/FRICTION COEFFICIENT]OF STEEL ON CONCRETE

---~ OR GROUT
By B. G.JRabbat, r ~Mr-ASeE a n d H. G. R U S S P U J * ^

ABSTRACT: £jj.n experimental investigation was conducted to determine the

coefficient of static friction between rolled steel plate and cast-in-place concrete
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

or grout. Fifteen tests were performed under conditions that represented the
interior and exterior bearing surfaces of a containment vessel. Test parameters • "\
included concrete blocks or grout blocks, wet or dry interface, and level of nor-
mal compressive stress. For conditions tested, the average effective coefficient
of static friction varied between 0.57 and 0.70. It is recommended that the coef-
ficient of static friction for concrete cast on steel plate and grout cast below steel
plate should be taken as 0.65 for a wet interface with normal compressive stress
levels between 20 and 100 psi (0;14»=atrd 0:6?MFd). For dry interface, the coef-
ficient of static friction should be taken as 0.57.

INTRODUCTION

The purpose of this investigation was to determine experimentally the

coefficient of static friction between rolled steel plate and cast-in-place
concrete or grout. For concrete specimens, the interface between con-
crete and steel plate was tested with wet and dry conditions. For grout
specimens, only the wet interface condition was tested.
Test Program.—Five sets of three similar push-off tests were con-
ducted. Specimens consisted of either concrete cast above, or grout cast
below a flat steel plate. During the tests, the interface between steel plate
and concrete or grout was either dry or wet. Different levels of normal
compressive stress, perpendicular to the interface, were applied.
Three sets of concrete specimens were tested with a wet interface at
normal stress levels of 20, 60, and 100 psi (0.14, 0.41, and 0.69 MPa).
One set of concrete specimens was tested with a dry interface at a nor-
mal stress of 60 psi (0.41 MPa). One set of grout specimens was tested
with wet interface at a normal stress of 60 psi (0.41 MPa). For each spec-
imen, nominal bond strength and coefficient of static friction at the in-
terface were determined.
Conclusions.—Results of the present investigation have shown the
following:

1. Bond strength for concrete specimens varied between 25 and 89 psi

(0.17 and 0.61 MPa). For grout specimens, bond strength was negligible.
2. For concrete specimens with wet interface, average effective coef-
ficients of static friction were 0.67, 0.65, and 0.64, at normal stresses of
20, 60, and 100 psi (0.14, 0.41, and 0.69 MPa), respectively. The corre-
'Principal Engr., Bridge Structures, Struct. Experimental Section, Construction
Technology
2
Laboratories, Portland Cement Assoc, Skokie, 111.
Dir., Struct. Development Dept, Construction Technology Laboratories, Port-
land Cement Assoc, Skokie, 111.
Note.—Discussion open until August 1, 1985. To extend the closing date one
month, a written request must be filed with the ASCE Manager of Journals. The
manuscript for this paper was submitted for review and possible publication on
April 12, 1984. This paper is part of the Journal of Structural Engineering, Vol.
Ill, No. 3, March, 1985. ©ASCE, ISSN 0733-9445/85/0003-0505/$01.00. Paper No.
19569.
505

J. Struct. Eng. 1985.111:505-515.

 sponding average peak coefficients of static friction were 0.70, 0.68, and
0.64, respectively. (Definitions of effective and peak coefficients of static
friction are given later, in the section titled "Shear Stress versus Slip
Relationship.")
3. For concrete specimens with dry interface and a normal stress of
60 psi (0.41 MPa), the average effective and average peak coefficients of
static friction were 0.57 and 0.69, respectively.
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

4. For grout specimens with wet interface and normal stress of 60 psi
(0.41 MPa), both average effective and average peak coefficients of static
friction were 0.68.

