Carpenter formalized his designs for Tom Vint in the seven sheets of
Drawing No. NP-PG-20 15 and the report he wrote accompanying the
REPORT TO THE CHIEF ARCHITECT
TO ACCOMPANY "PRELIMINARY STUDIES FOR ROAD ROAD-DESIGN,"
Drawing No. NP-PG-2015 (7 Sheets)
Thos. E. Carpenter
Deputy Chief Architect
Branch of Plans and Design
National Park Service
250 Federal Office Building
San Francisco, California
March 22, 1938
In our studies on road design, and working in
collaboration with the Bureau engineers, we have pioneered a number of
design features which the Bureau have adopted for National Park roads.
These include spiral curvature, flattening and rounding of cut slopes
and warping of fill slopes. Many other items are covered in our Special
Provisions for road projects.
The accompanying set of prints of Drawing No.
PG-2015, 7 sheets, represents, our latest studies on road design. The
plans are the following:
Sheet 1 - Perspective Sketch
Sheet 2 - Rounding Bottom of Cut Slopes
Sheet 3 - Flattening Fill Slopes
Sheet 4 - Rounding Top of Cut
Slopes on Horizontal to Plus Ground Slopes
Sheet 5 - Rounding Top of Cut Slopes on
Horizontal to Negative Ground Slopes
Sheet 6 - Flattening and Rounding Cut and Fill Slopes
Sheet 7 - Surfacing Cross Sections and Drain Inlets
The perspective sketch Sheet 1 illustrates a feature
that we have considered for some time and I feel would be one of the
most important steps that we could make in the advancement of road
design. This is the introduction of a curve at the bottom of the cut
slope in order to obtain a better transition from the lower part of the
slope into the roadbed, and thereby do away with the sharp angle that is
now obtained with ditch slopes such as 2 : 1 and 4 : 1 from the road
shoulder to the straight line of the cut slope in cross section. This is
accomplished through a "set back" of steep slopes which permits of
flattening and thereby stabilizing the lower part of the slope. The flat
gutter slope completes the transition from the slope to the road surface
proper. This relatively flat gutter is shown contrasted with the deeper
ditch slope, and a flatter fill slope is also shown for a low fill.
The perspective sketch illustrates what may be
expected in stabilizing the lower part of steep slopes. Native
vegetation will grow that otherwise can not become established on an
unstable slope. As trees grow in height from the bottom part of the
slope, they will also improve appearance of the upper part of the slope
as seen in perspective by those driving on the road. For desert and
semi-desert country, the main value would be stability and form of
slopes, but even here it would enable low native vegetation such as
sagebrush to become established.
Sheet 2 shows the cross sections for proposed slopes
in comparison with those used in present practice. Other sheets
illustrate proposed revisions in the present standards of cross section
design as pertains to flattening and rounding of cut slopes and also
advancement in the handling of fill slopes to a degree beyond prevent
standards, but primarily a better defined and simpler method for the
field engineer to follow in the construction of fill slopes and in
warping of fills. Sheets 2 to 5 are detailed to explain the design
features. All of the data is summed up and shown on Sheet 6 in the two
cross sections for cuts and fills with the slope tables. It is thought
that this sheet, or modifications of it that may be worked out with the
Bureau, will be included in the contract et of drawings to explain the
proposed design for cut and fill slopes. Modifications would be those of
road width and gutter details to fit specific road. projects.
Sheet 2 - Rounding the Bottom of Cut Slopes.
Taking the slope as a whole, curvature at the bottom
together with curvature in the reverse direction at the top, results in
an ogee. As a form, curvature at the bottom of the slope improves
perspective through the resultant feeling of stability expressed. Actual
stability of the slope is created in the 1 : 1 and 1-1/4 : 1 slopes,
with the introduction of a slope 1-1/2 : 1 and flatter at the bottom. It
is the kind of slope that nature makes at the bottom of steep slopes
through erosion, in reaching an angle of repose. This is necessarily
removed by maintenance, in keeping ditches open under the present
standards of cut slope design and construction.
