TECHNICAL REPORT C-69.4

USE OF EPOXY OR POLYESTER RESINCONCRETE IN TENSILE ZONE OF

COMPOSITE CONCRETE BEAMSby

H. 6. Geymayer

March 1969

APR 8 1969

Sponsored by C

Assistant Secretary of the Army (R&D)

Department of the Army

Con&dcted by

U. S. Army Engineer Waterways Experiment StationCORPS OF ENGINEERS

Vicksburg, Mississippi

THIS DOCUMENT HAS BO 2N APPROVED FOR PUBLUC RELEASEAND SALE, ITS DISTRIBUTION IS UNLIMITED

BEST AVAILABLE COPY

wmn m mC

W. AVAL M WM

Destroy this repor" when no longer needed. Do not return .:

it to the originator. ;

The findings in this report are not to 6• construed as an officialDepartment of the Army position unless so designated

by other authorized documents

/ ,I14

TECHN!CAL REPORT C-69-4

USE OF EPOXY OR POLYESTER RESINCONCRETE IN TENSILE ZONE OF

COMPOSITE CONCRETE BEAMSby

H. 6. Geymayer

March 1969

Sponsored by

Assistant Secretary af the Army (R&D)

Department of the Army

Conducted by

U. S. Army Engineer Waterways Experiment Station

CORPS OF ENGINEERSVicksburg, Mississippi

AUMY.MRIC VICK..UE.G MISS.

THIS DOCUMENT HAS BEEN APPROVED FOR PUBLIC RELEASEAND SALE; ITS DISTRIBUTION IS UNLIMITED

4(

I,

THE CONTENTS OF THIS RF`PORT ARE NOT TO BE

USED FOR ADVERTISING, PUJBLICATION, OR

PROMOTIONAL P•JRPOSES. CITATION OF TRADE

NAMES DOES NOT CONSTITUTE AN OFFICIAL EN-

DORSEMENT OR APPROVAL OF THE USE OF SLCH

COMMERCIAL PRODUCTS.

I ii.I

SFOREWORD

A research investigation, "Feasibility Study on Epoxy and Polyester

Resin and Portlend-Cement wiicrete Beams," sponsorcd by the A~sistant

Secretary of the Army (R&D), was authorized by a memorandum to the Chief,

Concrete Division, U. S. Arny Engineer Waterways Experiment Station (WES),

dated 10 November 1965, File WESVB, Subject: "In-House Laboratory Ini-

tiated Research Program, FY 66."

The work was performed at the WES Concrete Division during the period

January 1966 to December 1967, under the dire,:,ion of Messrs. Bryant Mather,

James M. Polatty, Dr. Helmut G. Geymayer, LP 5 William E. Walker, and

SP 4 William D. Smart. This report was prepared by Dr. Geymayer.

Directors of the WES during the investigation and the preparation of

this report were COL John R. Oswalt, Jr., CE, and COL Levi A. Brown, CE.

Mr. J. B. Tiffany was Technical Director.

Iv

II

V

i

I

CONTENTS

PageFOREWORD . . . . . . . . . . . . . . . . . . . . . . . . .. . . v

CONVERSION FACTORS, BRITISH TO METRIC UNITS OF MEASUREMMT . . . .. ix

SUMMAIVARY. . . . . . . . . . .................. . xi

PART I: INTRODUCTION. . .................... . 1

Background .......... ... .......................... 1Objective and Scope ........... ..... ...................... 4

PART II: TESTING PROGRAM ......... ....... ...................... 5

Materials and Techniques .... ..................... 6Fabrication of Test Beams ..... ................... . 10Test Procedures ........... ...................... ..... 12

PART III: TEST RESULTS ......... ....................... .... 16

Resin Concretes ........... ........................ ..... 16Results of Beam Tests with Lpoxy Resin Concrcte ........ ..... 17Results of Beam Tests with Polyester Resin Concretes . . . .. 21

PART IV: DISCUSSION OF TEST RISULTS ................. 23

Resin Concretes ........... ........................ .... 23Aggregates ............ ..... ......................... 24Resin Content ....... ........................ 25Engineering Properties of the 3elected Resin Concretes . . . 29Composite Feam Tests ........ ..................... ..... 34

PART V: CONCLUSIONS AND RECOfl-2:IEDATIONS ........ .............. 39

LITERATULR CITED .......... ........... .......................... hl

qABiLES 1-15P11OTbGRA1'1[S 1-19)

PLATES 1- 19

APfPL1IJ( A: ELASTIC ANALYSIS FOR CRACK11ri C'F RFKS3N C(1,CR!E . ... Al

TABLE Al

vii

CONVERSION FACTORS, BRITISH TO METRIC UNITS OF MEASUREMENT

British units of measurement used in this report can be converted to metric

units as follows:

____ __ip_ _ BY To Obtain

inches 2.54 centimeters

feet 0.3048 meters

square inches 6.4516 square centimeters

cubic feet 0.0283168 cubic meters

cubic yards 0.764555 cubic meters

quarts 0.946353 cubic decimeters

gallons (U.S.) 3.785412 cubic decimeters

pounds 0.45359237 kilograms

tons (2000 pourds) 907.185 kilograms

pounds per square inch 0.070307 kilograms per square centimeter

p'nmda peý cubic foot 16.0185 kilograms per cubic meter

foot-pound•s 0.138255 meter-kiiograms

nounds per cubic izch 27679.91 kilograms per cubic meter

Fahrenheit aegrees 5/9 Celsius or Kelvin degrees*

To obtain Celsius (C) temperature --e-dings from Fahre.nheit (F) readings,use the following formula: C - (5/9)jF - 32). To obtain Kelvin (K)readings, use: K (5/9)(F - 32) + 2,"-.15.

ix

This report dercribes the resultn of an investiL-tion into the feasi-bility of combining the high compreozive str'ýrath of portland cement con-crete and the superior tensile strenrth. of epoxy or polyester resin con-crete into a composite beam. This would increase the beam's flexuralstrenr;th and improve the corrosion protection for the reinforcement atiarg_ .1oi'lerz" ---- 1- -1!i-inating tp-c le Qr;,-ký..

The report describes in detail the ie%-io., * ent of high-strength resinconcrete mixtures and sux-mrizes the most imp' rtant engineering propertiesof the selected mixtures. Also !.nciuded are the rzzults of thir2-,'eintloading tests of 12 reinforced and unreinforced 2om-osite beams, with 1-1/2-and 3-in.-thick layers of epoxy and polyester resin cozrcretes. 7 .ese re-sults are compared with results of' tests of two reference neazms withoutresin corcrete layers and with analytIca1 results.

Tne study, led to the follcrv prinici al conrcluziens:

a. Properly desi,7ned resin conu-rete layerz at tne tenz;ion faceof' concrete bes czai he -- ed to -.Y> ieritely ircrease the.;trength Fand rigidity of reinforced coiirote beds, or toupgrade the flexural o;trenr4 tl. of uzireinforced i-eams by af'i.tor of two to threc.

b. M,1ore important tlla their in!,luence on .tren•th is theability of resin aor.or,. >r.er: t. rovile a no'r.crackirnmoisture barrier or rro.:icr rte :tocn vraiticittlly ur tobeam failure.

c. rhe epoxy resins itt-parul t-. be -Mure suitable .cr thl.- ap-plication than the ix'z-ter re..-irnZ iuxettic-ited. lue tolower ahrin-ka~ ;re! oxtir. u well as- iAný-r tni~AtrenClti and ten:ýIie - .

d. In projvrtioniinF r.:in ,'oon -rcte .xxtures:, enrl; ntttentonso~uld be 4ireote- t. 'ro;."rtIco ether t. snt,. (suchIi. rhrir~knec, exvVhcr- fiietof t ,rcicr• .••ep, se;•siti.,4ty to enr r ntai .ctc etc.).

USE OF EPOXY OR PGLYESTER RESIH CONC,-1 flI

TENSILE ZONE OF COMPOSITE CONCRETE BEANS

PART I: fliTRODUCTION

Background

1. During the past decade or two, resin concretes and resin mortars

have generated ever-increasing interest throughout the building comminity;

and as a result, considerable data on different resin concrete* mixtures

have been forthcoming from research laboratories all over the world. Al-

though a large variety of resin-hardener systems has been investigated for

this purpose, the vast majority of the work has been principally concen-

trated on three groups of resins:

a. Epoxy resins.

b. Polyester resins.

c. Furanic resins.

