travel time saved 113}. Davies, on the other hand, found it to he between 9 and 36 aircraft, depending on how optimistic one is (14]. The enormous differences among these studies stem from what one projects for the fare required to cover the aircraft’s total operating costs. It lakes a long time to sell one thousand aircraft. The first Boeing 747 began commercial flights in 1970; twenty-four years later one thousand 747$ had been deliv- crcd.

The challenge is to design, build, certify ami operate an SST while providing the air­lines a return on investment comparable to a similar investment in subsonic aircraft. This can only be accomplished with marginally increased fares over those for subsonic transport. The marginal increase in fares required, however, depends upon many factors, including aircraft price and operating cost.

Marginally increased fares – what does that mean? Assume such transport effectively saves the trasxlcr some fraction of a day. or at most, a whole day Whatever that traveler’s expenses would be for that day. or, correspondingly, whatever his income might be for that day. provides a reliable guide as to what he would be willing to pay to save a fraction of a day of business travel, or have as extra time for his vacation. This intuitive judgment agrees with stud­ies which predict little fall-off in ticket sales for a 10% surcharge 111|, [13].

As noted earlier, non-discoum passengers comprises 30% of the international market. To secure a significant fraction of this market an SST will need to provide three-class service. Current Boeing studies reflect this, hut show an SST with about 9% of the passengers in first class. 19% in business class, and 72% in economy. Can an SST succeed if it fills empty seats with discount coach passengers0 Can it succeed if it docs not?

A final comment is warranted on the growth of revenue passenger miles accorded air transports The "information highway” will reduce business travel needs. For a few hundred dollars you can buy the software needed for your group to discuss and share visual information by electronic mail. It is now possible, with more expensive software, to have the real-time image of each member in a working group displayed, hear their voices, and share visual infor­mation. A telecommunications vice president recently told me that he spent $23.000 on hard­ware and software and saved $100,000 in travel costs in the first year. The importance of this change was noted some years ago by Simpson in his remarks to the 1989 European Symposium on Future Supersonic-Hypersonic Transportation [15] When the information highway becomes an international highway, which it now nearly is. this will reduce the need for international busi­ness travel while simultaneously expanding the amount of international business li seems likely that these two effects will offset one another.

Technology has progressed steadily since the Concorde was conceived. But reduced energy efficiency, the sonic bang, engine emissions, and airport noise, remain deterrents to the economic success and acceptability of an SST. Let me now turn to the environmental barriers facing a future SST.