You are on page 1of 23

171

4.4 171

4.4 Design of Dowels and Joints

should be based on the full strength of the bar, namely,

t = 2( If s)

p.20

in which tis the length of the tie bar, (.1. is the allowable bond stress, A1 is the area of

one bar, and 20 is the bar perimeter. For a given bar diameter d, A1 = 7rd2/4 and

20 = 1rd, so Eq. 4.39 can be simplied to

400

The length I should be increased by 3 in. (76 mm) for misalignment.

It should be noted that many agencies use a standard tie-bar design to simplify

the construction. Tie bars 0.5 in. (13 mm) in diameter by 36 in. (914 mm) long spaced at

intervals of 30 to 40 in. (762 to 1016 mm) are most commonly used.

(4.39)

(4.40)

Example 4.11:
Same pavement as Example 4.10. Determine the diameter, spacing, and length of the tie bars
re-

quired, as shown in Figure 4.21.

24ft

Diameter, spacing and length I

of tie bars ?

FIGURE 4.21

Example 4.11 (1 in. = 25.4 mm,

1ft = 0.305 m).

Solution: Assume fs = 27,000 psi (186 MPa) for billet steel (see Table 4.1). With U = 12

ft = 144 in. (3.66 m), from Eq. 4.38, A, = 0.0868 x 8 x 144 x 1.507.000 = 0.00556 in.2lin.

If No.4 (0.5 in. or 1.2 mm) bars are used, from Table 4.2, the cross-sectional area of one bar is

0.2 in.2 (129 mmz). The spacing of the bar = 0.2/0.00556 = 36 in. (914 mm).

Assume that p. = 350 psi (24 MPa), from Eq. 4.40, t = 0.5 X 27,000 X 0.5/350 = 19.3 in.

(353 mm). After adding 3 in. (76 mm), t = 19.3 + 3 = 22.3 in. (use 24 in. or 610 mm).The design

selected is No. 4 deformed bars, 24 in. (610 mm) long and 3 ft (0.9 m) on centers.

DESIGN OF DOWELS AND JOINTS

The design of dowels and joints is mostly based on experience, although some theoretical

methods on the design of dowels are available. The size of dowels to be used depends on
the thickness of slab. Table 4.4 shows the size and length of dowels for different slab

thicknesses as recommended by PCA (1975). It can be seen that the diameter of dowels is

equal to one-eighth of the slab thickness In a recent edition of joint design, PCA (1991)

recommended the use of 1.25 in. (32 mm) diameter dowels for highway pavements less

172
Chapter 4

4.4.1

Stresses and Deflections in Rigid Pavements

TABLE 4.4 Recommended Dowel Size and Length

Slab thickness Dowel diameter Dowel length

(in) (in) (in)

5 g 12

6 3 14

7 g 14

8 1 14

9 1,1, 16

10 1% 18

11 1%; 18

12 1% 20

Note. All dowels spaced at 12 in. on centers, 1 in. - 25.4 mm.

Source. After PCA (1975).

than 10 in. (254 mm) thick and 1.5 in. (38 mm) diameter dowels for pavements 10 in.

(254 mm) thick or greater. A minimum dowel diameter of 1.25 to 1.5 in. (32 to 38 mm)
is needed to control faulting by reducing the bearing stress in concrete.

Design of Dowels

Dowel bars are usually used across a transverse joint to transfer the loads to the ad-

joining slab. The stress and deection at the joint are much smaller when the loads are

carried by two slabs, instead of by one slab alone. The use of dowels can minimize fault-

ing and pumping which has been considered by the Portland Cement Association

(PCA, 1984) as a factor for thickness design.

Allowable Bearing Stress Because concrete is much weaker than steel, the size and

spacing of dowels required are governed by the bearing stress between dowel and con-

crete.The allowable bearing stress can be determined by Eq. 4.41 (American Concrete

Institute, 1956),

4-d,

in which fb is the allowable bearing stress in psi, d is the dowel diameter in inches, and

f g is the ultimate compressive strength of concrete.