This country has taken the lead in the development and study of the first

two resin groups for civil engineering applications, while scrn European

laboratories have concentrated on the lower strength, but more economical,

f'uranic resins, apparently with fair success. 1 5

2. From the growing accirmulation of individual data, a technology

is nov be;,inning to evolve, comparable to the well-established conventional

cuncrete or asphalt technology; in fact, certain basic relations have

already been established. However, it appears that most studies to date

have been restricted ersentially to what one might call '"asic mixture

proportioning." For this r-ion, relatively little is ::novn about the per-

forrance of particular resin concretes or mortarz in the environments of

their pwtential use, especially, over any length of time. Several studien

have been undr rtiken on the inf urer.-x of resin modifier and hArdener type

&r...' content, as well nr aggregate rdner~log', shApe, moisture content, /ijd

The term resin concrete t ripplied to co.ncretes using resins in lieu ofportland cer.,nnt as a binder for ti'e ag~.regate ;articles.

grading, upon the static compressive strength of standard cubes and cy-

linderA, the elastic modulus (E), the rupture modulus, etc. Not until

fairly recently, however, was it realized that resin concretes can be

extremely sensitive to placing, curing, and testing conditions (e.g.

relative humidity, temperature, loading rate, and specimen shape and size),

obviously much more so than conventional concretes. This sensitivity makes

a comparison nrl evaltation of test results from different laboratories

difficult. Also, little is known about the important long-term behavior

of resin mortars in different environmental situations; thus, applications

of the new construction material in practice have been slow and very

limited in scope. This is, of course, largely a result of the still com-

paratively high costs of resins, especially epoxies. The majority of

practical applications of resin mortars to date has been in the field of

grouting7'8 and concrete repair,9"13 especially repair of concrete pave-

ment and bridges.4,5,14,15 True structural applications, i.e. important

load-carrying uses, of resin concretes or mortars have been scarce and

cautious and involved only small volumes of material. The few actual

structural applirations known to the author were made in the construction

of composite steel-concrete bridges where pure epcxy resins and resin

mortars have been used to bond oncrete decks to steel girders;16,17 also,

these resins have been used in prefabricated concrete construction to join

individual parts to ensure their monolithic action.18"23 The designers,

however, were usually careful not to rely entirely on the resin mortar

for the safety of their structitres and, more often than not, provided steel16,18

connectors fer good measure.

3. One can safely say that the potential of resin concrete as a

structural material is just beginning to be explored. Considering the

high tensile strength (fu ) to compressive strength ratio, the excellent

corrosion resistance, and the possible low permeability of these concretes

(in addition to high compressive strength, good bonding characteristics,

and rapid setting time) a wide field of structural applications can easily

be visualized--if it were not for thf. punishingly high costs. It should,

perhaps, be remembered that a large number of structures, in additien to

carrying external loads, ar', exposed to rather severe physicochemical

2

environments that necessitate very expensive auxiliary measures to protect

conventional construction materials (such as portland cement concrete,

masonry, or steel) from premature degradation. Provisions to protect the

load-bearing structure from an aggreý-ive environment are sometimes more

expensive than the structure itself. Therefore, it seems logical to search

for a material tha' combines the function of carrying the load and protect-

ing the otructure from the particular aggressive environment. Resin con-

crete appears very capable of fulfilling this double function in the ma-

jority of cases. In fact, it can well exceed conventional structural

materials in strength and at tLe same time surpass standard protective ma-

terials in corrosion resistance and impermeability. In addition, resin

concrete develops its full strength in a very short time, just the oppo-

site (,f conventional concrete.

4. From the above, it follows that resin concrete should not be

regarded as a potential substitute for all conventional concrete since

resins probably will always be more expensive than portland cement. But

in cases where conventional concrete is incapable of giving the desired

combination of strength and corrosion or moisture protection, resin con-

cretes could play an increasingly important role.

5. The study reported herein is an attempt to comoine conventional

concrete and resin concretes in a composite structural member so that the

advantages of both materials can be utilized to achieve an optimum solu-

tion from a technical and economic standpoint. The idea is to replace part

of the conventional concrete in the tensile zone of a beam with resin con-

crete, thereby taking full advantage of the higher f of this material1.1

without increasing the overall cost to an uanacceptable level. A resin

concrete layer at the bottom of a beam in which the tensile reinforcement

is embedded should help in bond and shear and also elimninate cracking

normally inherent in heavily loaded, reinforced concrete beans, thus sig-

nificantly redu-ing the threat of corrosion of the reinforcing steel,

particularly in highly aggressive environments (e.g. desalination plants,

maritime construction, etc.). On the underside of a slab, such a layer

could serve as an integral surface protective layer, shielding the struc-

ture from moisture and chemical attack while contributing to its strength.

3

Due to the considerably higher strength of resin concretes as compared to

conventional concrete (though usually at a relatively low E and high

creep), a significant increase in the load-carrying capacity could be

hoped for, possibly even allowing a reduction in the physical size or the

amount of reinforcement in the member. Reinforcing rods with poor bonding

characteristics (plain steel bars, fiberglass rods, etc.) could probably

be used also.

6. The concept of a reinforced, conmposite portland cement and resin

concrete member, if proven feasible and worthwhile, could certainly be

extended to other configurations (e.g. sandwich construction), different

applications (e.g., repair and strengthening of structures), and structural

elements. However, it is the flexural member that should, in theory, ex-

hibit the most benefic-.± effect. As a result, this feasibility study was

limited to simply supported beams with a rectangular cross section and two

fairly typical resin systems.

Objective and Scope

7. The objective of this pilot program was to investigate the fea-

sibility of using a layer of high-strength, corrosion-resistant, imperme-

able resin concrete in the tensile zone of reinforced concrete beams in

order to improve their strength (or make a reduction of reinforcement

possible), increase their resilience, and increase their resistance to

aggressive environments as well as to facilitate the use of reinforcmng

materials with poor bond characteristics and/or chemical compositio-is

incompatible with portland cement concrete. A secondary objective was

to develop resin concrete mixtures and procedures suitable for such

applications.

8. The scope of the investigation was restricted to two particular

resins, i.e. a two-component polysulfide-epoxy compound and a polyester

resin-methyl ethyl ketone peroxide catalyst system.

PART II: TESTING PROGRAM

9. The experimental program was conducted in the following three

phases.

a. The first phase consisted of the design of resin concretes

and the evaluation of their physical properties. With the

two resin-hardener systems chosen as binder materials, a

3/8-in.* maximum size, rounded quartz-chert aggregate was

selected for the main program. Several test series were

performed to determine an optimum aggregate gradation and

resin content for the two resin concretes. Eventually the

values of some of the most important physical properties

of both optimum mixtures (such as compressive strength,

flexural strength, tensile strength, stress-strain charac-

teristics, shrinkage behavior, etc.) were established. Ad-

ditional tests were undertaken with crushed limestone ag-

gregate. Excellent strength results for a polyester concrete

with limestone aggregate were obtained on standard laboratory

specimens; however, due to its excessive shrinkage, the pol-

yester concrete with limestone aggregate could not be used

successfully in the main beam program. (See paragraphs 37

and 62).

b. Reinforced and unreinforced concrete beams with epoxy

concrete layers were fabricated and tested. Nine 78- by 9-

by 4-in. simply supported beams with different reinforce-

ment and 1-1/2- and 3-in.-thick layers of epoxy concrete

(table 1) were made and tested under third-point loading.

The results of these tests were then compared with the re-

sults obtained from a conventional reference beam without

an epoxy concrete layer.

c. In the last phase, reinforced and unreinforced concrete beams

* A table of factors for converting British units of measurement to metric

units is presented on page ix.

5

were fabricated with polyester concrete layers and tested.

Part of the previous test series, involving a total of four

beams with resin concrete layers and a reference beam, was

repeated using the more economical polyester resin concrete,

and the results were compared with those of the reference

beam (table 1).

Materials and Techniques

Epoxy and polyester resins

10. The following two resin-hardener systems were used for this

a. A two-component polysulfide-epoxy compound (1:1 by volume)

having an amber color and a syruplike consistency (price:

about $13 per gallon or tl.15 per lb).

b. A two-part polyester resin-methyl ethyl ketone peroxide

catalyst diluted with 60% dimethyl phthalate and having

an almost waterlike appearance and viscosity (price: about

$0.38 per lb for small quantities).

Aggregate

11. Dry, clean quartz-chert aggregate was used for all but a few

resin concrete mixtures to ensure good bonding characteristics with the

resin matrix and minimum shrinkage. In order to obtain a minimum void

content between the aggregate particles and thus achieve the most economi-

cal resin concrete with a relatively high elastic modulus and good strength,

two test series were performed to select a maximum bulk density grading for

both a continuously graded 3/8-in. maximum size aggregate and a gap-graded

aggregate with the same maximum size.

12. Continuous grading. The aggregate was graded seven frac-

tions (3/8-in, to No. 4, No. 4 to No. 8, No. 8 to No. 16 16 to No. W0,

No. 30 to No. 50, No. 50 to Nu. 100, and passing the No. IuO sieve) and an

exponential sieve curve

A (d/D)n (reference 2 4 )

6|

ii

where

A = amount of material passing sieve opening d (in %)D = maximum size aggregate

d = variable sieve opening

Four different values (between 0.2 and 0.5) for the exponent n were

tried, and the bulk density of the resulting aggregate mixtures was de-

termined. It was found that the greatest aggregate compaction, i.e. mini-

mum void content, could be obtained with an exponent -f about 0.4, i.e.

slightly below the exponent that yields the familiar Fuller parabola

(n = 0.5) (see plate 1). This result agrees well .,ith earlier Lindirgs24

for alluvial sand-gravel mixtures. The grading used for all continu-

ously graded resin concretes is tabulated below.

Passing Retained PercentageSieve on Sieve by Weight

3/8-in. No. 4 25.2

No. 4 No. 8 17.3

No. 8 No. 16 13.9

No. 16 No. 30 10.6

No. 30 No. 50 8.0

No. 50 No. 100 6.5

No. 100 i. 5

13. Gap grading. According to the theory of packed spht;res, the

diameter, d , of a small sphere that will slip through the gaps between

densely packed larger spheres of a constant diameter, D , (in ictahedral

or tetrahedral configvxatios) cannot exceed

d = 0.l.-5D (reference 24)

24in practice, hcwever, it is recommended to reduce d to at least O.14D

since aggregates are not truly spherical and will be surrounded by a

binder matrix that further decreases %he size of the gaps. Thus, taxing

the minimum di .,ter of tVe coarsest aggregate fr,.:tion, 3/ 8 -in. to No. 4,

we obtain a theoretical maximum diameter for the neyt smaller fraction

o.14 x 0.187 = 0.0262 in.