The cross sections on Sheet 2 show the proposed
slopes in comparison with the as constructed cross sections of
Yellowstone roads, Old Faithful-West Thumb 26 ft. shoulder to shoulder
and Madison Junction-Old Faithful 28 ft. shoulder to shoulder, both with
a ditch 4 ft. wide on a 4 : 1 slope. The proposed cross sections are 24
ft. wide shoulder to shoulder in cuts on benched section and 26 ft.
wide in through fills. These widths are for finished grades. Even though
a relatively wide gutter cross section results, part of it being the
lower part of the transition slope, you will note that at the road
elevation there is less width (or not in excess) from the center line of
the road to the cut slope, than for either of the two sections in
In comparison with the Old Faithful-West Thumb
section the proposed : 1 slope is set back 3 feet in horizontal
measurement (less from the Madison Junction-Old Faithful section); the
proposed 1-1/4 : 1 slope is set back 2 feet; on 1-1/2 : 1 and flatter
slopes there is less excavation than on the two sections described. You
will note that part of the stability is gained with a slope set back and
part through the position of the toe of the slope, which is lifted in
the use of a relatively flat gutter. The set back for a 1 : 1 slope may
be called the criterion, because as the slope gets flatter - 1-1/4 : 1
an 1-1/2 : 1, it is not necessary to sieve the face of the cut back so
far in order to have it in a position where the bottom of the slope can
be flattened. There is a line at the toe of the slope or edge of the
gutter that is common to all slopes. This would likewise be the "toe"
line for rock cuts that are constructed on a 1/2 : 1 slope (District 2,
There are exceptions to the "set back" of slopes in
the case of extremely high cuts, especially where the ground slope above
is very steep. In some such cases it may be more economical to construct
a low retaining wall. This has been done under present standards.
Sheet 2 shows a certain amount of set back in the
steep cut slopes. It does not purport to be the final answer to this
subject, but is a big step in the right direction.
It should be explained that while comparison is made
here with sections of Yellowstone roads or with standards used by the
Denver District 3 Bureau Office, some modification in set backs would
have to be made if this type of design were used for sections of road
designed on the basis of the San Francisco District 2 Bureau Office, say
for Yosemite, where the ditch is less in width and depth than for
Increase in Excavation through use of the Proposed
Through contact with engineers of the Bureau District
2 Office here, we have obtained estimates of cost of the net excavation
increase on the basis of a 3 ft. set back throughout the cut slopes for
two road projects. The proposed cross sections discussed herein are
estimated to average about two-thirds of this cost per mile, because the
sections average less in set back than those estimated by the Bureau. On
a Forest Highway project in heavy type work, where road construction
would cost as high as $50,000 a wile, it was estimated that the new type
of cross section would increase the cost $800 (1,340 cu. yds.) for one
mile, and 2,24O (3,730 cu. yds.) for another mile, an average of $1,520
per mile, where the slopes are principally 1:1 and 1-1/4:1. This is
about 3%. For the Overton-Lake Mead project at Boulder Dam Recreational
Area, where the excavation was figured at 30¢ a cubic yard, it is
estimated that there would be an average increase in cost of $450 a
It is important to note that the increase in the
excavation quantities as given for these projects, is much less than the
gross excavation quantities that would be obtained by simply taking an
additional width of cut with the alignment and grade for the road as
designed. The figure is obtained by adjusting the roadbed, for example
raising a grade in a through cut, and shifting alignment right or left,
in order to reduce excavation. In other words, if this type of design is
adopted, the alignment and grade will, be established on a given project
to accommodate this change.
As near as can be estimated from data available now,
increase in cost of excavation may be 3% or 4% on heavier type
construction, but less on medium and low unit cost construction. Against
the whole road project, including structures, etc., the figure of 2%
should cover the increased cost over the average type of design now
used. This is exclusive of gutter surfacing which is described
Regardless of gutter or ditch types the type of cut
slopes shown in these plans is recommended. I feel that we are justified
in some increase in first cost for several reasons. An economic one is
the fact that maintenance will be reduced. There should be much less or
little surface sloughing of banks into ditches. There is exception of
course, for slides caused by certain ground water conditions in the out
banks. The paved gutter described in the following will likewise reduce
erosion that takes place in dirt ditches and reduce maintenance
With the present standards of flatter cut slopes,
adopted several years ago, we have accomplished much stability, in
slopes. The proposed slope standards will add stability to the steepest
slopes where it is needed, and likewise will permit of natural
restoration of native plant growth on a considerable area of these
slopes. This in itself should be sufficient justification for the
The Gutter and Cross Section Sheets Nos. 2 and 7
In my opinion the use of a relatively flat gutter
section will tend to promote driving closer to the shoulder than would
be the case with a steeper ditch section because there will be less
feeling that there is danger of running into a ditch. For this reason,
and also in order to obtain a proper transition where the end of the cut
section joins the fill, the fill is shown as 13 ft. width, in half
section. The 3 ft. width of shoulder on the fill is the same slope as
the same 3 ft. width in the cut section. This width is constant
throughout cuts and fills.
