In special cases, complete surveys of routes may be required finally to select the best route, but these instances are few in number.

=Road Surveys.=--When a road has been definitely selected for improvement, a careful survey is made to furnish information for the preparation of the plans. This will consist of a transit survey and a level survey.

The transit survey is made by running a line between established corners following the recorded route of the road, or if no records are available or the road is irregular in alignment, by establishing arbitrary reference points and running a line along the center line of the existing road or parallel thereto. The topography is referenced to this line in such completeness that it can be reproduced on the plans.

The level survey consists in taking levels on cross sections of the road at one hundred foot intervals, and oftener if there are abrupt changes in grade. Special level determinations are made at streams, railroad crossings, intersecting roads or lanes and wherever it appears some special features of the terrain should be recorded.

From the surveys and such other information as has been a.s.sembled relative to the project, a plan is prepared which embodies a design presumed to provide for an improvement in accordance with the best highway practice.

THE PROBLEM OF DESIGN

It will be convenient to consider separately the components of a road design, although in the actual design the consideration of these cannot be separated because all parts of the plan must fit together.

=Alignment.=--The alignment of the road is determined to a considerable extent by the existing right-of-way, which may follow section lines, regardless of topography, as is the case with many roads in the prairie states, or it may follow the valleys, ridges, or other favorable location in hilly country. In many places the roads of necessity wind around among the hills in order to avoid excessive grades. In designing an improvement, it is generally desirable to follow the existing right-of-way so far as possible. But the element of safety must not be lost sight of, and curves should not preclude a view ahead for sufficient distance to insure safety to vehicles. The necessary length of clear view ahead is usually a.s.sumed to be 250 feet, but probably 200 feet is a satisfactory compromise distance when a greater distance cannot be obtained at reasonable cost. To secure suitable sight distance, the curves must be of long radii, and where possible the right-of-way on the inside of the curve should be cleared of trees or brush that will obstruct the view. Where the topography will not permit a long radius curve and the view is obstructed by an embankment or by growing crops or other growth, it is desirable to separate the tracks around the curve to eliminate the possibility of accidents on the curve. This is readily accomplished if the road is surfaced, but if it is not surfaced, the same end is accomplished by making the earth road of ample width at the curve.

Relocations should be resorted to whenever they shorten distances or reduce grades sufficiently to compensate for the cost.

=Intersections.=--At road intersections, it is always difficult to design a curve that entirely meets the requirements of safety because there is not enough room in the right-of-way, and enough additional right-of-way must be secured to permit the proper design. It is not necessary to provide an intersection that is adapted to high speed traffic, where main roads cross, but, on the contrary, a design that automatically causes traffic to slow up has distinct advantages.

Where a main route, improved with a hard surface, crosses secondary roads, it is satisfactory to continue the paved surface across the intersecting road at normal width and make no provision for the intersecting road traffic other than a properly graded approach at the intersection.

=Superelevation.=--On all curved sections of road, other than intersections, account is taken of the tendency of motor cars to skid toward the outside of the curve. This tendency is counteracted by designing the cross section with superelevation.

[Ill.u.s.tration: Fig. 6]

In Fig. 6, _F_ represents the tangential force that tends to cause skidding. _W_ represents the weight of the vehicle in pounds, THETA = the angle of superelevated surface _c-d_, with the horizontal _c-a_.

_R_ represents the radius of the curve upon which the vehicle is moving. _w_ is the component of the weight parallel to the surface _c-d_, _v_ = velocity of the vehicle in feet per second. _m_ = ma.s.s of vehicle = _W/g THETA_

_w_ = _W_ tan _THETA_

_mv^2_ _wv^2_ _F_ = ------- = ------ _R_ _gR_

If _F_ = _w_ there will be no tendency to skid; hence the rate of superelevation necessary in any case is as follows:

_Wv^2_ _W_ tan _THETA_ = ------- _gR_

_v^2_ tan _THETA_ = ------- _gR_

The amount of superelevation required, therefore, varies as the square of the velocity and inversely as the radius of the curve.

Theoretically, the amount of the superelevation should increase with a decrease in the radius of the curve and should also increase as the square of the speed of the vehicle. On account of the variation in speeds of the vehicles, the superelevation for curves on a highway can only be designed to suit the average speed. At turns approaching ninety degrees, the curve is likely to be of such short radius that it is impossible to maintain the ordinary road speed around the curve, even with the maximum superelevation permissible. It is good practice to provide the theoretical superelevation on all curves having radii greater than 300 feet for vehicle speeds of the maximum allowed by law, which is generally about 25 miles per hour. Where the radii are less than 300 feet, the theoretical superelevation for the maximum vehicle speeds gives a superelevation too great for motor trucks and horse drawn vehicles and generally no charge is made in superelevation for radii less than 300 feet, but all such curves are constructed with the same superelevation as the curve with 300 foot radius.

The diagram in Fig. 7 shows the theoretical superelevation for various curve radii.

[Ill.u.s.tration: Fig. 7. Curves showing Theoretical Superelevation for Various Degrees of Curve for Various Speeds of Vehicle]

At the intersection of important highways, the problem is complicated by the necessity for providing for through traffic in both directions and for traffic which may turn in either direction and the engineer must provide safe roadways for each cla.s.s of traffic.

=Tractive Resistance.=--The adoption of a policy regarding the grades on a road involves an understanding of the effect of variation in the character of the surface and in rate of grade upon the energy required to transport a load over the highway. The forces that oppose the movement of a horse drawn vehicle are fairly well understood and their magnitude has been measured by several observers, but comparatively little is known about the forces opposing translation of rubber tired self-propelled vehicles.

The resistance to translation of a vehicle is made up of three elements: resistance of the road surface to the rolling wheel, resistance of the air to the movement of the vehicle and internal friction in the vehicle itself.