14. Based on these results, a quartz-chert aggregate (0.0059 to

0.0234 in.) supplied by the resin manufacturer for use with their epoxy

resins was considered suitable to fill the voids between the larger ag-

gregate particles. The percentage of the fine sand in the total aggre-

gate mixture was subsequently varied and the loose unit weight of each

aggregate composition determined in order to pinpoint a maximum aggregate

bulk density. Plate 1 shows that a minimum void content occurred at about

35% (by weight) fine sand content; this grading was subsequently maintained

for all gap-graded mixtures.

15. Warren25 reported an optimum strength for gap-graded resin mor-

tars when the diameter ratio, the ratio of the mean values of the mesh

numbers defining the coarse and the fine sand, was about 1:14 and the per-

centage of the fine sand ranged between 30 and about 50% of the coarse

sand (by weight). This empirical result seems to confirm the validity of

the theoretical considerations that led to a diameter ratio of about 1:20

and a weight ratio of nearly 2:1 between coarse and fine aggregates.

16. Additional trial series with limestone aggregate. Crushed lime-

stone aggregate from Tennessee with a continuous grading, as described

above, was used instead of siliceous aggregate for a seriels of trial tests

with the polyester resin binder. Since the results of thcse tests were

considered unsatisfactory (see paragraph 37), limestone aggregate was later

abandoned in the beam test series.

Resin concrete mixture prorortioning

17. Using each quartz-chert aggregate mixture, gao-graded and con-

tinuously graded, a series of mixtures was made with both resins, varying

the resin content between 10 and 20% with respect to the total aggregate

weight. Prismatic and cylindrical test specimens were fabricated from

each mixture, and the unit weight, comprescive strength, E , and mn-dulus

of rupture were determined after 7 days of curing at room temperature.

Plates 2 and 3 and tables 2 and 3 swunari-e the results of these tests.

A similar series was subsequently made with the polyester resin and a

8

continuously graded limestone aggregate; the results are shown in table 4and plate 4.

18. Two mixtures were finally selected on the basis of strength,

shrinkage characteristics, workability, and economy. It was decided that

a gap-graded mixture with a 16% epoxy resin content and a continuously

graded mixture with a 10% polyester resin content would be used for the

main test series. The decision to use a gap-graded aggregate in connec-

tion with the epoxy resin and a -'intinuously graded aggregate for the

polyester concrete was prompted by the different viscosities of the two

resins, which resulted in distinctly different workability and bleeding

characteristics. One limestone aggregate-polyester resin mixture (12%

resin content) that showed excellent strength characteristics but seemed

to shrink excessively was also subjected to further testing.

19. All resin concretes were mixed in a 1 .u ft verticai mixer with

additional hand mixing to ensure thorough homogenization. After mixing,

the resin concrete was placed into the molds in 1-1/2- to 2-in.-thick lay-

ers and compacted with regular tamping rods (polyester concretes) or me-

chanical tampers (epoxy concretes). Laboratory temperatures during the

mixing and placing, as well as during the subsequent curing and testing

period, stayed between 70 and 90 F with the relative humidity ranging be-

tween 50 and 90%. The lack of close humidity and temperature control in

the laboratory is believed to have caused some of the variations in the

test results.

Portland cement concrete data

20. Unintentionally, two different portland cement concirete mixtu':es

were used in the epoxy and the polyester resin concrete beam series. J>oth

concretes were proportioned with J!3•-1*z. maximum size crushed limestrne ag-

gregate to have a slrup of 2 + 1/2 .n. and coxpresssive strengths of ap-

proximately hO00 and 13000 psi, rc:p-.tively, at 2,4 da'.s. Mixture data and

physical properties of the two ý'_ncretcz are coniled in ý.ble ). It is

felt that the use of two diffe'-ent portlan! cement concretes :n the two

test series, although unintertionrl, ili not i dvai'iate a ciMpar.'on of

results in the two series zinze all lcazr,. (vxcep.. A.) werie !xtre•mel- under-

reinforced and beam failu:es were tli',tated by The tensile ztren•th of khe

r reinforcement ana of the resin concrete layer rather than by the strength

of the portland cement concrete itself.

Properties of reinforcing materials

21. The following reinforcing materials were used in the two com-

posite beam series.

a. Deformed high-strength No. 4 steel bars (nomimal 1/2-in.

diameter), obtained in Mississippi, were uLed as longi-

tudinal reinforcement. The steel had a yie Ld strength

(f ) of 53,500 psi (plate 5), an f u of 69,000 psi, and

an E of 29.9 x 10 psi. Its stress-strain curve up to

10,000 microstrain* was essentially bilinearly

elastoplastic.

b. Stirrups were undeformed No. 2 steel bars, also obtained

in Mississippi (f = 43,000 psi, fu = 76,500 psi,6 y

E = 29.3 x 10 psi).

c. One-half-in.-diameter deformed polyester-fiberglass

rods were also tested. These rods had an E of

6.7 x 106 psi, a linear stress-strain curve up to failure,

end an f in excess of 87,000 psi (used in beam 5A).

Fabrication of Test Beams

22. A total of fourteen 78-in.-long beams with r.ýtangular cross

sections (4 by 9 in.) were cast during the resin concrete composite beam

program. In casting the beam in an inverted position, the !'nventional

concrete was first placed and consolidated with internal vibrators. The

forms were removed after two days, and moist curing was continueO tc a

concrete age of 7 days, whereupon the specimens were stored in the labo-

ratory air. Preceding the application of the resin concrete layers at

21 days age, all concrete and rein-forcement surfaces to be In contact

with the resin concrete were first sandblasted and then painted with pure

resin. Finally the resin c-oncrete was placed in 1-1/2- or 3-in.-thick

.10 -6in./in.

10

layers and compacted with tamping rods or mechanical tempers, as mentioned

earlier. The total number of beams were fabricated in one ten-beam series

and in one four-beam series.

23. This first series (Series A) consisted of the following beams.

a. Beam 1A (reference beam). This beam was a conventional

concrete bean with two regular No. 4 deformed steel rein-

forcing bars and vertical tiev (No. 2 bars at 4-in. spaci%,

except for a 16-in.-wide portion in the teatm center that

contained no ties). The reinforcement arranWement and exact

beam dimensions are shown in table 1 and plate 6 for all

beams.

b. Beam 2A. This comuposite beam had 7-1/2 in. Of portland

cement concrete and a l-i/2-in.-t*_ck bottom Lt'ycr of epcxy

resin concrete. The beam was reinforced with s~'il bars

and vertical stirrups as in beam MA.

c. Beam 3A. A composite beam similar to beam 2A, except that

beam 3A had only one reinforcing bar.

d. Beam 4A. Similar to beams 2A and 3A, except that beam 4A

did not contain any longitudinal reinforcement.

e. Beam 5A. The. same as beam 2A except that this beam was re-

inforced with two N1o. 4 deformed fiberglass rods instead of

regular reinforcing bars.

I, Beam 6A. A composite beam having 6 in. cf portland cement

concrete and 3 in. of epoxy resin concrete on the bottom.

The beam wi., reinforce:1 with two steel bars and vertical

stirrups iientical with t.ose used in beam 2A.

f. Beam 7A. The same as beam. 2A except that beam 7A had a

'-in.-thick layer of epmzzy concrete in the tension zone.

Bh. Beam 7'. 4.e st•e as beam 4,A except that team ,IA was cast

with a .- in.-thi_ layer of ,:pmy concrete.

i. E-eam )A. The s•a;n bea% .A, exce•,t tn-.t beam 9A cortai.•ed

nc shear reinfrceernt.

P. aeam 1iA. The z%-= a: beam =A cx,-ept that beam11 A con-

tained sc stirrups.

11

24 . Due to disappointing results in the polyester concrete pretest

series and to limited funding, orly four beams were cast and tested in the

polyester resin concrete phase of the comrosite beam program. This series,

Series B, consisted of the following beams.

a. Beam ]. (reference beam). The same as beam IA of epoxy

concrete series.

b. Beam 6B. The s-e as beam 6A of epoxy concrete series ex-

cept that polyester resin, quartz-chert aggregate concrete

was used instead of epoxy concrete.

c. n" ._B" The same as beam 7A (using polyester concrete in-

stead of epoxy concrete).

d. Beam 8B. The same as beam 8A (using polyester concrete in-

3tead of -pcxy concrete).

Two additional beams, identical with beams 2A and 3A of the epcxy resin

concrete series, were fabricated using the Iolyester resin-limestone ag-

gregate concrete mixture described in table 7 for the 1-1/2-in.-thick resin

concrete layer. During setting, however, the polyester-limestone concrete

developed numerous shrinkage cracks (photograph 5), and the beams were not

tested.

