,
" •
SAFETY TREATMENT OF ROADSIDE DRAINAGE STRUCTURES
by
Hayes E. Ross, Jr. Dean Sicking T. J. Hirsch
Texas Transportation Institute
Harold D. Cooner John F. Nixon Samuel V. Fox
Texas State Department of Highways and Public Transportation
C. P. Damon Federal Highway Administration
Submitted to
TRB Committee A2A04 for
review and possible presentation and publication by TRB
August, 1981
,
H. E. Ross, et al 1
ABSTRACT
The purpose of the research was to develop traffic-safe end treatments
for (1) cross-drainage structures and (2) parallel-drainage structures that
would not appreciably restrict water flow. Cross-drainage culverts are used
to convey water under the highway. Parallel-drainage culverts are used to
convey water under driveways, side roads, ramps, or median crossovers that
abut the highway.
Preliminary designs were first evaluated by computer simulation, by use
of a test pit in which the clear open space and grate spacing could be var
ied, and by use of an earth berm similar in geometry to a driveway. Prom
ising designs were then subjected to full-scale prototype testing with both
subcompact and full-size automobiles.
Traffic-safe culvert end treatments can be achieved as follows:
Cross-drainage structures - (a) All culvert ends should be made to match the
existing side slope with no protrusion in excess of 4 in. (10.2 cm) above
grade; (b) culverts with clear openings 30 in. (76.2 cm) or less need no
safety treatment other than as mentioned in (a); (c) culverts with clear
openings greater than 30 in. (76.2 cm) can be made traffic-safe by grate mem
bers placed on 30 in. (76.2 cm) centers oriented parallel to the flow and in
the plane of the side slope.
Parallel-drainage structures - (a) The roadway side slope (or ditch slope)
shoul d be 6 to 1 or fl atter in the vi ci ni ty of the dri veway; (b) the dri veway
slope should be 6 to 1 or flatter; (c) the transition between the side slope
and the driveway slope should be rounded; and (d) safety treatment of the
culvert opening should include an end section cut to match the driveway slope
with cross members (grates) spaced every 24 in. (61.0 cm) perpendicular to
the direction of flow.
o
•
... :,
•
H. 1:. KOSS, et al 2
Application of these findings will result in improved safety for the
motorist. They will also result in more hydraulically efficient and more
economical safety treatments than have been used in the past.
I NTROD U cn ON
In designing drainage culverts, the primary objective is to properly
accommodate su rface runoff along the highway ri ght-of-way. However, a second
important goal should be to provide a traffic-safe design that would be tra-
versable by an out-of-control vehicle without rollover or abrupt change in
speed.
Guidelines for designing traffic-safe grates have been very limited.
NCHRP published guidelines for traffic-safe drainage structures in 1969 UJ. The recommendat ions dealt prima ri ly with the geometry of adjoi ni ng slopes.
Computer simulations have also been used to futher investigate the dynamic
behavior of automobiles traversing various slope and ditch configurations
near driveways and median crossovers (1,1). Criteria for the structural de
sign of inlet grates was published in 1973 (1). However, the study did not
address the problem of grate design as related to safety.
Recent field reviews of drainage culverts in Texas revealed that im-
provements and some modification of design details could improve both drain-
age and safety (~). Many of the safety grates used in the past to cover the
open ends of culverts have small openings and the grates are easily clogged
with debris, causing water to back up and flow over the roadway, the ditch
crossing, or adjacent property. In some cases safety grates do not possess
enough strength to be effective or they are used on small pipe culverts which
need no safety treatment.
I I I
I I I I I
I I i I 4 ,
.' •
H. E. Ross, et al 3
The objective of this study was to develop guidelines for safety treat
ment of both cross-drainage and parallel-drainage structures that (1) can be
safely traversed by an errant vehicle and (2) will exhibit desirable hydrua
lic behavior. Although no hydraulic analyses were made it was assumed that
hydraulic efficiency increases as the number of grate members decrease. It
was therefore a goal of the research to meet safety requirements with as few
grate members as possible.
This paper summarizes the findings of two research studies, one conduc-
ted in 1979 (i) and the other in 1980 (2). Reference should be made to the
cited literature for complete details of the studies.
EVALUATION CRITERIA
A review of the literature showed that there are no nationally recog-
nized safety performance standards for roadside drainage structures. Decele
ration and stability of a vehicle during and following impact are the two
primary measures of performance for safety appurtenances such as guardrail s,
crash cushions, etc. (~). For the cross-drainage structures, performance was
judged satisfactory if the vehicle smoothly traversed the culvert and the ad
joining ditch slope without rollover for speeds from 20 mph (32.2 km/h)
through 60 mph (96.5 km/h).