(4.41)

Bearing Stress on One Dowel If the load applied to one dowel is known, the maxi-

mum bearing stress can be determined theoretically by assuming the dowel to be a

beam and the concrete to be a Winkler foundation. Using the original solution by Tim-

oshcnko, Friberg (1940) indicated that the maximum deformation of concrete under

the dowel, as shown by yo in Figure 4.22, can be expressed by

_R@+Bd

y" 4B3 Edi.


in which yo is the deformation of the dowel at the face of the joint, R is the load on one

dowel, z is the joint width, Ed is Youngs modulus of the dowel, Id is the moment of

(4.42)

173
4.4 Design of Dowels and Joints 173

Dowel

FIGURE 4.22

,2 Dowel deformation under load.

inertia of the dowel, and B is the relative stiffness of a dowel embedded in concrete.

Note that

=_44

Id i I 1rd (4.

and

Kd

= 4 .44

B 4Ed1d (4 )

in which K is the modulus of dowel support, which ranges from 300,000 to 1,500,000 pci

(81.5 to 409 GN/m3), and d is the diameter of dowel. The bearing stress 0., is propor-

tional to the deformation:


KPKZ + 32)

433Ed1d

The bearing stress obtained from Eq. 4.45 should compare with the allowable bearing

stress computed by Eq. 4.41. If the actual bearing stress is greater than allowable, then larg.

er dowel bars or smaller dowel spacing should be used. Recent studies have also shown

that the bearing stress is related to the faulting of slabs, as described in Section 12.1.4. By

limiting the bearing stress, the amount of faulting can be reduced to the allowable limit.

(Tb = Kyo =

Dowel Group Action When a load W is applied on one slab near the joint, as shown in

Figure 4.23, part of the load will be transferred to the adjacent slab through the dowel

group. If the dowels are 100% efcient, both slabs will deect the same amount and the re-

active forces under both slabs will be the same, each equal to 0.5W, which is also the total

shear force transferred by the dowel group. If the dowels are less than 100% efcient, as in

the case of old pavements where some dowels become loose, the reactive forces under the

loaded slab will be greater than 0.5 W, while those under the unloaded slab will be smaller

1 1 FIGURE 4.23

Load transfer through dowel

W/Z WIZ group.

174
Chapter 4 Stresses and Deflections in Rigid Pavements
than 0.5W. As a result, the total shear force on the dowels is smaller than 0.5 W. Therefore,

the use of 0.5 W for the design of dowels is more conservative.

Based on Westergaards solutions, Friberg (1940) found that the maximum nega-

tive moment for both interior and edge loadings occurs at a distance of 1.88 from the

load, where 8 is the radius of relative stiffness dened by Eq. 4.10. When the moment is

maximum, the shear force is equal to zero. It is therefore reasonable to assume that the

shear in each dowel decreases inversely with the distance of the dowel from the point

of loading, being maximum for the dowel under or nearest to the point of loading and

zero at a distance of 1.86. The application of the above principle for dowel design can

be best illustrated by the following examples.

Example 4.12:

Figure 4.24 shows a concrete pavement 8 in. (203 mm) thick having a joint width of 0.2 in. (5.1

mm), a modulus of subgrade reaction of 100 pci (27 kN/m3), and a modulus of dowel support
of

1.5 X 106 pci (407 GN/m3). A load of 9000 lb (40 RN) is applied over the outermost dowel at a

distance of 6 in. (152 mm) from the edge. The dowels are ,1: in. (19 mm) in diameter and 12 in.
(305

mm) on centers. Determine the maximum bearing stress between dowel and concrete.

9000 lb

n: n: a: n: a:

3. 3. '43 R. 8.

OCoOO
P l I D:

Sum = 3.27 P,

FIGURE 4.24

Example 4.12 (l in. = 25.4 mm, 1 lb = 4.45 N).