Test Procedures

Tests on resin concrete mixtures

25. The following series of preliminary tests was performed on spe-

ciallý prepared specimens.

a. Modulus of rupture &-Ad flexural elastic modulus (7-day

tests). To determine the rmoduluz of rupture amd the

flural elastic modulus, two 2- by 2- by 11-l/b-Ln.

prisim (photoraph 1) maae fro each rixture were put on

roller supports 10 in. ap&rt with their finished side up

and third-point lisdei nt a rate of appre•imitely . in./

min. Diel gs~ge: rcasured mi-span -erlecticn ur-dt.- in-

creasing lcir 42 )trap ) Ba:ic li:•eajr elastic" equa-

tions werc used tc calculite •:;e .ila t.. nxoujluz and tAe

l.n of rurture.

I2

b. Compressive strength at 7 days. For these tests 3- by 6-in.

cylinders, 2-in. cubes s.wed from the remain• of the pris-

matic specimens after completion of the flexural tests, or

both, served to determine the compressive strength of all

trial mixtures at 7 days.

c.Unit weit. All tesi specimens were weighed and measured

before loading to determine their unit weight.

d. Visual bleeding and shrinkage observations. After fabrica-

tion, the specimens were repeatedlv observed for bleeding

and for development of visual gaps between the specimens

and the molds that would indicate exceazi,,e shrinkage.

26. After selecting the most suitable resin concrete mixtures, a

more comprehensive test program was conducted on these mixtures to deter-

mine the values cf their most important engineering properties.

a. Tensile splitting test. A group of 3- by 6-in. cylinders

served to obtain the tensile splitting .. Lrength of selected

resin concretes in accordance with methca CRM-C 77-61 (26).

b. Direct tensior test. Using 2- by 6- by 1/4-in. plates as

inserts in regular 2- by 2. by 11-1/4-in, prism rolds,

necked specimens were obtaired on which direct tension tests

were performed at 7 days Ege (photcgraph 3). These speci-

mens were instrumented with 1-in. strvJ.n sages to obtain

stress-strain curves and Poisson's ratios in tension. The

loading rate was about 0.05 in./min.

c. Stress-3trair. curves. n addition to stress-strain curves

and Foisson's ratios in tension, as described above, regu-

lar compressienal stress-strain -,.n-'ves anu2 Poisson's ratios

were detertmined cn strain-gaged 3- by t.-in. cylinderi also

at 7 days a&e. The loading ra's for t-.ese tests was Vgain

appraximately 0.05i in.,r.in.

1. Shrinkage characteris:ic,. l-, . j r -r

of resin concretes were m:asure-d with s-in. me--haical strain

gafes by inserting the s li. :s int t sur~fce r-

2- by 2- "y Il-1i.-in. pri-v. •-n: taklirw continuoul readlirns

as soon as the resin had set up enough to allow such

measurements.

e. Strengtdetelopment with time. Using 3- by 6 -in. cylinders,

the change in the compressive strength with time was deter-

mined for all three resin concrete mixtures.

27. The length changes of two 2- by 2- by 11-i/4-in, prisms during

five temperature cycles between 75 and 150 F (leaving the specimens exposed

to each temperature until no further changes were observed) were meisured

qith b-in. mechanical strain gages.

28. Thermocouples in the center of l-qt vacuum bottles filled with

fresh resin concrete (which was rodded for compaction) served to obtain

temperature rise curve. for the three mixtures.

Beam tests

29. Loading apparatus. A rigid steel testing frame, shown in photo-

graph 4, was used for all beam tests. The beams were supported on a full

rocker cystem on one side and a half rocker system on the other side, pro-

riding a span of 6 ft. Third-point loads were applied, by two calibrated

hydraulic jacks resting on ball bearings. One-inch-wide pads between the

rollers and the beams served to distribute loads and support reactions.

30. Test measurements and instrimentation. Longitudinal strains in

the tensile reinforcement were measured in the center of each beam by

1/4-in. resistance strain gages glued to the reinforcing bars. Concrete

surface strains at the top and the sides -Df the be;am center were measured

with a 2-in.-long mechanical strain gage (for the location of surface strain

meesurements see plate 7). One-in. dial gages mounted on an independent

scaffold, thus unaffected by deformations of the testing frame, measured

beam deflections at five points along the span. A hydraulic system con-.

si3ting of two 20-ton jacks, a control panel, and a 2500-psi precision

pressure gage (calibrated before and after the test series) was used to

apply and measure loads.

31. Test procedure. In placing the beam in the testing frame, par-

ticular attention was given to the exact alignment of the test beam, the

supports, and the loading assembly to ensure true axisymmetric bending.

The five dial gages, mounted on a separate scaffold, were zeroed against

14

the underside of the beam. Strain gages were connected with the WIdeat-

stone Bridge, and initidl (electrical and mechanical) strain readings were

taken.

32. Loads were applied in 500-lb increments (total load), and beam

deflections -iere read after each increase in load. At intervals of 1000 lb

(total load) a full set of mechanical, concrete surface strain and electri..

cal reinforcement strain readings were taken. Upon completion of the

strain measurements, deflections were read for a second time under the same

load, generally about 3 min after the first reading.

33. Loads were completely released at intervals of 3000 lb to check

nonelastic deformations (deflections and strains) of the beam prior to the

continuation of loading. A full load-unload cycle, leading to a total load

3000 lb higher than the maximum loal achieved in the previous cycle, usu-

ally took about 15 to 20 min to complete.

34. Cracks were observed throughout the test and all hairline cracks

were marked with ink. Upon any significant change in the crack pattern a

photograph was taken. All beams were tested to failure.

15

PART III: TEST RESULTS

Resin Concretes

Results of preliminary tests to se-

lect optir.mm resin concrete mixtures

35. Tests concerning the selection of aggregate type and grading

were described in paragraphs U1-]6; results are shown in plate 1. Two

types of aggregates, a quartz-chert sand and gravel aggregate and a crushed

limestone aggregate, with identical gradings were tried during the poly-

ester resin concrete mixture proportioning series. It is interesting to

note that the limestone aggregate gave a considerably higher strength than

the equivalent quartz-chert aggregate mixtures (tables 3 and 4). However,

short pot life, rapid setting, and excessive shrinkage of the polyester-

limestone concrete prevented its successful use in the main beam test

series. (For more details see below and paragraph 62.)

Trial mixtures

36. Epoxy resin concretes. Results of two series of trial mixtures

(using a continuously graded and a gap-graded quartz-chert aggregate and

varying the epoxy resin content between 10 and 20% of the total aggregate

weight) are sumnarized in table 2 and plotted in plate 2. Based on the

results of these trial mixtures, which indicated a better workability and

a somewhat higher strength for gap-graded aggregate mixtures, a 16% resin

concrete with gap-graded aggregate was finally selected for the main epoxy

concrete beam test series.

37. Polyester resin concretes. As mentioned earlier, two different

types of aggregates, i.e. quartz-chert sand and gravel and crushed lime-

stone, were investigated during the polyester resin concrete trial mixture

tests. Results of two series with quartz-chert aggregate (again usin& a

gap-graded and a continuously graded aggregate and varying the resin con-

tent between 10 and 16%) are compiled in table 3 and plotted in plate 3,

while the results for the third series, using continuously graded limestone

aggregate, are summarized in table 4 and plate 4. From these tables and

plates it can be seen that the limestone aggregate series showed a

16

considerably higher 7-day strength than both quartz-chert aggregate series.

However, all limestone mixtures also exhibited short pot life, flash set-

ting, and excessive shrinkage. Additional series on limestone aggregate-

polyester resin concrete mixtures were subsequently conducted, reducing

the catalyst content from 1 to 1/2, 1/4, and 1/8% in an attempt to elimi-

nate the undesirable flash setting and reduce shrinkage. However, even

a drastic reduction in the catalyst did not satisfactorily eliminate the

problems. Precooling of the aggregate and the resin provided a somewhat

longer pot life; however, the setting was still very rapid, resulting in

high concrete temperatures and excessive shrinkace.

Engineering properties of theselected resin concrete mixtures

38. Summarized in tables 6 to 8 are the following: unit weights,

7-day compressive strengths of 3- by 6-in. cylinders and 2-in. cubes,

moduli of rupture of third-point loaded 2- by 2- by 11-1/4-in, prisms,

tensile splitting and direct tensile strengths, elastic moduli in

tension and compression, 7-day shrinkage values, average coefficients

of thermal expansion (between 75 and 150 F), and approximate pot life

of the selected epoxy resin concrete and of the two polyester resin con-

cretes (a continuously graded quartz-chert aggregate mixture with 10%

resin content and an identically graded crushed limestone aggregate mix-

ture with 12% resin content). Average stress-strain curves in tension

and compression for the three resin coo'cretes are shown in plate 8.

Plate 9 and table 9 depict the strength-time relations, plate 10 the

shrinkage curves, and plate 11 the exothermal temperature rise for the

three mixtures.

Results of Beam Tests with Epoxy Resin Concrete

39. Results of the 10 beam tests in Series A are summarized in

table 10, which indicates the cross-sectional geometry of individual beams

(for exact dimensions see table 1) and lists their calculated and measured

ultimate loads, cracking loads, and midspan deflections under various load

levels.

17

Cracking and failure of beams

40. Beam IA (reference beam). The first hairline cracks in the

portland cement concrete appeared at about 4000-lb total load and in-

creased in size and number as the load increased (photograph 6). At about

12,800-lb total load the reinforcement began to yield. This caused rapidly

increasing beam deflections and led finally to a compressio•Aal failure of

the concrete at the top of the beam at an ultimate load of 13,000 lb.