Previous research (£&) indicated that a very flat ditch slope, a very
flat driveway slope, and a very long culvert would be necessary to satisfy
the above criteria for parallel-drainage structures. In view of the economic
and hydraul i c imp 1 i cat ions of such a desi gn it was conc 1 uded that tradeoffs
would be necessary to achieve an acceptable balance between the controlling
elements. Performance of parallel-drainage structures was therefore judged
acceptable if the vehicle smoothly traversed the adjoining slopes and culvert
j I i 1 !
I
I I i I ! , ~ i I ~
I I
I I I I , ~ ! ~ !
~
H. E. Ross, et al 4
without rollover for speeds from 20 mph (32.2 km/h) through 50 mph (96.5
km/h) •
RESEARCH APPROACH
A three-phase approach was taken in the development of safety treatments
of both cross-drainage and parallel-drainage structures. In the first two
phases computer simulations in combination with a preliminary test program
were used to develop tentative design concepts. In the latter phase proto-
types were constructed using the rl'!sults of the preliminary studies and
tested under representative roadside configurations.
Cross-Drainage Structures
Simulation studies - A computer simulation study was conducted, using
the Highway-Vehic1e-Object-Simu1ation-Mode1 (HVOSM) (1) , to evaluate wheel
drop into various culvert openings on flat terrain. HVOSM was also used to
investigate the effect a ramp at the leading edge of the culvert opening
would have on vehicle behavior. Ramps having the following dimensions were
eva 1 uated:
Hori zonta 1 Ve rt i ca 1
Ramp Dimension (i n.) Dimension (in.)
1 3.0 3.0
2 6.0 3.0
3 6.0 6.0
4 12.0 6.0
A 1974 Honda Civic was simulated in each of the computer runs since it was
assumed a mini-size automobile would be more critical than a larger vehicle
for the given conditions. A speed of 20 mph (32.2 km/h) was used in each run
since it was deemed a critical speed. At higher speeds it was felt the
, I i
I I
I I ! i
I I I , ! I , ! ; I
I !
•
H. E. Ross, et al 5
vehicle would clear the opening easier. At lower speeds, although the vehi
cle would tend to drop more, velocity changes would be tolerable.
Preliminary tests - In the second phase a test pit was constructed on
flat terrain as shown in Figure 1 to study the behavior of a vehicle as it
traversed various openings. The objectives of these tests were to determine
preliminary values for (1) the maximum clear opening permissible on a non-
grated culvert end and (2) the maximum spacing permissible when grates are
necessary. All runs were live-driver tests at various speeds and encroach-
ment angles. Figure 2 is a photograph of the test pit after installation. A
total of 31 runs were made to determine the maximum clear opening. Test
speeds ranged from 5 mph (8.0 km/h) to 35 mph (56.3 km/h);encroachment angles
varied from 0 degrees to 15 degrees; and clear openings ranged from 12 in.
(30.5 cm) to 36 in. (91.4 cm). All tests were with a 1974 Honda Civic having
a curb weight of approximately 1800 lb (817.2 kg). Limiting values were
determined by the severity of the ride as judged by the driver. The driver
was a Texas Transportation Institute (TTl) technician with a nonprofessional
driving history. Sequential photos of a 20 mph (32.2 km/h) run with a 30
in. (76.2 cm) clear opening are shown in Figure 3.
Upon completion of the clear opening tests the pit was used to determine
maximum permissible grate spacing. A total of 22 live-driver tests were con-
ducted for this purpose. Test speeds ranged from 5 mph (8.0 km/h) to 25 mph
(40.2 km/h); encroachment angles varied from 0 degrees to 30 degrees; and
grate spacing varied from 16 in. (40.6 cm) to 30 in. (76.2 cm). The grates
were 3 in. (7.6 cm) schedule 40 steel pipe anchored to a steel beam with pro-
vision to allow adjustments of the pipe to any desired spacing. Figure 4
shows the pit setup for a 16 in. (40.6 cm) grate spacing. Each grate config-
uration was evaluated with the 1974 Honda Civic. A 1975 Plymouth Fury
I I I I
I i i ! I I I ~
I I , i I ! i i i I ~ i
I ! I I i ! ! ! ~ , !
I I I
I I I I I r· ~ ! I ~
•
II. ~. I'\U;:';:', ~t.. ell 6
" .....
.. .;.,.....
10'-0" Clear
Fl at Terra i n~Ii--_.~\' I . "':
S-o ttl
I OJ - ~ U') U
3"<P Std. Pipe Grating; Spacing Varies
Edge of Pavement •... ~ .. (: .. <:; • ...
'. . ' ". ".