With the old 18 ft. standard, 3 ft. shoulders were
used either side of 9 ft. paved lanes. Likewise a 3 ft. width of
shoulder is used in this design with the 10 ft, width of paved lanes,
which are current practice in surfacing park roads.
A gutter of the type shown, will permit cars to turn
off the traveled lane and park if desired, in order to change a tire or
for other purposes and will supplement parking spaces made by turn-outs
and daylighting of outside cuts. This will become more important as
traffic increases in the parks, and I feel will offset the present
desire to design wider roads such as 28 ft. shoulder to shoulder
throughout, in order to have a 4 ft. shoulder for parking but with, in
addition, a 4 ft. width of ditch which is now sloped too steeply to use
In our studies of the subject and in discussion with
Bureau engineers we have endeavored to have road cross sections designed
with shallower ditches. The engineers have contended that it is
necessary to have the ditch deep enough that run-off water will be lower
than the subsurfacing of the road in order not to disintegrate this
The detailed cross sections on sheets 2 and 7 show
gutters of bituminous surfacing to seal out the water and protect the
road surfacing proper. The gutter shoulders are designed for a heavier
or coarser aggregate than the 20 ft. wide zeal coat, in order to obtain
a texture that will define the limit of the through travel lanes. It is
expected that there will also be a little difference in color which will
add to the definition. The surfacing design is the function of the
engineer. The sections shown in the plane are based on discussion with
them, and are shown as an example of what may be required in order to
obtain a shallow gutter and add to the safety of the road.
While the gutter is shown in the cross sections as 8
inches deep, the surfacing is indicated for a paved capacity of 4 inches
depth and approximately 4 ft. wide, which should carry normal run-off.
The reason for not paving for the full depth of 8 inches is to reduce
the width of the gutter paving in order to avoid having a cross section
too asymmetrical in the cuts in contrast to that in the fills. Ground
cover vegetation should grow at the back of the gutter in the flat slope
and there should be scarcely any more scour there than there would be if
the gutter paving were carried 4 inches higher, i.e., to the elevation
of the road shoulder.
One question that has come up in regard to the use of
paved bituminous gutters on roads has to do with how the material will
stand up. It is stated that unless bituminous pavement has wearing use
it may not stay resilient, but that this can be taken care of by an
application of oil. We have an example of paved gutter (bituminous) at
Zion, constructed by the B.P.R. District 12, that has stood up very well
without any subsequent oil treatment.
The paved gutter will carry more drainage than one of
the same cross section area constructed in dirt. This is obvious because
there is less friction and the water is carried faster. With paved
gutters the erosion of ditches is eliminated. From contact with District
II of the Bureau we have data showing estimated capacities of road
ditches and gutters. Without going into much detail on this, it can be
said that the gutter cross sections shown on Sheet 2 would have, without
surfacing, at least the same capacity as the road ditches in Yosemite
and Sequoia. If the gutters are paved, they have a greater water
Data is available on the annual precipitation and
snowfall in Yosemite (and other Sierra Nevada parka), Yellowstone and
Rocky Mountain National Parks. Rainfall and snow depth are much greater
in Yosemite than in Yellowstone, and Yosemite has much more snow than
Rocky Mountain. Outside of a deluge, causing flood conditions which are
not provided for in normal gutter design because of excess cost, the
spring run-off (rain with melting snow) is probably the greatest problem
in surface water drainage. This problem is greater in Yosemite and the
Sierras than in Yellowstone where the run-off is slower.
Yet the road ditches for Yellowstone projects are
designed for greater depth and width than those in Yosemite and other
Sierra parks. The composition of the soil and how it drains may have
something to do with this difference. It would appear that much of the
difference in ditch standards may be due to a difference in engineering
practice in different sections of the West.
On Sheet 7 there is shown a type of drain inlet for
use with paved gutters. Since cars will, on occasion, turn into these
gutters, it will necessitate an inlet or catch basin of heavy
construction to withstand the weight. In our study of road inlets we
have endeavored to design a type that is both serviceable and as
economical as possible. The enclosed print of a pipe drain inlet was
prepared as a result of discussion on this subject with the Bureau and
with representatives of a culvert pipe company.
We are informed that a number of pipe culverts of
this type were installed on the Redwood Highway in the northern part of
the State by the maintenance division of the California Highway
Commission and that these worked very satisfactorily in handing the
drainage water during the heavy precipitation of the last to mouths.