41. Beam 2A. The first hairline cracks in the portland cement con-

crete appeared at the 6500-lb total load; they dia not, however, extend

into the epoxy resin concrete layer (photograph 7). Beam deflections at

all load levels were considerably smaller than those of the equivalently

reinforced reference beam. Increasing loads subsequently caused an in-

crease in the number and size of the portland cement concrete cracks, but

the epoxy concrete layer in which the tensile reinforcement was embedded

remained uncracked up to a total load of 15,000 lb. At this load the

l-l/2-in.-thick resin concrete layer suddenly developed a single major

crack and the steel reinforcement started to yield, causing rapidly in-

creasing beam deflections. Compressional concrete failure finally occurred

at an ultimate load of 15,300 lb.

42. Beam 3A. Here the first hairline cracks in the concrete were

observed at 3000 lb (photograph 8). Again the epoxy concrete layer re-

mained uncracked while a total of nine cracks gradually developed in the

portland cement concrete as loads increased. At 8100-lb total load the

epoxy concrete layer failed in tension. Rapid yielding of the single re-

inforcing bar resulted in compressional concrete failure.

43. Beam 4A. As might be expected of this unreinforced beam, sudden

failure occurred due to simultaneous cracking of the tensile (epoxy and

portland cement) concrete layer and was not preceded by the formation of

visible hairline cracks in the portland cement concrete (photograph 9).

However, this mode of failure was distinctly different from that of the

other unreinfrrced beams, which did develop tensile cracks in the port-

land cement concrete long before the cracking of the resin concrete led

to sudden failure.

1414. Beam 5A (fiberglass-reinforced).. The first visible cracks under

18

the 6000-lb total load extended from the bottom of the beam through the

epoxy concrete layer up into the top third of the beaer. Subsequent in-

creases in the load caused a gradual increase in the size and number of

cracks (photograph 10) until at 16,600 lb the beam suddernly failed in a

compressional mode.

45. Beam 6A. Cracks were first obserred urder the 8500-lb load.

They subsequently became larger and more iumerous, but d!..d riot start to

extend into the epoxy concrete layer until the total loal reached 12,000 lb

(photograph 11). At 14,0oo lb, the reinforcement began Ito yield, and com-

pressional concrete failure occurred at 14,51O lb.

46. Beam 7A. Hairline cracks .3tarted to appear at 5500 lb, and the

first crack in the 3-in.-thick epoxy concrete layer was objerved under

8000-lb total load (photograph 12). Under the 9000-lb load the reinforce-

ment started to yield, causing a rapid incrtase in deflection without fur-

ther increases in load.

47. Beam 8A. Photographt 13 shows a number of hairline cracks that

formed in the portland cement concrete under 4000-lb total load while the

epoxy concrete layer remained iitac:t. With increasing loads, the cracks

in the portland ceme•at concrete increased in size and number until at

5500 lb one of the cracks finally propagated into the epoxy. concrete layer,

causing sudden failure.

48. Beam 2A. Similar to the preceding teot, the portland cement

concrete developed a tctal of six cracks under a load of 3000 lb (photo-

graph 14), thereby transferring tensile forces to the l-1/2-in.-thick

layer of epoxy concrete. The epoxy concrete layer remained intact until

B300 lb wher one of the concrete cracks propagated into that layer, ini-

tiating sudden failure.

49. Beam IIA. Again, hairline cracks in the portland cement con-

crete started to form at a relatively low load (about 2',100 lb) and became

larger and more numerous as the load increased (photog-raph 15); however,

not until a total load of 4200 lb di,! one of the cracks propagate into the

epcxy concrete layer, causing failure.

Beam defl ection-

'•0. Lo<ad-deflectlon curves for all 10 beams of the epaxy concrete

l1

series are presented in plate 12. Individual deflection readings at five

points along the span are compiled in table 11, and midspan deflections

for each beam during all loading and unloading cycles are plotted in

plate 13.

Strain readings

51. Average readings, under various incremeuts of load, from two

1-in. bonded resistance wire gages cemented to the reinforcing bar(s) and

waterproofed are shown in plate 14. The same plate also summarizes aver-

age mechanical strain measurements on the concrete surface, namely, com-

pressive strain measurements on the top beam surface (average of four

readings at locations 1, 2, 3, and 4 in plate 7); lateral measurements on

the same top surface (average of three readings at locations 5, 6, and 7

in plate 7); and tensile strain measurements at two different elevations

in the lower portion of the lateral beam surfaces (i.e. average of four

readings at the bottom of the beam, locations 10, 11, 14, and 15 in

plate 7); and average of another four readings about 1 in. abcve the bottom

of the beam, locations 8, 9, 12, and 13 in plate 7. It must be emphasized

that in the case of the last two mechanical tensile strain measurements the

average plotted in plate 14 was repeatedly obtained from widely varying

individual readings, since some of the 2-in.-long measuring distances in-

cluded a crack while others did not. Obviously, the strain readings in

cracked sections were very high, while the neighboring uncracked sections

hardly showed any strain at aUl following the formation of a crack in the

adjoining section. In some instances uncracked sections close to a major

crack showed small compressive strains (up to a maximum of 150 micro-

strains (10-6 in./in.)). The average mechanical tensile strain measure-

ments therefore represent an average strain over a 4-in.-long distance in

the beam center that at higher loads frequently included at least one

visible crack.

Moment-curvature relation

52. Midbeam curvatures under different loads were calculated from

the average compressional strain reading taken on the top beam surface,

from two or three different tensile strain rceading:, i.e. electrical strain

readings on reinforcing bars, and from . echanical :train readings at two

20.

elevations on the lateral beam surface. Two or three independent curvature

values were thus obtained for each beam and loading condition (table 12).

The agreement between the three values was usually surprisingly good, witn

most individual values staying within less than +i0l1 of their cmummon mean.

Only for very small curvature values (after unloading) and in some instances

at very large curvature values (preceding failure) did individual values

deviate considerably more than 10% from their mean--except beam 3A where

a significant difference between electrical and mechanical strain readings

occurred throughout most of the test.

53. To facilitate comparison, average noment-curvature curves (omit-

ting all unloading phases) for the 10 beams of the epoxy concrete series

are presented in plate 15a.

Results of Beam Tests with Polyester Resin Concr2tes

54. It was originally planned to duplicate the whole epoxy concrete

beam series in the polyester resin concrete series; however, due to disap-

pointing results with the polyester resin concrete and limited funding, it

was decided to curtail the polyester concrete beam program. The results of

the four beam tests in the abbreviated polyester resin concrete series,

Series B, are suzmrized in table 13, which lists the cross-sectional geom-

etry of individual beams, calculated and measured ultimate loads, cracking

loads, and midspan deflections under vwrious load levels. Two additional

beams, with cross sections identical with thos.e of beams 2A and 3A in the

epoxy concrete series, were cast using the limestone aggregate polyester

resin mixture described in paragraph 18 for the resin concrete layer.

However, during setting, numerous large cracks developed in the resin con-

crete layer (photograph 5) and the two Levuns were not tested.

Cracking and failure of beams

55. Beam lB (refer-!nce ben. The first hairline cracks were ob-

served at a -OW-lb total load and in-reaseA in size and number until at

12,00O-li total load the reinfcrcement Artrted to yield, causing a compre.-

sional fuilure of the top concrete at 12,'tD bl (photogirftph 1 ). The

crackirn and ultimate moment at a -om.,w,.at lower loadin4 obtt-1ned on t)his

21

beam, as compared with reference beam IA for the epoxy series, was caused

by a somewhat smaller beam width (3.90 in. versus 3.97 in.) and a lower

compressive strength for the polyester resin beam series.

56. Beam 6B. Cracks in the polyester concrete layer appeared at

9000 lb (photograph 17). At a 10,000-lb total load, the por;land cement

concrete started to crack. ¶he leinforcement began yi0r.Laing et about

11,600 ib, leading to compressional concrete failure at the 12,000-lb total

load.

57. Beam 7B. A miuspan crack in the 3-in. polyester resin concrete

layer and in the lower part of the portland cement concrete appeared at

6000-lb total load (photograph 18). Subsequent increases in loading caused

a rapid growth of this midspan crack and the formation of several other

cracks. At about 6400 lb the reinforcement began to yield and the beam

reached its ultimate load-carrying capacity and failed in compression at

6840 lb.

58. Beam 8B. No visible cracks appeared in this unreinforced beam

prior to its sudden failure under 3580-lb total load (photograph 19).

Beam deflections

59. The load-midspan deflection curves for the four tested beams of

the polyester concrete series are presented in plate 16 and are plotted

individually in plate 17. Table 14 summarizes all deflection measurements.

Strain readings and

moment-curvature relations

60. Average electrical and mechanical strain readings and computed

moment-curvature relations for three of the beams of this series are com-

piled in table 15 and plates 18 and 15b in the same manner as for the beams

in the epoxy concrete series. Reference beam 1B exhibited a drastic dif-

ference between electrical strain readings on the reinforcement rods and

mechanical strain readings at the concrete surface, possibly caused by an

early hairline crack in the measuring distance.

22

PART IV: DISCUSSION OF TEST RESULTS

Resin Concretes

61. The two binder systems used were arbitrarily chosen as fairly

representative examples of a polysulfide-epoxy and a rigid polyester resin

system. It must be realized that other epoxy or polyester resin-hardener

systems will perform differently. For this reason, the results of a

limited investigation should not be generalized as some -f the problems

encountered in a pilot study could certainly be overcome by a systea, ie

investigation of various resin systems and subsequent selection of th ýost

suitable hinder for any particular application.