18" Deep Concrete Pit
~_ Adjustable Cover Plate to Create Clear Opening
Variable Clear Opening
.' .. ;. ~ ..
\ \
'. . .... . '.." ~
Vehicle ~; Encroachment Angle Varies
Figure 1. Plan View of Culvert Test Pit.
ANGLE: IS 5 PEE O:l0l: OPN'G 22 PL,
Figure 2. Test Pit Installation.
•
I
I I I i I I
I ~ I
I I I 1
I , I !
I * ! .~
~ 1 1 ~ I I I I I j !
I !
i I j
I I ~.
i
! I I ~ I
I I ! , I , !
I
H. E. Ross, et al
0.000' sec.
0.060 sec.
0.135 sec.
ANGLE: 0
SPEEO:20,: OPN°G;J.O Pl.
0.030 sec.
0.105 sec.
Figure 3. Sequential Photos of Nongrated Culvert Test, 30 in. Clear Opening, 1974 Honda Civic
7
I ! I I I ,
I i I , I I
I I
I I ~ !
I ~ , , ~ I ~
I I I i
i I I I [ I
I i ~ I
H. t. KOSS, et al
• Figure 4.
"
ANGLE: 0 SPEED:lDf DPN'G,ISPIPE
Test Pit With 16 in. Grate Spacing.
8
H. E. Ross, et al 9
weighing about 4500 Ib (2043 kg) was also used to evaluate the larger grate
spadngs.
As part of the second phase of the study a limited number of live-driver
tests were conducted to further evaluate the effects of a ramp at the leading
edge of the cul vert openi ng. Based on the HVOSM results, a ramp with a hori
zontal dimension of 12 in. (30.5 cm) and a vertical dimension of 6 in. (15.2
cm) was selected and constructed. HVOSM indicated this combination would
produce the greatest wheel hop of all combinations considered. The 1974
Honda Civic and the 1975 Plymouth Fury were used in the ramp test. Each test
was conducted at 20 mph (32.2 km/h).
Prototype tests - Based on results obtai ned from the prel imi nary stud
i es, two cul vert structures were constructed for full-scale testing. They
consisted of a 30 in. (76.2 cm) diameter corrugated steel pipe culvert and a
5 ft (1.5 m) wide by 3 ft (0.92 m) high concrete b.ox culvert with adjoining
head and wing walls. Grate members on the box culvert consisted of 3 in.
(7.6 cm) schedule 40 steel pipe on 30 in. (76.2 cm) centers. Photos of both
installations are shown in Figure 5.
General details of the six tests conducted are shown in Figure 6. Note
that the culverts were subjected to tests with both the mini-size and full-
size automobiles. In each test, with the exception of test 5, all four
wheels of the test vehicle crossed the sloped culvert opening. In test 5 the
vehicle straddled the cross member at the end of the box culvert, allowing
the left side wheels to drop approximately 1.5 ft (0.46 m) to the ditch·
bottom. Sequential photos of test 3 are shown in Figure 7.
Analysis of the strength requirements of grate members indicated that a
3 in. (7.6 cm) J.D. schedule 40 pipe was adequate for spans up to 12 ft (3.7
m). Since grate spans on many box culverts would exceed 12 ft (3.7 m) it was
, I I I ! I ! , !
I i ~ I I I I , , I I i ! , i ,
H. E. Ross, et al 10
a) Corrugated Steel Pipe Culvert.
b) Grated Box Culvert.
Figure 5. Prototype Test Installations.
•
H. E. Ross, et al
Make Model Year Test Weight (1 b) Test No . Velocity (mph)
, .•...
.~
Cl
'-4J , I
Y
VEHICLE DATA
Car 1
Honda CVCC 1974 1800 2, 3, 5 20
'j ..
Car 2
Plymouth Fury 1975 4500 1, 4, 6 20, 60
Metric Conversions:
1 lbm = 0.454 kg
1 mph = 1.609 km/h
, . "-:.;.~"~ ." Edge of Run~/ay , ..
):': '. Ditch Front S19pe Approx. 5: 1 J .~ • ,It _. : .. ~~ "
1~~12;-~Pipe Culvert , .
. ' ... . .. ,; .,'r-:'"
20 mph, Tests 1 and '.2
Box Culvert with Grating '. '.,
.. ':: -:: . , I
~vehicles towed with cable
t V = 20 mph, Tests 3', 4, and 5. r-- 60 mph, Test g.
Figure 6 Plan View of Site for Prototype Tests.
11 !
i I I ~ I I
I I I !
i I I I I I I ,
I I ! I I I
! I I I !
! [
I
I i I
I I I I ! I I ~ I I
I I I ! I I
I I ~ I ,
I~
0.000 sec 0.092 sec
0.174 sec 0.353 sec
0.530 sec
Figure 7. Sequential Photos, Test3.