Recommendations.—Based on the results of fifteen push-off tests,

performed as described in this paper, it is recommended that:

1. The coefficient of static friction for concrete cast on steel plate and
grout cast below steel plate should be taken as 0.65 for a wet interface
with normal compressive stress levels between 20 and 100 psi (0.14 and
0.69 MPa).
2. The coefficient of static friction for concrete cast on steel plate and
grout cast below steel plate should be taken as 0.57 for a dry interface
with normal compressive stress levels between 20 and 100 psi (0.14 and
0.69 MPa).

EXPERIMENTAL PROGRAM

Test Specimens.—Test specimens represented conditions at either a

steel-concrete or a steel-grout interface in the lower support region of a
containment vessel. To simplify testing, the friction surface was simu-
lated as a flat plane.
A test specimen is shown schematically in Fig. 1. The test surface was
13.5 in. (0.34 m) wide and 24 in. (0.61 m) long. These dimensions gave
a test area of 324 sq in. (0.21 m2).
Test Variables.—Fifteen specimens were tested in this investigation.
All specimens were tested in sets of three. Test variables included: (1)
Concrete blocks or grout blocks; (2) wet or dry interface; and (3) level
of normal compressive stress.
The experimental program as summarized in Table 1 was comprised
of the following three series:

Concrete or Grout Block

V 1 in. = 25.4 mm

FIG. 1,—Test Specimen

506

J. Struct. Eng. 1985.111:505-515.

 TABLE 1.—Experimental Program
Normal compressive
Test Specimen stress, in pounds
number Series identification" per square inch Interface condition
(1) (2) (3) (4) (5)
1 I CWB-1 60 Concrete to steel
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

2 CWA-1 100 (wet)

3 CWC-1 20
4 CWB-2 60
5 CWA-2 100
6 CWC-2 20
7 CWB-3 60
8 CWA-3 100
9 CWC-3 20
10 II CDB-1 60 Concrete to steel
11 CDB-2 60 (dry)
12 CDB-3 60
13 III GWB-1 60 Grout to steel
14 GWB-2 60 (wet)
15 GWB-3 60
"Specimen identification is as follows: among first letters, C denotes concrete,
and G denotes grout; among second letters, W denotes wet, and D denotes dry;
and among third letters (load level), A = 100 psi, B = 60 psi, and C = 20 psi.
The last digit is for first, second, or third test of each set.
Note: 1 psi = 6.895 kPa.

1. Wet concrete-to-steel interface.—Series I consisted of nine tests. Three

tests were performed at each of three levels of normal compressive stress.
After breaking the bond, the concrete-to-steel interface was intentionally
saturated with water for the friction tests.
2. Dry concrete-to-steel interface.—Three tests were performed in Se-
ries II. Normal compressive stress was held constant at an intermediate
level. The concrete-to-steel interface was not wetted after breaking the
bond.
3. Wet grout-to-steel interface.—Series III consisted of three tests at
the same level of compressive stress as in Series II. However, after
breaking the bond, the grout-to-steel interface was saturated with water.

Manufacture of Specimens.—Concrete and grout specimens were

manufactured according to procedures described below. In all speci-
mens, the steel plate was ASTM Designation A 516, Grade 70 steel (6),
1 in. (25 mm) thick and 14 x 28 in. (0.36 x 0.71 m) in size.
Steel Plate Cleaning.—The steel plate was cleaned in accordance with
the Steel Structures Council Specifications SSPC-SP2-63, Hand Tool
Cleaning (7). Oil and grease were removed using acetone solvent. Rust
scale was removed by hand hammering. Loose mill scale was removed
by hand wire brushing.
Concrete-to-Steel Specimens.—After cleaning was completed, wood
formwork was clamped to the top of the steel plate. The concrete mix
was designed to yield a compressive strength between 3,000 and 5,000
507