62. Nonetheless, a few findings in this study appear to be of gen-

eral importance, particularly those involving some negative results. The

results, for instance, indicate clearly that an evaluation of binder sys-

tems (as well as of aggregate and mixture compositions) that is entirely

based on routine strength tests of standard small laboratory specimens

(such as conventional compressive, flexural, and tensile tests) can lead

to entirely erroneous conclusions as to the suitability of a particular

resin concrete mixture. Besides not reflecting actual field conditions

that may rather drastically affect the performance of polymer binders,5'27

these tcsts may also overlook another factor of potentially great impor-

tance, namely, the effect of specimen size. A small laboratory strength

specimen usually will be relatively unaffected by exothermal heat release

and volume changes that may lead to very serious problems in actual con-

struction, sometimes making a mixture with excellent laboratory strength

results entirely useless for practical applications. The practical sig-

nificance of strength data obtained from small laboratory specimens under

closely controlled conditions is always debatable; however, with resin

concrete this appears to be a truly critical question. For instance, ex-

cellent strength results were obtained during this investigat-. n on small

specimens of a poi-e!ter resin limestone aggregate mixture; yet when the

same mixture was used in the main beam program the resin concrete developed

large cracks and turned out to be entirely unsuitable for the contemplated

23

application. The problem was caused by the excessive shrinkage of this

mixture. This condition had been anticipated in this particular casebecause unrestrained shrinkage measurements had been made together with

the routine strength tests. However, frequently such shrinkage measure-

ments will not be made; and, based on excellent strength results alone,

the erroneous conclusion could then be drawn that a very suitable mixture

had been developed. Similar problems can be encountered with exotherr-Ic

heat, humidity effects, coefficient of thermal expansion, and other ef-

fects whose importance is governed by shape, size, and boundary conditions

of the resin concrete body.

63. The results also show that even under laboratory conditions with

closely controlled manufacturing and testing procedures and only moderate

environmental changes (the laboratory rooms used were not climate controlled

and had temperature variations between 70 and 90 F and the relative humid-ity ranging between 50 and 90%) a rather large variation in test results can

occur. This seems to indicate the sensitivity of resin binders to environ-

mental factors.

Aggregates

664. The tests to develop optimum aggregate gradings and their re-

sults were described and discussed in paragraphs 11-15. -Two different

aggregates, a well-rounded quartz-chert and crushed limestone with iden-

tical grading, were used during the polyester resin mixture design series.

As emphasized before, the polyester resin limestone aggregate concretes

developed a considerably higher strength than all polyester resin quartz-

chert aggregate mixtures. However, the pot life was shorter and the

temperature rise and shrinkage considerably greater with the limestone ag-

gregate. The author theorizes that this phenomenon was caused, at leastin part, by the lower diffusivity of limestone as compared to quartz-chert,

which resulted in a faster and higher temperature rise bringing about a

quicker and more complete polymerization (as well as greater thermal

shrinkage) for the limestone aggregate polyester resin concrete. However,

other factors such as particle shape, surface texture, specific surface,

24

etc., may have been contributing factors. There is also a possibility of a

chemical reaction between the limestone aggregate and the polyester resin(which has a pH of' about 3.8). However, a brief additional test series,

in which 100 g of polyester resin and 1 g of 60% methyl-ethyl ketone per-

oxide diluted in dimethyl phthalate were mixed with 50 g of both the quartz-

chert and the limestone aggregate that passed the No. 200 sieve and the

temperature rise of these two mixtures was compared with the temperature

rise of the TrAre resin-hardener system, failed to indicate any such reaction

(plate 19). The temperature measurements showed only a minute difference

between the limestone and the quartz-chert aggregate which can be explained

by the different diffusivity of the two materials. The author strongly

feels that the observed phenomenon warrants further investigation; however,

the scope of this program unfortunately did not allow a systematic study.

It might be mentioned here that Stamenov, Goudev, and Malcev 1 9 have also

reported higher strength results for polyester concretes using a basic

rock (diabase) rather than quartz as fine aggregate. J. Michie* reported

excellent strength results for a polyester resin concrete with basalt ag-

gregate used in a Nevada field test.

Resin Content

65. Following the selection of resin binder systems, aggregates, and

aggregate gradings, the next step was the development of optimum mixture

proportions. Tables 2-4 and plates 2-4 show the results of the trial

series with varying amounts of epcocy or polyester resins.

Epoxy concrete series

66. Both the gap-graded and the continuously graded quartz-chert

aggregate series reached a maximum unit weight at a resin content of about

1II to 161 of the total aggregate weight. This result is consistent with

theoretical considerations (it being roughly the amount of resin necessary

to fill the 21 and 24% voids between the densely packed aggregate patti-S~25cles), and with results of earlier investigations. The modulus of

* Private communication with Mr. J. Michie, Southwest Research Institute,Oct 1966.

25

rupture for both serics incre.seld with increasing resin content, which,

too, is a normal result for rusin concretes. 6 ,14,19 The range of modu1-

lus of rupture values between 1795 and 2995 psi can be considered typi-

ral foi epoxy concretes, as can the modulus of rupture for the pure

resin (5395 psi). Somewhat different results were obtained for th,! two

series with regard to conipressive strength. While the gap-graded series

showed a distinct increase in compressive strength with increasing resin

content throughout the tested range, the continuously graded series de-

veloped a strength maximum at 14% resin content with values for higher

and lower resin contents falling considerably below that maximun. Sev-

eral earlier investigators 1' 3 ' 1 5' 2 7 have reported that the compressive

strength of resin concretes reached a maximum with a resin content of

about 15 to 20%, depending on the type of binder end aggregate used.3Bares juggested recently that two maxima for the compressive strength

as a function of the resin content occur in resin concretes, one at a

very high resin content (approaching the case of pure resin) and another

at a resin content of the above-indicated magnitude. It may thus be that

the gap-graded mixture -would have reached its first maximzi slightly above

20% resin content (by total aggregate weight). ThB compressive strength

of the pure resin was 12,500O psi, almost identical with the highest value

obtained with the gap-graded aggregate at a resin content of 20%. The

range of compressive strength values detenained for the epoxy resin trial

mixtures, in general, corresponds to those usually fotund in the literature

for resin concretes, 6 '9'15 though occasionally much higher values have

been reported.28

67. In his conmments in reference 25, Bares also presents a plot

showing the effect of resin content upon 6he Young's modulus of resin

-oncretes. His diagram indicates that for sn epoxy resin concrete a

maximum elastic modulus was obtained with a resin content slightly less

than that necessary to produce a maximum compressive strength, i.e. about

14% resin content with respect to the total aggregate weight. 'while this

was found to be true in this study for the continuously graded aggregate,

the flexural elastic modulus of the epoxy resin concrete with gap-graded

aggregate stayed nearly the siine for various resin contents (around

26

I

1 x 10" psi) except for the mixture with 20% resin content, which developed6

a considerably higher modulus (1.6 x 10 psi). The magnitude of values in

general corresponds to those cited by Bares for epoxy resin concretes.

However, it should be emphasized that the method used to determine the flex-

ural elastic modulus in the trial series of this program is rather crude

and cannot be expected to yield very accurate results.

68. A few remarks remain to be made about the effect of the resin

co.itent upon the workability of the tested epcxyf concrete mixtures. It

seems that an. optimum workability was obtained with a resin conient of

about 14 to 16%. Iower resin contentq resulted in dry, harsh mixtures

with poor workability; higher resin contents tended to cause serious

bleeding.

Polyester resin concretes

69. Within the tested range of resin contents (10 to 20%), the con-

crete unit weight tended to decline with increasing resin content in all

three test series (i.e. gap-graded quartz-chert, continuously graded quartz-

chert, and limestone aggregate). Due to the lower density of the poly-

ester resin (approximately 1.03 g/cu cm) a resin content of roughly 10%

(continuously graded) And 12% (gap-graded) theoretically sufficed to fill

the voids between the densely packed aggregate particles, The lower vis-

cosity of tne polyester resin, as compared with the epoxy, made it easier

to approach a 100% compaction with a low resin content.

70. The modulus of rupture of the mixtures with gap,-graded quartz-

chert aggregate stayed around 1900 psi for all three resin contents tested

in this series. Both continuously graded aggregate series exhibited a maxi-

mum modulus of rupture with 12% resin content, the values for the limestone

aggregate concretes being considerably higher thEn those for the quartz-

chert aggregate concretes (e.g. 2316 psi versus 1703 psi for 12% resin

content). Practically all polyester resin concrete mixtures in the trial

series, however, developed. a lower flexural strength than comparable epoxy

resin mixtures. The two continuously graded aggregate series also showed

a compressive strength maximum at .4% resin content, while the compressive

strength of the gap-graded aggregate series was not clearly affected by

the resin content. Again, the strength values for the limestone aggregate

27

mixtures were much higher than those for the two quartz-chert aggregate

series. In the majority of cases they even exceeded the compressive

strength obtained on equivalent epoxy resin concrete mixtures.