=Rolling Resistance.=--When the wheel of a vehicle rolls over a road surface, both the wheel and the surface are distorted. If the wheel has steel tires and the road surface is plastic, there will be considerable distortion of the road surface and very little of the wheel. A soft rubber tire will be distorted considerably by a brick road surface. Between these extremes there are innumerable combinations of tire and road surface encountered, but there is always a certain amount of distortion of either road surface or wheel, or of both, which has the same effect upon the force necessary for translation as a slight upward grade. When both the tire and the road surface strongly resist distortion (as steel tires on vitrified brick paving), the resistance to translation is low but the factor of impact is likely to be introduced. Where impact is present, energy is used up in the pounding and grinding of the wheels on the surface, and this factor increases as the speed of translation, and may be a considerable item. Impact is especially significant on rough roads with motor vehicles, particularly trucks, traveling at high speed.

These two factors (impact and rolling resistance) combined const.i.tute the major part of the resistance to translation for horse drawn vehicles.

=Internal Resistance.=--For horse drawn vehicles, the internal resistance consists of axle friction, which is small in amount. For self-propelled vehicles, the internal resistance consists of axle friction and friction in the driving mechanism, of which gear friction and the churning of oil in the gear boxes is a large item.

Internal friction is of significance in all self-propelled vehicles and especially so at high speeds.

=Air Resistance.=--At slow speeds, the resistance of still air to translation is small, but as the speed increases, the air resistance increases rapidly and at the usual speed of the pa.s.senger automobile on the road becomes a very considerable part of the total resistance to translation. This factor has no significance in connection with horse drawn vehicles, but is to be taken into account when dealing with self-propelled vehicles at speeds in excess of five miles per hour.

Many determinations of tractive resistance with horse drawn vehicles have been made from time to time and these show values that are fairly consistent when the inevitable variations in surfaces of the same type are taken into account. Table 4 is a composite made up of values selected from various reliable sources and Table 5 is from experiments by Professor J. B. Davidson on California highways.

TABLE 4

AVERAGE TRACTIVE RESISTANCE OF ROAD SURFACES TO STEEL TIRED VEHICLES

Surface Tractive force per ton

Earth packed and dry 100 Earth dusty 106 Earth muddy 190 Sand loose 320 Gravel good 51 Gravel loose 147 Cinders well-packed 92 Oiled road--dry 61 Oiled road--wet 108 Macadam--very good 38 Macadam--average 46 Sheet asphalt 38 Asphaltic concrete 40 Vitrified brick--new 56 Wood block--good 33 Wood block--poor 42 Cobblestone 54 Granite tramway 27 Asphalt block 52 Granite block 47

TABLE 5

TRACTIVE RESISTANCES TO STEEL TIRED VEHICLES[1]

----------+-----------------+-----------------+-----------+----------- | | Condition | Tractive | Resistance Test No. | Kind of Road | of Road | Total lb. | per ton lb.

----------+-----------------+-----------------+-----------+----------- 29-30-31 | Concrete |Good, excellent | 83.0 | 27.6 | (unsurfaced) | | | [2]11-12 | Concrete |Good, excellent | 90.0 | 30.0 | (unsurfaced) | | | 26-27-28 | Concrete 3/8-in.|Good, excellent | 147.6 | 49.2 | surface | | | | asphaltic oil | | | | and screenings| | | 13-14 | Concrete 3/8-in.|Good, excellent | 155.0 | 51.6 | surface | | | | asphaltic oil | | | | and screenings| | | 9-10 | Macadam, |Good, excellent | 193.0 | 64.3 | water-bound | | | 22-23 | Topeka on |Good, excellent | 205.5 | 68.5 | concrete | | | 8 | Gravel |Compact, good | 225.0 | 75.0 | | condition | | [3]45-48 | Oil macadam |Good, new | 234.5 | 78.2 [4]46-47 | Oil macadam |Good, new | 244.0 | 81.3 38 | Gravel |Packed, in | 247.0 | 82.3 | | good condition | | 18-19-20 | Topeka on plank |Good condition, | 265.0 | 88.3 | | soft, wagon | | | | left marks | | 34 | Earth road |Firm, 1-1/2-in. | 276.0 | 92.0 | | fine loose dust| | 24-25 | Topeka on plank |Good condition, | 278.0 | 92.6 | | but soft | | 1-2-5 | Earth road |Dust 3/4 to 2 in.| 298.0 | 99.3 3-3 | Earth |Mud, stiff, firm | 654.0 | 218.0 | | underneath | | 6-7 | Gravel |Loose, not packed| 789.0 | 263.0 ----------+-----------------+-----------------+-----------+-----------

[1] Prof. J. B. Davidson in _Engineering News-Record_, August 17, 1918.

[2] Graphic record indicates that the load was being accelerated when test was started.

[3] Drawn with motor truck at 2-1/2 miles per hour.

[4] Drawn with motor truck at 5 miles per hour.

Comparatively few data are available showing the tractive resistance of motor vehicles, but the following tables are based on sufficient data to serve to ill.u.s.trate the general trend.

These data on the tractive resistances of an electric truck with solid rubber tires on asphalt and bitulithic, wood, brick and granite block, water-bonded and tar macadam, cinder and gravel road surfaces were obtained by A. E. Kennelly and O. R. Schurig in the research division of the electrical engineering department of the Ma.s.sachusetts Inst.i.tute of Technology, and are published in Bulletin No. 10 of the division.

An electric truck was run over measured sections, ranging from 400 to 2600 feet in length, surfaced with these various materials, at certain speeds per hour, ranging from about 8 to about 15.5 miles per hour.

The result of the observations of speeds, tractive resistances, conditions of surfaces, etc., were collected and studied in various combinations.

TABLE 6