J. Struct. Eng. 1985.111:505-515.

 psi (20.7 and 34.5 MPa) at 28 days. The mix was made using Type I
Portland Cement, Elgin Sand, and Elgin Gravel with a maximum ag-
gregate size of 1-1/2 in. (38 mm).
Concrete was consolidated using an internal vibrator. From each batch,
six 6 x 12 in. (152 x 305 mm) standard concrete cylinders were taken.
These cylinders were used to determine the concrete compressive strength.
Concrete was cured for 7 days under polyethylene sheeting. Form-
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

work was then stripped and specimens kept in the laboratory at a tem-
perature of 73° F (23° C) and a relative humidity of 50%. Specimens were
tested at an age of 28-30 days.
Grout-to-Steel Specimens.—These specimens were prepared in the
setup shown in Fig. 2. The setup was designed to simulate field con-
ditions of placing grout at the exterior bearing surface of a containment
vessel. Formwork was clamped to the underside of the steel plate. The
specimen was inclined during grouting. The difference in height be-
tween the bottom horizontal edges of the grout block was 4 in. (102
mm).
The grout mix specifications called for cement that was Type II, ASTM
designation C 150 (5), with no admixtures; a fine aggregate of Natural
Elgin Sand, ASTM designation C 33 (3); a cement to sand ratio of 1:2
by volume; and a water to cement ratio of 0.5 by weight. Flow of this
mix was measured on a flow table after five drops in 3 sec.
Grout was placed through a funnel secured to the top of the right-
hand side pipe shown in Fig. 2. This pipe had a 2 in. (51 mm) diameter.
Because of its consistency, the grout filled the form and flowed upward
in the left-hand pipe. The height of the left-hand side pipe corresponded

FIG. 2.~Setup for Preparation of Grout Specimens

508

J. Struct. Eng. 1985.111:505-515.

 to 5 ft (1.52 m) head measured from the center of the grout block.
To permit bleeding of air trapped under the plate, holes were pro-
vided around the top perimeter of the form. These holes were sealed as
soon as air bleeding was completed.
The pipe was sawed off after 7 days. The specimens were then stripped
and turned upside down. They were kept in the laboratory at a tem-
perature of 73° F (23° C) and a relative humidity of 50% until test time.
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

Grout compressive strength was measured using six 2 in. (51 mm) cubes.
Test Setup.—The test setup is shown in Figs.. 3 and 4. This setup is
capable of applying vertical loads normal to the shear plane and hori-
zontal loads parallel to the shear plane. Horizontal rams were used to
apply in-plane shear loads. Vertical rams were used to apply normal
stresses.
Instrumentation.—All loads were measured using calibrated load cells
(1) acting in series with the hydraulic rams. Slip of the concrete test block
relative to the steel plate was measured with two potentiometers.
Test Procedure.—Bond between concrete and steel was broken by ap-
plication of an in-plane shear force applied at the shear-friction interface.
No normal compressive stress other than the weight of the upper block
and loading equipment was applied at this stage. This weight was equiv-

FIO. 3.—Test Setup

-Load Cell

--Rom for Normal Stress

^Structural Tube

Sfeel
Ram for -Test Role
Push-Off Specimen il | H
Load

HALF-SECTION k-A
~t
HALF-SECTION B-B

FIG. 4.—Schematic of Test Setup

509

J. Struct. Eng. 1985.111:505-515.

 60

Shear
40
u^www^#^
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

Stress,
psi

zo

I in. = 25.4 mm
I psi' 6.895 k Pa
0
0 0.1 0.2
Slip, Inch

FIG. 5.—Representative Continuous Record of Shear Stress versus Slip

alent to a nominal stress of 8.18 psi (56 kPa). The shear stress required
to cause first slip was noted. The concrete block was then moved back
to its original position on the steel plate.
For tests requiring a wet friction surface, the interfaces were separated
after bond was broken. The interfaces were then ponded in water to
ensure a wet surface during testing. Interfaces remained saturated dur-
ing subsequent testing. During each test, normal stress was held con-
stant.
After application of normal compressive force, the shear load was ap-
plied slowly. A continuous record of the horizontal load versus slip at
the shear-friction interface was obtained. A representative shear stress
versus slip curve is shown in Fig. 5. It can be seen that slip occurred in
increments. Each slip increment occurred within a fraction of a second
and was accompanied by a drop in shear stress. The drop was a result
of the stiffness of the loading system. The shear stress versus slip curves
reported in the results section were idealized by a curve joining the peaks
of the saw teeth of Fig. 5. Note that the testing rate was adjusted to
obtain 0.1 in. (2.5 mm) slip within 3-5 min. The test was stopped when
a slip of 0.5 in. (12.7 mm) was reached. At a slip of 0.5 in. (12.7 mm),
the horizontal component of the vertical loading system was negligible.