I I i
I I I
I I i
I I , , ~ ~
~ I ~ ~
~ i I I I L I I ! ! I I I , I ; ! i
I I I I I ! I ~ I I ! ~ i i ~ ;1 ,
H. Eo Ros s, et a 1 13
conclude.d that a limited test program should be undertaken to determine pipe
size requirements for larger spans. To accomplish this another test pit was
constructed on flat terrain. The pit was of 20 ft (6.1 m) long, 10 ft (3.1
m) wide, and 1.5 ft (0.46 m) deep. A total of four full-scale vehicular
tests were conducted with a 4500 1b (2043 kg) vehicle, each at 20 mph (32
km/h) and each at a head-on approach, perpendicular to the 20 ft (6.1 m)
dimension of the pit. Further details of each test are given in Table 1, in
cluding the permanent deformations noted after each test. With the exception
of test 4 the grates had a 20 ft (6.1 m) clear span. In test 4, vertical
supports consisting of 3 in. (7.6 cm) 1.0. schedule 40 pipe were placed at
midspan of each of the three grate members. The grates were attached to the
walls of the pit with a pin connection, constructed according to TSDHPT stan
dards.
Parallel-Drainage Structures
Simulation studies - Design of a traffic-safe parallel drainage struc
ture not only involves the culvert itself but adjoining slopes as well. In
fact, the slopes can in many cases be a greater hazard than the culvert
structure. Studi es of medi an cross-over geometry poi nted to the need for re
latively flat slopes to minimize vehicle rollover (~,l). To gain further in
sight, HVOSM was used to examine the behavior of·a vehicle traversing various
driveway conditions. Parameters investigated included departure angle,
departure speed, and the path of vehicle encroachment; the side slopes of
both the ditch and the driveway; the type of transition zone between the two
slopes; depth of the ditch; and vehicle size. These parameters are illus
trated in the definition sketch of Figure 8.
Following is the range of each parameter evaluated:
DEPARTURE ANGLE: .150 and head-on
H. E. Ross, et a1 14
Table 1 . Cross Member Deflections of Box Culvert Grating Strength Tests
TEST PIPEa GRATE EST. NO. 1.0. MEMBERb VERTICAL DEFLECTION HORIZONTAL DEFLECTION
( in. ) (in. ) (in.)
{ First -0 0
1 5 Second -0 0
Third -15/16 3/8
{ First -1/8 0
2 4 Second -1/2 1/4
Third -3 1 7/8
{ First -1 3/4 2 7/8
3 3.5 Second -4 3/4 3 1/16
Third -4 1/8 1 7/16
{ First -0 3/4 1 1/2
4c 3 Second +0 1/2 1 7/8
Third +0 1/8 4 3/4
a. Schedule 40 Steel Pipe.
b. Grate Members Spaced on 30 in. Centers.
c. Midspan Vertical Supports Used On Each Grate.
H. E. Ross, et a1
. SHOULDER
,.;. SIDE
SLOPE)
"--,
. DITCH J
/' I // )-~/~--------~
/ I
/ / / / DRIVEWAY /
/
/~--/~/------------------------~-------4 ( ./
TRANSITI0!;J, ... ~ ./ ZONE SEE DETAILS", a ~ (TYPICAL FOUR , PLACES) ~'----~--------------------~r-----~
DRIVEWAY~
ANGLED PATH (I)
ANGLED PATH (2)
DEPARTURE ANGLE
'~
PLAN VIEW
Figure 8. Definition Sketch
15
H. E. Ross, et a1
(
DITCH DEPTH
SECTION ''A-All
SECTION "8-811
LINES OF EQUAL ELEVATION
ABRUPT TRANSITION
DETAIL SMOOTH TRANSITION
DETAIL 2
Figure 8. Definition Sketch (continued)
16 ! ~ I ~ ;
I 1 I
1 I I 'l I I ~ , 1
! I I I
I I , I
I I I I ! I I I
I i I I I ~ I ~ ~ ~
I i I I I ! I I i ~
I I ] !
I
I I 1 I ~ , j I ~ I I ~ I I 1 I ! I !
o
H. t. KOSS, eo ctl
ters.