As shown in the sketch, the pipe culvert is set in
the bank with the front face at the edge of the gutter line. Since there
is no weight to hold up, as in the case of an inlet under a paved
gutter, a metal pipe is strong enough. The sketch shown an opening from
the gutter 6 inches in height by 18 inches wide for a 2 foot pipe. In
case it is not possible to predetermine the exact height of the drop
inlet for vertical pipe, the opening in the drop inlet can be cut with a
torch in the field and then slipped down over the road culvert. As an
example of cost, with the requirements of a drop inlet 24 inches in
diameter by 3 feet deep, connected to an 18 inch diameter road culvert,
the following would be the approximate cost, exclusive of wood or cast
iron cover, or bottom for inlet:
|3' -- 24" #16 gauge Corrugated Pipe -- at -- $1.50 --||$4.50|
|2' -- 18" #16 gauge Corrugated Pipe -- at -- $1.30 --||2.60|
|Labor cutting and welding Tea -----------------------||7.37|
|($7.37 is the estimated cost of cutting and welding in the
shop. The contractor on the job would probably do it for very
|Labor cutting 6" x 18" opening in drop inlet -------||.75|
We believe that this type of inlet has good
possibilities for adoption in road drainage. We expect to receive some
photographs and further information regarding the experimental
installations that have been made on the Redwood Highway.
When there is a flow or seepage of subsurface water
from cut banks to an amount that is measurably harmful to the roadbed,
it is necessary in most cases to remove this water by underground
drainage, separate from gutter surface drainage.
As an example of subsurface drainage, on sections of
a major road in Sequoia National Park and on the Sequoia-General Grant
Approach Road in elevations between 6,000 and 7,500, on some 16.5 miles
of road, the total cost for subdrainage was $21.900. A total of 8,000
lineal feet of 8 inch perforated pipe was laid. The coat includes trench
excavation, back-fill and seal consisting of a bituminous treatment of
thin oil mix. Of the above construction, the heaviest section for
drainage was one of 81 stations or about 1-1/2 miles where a total of
3,600 linear feet of pipe was laid. Most of this construction was done
with a surfacing project, a smaller amount during grading. Post
construction showed the need for it. Hoist of the pipe was laid 4 feet
deep. This was for ground water that had to be taken away.
The point of this explanation is that to the larger
extent, the handling of subsurface drainage is separate and distinct
from the handling of surface water, the latter being the function of the
ditch or gutter. Data would appear to indicate that many of the "deep"
ditches are not deep enough to catch subdrainage, but are much deeper
than necessary to handle ordinary surface water drainage.
By a set back of the cut slopes, an important
advantage would be gained in subsurface water conditions as affecting
the roadbed. Ground water under the slope would be carried to a lower
elevation under the gutter and roadbed, and therefore there would be
lees danger of capillary action of the water affecting the subsurface of
Engineers are now giving considerable study to the
subject of road subsoils. Some soils hold more water than others and
therefore affect differently the subsurface drainage problem as well as
the road surfacing to be laid over them. For example, the Bureau have
taken soil samples of the Old Faithful-West Thumb road and those are
being tested. The result will show the possibilities of bituminous
reinforcement and determine road surfacing design and cost.
Flattening Fill Slopes - Sheet 3
The drawing is self-explanatory in that it shown
proposed fill slopes based on two factors; one the height of the fill at
the road shoulder and the other the percentage of the ground slope.
Present practice with the Bureau is to construct fill
slopes on a 4:1 where the fill is 2'6" in height at the shoulder and
sometimes up to a height of 5' or 6' whore the ground is flat. All the
data we have seen on this subject shows cross sections of fills
superimposed upon practically level ground. The need for data shown on
Sheet 3 is self-evident.
We have been using Special Provisions in contracts
but with a plan as instruction to engineers, for "Warping the End of
Fill Slopes." Also, we have Special Provisions and plan for "Filling
Inside Pockets." The proposed cross sections will eliminate both of
these, because this data is incorporated. A review of those sections
with the District II Bureau office shows that the sections do not depart
very much from what has been done under one circumstance or another of
fill construction. The cross sections are prepared to fill the need for
definite instruction to contractors and engineers. Fill slopes based on
this design will result in warping the ends of fills better than has
been done heretofore. Flatter slopes on the lower fills will add safety
to the road and in some cases eliminate necessity for guardrail, thereby
economizing in construction.
The cross section and note for rounding the bottom of
fill slopes where the positive ground slope is 3:1 and steeper is
similar to the cross section shown in the plan for "Filling Inside
Pockets." Construction similar to this design was carried out on a
Yosemite road project in 1937.
Rounding Top of Cut Slopes on Horizontal to Plus Ground Slopes -
This sheet shows moderate revision of the present
standard of rounding. The proposed slope results in a parabola curve and
makes a better transition than the present practice under which an equal
distance either side of the top of slope stake is used for rounding.