TEST RESULTS

Results presented in this section include material properties, bond

strengths at interface, shear stress versus slip relationships, and coeffi-
cients of static friction. The effects of normal stresses, wet interface, grout
versus concrete, and concrete strength on the coefficient of static friction
are examined.
Concrete Strength.—At test age, six concrete cylinders were tested to
determine concrete compressive strength. Average concrete strengths are
listed in Table 2. Ages of concrete at time of testing are also listed.
510

J. Struct. Eng. 1985.111:505-515.

 TABLE 2,—Concrete Strength
Test Test age, Average concrete strength,
number Specimen in days in pounds per square inch
(1) (2) (3) (4)
1 CWB-1 28 3,770
2 CWA-1 29 3,510
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

3 CWC-1 28 3,275
4 CWB-2 28 3,510
5 CWA-2 28 3,780
6 CWC-2 28 3,525
7 CWB-3 28 3,030
8 CWA-3 28 3,800
9 CWC-3 28 3,270
10 CDB-1 28 3,620
11 CDB-2 28 3,720
12 CDB-3 28 3,710
Note: 1,000 psi = 6.895 MPa.

Grout Properties.—Grout strength was measured using 2 in. (51 mm)

cubes. The average strengths of six cubes are listed in Table 3. Flow
measured on the flow table after five drops in three sec is also listed.
Bond Strength at Interface.—Bond strength is the nominal shear stress
required to break bond between the concrete or grout block and the steel
plate. It is equal to the in-plane shear force that broke the bond divided
by the area of the interface (324 sq in.) (0.21 m2).
Measured bond strengths are listed in Table 4 where the specimens
have been grouped in sets of three similar tests. For concrete specimens,
the bond strength varied between 25 and 89 psi (0.17 and 0.61 MPa).
For grout specimens, bond strength was negligible. Inspection of the
interface after testing indicated that after placing the grout, bleeding oc-
curred inside the formwork. Sand particles settled and water laitance
moved up, with a larger accumulation at the highest point of the form-
work. As cement hydrated, water was absorbed. As a result, air gaps
formed between the steel plate and grout block. Moreover, air bubbles
had formed at the interface as shown in Fig. 6. The left-hand side was
the highest point as cast. In Specimens GWB-2 and GWB-3, the bond
had broken during handling of the specimens before testing.

TABLE 3.—Grout Properties

Test Test age, Average cube strength, in
number Specimen in days pounds per square inch Flow"
0) (2) (3) (4) (5)
13 GWB-1 28 7,680 142
14 GWB-2 30 7,955 121
15 GWB-3 28 8,390 120
a
ASTM designation C 230 (4).
Note: 1,000 psi = 6.895 MPa.

511

J. Struct. Eng. 1985.111:505-515.

 TABLE 4.—Bond Strength and Coefficient of Friction
Bond
Shear Stress,
strength,
in Pounds per Coefficient of Friction
in pounds
per square Square Inch Average of Average of
Specimen inch Peak Effective Peak each set Effective each set
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

(1) (2) (3) (4) (5) (6) (7) (8)