DEPARTURE SPEED: 30 mph (48.3 km/h), 40 mph (64.4 km/h), 50 mph (80.5
km/h), and 60 mph (96.6 km/h)
PATH: 150 angled path across transition (path 1), 150 angled path
across ditch bottom (path 2), and head-on path into driveway
slope (path 3)
ROADWAY SLOPE: 4:1 and 6:1
DRIVEWAY SLOPE: 4:1,5:1, and 6:1
TRANSITION TYPE: Abrupt and rounded
DITCH DEPTH: 2 ft (0.61 m) and 3 ft (0.92 m)
VEHICLE SIZE: 2250 lb (1022 kg) and 4500 lb (2044 kg)
A total of 68 computer runs were made to eva 1 uate the va ri ous pa rame-
Preliminary tests - Ten full-scale vehicular tests were conducted to
(1) evaluate vehicle response as a function of the driveway slope and (2) to
develop a tentative safety treatment for parallel-drainage structures. The
test vehicles were 1974 and 1975 Chevrolet Vegas weighing approximtely 2250
lb (1022 kg). In each test the vehicle was towed to the test site along a
guidance cable, released, and then allowed to traverse the test area in a
free-wheel (no steer input), no-braking mode. A summary of the 10 tests is
given in Table 2, tests 1-1 through 7-6. Tests 1-1 through 5-1 were designed
to evaluate the relative hazard of the driveway slope. An earth berm was
constructed to simulate the driveway. The berm for tests 1-1 through 1-4 had
a 3.8 to 1 slope, was approximately 3 ft (0.92 m) high, and was approximately
20 ft (6.1 m) wide at the top. Sequential photos of test 1-4 are shown in
Figure 9.
After test 1-4 the berm slopes were flattened to the dimensions shown on
the first page of Figure 10. In this case the slope on the approach side was
j 1
I , r ! i I I I I
I I t
I I
I I ! ! ! [
I I ~ I !
"
H. E. Ross, et al 18
6.7 to 1. It was obvious from test 1-3 that an automobile could traverse the
6.7 to 1 slope at speeds in excess of 40 mph (64.4 km/h) without roll i ng
over. Hence, test 5-1 was conducted at 50 mph (80.5 km/h) with the automo
bile approaching from a head-on path. Although the vehicle was airborne for
approximately 75 ft (22.9 m) it remained upright with no appreciable
pitching.
The next seri es of tests (7-1 through 7-6) were conducted to determi ne
if safety treatment of the culvert end was needed in addition to the sloped
end treatment. The 6.7 to 1 driveway slope was used in each test. It was
assumed that a head-on path into the driveway culvert would be as critical,
or more critical, than any other path regarding the culvert itself. Based on
this assumption, a 24 in. (61.0 cm) diameter corrugated steel pipe culvert
with a sloped end was installed in the earth berm as shown on the first page
of Figuere 10. This culvert size was selected since the diameter of most
driveway culverts in Texas are equal to or less than 24 in. (61.0 cm). Vehi
cle impact point for this series of tests was selected such that the right
side wheels of the test vehicle traversed the center of the culvert end.
Details of the culvert configuration for each of the culvert tests are
given in Figure 10. Test 7-1 was conducted at 50 mph (80.5 km/h) with an
open culvert, i.e., no grate members. Photos of the installation are given
in Figure 11 and sequential photos of the test are given in Figure 12. Large
pitch and roll rates occurred after impact with the culvert, and the vehicle
rolled over. In test 7-2 a single grate member was placed across the culvert
as shown in details 3 and 4 of Figure 10. Very little improvement in vehicle
behavior was realized and rollover again occurred.
Analysis of test 7-2 showed that grates spaced approximately on 2 ft
(0.61 m) centers was needed to avoid excessive wheel drop and wheel snagging.
H. E. Ross, et al 19
The next treatment therefore incorporated thi s featu re as shown in detail s 5
and 6 of Figure 10. Grate members consisted of 2 lb/ft (2.98 kg/m) steel
flanged channel sections. The channel section was chosen since it is widely
used as a del ineator post by TSDHPT and woul d therefore be readily avail
able. The first test on this treatment, test 7-4, was conducted at 20 mph
(32.2 km/h) and the results were acceptable. Test 7-5 was conducted at 50
mph (80.5 km/h) and rollover occurred due to structural failure of the
grates.
In test 7-6, 2-1/2 in. (6.35 cm) 1.0. standard steel pipe (schedule 40)
were used as a grate member. Details 7 through 10 of Figure 10 show how the
pipe was attached to the culvert. Although .the vehicle was airborne approxi
mately 65 ft (19.8 m) it remained upright and the test was deemed accept
able. The culvert was only slightly damaged.
Prototype tests - The final two tests, 9-1 and 9-2, were selected to
verify the tentative conclusions r.eached as a result of the simulation work
and the full-scale slope and culvert testing. A full-scale prototype of a
ditch-dri veway confi gu rat i on was constructed as shown in Fi gu re 13 and the
photos of Figure 14. Test 9-1 was conducted at 40 mph (64.4 km/h) and the
approach path into the driveway was as shown in Figure 13 such that the left
side wheels crossed the culvert. No adverse vehicle behavior occurred during
the test and the results were considered acceptable.