Solution: From Eq. 4.10, e = [4 x 106 x 512/( 12 x 0.9775 x 100)]125 = 36.35 in. (427

mm). If the dowel directly under the load is subjected to a shear force P., the forces on the
dow-

els within a distance of 1.88, or 66 in. (1.68 m), can be determined by assuming a straight-line

variation, as shown in Figure 4.24. The sum of the forces on all dowels is 3.273, which must be

equal to one-half of the applied load based on 100% joint efficiency, or R = 4500/3.27 = 1376 lb

(6.1 kN). From Eq. 4.43, I,l = (075)4/64 = 0.0155 in.4 (6450 mm). From Eq. 4.44,

3 = [1.5 x 106 x 0.75/(4 x 29 x 106 x 0.0155)]"-25 = 0.889 in. (22.6 mm). From Eq. 4.45,

ab = 1.5 X 106 X 1376(2 + 0.889 X 0.2)/(4 X 0.703 X 29 x 106 X 0.0155) = 3556 psi (24.5

MPa). For a 3000psi (20.7-MPa) concrete, the allowable bearing stress obtained from Eq. 4.41
is

fb = (4 - 0.75) X 3000/3 = 3250 psi (22.4 MPa). Because the actual bearing stress is about

10% greater than the allowable, the design is not considered satisfactory.

In this example, only the left-wheel load near the pavement edge is considered. The right-

wheel load is at least 6 ft (1.83 m) from the left-wheel load, which is greater than 1.86, so the
right

175

4.4 Design of Dowels and Joints 175

wheel has no effect on the maximum force R on the dowel near the pavement edge. If the slab
is

thicker and stronger and the foundation is weaker, 8 may become much larger and both
wheels
must be considered in determining the force R on the most critical dowel.

Example 4.13:

Figure 4.25:: shows a 9.5-in. (241-mm) slab resting on a foundation with k = 50 pci

(13.6 MN/m3). Twelve dowels at 12 in. (305 mm) on centers are placed at the joint on the 12-
ft

9000 lb 9000 lb

9.

1.0

o 86

0.73

0.59

00

0.45 .5

0.32

o 18

o 05

:i

H
co

:2

.3;

2|

('0

O-

U3

0 7.08 Effective dowels

(c) Dowel Forces due to Load at B

1191

Force(1b)

(d) Dowel Forces due to Both Loads

FIGURE 4.25

Example 4.13 (1 in. = 25.4 mm, 1 lb 8 4.45 N).

176
Chapter 4

4.4.2
Stresses and Deflections in Rigid Pavements

(3.66-m) lane. TWO 9000-lb (40-kN) wheel loads are applied at points A and B. Determine the

maximum load on one dowel.

Solution: From Eq.4.10, = [4 x 106 x (9.5)3/(12 X 0.9775 x 50)]"'25 = 49.17 in. (1.25 m),

so 1.88 = 88 in. (2.24 m). First, consider the 9000-lb (40-kN) load at A. If the dowel at A has a

load factor of 1, the load factors at other dowels can be determined from similar triangles, as

shown in Figure 4.25b. The sum of these factors results in 4.18 effective dowels, so the load
car-

ried by the dowel at A is 4500/4.18 or 1077 lb (4.8 kN). The loads carried by other dowels can
be

determined by proportion. Next, consider the 9000-lb (40-kN) load at B. If the dowel at B has a

load factor of 1, the load factors at other dowels can be determined from the triangular
distribu-

tion, as shown in Figure 4.25c. The sum of these factors results in 7.08 effective dowels. Note
that

the dowels on the other side of the longitudinal joint are not considered effective in carrying
the

load. The load carried by the dowel at B is 4500f7.08 or 636 lb (2.8 RN), and those carried by

other dowels can be determined by proportion. Figure 4.25d shows the forces on each dowel
due

to the combined effect of both loads. It can be seen that the dowel nearest to the pavement
edge

is the most critical and should be used for design purposes. The load carried by this dowel can
be

determined directly by Pl = 4500/4.18 + 0.18 X 4500/7.08 = 1191 lb (5.3 kN).

The above examples are based on the assumption that the maximum negative mo-

ment occurs at a distance of 1.86 from the load. Recent studies by Heinrichs et a1. (1989)

have shown that the maximum negative moment occurs at 1.06, so the load carried by

the most critical dowel should be larger than those shown in the examples This has been
proved by comparing the results with the KENSLABS computer program.