71. The observation that polyester resin concretes can wovelop a

higher compressive strength than epcocy concretes and at the same time be

lower in flexural strength was also made by Bares.6 In turn, Cirodde1 5

reported that the opposite can also be true. L'Hermite,28 in a very re-

cent paper on high-strength resin concretes (he reported compressive

strengths up to above 20,000 psi for epoxy resin concretes and an average

compressive strength of 12,000 psi for polyester concretes), gives almost

the same ratio between tensile and compressive strengths for both types of

resin concretes (i.e. about 0.125 to o.14).

72. The magnitude of values for the compressive strength aiid modu-

lus of rupture obtained in all three polyester resin concrete series was

similar to those indicated by most other investigators; 6 ',25,29,Varnell*

however, Stamenov, et al.,19 reported compressive strength values up to

about 17,000 psi and a modulus of rupture as high as 5200 psi for a poly-

ester resin with a diabase and quartz-chert aggregate mixture with approxi-

mately 19% resin content (with respect to aggregate weight).

73. Finally, Young's modulus in flexure for the two continuously

graded aggregate series declined with increasing resin content, and again

the limestone aggregate polyester resin concretes developed much higher

moduli than their counterpart quartz-chert aggregate mixtures (e.g.,

1.45 x 106 versus 1.14 x 106 psi at 12% resin content). The flexural

elastic modulus for the gap-graded quartz-chert aggregate series stayed at

a constant 1.06 x 10 psi for the three resin contents tested. Since

rigid polyester resins usually have a higher elastic modulus than epoxy

resirn.,100 polyester resin concretes are normally expected to show a

higher elastic modulus than epoxy concretes.25 While this was found to

be true in this investigation as far as the secant modulus in compression

is concerned, the flexural elastic modulus and the tangent modulus in ten-

sion of the quartz-chert aggregate concretes did not change -much when

* Personal communication with Mr. W. R. Varnell, Concrete Development

Corporation, Sap Antonio, Tex., 26 Oct 1966.

28

polyester resin was used instead of epoxy resin. Due to the high exotherm

of the polyester resin, an attempt to cast 2- by 2- by 11-1/4-in, pris-

matic and 3- by 6 -in. cylindrical specimens of pure polyester resin failed

and the intended comparison of their properties with those of similar pure

epoxy resin specimens could not be made.

74. Cast polyester resi.is are known for their tendency to shrinkS. ... 1,18,31,Varnell* ( hnmnntasignificantly during polymerization (a phenomenon that

18,32usually poses no problems with epoxy resins8 ). Such shrinkage was

indicated during the trial series by the development of small gaps between

the molds and the polyester concrete specimens during the setting process.

Particularly drastic shrinkage could be observed in the limestone agre-

gate series and in all series the shrinkage appeared to increase as the

resin content increased.

75. The workability of the continuously graded quartz-chert

aggregate polyester resin mixture series reached an optimum at approximately

10 to 12% resin content, and that of the continuously graded limestone ag-

gregate and the gap-graded quartz-chert aggregate mixtures at about 12 to

14% polyester resin content (of total aggregate weight). Higher resin con-

tents quickly led to excessive bleeding of the concrete, while lower resin

contents resmited in harsh mixtures with poor compactability.

76. After the conclusion of the trial series, a continuously graded

quartz-chert aggregate mixture with 10% polyester resin content a~id a

crushed limestone aggregate mixture with identical grading and 12% resin

content were selected for the beam test program. Engineering properties

of these mixtures are summarized in tables 6, 7, and 8 and are discussed

below.

Engineering Properties of the Selected Resin Concretes

Compressive strength

77. After 7 days of curing at room temperature, the three selected

mixtures developed the following compressive strengths.

* See footnote, page 28.

29

Sb

SCompressive Strength, psi Ratio of Cube3- by 6-in. to Cylinder

Mixture 2-in. Cubes Cylinders StrengthGap-graded qua.rtz-chert aggre- 10,52h 6,556 1.60

gate with 16% epoxy resin (10,890)*

Continuously givaded quartz. 6,386 6,016 1.06chert aggregaLte with 10% (7,127)*polyester re,:Iin

Continuously graded crushed 13,920 13,4,78 1.03limestone aggregate with 12% (1-1,597)*polyester resin

SValues in parentheses indicate results obtained in t:ie trial mixtureseries.

The ratio of cube to cylinder strength was thus extremely high for the

epoxy concretes and very low for the polyester concretes. Presumably this

difference is caused by the lower elastic modulus of the epoxy resin, which

perhaps makes the epoxy concrete more sensitive to end restraints.

Tensile strength

78. Three methods were used to test the tensile strength of the se-

lected resin concretes: flexural tests (modulus of rupture), tensile

splitting tests, aund direct tenmion tests. Their results are compared

below:

Modulus Directof Splitting Tensile

Rupture Strength Strength RatiosMixture (R), psi (S), psi (T), psi L. S77 1ý.

Gap-graded quartz- 2613 1140 1600 1.63 0.71 0.244chert aggregate with (2700)**16% epoxy resin

Continuously graded 1254 1162 630 1.99 1.84 Q.lo:quartz-chert aggre- (1366)**gate wIth 10% poly-ester resirn

Continuously gradea 2658 2152 12 2.0, 1.! 0C.0"crushed limestone (2316)**aggregate with LKpolyester resin

* CS, cylinder compressivLt strerkt.

** Values in parentheses indicate rezultz obtarine in trihil L.crie.

It is thought that the low tensile splitting strength for the epoxy resin

concrete (leading to the unusually low ratio of 0.71 between splitting and

direct tensile strength) is also a result of the low elastic modulus, rel-

atively high tensile strength, and pronounred j,-vsturc 4! tLc stress-

strain curve of the epay resin concrete. The tensile splitting test is,

after all, strictly valid for brittle and fairly linearly elastic mate-

rials only; and since the epoxy resin concrete tested was not such a mate-

rial, it is not surprising that its S/T ratio did not fall within the

usual range. For the polyester resin concrete with strength properties

more similar to those of ordinary concrete, the ratio resembled that usu-

ally found for portland cement concrete.

79. The ratio of direct tensile to compressive strength for the two

polyester resin concretes is only about 30 to 50% higher than the ratio

that would be expected for a conventional concrete of comrparable c',=pres-

sive strength, whereas the epoxy resin concrete developed an extremely high

ratio of 0.244, which is about three times as high as that for comparable

portland cement concrete.

Elastic modulus and

stress-straln curves

80. Stress-strain curves in compression and tension were determined

on 7-day-old specimens of all three resin concrete mixtures, and the secant

moduli between 0 and 5000 psi in compression (E c) and between 0 and 1000 psi

in tension (Et) were computed as follows:

E EtC

Mixture 106 psi 106 psi E t/Ec

3ap-graded quartz-chert aggregate 0.969 1.935 2.0with 16 epoxy resin

Continuously graded quartz-chert 1.50* 2.07" 1.38aggregate with 10% polyesterresi••

Continuously graded crushed limestone 3.56 3.34 0.94aggregate with 12A polyester resin

0 to 4000 psi.• 0 to 600 psi.

j 31

I

Plates 8a and 8b show that the stress-strain curve of the polyester resin

lime.stone aggregate mixture was nearly linear over a wide range of stresses

with a sharp curvature close to the ultimate compressional stress, while

the stress-strain curve of the epoxy concrete had an appreciable and fairly

constant curvature throughout. This explains the marked difference in the

Et/Ec ratios. The shape and position of the stress-strain curve of the

polyester resin quartz-chert aggregate concrete falls between the other two.

It should also be noted that both the ultimate compressional and tensile

strains of the epcocy concrete were much higher than those of both polyester

concretes (i.e. 10,400 versus 6000 psi in compression and 1300 versus about

400 psi in tension).

Volume changes

ý1. Total linear shrinkage curves (after setting) were obtained on

2- by 2- by 11-1/4-in, prisms. It is thought that the total shrinkage con-

sists of:

a. Thermal volume changes due to thermal contraction during

the cooling off period after the hardening of the resin.

b. Isothermal volume changes covering all volume changes not

due to temperature variations, such as volume (or density)

changes caused by polymerization.

The total (autogenous) linear shrinkage of the epoxy concrete was less than

1 X 10-4 in./in. or about a third to a fifth of the values normally en-

countered with portland cement concrete; both polyester concretes developed

much higher shrinkage. The selected polyester resin quartz-chert aggregate

mixture showed a final shrinkage (after 7 days) of about 9 X 10"4 in./in.

and the polyester resin limestone aggregate shrinkage increased to an ex-

cessive 49 X 10 4in./in. Values of the same ragnitude were reported by

Kreijger 3 2 for thin films of polyester resins. It is obvious that shrink-

age of this magnitude will pose serious problems, especially in composite

construction.

82. The question of shrinkage of polyester resin concretes is still33 an usqe~l tes34,35

controversial. In l902, Franz and Bossler, and subsequently others,

reported that "shrinkage values of polyester concretes correspond to those

of regular concretes." However, it appears that Franz and Bossler did not

32

start their shrinkage measurements until the day after the fabrication of

their specimen, and consequently measured only a portion of the total

shrinkage (see plate 10).

Exothermal heat

83. Temperature rise curves obtained from resin concrete in l-qt

vacuum bottles for the three selected mixtures are shown in plate Ul.

Due to the relatively low initial mixture temperature (65 to 69 F), two of

the three resin concretes were unusually slow to react. Polymerization

did not get under way until about 2 hr after mixing for the two mixtures

containing quartz-chert aggregate. In the case of the polyester resin con-

cretes, a new shipment of resin was used that may also account for the

much slower hardening than observed during all other parts of the program.