Test 9-2 was identical to test 9-1 except the speed was increased to 50
mph (80.5 km/h). Sequential photos of the test are shown in Figure 15. The
vehicle remained upright and sustained only minor damage. The culvert was
only slightly damaged and could have been used without repair.
! I I
I I I
I I I ~
I 1
I I I ~ j
I i I I ! I
I
I I ~ I j
·1 1 ! !
~ tl'
TABLE 2. SUMMARY OF FULL-SCALE TEST RESULTS
TEST VEHICLE VEHICLE DRIVEWAY DITCH CULVERT NO. SPEED PATH SLOPE SLOPE CONFIGURATION RE SUL TS
(mph) (See Fig. 1) ------- --
1 - 1 30 3 3.8; 1 N/A . No Cu1 vert Satisfactory - no ro1iover
1-2 35 3 3.8; 1 N/A No Culvert Satisfactory - no rollover
1-3 40 3 3.8; 1 N/A No Cu1 vert Satisfactory - no rollover
1-4 50 3 3.8; 1 II/A No Cu1 vert Unsatisfactory - vehicle pitched over
5- 1 50 3 6.7;1 N/A No Cu1 vert Satisfactory - no rollover
7-1 50 3 6.7;1 N/A (See Fig. 10) Unsatisfactory - vehicle rolled over
7-2 50 3 6.7; 1 N/A (See Fig. 10) Unsatisfactory - vehicle rolled over
7-4 20 3 6.7; 1 N/A (See Fig. 10) Satisfactory - no rollover
7-5 50 3 6.7: 1 N/A (See Fig. 10) Unsatisfactory - vehicle rolled over
7-6 50 3 6.7;1 N/A (See Fig. 10) Satisfactory - no rollover
9-1 40 2 6.5;1 6.8: 1 (See Fig. 13) Satisfactory - no rollover
9-2 50 2 6.5; 1 6.8; 1 (See Fig. 13) Satisfactory - no rollover
Metric Conversions: 1 mph = 1.609 km/h
::J:
fT1
;;0 0 VI VI . CD <+
'" ~
'" o
H. E .Ros s, et a 1
0.091
0.402
r
1.018
.-, .
1.777
figure 9. Sequential Photos, Test 1-4.
21
0.332
I ntr
0.775
1.395
I f
I I i
I i I ! I
I I I i f,
" I
i I I I I
I I i ~ i I !
H. E. Ross, et al
3 - RAIILRC)AD ____ T1ES--
PLAN ..
22
24'-5"
BERM
FOR TEST 7-1 SEE DETAILS I AND 2 FOR TEST 7-2 SEE DETAILS 3 AND 4 = FOR TEST 7-48& 7-5 SEE DETAILS 5 a 6 I
_~::--~FOR TEST 7-6 SEE DETAILS 7 AND 8 '-t ! =jjll-=i1i~II§ilil;~. i~ 1Ii1- A ltllggojill i ----==J1I1~~-- -. - RAILROAD 1I11'§JlTI -
-Sllil -.. TIES 1111_
SECTION "A II
3/4" GAL\/. CHANNEL I CRIMPED a SPOT WELDED 7'-0· (TYPICAL . TEST) ~;;-:::S-__
t . Ill! llil~- i1 ==.L- 1 = 1 ~\l1
~ .:::::1111=_1 \I llli =
, "2" 7.::oTIlIUIII ~IIII IIIF" DETAIL I
24" OIA. GALV.
~L~~...-lil
ti~~kr:-;-7-:-7· 770'l/ / /./ I' ; I
/ I· / ,
,//1/ ,1//
.. /i III· / //.///.;/ ill / .// I ,/ / /' , .' I
CORRUGATED PIPE---
3/4" GAL\/. ANGLE CRIMPED a SPOT WELDED (TYPICAL EA. TEST)
DcTAIL 2
Figure 10. Berm and Culvert Details, Tests 5-1 through 7-6.
13
,.
H. E. Ross, et al
7'-0"
===:1\= - Jill
-=D=E=TA:-:I:--L--:3=---:2 1/2" DIA. STEEL .... ,~~'"" (SCHEO. 40)
7'-0·
.:D:E1i::~:;;-;I~L:;5~~ 4 - DELINEATOR POSTS EQUALLY SPACED
NOTE: SEE DETAIL 9 1-____ 8~'-::::.0_., ___ ___I. / FOR WELD LOCATiONs
Figu!'e 10.