CWA-1 52.8 62.0 62.0 0.62 0.64 0.62 0.64
CWA-2 48.6 63.5 63.5 0.63 0.63
CWA-3 51.4 68.0 67.8 0.68 0.68
CWB-1 55.0 42.0 40.7 0.70 0.68 0.68 0.65
CWB-2 25.0 40.4 39.6 0.67 0.66
CWB-3 89.0 40.5 36.1 0.68 0.60
CWC-1 61.2 13.8 12.8 0.69 0.70 0.64 0.67
CWC-2 83.6 14.2 14.0 0.71 0.70
CWC-3 78.4 13.9 13.5 0.69 0.67
CDB-1 67.3 40.5 33.7 0.68 0.69 0.56 0.57
CDB-2 53.0 46.2 35.0 0.77 0.58
CDB-3 58.2 36.9 34.0 0.62 0.57
GWB-1 8.0 41.2 41.2 0.69 0.68 0.69 0.68
GWB-2 — 40.9 40.9 0.68 0.68
GWB-3 — 40.0 40.0 0.67 0.67
Note: 1 psi = 6.895 kPa.

Shear Stress versus Slip Relationhip.—Shear stress versus slip curves

for Series I, II, and III are plotted in Figs. 7, 8, and 9, respectively. Peak
and effective shear stresses are listed in Table 4. Peak shear stress is
defined as the highest shear recorded on the shear stress versus slip
curve. Where the peak shear stress occurred at a slip less than 0.5 in.
(12.7 mm), an effective shear stress was determined. Effective shear stress
is defined as the lowest shear stress occurring after the peak is reached.
Where the peak shear stress occurred at 0.5 in. (12.7 mm) slip, the ef-
fective shear stress was considered equal to the peak shear stress.
Coefficient of Friction.—Coefficient of friction is defined as the shear

FIG. 6.—Friction Interface of Specimen GWB-1 After Testing

512

J. Struct. Eng. 1985.111:505-515.

 Normal Compressive
Stress, !00pst
CWA:3
•"" ' CWA-2
60
Shear
Stress,
psi
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

_ _ CWB-I . *>&
fS / " • —- . CWB-2
CWB-3

20
rcwc-2 CWC-3TE0PSI
r " C-cwcT'
lin.=25.4mm
1pst=6B95kPa
0
0 0.1 02 0.3 0.4 0.5
Slip, inch.

FIG. 7.—Shear Stress versus Slip of Series I

Shear
Stress,
psi

Normal Compressive
Stress, 60 psi

I in. = 25.4 mm
lpsi=6.895kPo

0 0.1 02 Q3 0.4 0.5

Slip, hch.

FIG. 8.™Shear Stress versus Slip of Series II

Shear
Stress,
psi
fGWB-3 GWB-I
Normal Compressive
Stress, 60psi

i in.= 25.4mm
lpsi=6.895kPa

Slip, inch.

FIG. 9.—Shear Stress versus Slip of Series III

513

J. Struct. Eng. 1985.111:505-515.

 stress divided by the normal stress. Coefficients of friction correspond-
ing to each of the peak and effective shear stresses are listed in Table
4. An average coefficient of friction for each set is given in the same
table. The shear-friction design method given in the recent ACI Building
Code (1) recommends a coefficient of friction of 0.70 for normal weight
concrete anchored to as-rolled structural steel by headed studs or by
reinforcing bars. Measured coefficients for friction were slightly lower
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