Design of Joints

Joints should be provided in concrete pavements so premature cracks due to tempera-

ture or moisture changes will not occur. There are four types of joints in common use:

contraction, expansion, construction, and longitudinal.

Contraction Joints Contraction joints are transverse joints used to relieve tensile

stresses. The spacing of joints should be based on local experience since a change in

coarse aggregate types may have a significant effect on the concrete thermal coef-

cient and consequently the acceptable joint spacing. As a rough guide, the joint spacing

in feet for plain concrete pavements should not greatly exceed twice the slab thickness

in inches For example, the maximum joint spacing for an 8-in. (203-mm) slab is 16 ft

(4.9 m). Also, as a general guideline, the ratio of slab width to length should not exceed

1.25 (AASHTO, 1986).

Figure 4.26 shows typical contraction joints. In Figure 4.26a, a dummy groove is

formed by placing a metal strip on the fresh concrete, which is later removed, or by

sawing after the concrete is set. The groove is then sealed with a plastic material. If the

joint spacing is small, the load transfer across the joint can be achieved by the aggre-

gate interlock and no dowels may be needed. However, dowels are needed if the joint

spacing is large or if the short panels are located near the end of the pavement. In such

cases, the joint may open up and the load transfer through aggregate interlock may be

lost. In lieu of a dummy groove, joints can be formed by placing a felt, asphalt ribbon,

or asphalt board strip in the fresh concrete and leaving it there permanently, as shown

in Figure 4.26b.

Seal with

plastic material
hl4

h to

his

177
4.4 Design of Dowels and Joints 177

Seal with

plastic material

Lubricat

smooth

dowel /

bar

(a) Dummy Groove

Felt, asphalt ribbon

or asphalt board strip

h/4

to

h 11/3

bar

Felt, asphalt ribbon

or asphalt board strip

Lubricated I

ooth 11/2
dowel /

(b) Premolded Strip

FIGURE 4.26

Typical contraction joints.

The sealant used in the joints must be capable of withstanding repeated extension

and compression as the temperature and moisture in the slabs change. Sealants can be

classied as eld molded and preformed. Field-molded sealants are those applied in liquid

or semiliquid form, and preformed sealants are shaped during manufacturing. Figure 4.27

shows the design of joint sealant reservoir for eld-molded sealants To maintain an effec-

tive eld-molded seal, the sealant reservoir must have the proper shape factor or depth-

to-width ratio. The common practice is to have the ratio between 0.5 to 1. Table 4.5 shows

the reservoir dimensions for eld-molded sealants, and Table 4.6 shows the joint and

sealant widths for preformed seals as recommended by PCA (1975).

1/ " below Surface Sealant Material

Tape to Prevent

Bond between
Sawed Joint Face

Initial Saw Cut

or Inserted Strip

1/4" below Surface Sealant Material

Rope or Rod

Back-up Material

Initial Saw Cut

or Inserted Strip

Sawed Joint Face

Sealant and Concrete

FIGURE 4.27

Design of joint sealant reservoir (1 in. = 25.4 mm).


(After PCA (1975).)

178
Chapter 4 Stresses and Deflections in Rigid Pavements

TABLE 4.5 Reservoir Dimensions for Field-Molded

Sealants

Joint spacing Reservoir width Reservoir depth

(ft) (it!) (ill)

15 or less i % minimum

20 g % minimum

30 minimum

55

40 s s

Nore.1 ft = 0.305 m, l in. - 25.4 mm.

Source. After PCA (1975).

TABLE 4.6 Joint and Sealant Width for Preformed

Seals

Joint spacing Joint width Sealant width

(it) (in.) (in.)

20 or less i- 17;

30 g g

3
40 r7; 4

50 ,1,

Note. 1 ft = 0.305 m, 1 in. = 25.4 mm.

Source. After PCA (1975).