=~~t ~/:..--:.---~~ ;
. DETAIL 7
i . , ! / ' ;'
, , / ,
II =\ -1\\\_ 1\\1 1111----
1111 4- 2 1/2" DIA. STEEL PIPE (SCHED. 40) EQUALLY SPACED
Berm and Culvert Details, Tests 5-1 through 7-6. 14
23
DETAIL 4
DETAIL 6
DETAIL 8
(continued)
, I I
I I
I I I I I I !
I . i I I I I ~ I I I I
, , i ~ i1 I R
I I ; J !
I I I , i
I I I , I I !
I ~ i
i , ~ i I I , ~
I I
H. E. Ross. et al
I=-~---=::-=~ ~ - - - - - - -I I I I I I I I I I I I I
(/) w a:: ~
I I I
I I I
I I I
I
24
_--- 2 1/2" DIA. STEEL PIPE (SCHED. 40)
3/4" GAL\/. CHANNEL
~ I I
i:=;.=-~LjE:~=========ti===:::::==:~
PLAN
DETAIL 9
-DETAIL 10
1/4
/
TYP. EA. FOR ALL
/
END PIPES
Figure 10. Berm and Culvert Details. Tests 5-1 through 7-6. (continued) 15
H. E. Ross,et al 25
Figure-ll~ Test Installation Before Test 7-1.
Ii , " ! ~ i !
I I I i I I I , l ~ ~
I ! I I I ~ ~
I , i
i I ~ ffi ',j
I I
H. E. Ross, et a1 26
0.000 0.177
0.430 0.531
0~911 1.214
•
1.518 2.100
Figure 12. Sequential Photos, Test 7-1.
. .
H. E. Ross, et a1
24" DIA. GAL\l CORRUGATED PIPE WITH 4 - 2 1/2" DIA. STEEL. PIPES (SCHED 40)
=
PLAN
SECTION "A"
AVERAGI:: SLOPE 6.54· I
= -'
.:,
AVERAGE SLOPE 7.14' [
27
AVERAGE SLOPE 6.75. [
Figure 13. Test Site Conditions, Tests 9-1 and 9-2.
I
I I i
I I
I I I I I
I i I i
H. E. Ross, et a1 28
•
Figure 14. Test Site, Tests 9-1 and 9-2.
I , I
I I .I I i !
I I I I ! ! ! ! J I ~
I , j
I I I I !
I I I
i , I ! ~
I i ~
I I , !
I j j
~ I I w
i " 1 @
r
H. E. Ross, et al 29
r"
Figure 15. Sequential Photos, Test 9-2.
•
H. E. Ross, et al 30
FI ND INGS
Cross-Drainage Structures
Based on the computer simulations and the preliminary test program, it
was shown that clear openings of at least 30 in. (76.2 em) could easily be
tra ve rsed at a speed of 20 mph (32.2 km/h). A 36 in. (91.4 em) spaci ng was
easily traversed at 25 mph (40.2 km/h). For clear openings in excess of 30
to 36 in. (76.2 to 91.4 em) it was shown that grates spaced on 30 in.
(76.2 em) centers would provide satisfactory safety treatment. These find
ings were in fact borne out through six full-scale prototype tests. Tests of
a 30 in. (76.2 em) diameter corrugated steel pipe culvert end, cut to match a
5 to 1 side slope, were successfully conducted. The culvert opening was
readily traversed by both a full-si ze automobil e and a mi ni -s i ze automobil e
at 20 mph (32.2 km/h). Tests of a relatively large box culvert constructed
to match the existing 5 to 1 side slope also verified that grates spaced on
30 in. (76.2 em) centers provide a satisfactory safety treatment. Tests of
this treatment at 20 mph (32.2 km/h) and 60 mph (96.5 km/h) by both full-size
and mi ni -s i ze automobil es were conducted. I twas al so shown that the grates
should be extended and anchored at the flow line to avoid any appreciable
dropoff at the end of the culvert treatment. In one test, vehicle rollover
occurred when the left side wheels dropped off an 18 in. (45.7 em) opening at
the end of the culvert.
Preliminary tests and the prototype tests showed that 3 in. (7.6 em)
I.D. schedule 40 steel pipe grates were of sufficient strength to support a
full-size automobile for simple-supported spans up to approximately 12 ft
(3.7 m). Additional full-scale tests were conducted with a test pit to
determine pipe size requirements for larger spans. Results of these tests
provided the following guidelines:
I , I i !
I I I I i , I I I ! I I I~
~ I ! I
I ! !
I I ill
>
H. E. Ross, et al 31
Suggested Standard
Schedule 40 Pipe Size
Sean Length (ft) 1.0. (in.)