This is borne out by field construction and clay models. No change is
made in the present standard of slope flattening on horizontal to
plus ground slopes.
Rounding and Flattening Cut Slopes on Horizontal to Negative
Ground Slopes - Sheet 5.
The present standards for rounding top of cut slopes
have been inadequate when applied to most negative ground slopes. There
has been too sharp an "angular" condition at the top. In many cases this
has been taken care of by checking the individual cuts with the
engineer. The proposed rounding will flatten the curve at the top of the
It is also desirable to modify the sharp ridge effect
that often obtains, as seen in cross section through the outside cut
sections. It is proposed that the cut slopes be made flatter where the
ground slopes are negative. The excavation quantities involved are not
Flattening and Rounding Cut and Fill Slopes - Sheet 6
As explained earlier, all of the data shown on Sheets
2 to 5 inclusive (with the exception of the detailed gutter surfacing
section on Sheet 2) is covered on Sheet 6 in the two cross sections and
(1) A setback of the steep cut slopes, enables
rounding of the lower part of the slope. This adds stability to the
slope and improves appearance through a design for transition from the
cut slope to the roadbed. It is complementary to rounding the top of the
cut slopes and this together with slope flattening serves to further the
appearance of the road fitting the natural terrain.
(2) Stabilizing the slopes will increase vegetative
cover on the steep cuts where it is needed so badly, and more plant
growth that is lost through clearing, can thereby be restored. This is
in the direction of making the least obtrusive development necessary to
road construction in the parks.
(3) A relatively low percentage of increase in first
cost is involved with transition cut slopes. This would be offset by a
decrease in maintenance. Even less increase in excavation quantities is
involved when the proposed cross sections are compared with sections of
the highest road standard (28 ft. shoulder to shoulder) for an entire
project, that has been constructed in the parka by the Bureau.
(4) The flat gutter section will improve appearance
in connection with the transition cut slope. This type of gutter will
give additional safety in the road through providing places where cars
can turn off the through lanes either in case of emergency or in regular
use for the purpose of stopping to admire natural features.
It is proposed that gutters and road shoulders be
constructed of coarser aggregate than the center seal coat pavement of
the road. The paved gutter will eliminate ditch erosion, and protect the
road surfacing, through sealing out surface water, thus eliminating
necessity for deep ditches designed to keep surface water at a lower
elevation than the bottom of the road surfacing material. With paved
gutters less cross section area of gutter is needed to carry surface
water and hence a shallower gutter can be used.
(5) Two types of drain inlets are presented for use
with paved gutter sections.
(6) To the larger extent the handling of subsurface
drainage is distinct from the handling of surface water. This affects
ditch design. The proposed cut slope cross sections are advantageous to
subsurface water conditions as affecting the roadbed.
(7) Flattening fill slopes in relationship to the
height of fill and percentage of ground slope will improve appearance in
fitting the road to the natural terrain.
More safety is added to the road, and in the case of
lower fills, flattening may permit of eliminating guard rail. Bottom of
fill slopes are to be rounded with steeper positive ground slopes. The
drawing will fill a need for definite instruction to contractors and
engineers in fill slope construction.
(8) For Improvement in appearance at moderate
revision in proposed for rounding the top of cut slopes. Flatter cut
slopes are proposed for construction in negative ground slopes.
(9) The design standards exclusive of surfacing cross
sections are shown on one sheet. It is proposed that this sheet or its
equivalent to fit a particular project, be included in the contract set
of drawings to explain the proposed design for cut end fill slopes.
- - - - - - - - - -
The new features of road design shown in the drawings
and described in this report are an approach to the kind of design that
landscape architects have done in private practice, and in public
projects when opportunity was presented to do this. Advancement has been
made in road design, and the engineering profession has a better
appreciation and recognition of standards for better appearance.
Likewise, we have learned from them.
Much of the design we have accomplished, has been
educational. Not only has the Bureau adopted those features in National
Park roads, but are also using many of them on Forest Highways. Some
have been used by State Highway departments in the West as roadside
development has gradually advanced to where it is now recognized as an
integral part of highway planning.
Of the several features described, it is felt that
the transition cut slope with it possibilities for further stabilization
and for restoration of plant growth, and the flatter gutter section are
most important. The other features are for the most part modifications
for improvement in standards that are functioning at present, and can be
more readily adapted.
These new design standards are submitted for your
review, and it is recommended that they be used as the basis for
improving design in the National Park roads.
Thos. E. Carpenter
Deputy Chief Architect