than the value specified in the ACI Building Code for design of shear
friction reinforcement.
Effect of Normal Stresses.—The effect of normal stress on coefficient
of friction can be determined from the results of Series I, concrete-to-
concrete specimens with wet interface. At normal stresses of 20, 60, and
100 psi (0.14, 0.41, and 0.69 MPa), the average peak coefficients of static
friction were 0.70, 0.68, and 0.64, respectively. The corresponding av-
erage effective coefficients of static friction were 0.67, 0.65, and 0.64,
respectively. Average coefficients of static friction tended to decrease as
normal compressive stress increased.
Effect of Wetting Interface.—The effect of wet interface on coefficient
of static friction can be seen when comparing average results of CWB
and CDB tests in Table 4. Corresponding to peak shear stresses, the
average coefficients of static friction were 0.68 and 0.69 for wet and dry
interfaces, respectively. These coefficients were fairly close. However,
the average coefficients corresponding to effective shear stress were 0.65
and 0.57 for wet and dry interfaces, respectively. Therefore, it appears
that the average coefficient of friction corresponding to effective shear
stress was about 12% lower for the dry interface.
Grout versus Concrete Interface.—Comparison between coefficient of
static friction for grout and concrete against rolled steel is based on re-
sults of GWB and CWB tests. The average peak coefficient of static fric-
tion was the same in both cases although, in the grout series, not all of
the interface area was in contact with the steel plate. This interface con-
dition with air gaps and bubbles can be expected in the field. However,
variation of the grout surface condition did not appear to affect the mea-
sured coefficient of friction.
The average effective coefficient of static friction for concrete and grout
specimens was 0.65 and 0.68, respectively. The observed coefficient for
the grout specimens was about 5% higher than for the concrete speci-
mens.
Effect of Concrete Strength.—Concrete strengths are listed in Table 2,
and coefficients of static friction are listed in Table 4. Based on the test
results of Series I, the coefficient of static friction apparently increased
with increased concrete strength. However, the range of concrete strengths
was too small to state definitively the effect of concrete strength on the
coefficient of static friction.

ACKNOWLEDGMENTS

This investigation was sponsored by Fluor Power Services, Inc., Duke

Power Company, and Tennessee Valley Authority. The tests were car-
ried out in the Structural Experimental Section of the Construction Tech-
nology Laboratories, a division of the Portland Cement Association un-
514

J. Struct. Eng. 1985.111:505-515.

 der Dr. W. G. Corley, Divisional Director, Engineering Development
Division. Particular credit is due to W. T. Fasig a n d W. Hummerich, Jr.
for their assistance in manufacture and testing of the specimens.

DISCLAIMER

This paper is based on the facts, tests, and authorities stated herein.
Downloaded from ascelibrary.org by NDSU LIBRARY on 05/26/14. Copyright ASCE. For personal use only; all rights reserved.

The Construction Technology Laboratories disclaims any a n d all re-

sponsibility for application of the stated principles or for the accuracy of
any of the sources other than work performed or information developed
by Construction Technology Laboratories. The contents do not neces-
sarily reflect the official views or policies of Fluor Power Services, Inc.,
Duke Power Company, or Tennessee Valley Authority.

APPENDIX.—REFERENCES

1. Building Code Requirements for Reinforced Concrete, ACI Committee 318, Amer-
ican Concrete Institute, Detroit, Mich., 1983.
2. Hognestad, E., et al., "Facilities and Test Methods of PCA Structural Labo-
ratory," Journal of the PCA Research and Development Laboratories, Vol. 1, No. 1,
Jan., 1959, pp. 12-20, 40-45; Vol. 1, No. 2, May, 1959, pp. 30-37; and Vol.
1, No. 3, Sept., 1959, pp. 35-41. Reprinted as PCA Development Department
Bulletin D33, Portland Cement Association, Skokie, 111.
3. Standard Specification for Concrete Aggregate, ASTM C 33-82, American Society
for Testing and Materials, Section 4, Vol. 04.02, Philadelphia, Pa., 1983.
4. Standard Specification for Flow Table for Use in Tests of Hydraulic Cement, ASTM
C 230-80, American Society for Testing and Materials, Section 4, Vol. 04.02,
Philadelphia, Pa., 1983.
5. Standard Specification for Portland Cement, ASTM C 150-83, American Society
for Testing and Materials, Section 1, Vol. 04.02, Philadelphia, Pa., 1983.
6. Standard Specification for Pressure Vessel Plates, Carbon Steel, for Moderate- and
Lower-Temperature Service, ASTM A516-82, American Society for Testing and
Materials, Section 1, Vol. 01.04, Philadelphia, Pa., 1983.
7. Surface Preparation Specifications—No. 2 Hand Tool Cleaning, SSPC-SP2-63, Steel
Structures Painting Council, Oct., 1963.

515

J. Struct. Eng. 1985.111:505-515.