The preformed sealant is the type most recommended to achieve long-term perfor-

mance. Preformed sealants can do an excellent job of keeping out incompressibles over a

long period of time, but might not be completely water tight compared to the field-molded

sealants. The preformed sealants should be so designed that the seal will always be com-

pressed at least 20% in the joint. 'Ihe maximum allowable compression of the seal is

50%. Thus, the seal working range is 20 to 50% (Darter and Barenberg, 1977).

Example 4.14:

A concrete pavement 15 ft (4.6 m) long is placed on a gravel subbase. If the joint width is iin.

(6.4 mm), the design temperature range AT is 100F (556C), the coefficient of thermal expan-

sion a, is 5 x 105/F (9 x 10'6/C), and the drying shrinkage coefficient 6 is 1.0 x 10, deter-

mine the width of preformed sealant required.

Solution: From Eq. 4.36, the joint opening due to temperature change is AL = 0.65 X

15 X 12(0.000005 X 100 + 0.0001) = 0.07 in. (1.8 mm). Try {3 or 0.4375 in. (11.1 mm) sealant

installed in summer, so the joint would not be further compressed. Check maximum
compression

of sealant: (0.4375 - 0.25)/0.4375 = 0.43 < 50%, OK. Check minimum compression of sealant:

(0.4375 - 0.25 - 0.07)/0.4375 = 0.27 > 20%, OK. Therefore, the use of {gin (11.1 mm)

sealant for a -in. (6.4-mm) joint is satisfactory, as shown in Table 4.6.

179
4.4 Design of Dowels and Joints 179

Contraction joints are usually placed at regular intervals perpendicular to the center

line of pavements. However, skewed joints with randomized spacings, say 13-19-1812 ft

(4.0-5.85.5-3.7 m), have also been used. The obtuse angle at the outside pavement

edge should be ahead of the joint in the direction of traffic, since that corner receives

the greatest impact from the sudden application of wheel loads. The advantage of

skewed joints is that the right and left wheels do not arrive at the joint simultaneously,

thus minimizing the annoyance of faulted joints. The use of randomized spacings can

further reduce the resonance and improve the riding comfort.

Expansion Joints Expansion joints are transverse joints for the relief of compressive

stress. Because expansion joints are difficult to maintain and susceptible to pumping,

they are no longer in use today except at the connection between pavement and struc-

ture. Experience has shown that the blowups of concrete pavements are related to a

certain source and type of coarse aggregates. If proper precaution is exercised in

selecting the aggregates, distress due to blowups can be minimized. The plastic flow of

concrete can gradually relieve compressive stress, if any, so it is not necessary to install

an expansion joint except at bridge ends.

Figure 4.28 shows a typical expansion joint. The minimum width of joint is gin.

(19 mm). Smooth dowel bars lubricated at least on one side must be used for load

transfer. An expansion cap must be installed at the free end to provide space for dowel

movements. Nonextruding fillers, including fibrous and bituminous materials or cork,

must be placed in the joint and the top sealed with a plastic material.

Construction Joints If at all possible, the transverse construction joint should be placed

at the location of the contraction joint, as shown by the butt joint in Figure 4.29a. If the

work must stop due to an emergency or machine breakdown, the key joint shown in

Figure 4.2% may be used.This joint should be placed only in the middle third of a normal
joint interval. Key joints have not performed well and many failures have occurred.

Longitudinal Joints Longitudinal joints are used in highway pavements to relieve

curling and warping stresses. Different types of longitudinal joints are used, depending

on whether the construction is full width or lane-at-a-time.

In the full-width construction, as shown in Figure 4.30, the most convenient type

is the dummy groove joint, in which tie bars are used to make certain that aggregate

interlock is maintained, as shown in Figure 4.30a. These bars may be shoved into the

wet concrete before the nal finishing and placement of the dummy groove. The joint

3/4 in. minimum

Seal with plastic material

3 Expansion cap

Smooth, lubricated dowel bar

fixed

FIGURE 4.28

Non-extruding filler Expansion joint (1 in. = 25.4 mm).


180
Chapter 4 Stresses and Deflections in Rigid Pavements

First slab First slab

Smooth, lubricated

dowel bar

"60 429 (a) Butt Joint at (b) Key Joint for

Construction joints Contraction Joint. Emergency.

steel plate

Deformed bar Deformed bar Deformed bar

(a) Dummy Groove. (b) Ribbon or Premolded (c) Deformed plate.