Up to 12 3.0
12 to 16 3.5
16 through 20 4.0
If midspan vertical supports are used, 3.0 in. (7.62 em) 1.0. standard
schedule 40 pipe can be used for spans up to 20 ft (6.1 m). Other sections
having equivalent strengths could of course be used. Reference may also be
made to an FHWA report (1) for strength requi rements of gr.ates.
Results of the study to evaluate the effect of a ramp at the leading edge
of a culvert opening were inconclusive. HVOSM results indicated that appre
ciablewheel hop could be achieved by a small ramp, thus enabling the vehicle
to clear larger culvert openings. An attempt to verify these findings via a
full-scale test program was made. However, due in part to the test proce
dure, the tests did not provide sufficient data to reach any firm conclu
sions. To minimize damage to test vehicles the area behind the ramp was not
excavated and as a consequence the total wheel drop that would have occurred
otherwise was unobtainable. Further evaluation and testing of ramp treat
ments appear warranted.
Parallel-Drainage Structures
Based on the computer simulations and the preliminary test program it was
shown that the driveway slope should be 6 to 1 or flatter to avoid vehicle
rollover for speeds up to 50 mph (80.5 km/h). The computer simulations indi
cated that the ditch side slope should also be 6 to 1 or flatter. Even at
these relatively flat slopes a vehicle traveling at 50 mph (80.5 km/h) will
become airborne for approximately 65 ft (19.8 m). The computer simulations
H. E. Ros s, et a 1 34
4. Grate members should extend to and be anchored at the flow line. Drop
offs at the end of the culvert should be avoided.
5. Necessary grate member sizes will depend on the span of the grates, the
manner in which the grates are supported and the design vehicle weight.
To support a full-size automobile the following sizes or their equivalent
are adequate.
Suggested Standard
Schedule 40 Pipe Size
S~an Len~th (ft 1 1.0. (i n.l
Up to 12 3.0
12 to 16 3.5
16 through 20 4.0
A 3.0 in. (7.62 cm) 1.0. standard schedule 40 pipe can be used for spans
up to 20 ft (6.1 m) if a midspan vertical support is used.
Parallel-Drainage Structures (for driveways, median crossovers, ramps, etc.)
1. The roadway side slope (or ditch slope) in the vicinity of the driveway
slope should be 6 to 1 or flatter.
2. The driveway slope should be 6 to 1 or flatter.
3. The transition area between the roadway side slope and the driveway slope
shoul d be rounded or smoothed as opposed to an abrupt trans it ion.
4. Safety treatment of the culvert opening should include an end section cut
to match the driveway slope with cross members (grates) spaced approxi
mately every 24 in. (61.0 em) perpendicular to the direction of flow.
5. The cross members should be designed to support a concentrated wheel load
of approximately 10,000 lb (44,480 N) applied at midspan.
I I I I
I I I I I g
I I I
, I I I ! I I I
I I I '~
ii
I I I i I I I ~ ! ,~
i ~ ~' .
: ~ ! ~
<I
t
•
l.
2.
REFERENCES
"Traffic-Safe and Hydraulically Efficient Drainage Practices," National
Cooperative Highway Research Program -- Synthesis of Highway Practice 3,
Highway Research Board, 1969.
Ross, Hayes E., Jr., and Post, E. R •• "Criteria for the Design of Safe
Sloping Culvert Grates," Research Report 140-3, Texas Transportation
Institute, Texas A&M University, August 1971.
3. DeLeys, N. J., "Safety Aspects of Roadside Cross Section Design," Report
No. FHWA-RD-75-41 Federal Highway Administration, February 1975.
4. Ballinger, C. A. and Gade, R. H., "Evaluation- of the Structural Behavior
of Typical Highway Inlet Grates, with Recommended Structural Design Cri
teria," Report No. FHWA-RD-73-90, Federal Highway Administration, Decem
ber 1973.
5. "Improving Safety of Drainage Facil ities," Administrative Circular No.
8-79, Texas State Department of Highways and Public Transportation, Janu
ary 1979.
6. Ross, Hayes E., Jr., Hirsch, T. J., Jackson, Benito, Jr., and Sicking,
Dean, "Safety Treatment of Roadsi de Cross-D rai nage Structures," Research
Report 280-1, Texas Transportation Institute, Texas A&M University, March
1981.
7. Ross, Hayes E., Jr., Hirsch, T. J., Sicking, Dean, "Safety Treatment of
Roadside Parallel-Drainage Structures," Research Report 280-2F, Texas
Transportation Institute, Texas A&M University, June 1981 •
8. "Recommended Procedures for Vehicle Crash Testing of Highway Appurte-
" nances," Transportation Research Circular 191, February 1978.
9. Segal, David J., "Highway-Vehicle-Object-Simulation Model -- 1976," Re
port No. FHWA-RD-76-164, Federal Highway Administration, February 1976.