Strip.

FIGURE 4.30

Longitudinal joints for full-width construction.


Groove 1/8" to 1/4 " wide

and 1" deep

FIGURE 4.31

Longitudinal joints for lane-at-a-time

construction (1 in. = 25.4 mm).

can also be formed by inserting a premolded strip into the fresh concrete and leaving it

there permanently as an integral part of the warping joint, as shown in Figure 4.30b.

Another method is to install deformed steel plates and tie bars at the center line before

the pour of concrete, as shown in Figure 4.30c.

Lane-at-a-time construction is used when it is necessary to maintain traffic on the

other lane. To insure load transfer, key joints are usually used, as shown in Figure 4.31.

In most cases, the keyed joints are tied together with tie bars. However, tie bars may be

omitted if the longitudinal joint is at the interior of a multilane pavement and there is

very little chance that the joint will be wide open.

Butt joints have also been used for lane-at-a-time construction. Current practice

prefers the use of butt joints over keyed joints because keyed joints usually do not per-

form well due to the occurrence of cracks along the key and they are also difcult to

construct with slipform paving.

SUMMARY

This chapter discusses the stresses and deections in rigid pavements, on the basis of

Westergaards theory. Westergaard viewed pavement as a plate on a liquid foundation


181
Summary 181

with full subgrade contact. Analyses based on partial contact and other types of foun-

dation are presented in Chapter 5.

Important Points Discussed in Chapter 4

1. Curling stresses in an infinite slab are caused by the restraining effect of the slab

2.

3.

and can be determined easily from Hookes law by assuming plane strain.

Curling stresses in a finite slab are caused by the curling of the slab and are diffi-

cult to compute. Using Westergaards theory, Bradbury developed a simple chart

for determining the maximum warping stress in the interior and at the edge of a

finite slab.

Whether curling stresses should be considered in rigid pavement design is contro-

versial. The Portland Cement Association does not consider curling stresses in

fatigue analysis because the very few number of stress reversals does not con-

tribute to fatigue cracking and also the curling stresses may be added to or sub-

tracted from the loading stresses to neutralize the effect. Others think that curling

stresses should be combined with loading stresses. because past experience has

shown that longer slabs with greater curling stresses always result in more cracking

of the slab. A more reasonable approach is to consider the fatigue damage due to

curling separately from that due to loading and then combine them.
Westergaards closed-form formulas can be used to determine the maximum

stresses and deflections in a concrete slab due to a circular loaded area applied at

the corner. in the interior. or near to the edge. if the load is applied over a set of

dual tires. the formulas can still be applied by using an equivalent circular area.

The stresses and deections due to interior and edge loadings can also be deter-

mined by influence charts. When inuence charts are employed. the actual tire

imprints should be used, instead of assuming the imprints to be circular areas.

The design of steel reinforcements and tie bars is based on the stresses due to

friction. These steel reinforcements, such as wire fabric and bar mats. do not

increase the structural capacity of the slab but are used to increase the joint spac-

ing and to tie the cracked concrete together to maintain load transfers through

aggregate interlock.

The design of dowels is mostly based on experience. One rule of thumb is that the

diameter of dowel be equal to 21; of the slab thickness. However. a theoretical

method is available to determine the bearing stress between dowel and concrete

and check against the allowable bearing stress.

Joints should be provided in concrete pavements so that premature cracks due to

temperature or moisture changes will not occur. As a rough guide. the joint spac-

ing in feet for plain concrete pavements should not greatly exceed twice the slab

thickness in inches, and the ratio of slab width to length should not be greater

than 1.25. Contraction joints are usually placed at regular intervals perpendicular

to the center line of. pavements. However. skewed joints with randomized spacings

have also been used. Expansion joints are used only at the connection between

pavement and structure. Longitudinal joints are used to relieve curling and warp-

ing stresses. and different types may be used. depending on whether the construc-

tion is full width or lane-at-a-time.

You might also like