84. At any rate plate 11 shows that:

a. Both polyester resin concretes had a considerably steeper

temperature rise and a higher peak temperature than the

epoxy concrete, indicating a larger and faster release of

exothermal heat.

b. The use of limestone aggregate somenow accelerated the

polymerization of the polyester resin concrete as evidenced

by the shorter pot life, the somewhat steeper slope of the

temperature rise curve, and the higher peak temperature.

Linear coefficientof thermal expansion

85. Both polyester resin concretes exhibited rather small coeffi-

cients of thermal expansion (6.6 and 7.5 x l0-6i/F), which were somewhat

below the range of 8.3 to 12.0 x 10 6/OF reported by Liesegang34 '35 for33

polyester concrete. Franz and Bossler, however, have mentioned values

between 7.2 and 7.8 x 10 0 /oF for room-cured polyester concretes. The

limestone aggregate concrete showed a -maller coefficient than the quartz-

chert aggregate concrete, probably due to the lower coefficient of thermal

expansion of limestone.

8(. Surprisingly, the coefficient of thermal expansion was consider-

ably higher for the epoxy resin concrete (13.2 x 1/O6i°F) than for both

polyester resin concretes. Based on the ccefficients given in the

W-

literature for pure epoxy and polyester resins ,36,37 the opposite would

have been expected.

Composite Beam Tests

Epoxy concrete composite beam series

87. The replacement of a 1-1/2- and a 3-in.-thick layer of portland

cement concrete on the tension side of unreinforced and reinforced concrete

beams by epoxy resin concrete resulted in a distinctly increased ultimate

moment, as expected. However, this contribution to the flexural (and prob-

ably the shear) strength, which in the case of typically reinforced beams

was in the order of 10 to 20%, would hardly justify the additional expense

of an epoxy resin concrete layer since the same or a higher increase in

strength can normally be realized more easily and cheaply by conventional

means, e.g. additional reinforcement, larger cross sections, etc. However,

in the case of unreinforced beams a strength increase of about 100 to 200%

was realized through the use of epoxy resin concrete layers.

88. Thus, perhaps more significant than its contribution to the

strength is the ability of an epoxy resin concrete layer to provide a

noncracking, corrosion-resistant, impermeable cover protecting the embedded

reinforcement from corrosion even in highly aggressive environments and in

situations where ordinary portland cement concrete would have cracked long

before the epoxy, exposing the reinforcement to the environment. Whether

or not equivalent corrosion protection can also be obtained more cheaply

by other means or whether the corrosion protection plus the moderately in-

creased strength and stiffness justify the additional expense of an epoxy

concrete layer are questions that though important, cannot be resolved

within the scope of this feasibility study.

89. Ccuarison with analysis and individual results. The cracking

loads for all beams with a resin concrete layer were derived in an elastic

analysis that assumed an ultimate tensile strain capacity of 1300 x 10-6

in./in. for the epoxy resin concrete (Appendix A). In addition, the yiela

moment of all reinforced concrete beams was computed using ACI Code d equa-

tion 16-1 with • = 1 and disregarding the contribution of the resin

34

concrete layer, which was considered cracked before yielding of the rein-

forcement started.

90. For the fiberglass-reinforced beam, 5A, an analysis similar to

that described by Sinha and Ferguson,39 but again setting • - 1 , was

used to obtain the ultimate moment.

91. The higher of the two computed moments (i.e. the moment at

which the resin concrete cracked or the moment at which the reinforcement

yielded or the concrete failed) was taken as the ultimate moment and trans-

formed into the ultimate load-carrying capacity and compared with the

measured ultimate load. The agreement between calculated and measured

cracking and ultimate loads was fair for the majority of beams; however,

three unreinforced and one reinforced beam with epoxy resin concrete layers

showed large differences between calculated and measured ultimate loads.

It is felt that variations in the resin concrete are responsible for the

discrepancy. The individual results are as follows.

a. Beam lA (reference beam). As usual, the measured ultimate

moment was in very close agreement with the calculated

ultimate moment (ACI Code 318-6338 equation 16-1, using

S= 1), the difference being less than 1%.

b. Beam 2A. The epoxy resin concrete layer was uncracked up

tc a total load of 15,000 lb (while the concrete above it

showed first visible tensile cracks at 6500 lb) or up to a

load about 15% higher than the ultL~tc load cf the refer-

ence beam and 3?c higher than the cowputed cracking load.

This indicated that the actual tensile strength of the resin

concrete layer in this beam was considerably higher than

assumed in the analvysis. Failure of the beam occurred

shortly after the resin concrete cracked at a total load

of 15,300 lb. almost 20% above the ultimate load of the

reference Ie~am.

c. Beam 3A. Cracking of the epoxy concrete layer occurred at

8100 lb (the first concrete cracks were observed at 3000 lb),

and was presently followed by the failure of the beam. The

calculated cracking or ultimate moment was within 3% of the

35

test result, which was about 20%1' higher than the theoretical

'itilnate Mrment f'or the same beam without a resin concrete

layer.

d. Beam 4A (unrei½forced). The calculated ultimate moment was

some 142) below the test result, which, in view of' the pos-

sible variation in the epoxy concrete, was an acceptable

agreement. The determined ultimnte moment is thus about

three times the ultisiate moment that would be expected for

a plain unreinforced concrete beam of the samne cross section

(assuming 400-psi tensile strength of the concrete).

e. Beam 5A (fiberglass-reinforced). Both the calculated crack-

ing and ultimate moments were in very close agreement with

the test results. Due to the low elastic modulus of the

fiberglass reinforcement, the resin concrete layer cracked

under rather low loads, comparable to those of unreinforced

beam 4A.

f. Beam 6A. Due to a lower strength of the resin concrete,

possibly caused by temperature and shrinkage stresses, this

beam developed a lower ultimate moment than beam 2A despite

its 3-in.-thick epoxy concrete layer. The resin concrete

layer cracked under 12,000 lb (calculated cracking load

12,980 lb, or about 8% higher); the first concrete cracks

were observed at 8500 lb, and the ultimate load was reached

at 14,500 lb, i.e. 12% above the calculated ultimate flex-

ural load. In this case, as in the case of beams 2A and 7A,

it is difficult to explain why the measured ultimate moment

was much higher than the computed ultimate moment despite

the fact that the resin concrete had already cracked at

loads significantly below the ultimate. The author suggests

as a hypothetical explanation that the excellent bond be-

tween the epoxy concrete and the reinforcing steel may have

caused a restraint of the transverse contraction of the re-

inforcing bar at the crack section, resulting in an in-

creased yield strength.

36

i"

•. Beam 7A. The calculated cracking or ulimate load of

9240 lb was 2.7W above the measured tximate load (9000 ib)

and 15% above the observed cracking Acad. (8000 AL). Since

the computed yield load (6800 ib) was much lower than the

mneasured ultimate load and since cracking of the epoxy con-

crete at 8000 lb did not resu.lt ii: immediate yielding of the

reinforcement, the hypothetical explanaticn mentioned

in subparagraph f is again inferred.

h. Beaams 8A through 11A. Apparently due to a lcwer strength of

the resin concrete layer, these three unreinforced beams

developed only 52 to 74% of their predicted flexural capac-

ity. However, their failure moment was still almost two to

three times as high as what was expected for an unreinforced

portland cement concrete beam of equivalent dimensions.

92. Deflections and curvature. Since the epoxy resin concrete. layer

did not crack until ..e beams approached failure (except beam 5A with fiber-

glass reinforcement), the cross-sectional moment of inertia remained higher

than in conventional reinforced beams ; consequently, the curvature and de-

flections of beams with epoxy resin concrete layers were significantly

smaller. Plate 12 and tables 9 and 10 show that the midspan deflections of

reference beam 1A were between approximately 20 and 50% higher for any

given load than those of beams 2A and 6A, which had 1-1/2- and 3-in. epoxy

resin concrete layers. The difference between beam 2A with a 1-1/2-in.

layer and beam 6A with a 3-in. layer of epoxy resin concrete was relatively

small as far as their deflections were concerned, but was more distinct

with respect to curvature (plate 15).

Polyester concrete

composite beam series

93. Due to the lower tensile strength and much lower ultimate ten-

s.Ie strain of the polyester resin quartz-chert aggregate concrete used,

the contribution of this concrete layer to the flexural. strength of rein-

forced beams was small. The developed ultimate strength of reinforced

beams with 1-1/2- or 3-in.-thick polyester concrete layers was only 4%

higher than the strength predicted by the ACI Code for a conventional beam

37

of equal cross section and reinforcement. Thus, it can be said that the

polyester concrete layer contribut-d very little to the ultimate strength

of reinforced beams. However, the layer did -icrease the cracking, load

of the composite beams to about twice the cracking load of similar conven-

tional reinforced beams. Consequently, beams with polyester resin con-

crete layers exhibited somewhat smaller deflections and curvatzres in the

lower load range than conventional beams, but generally it must be con-

zidered that the investigated polyester resin concretes showed little

promise for application in this type of composite structure. Tne total

failure experienced when trying to use the polyester resin lirne-tone ag..

gregate concrete (which bad chown good tensile strength in routine tests)

was described earlier.

PART V: CONCLUSIONS AMD RECOMMENDATIONS

94. The results of this investigation showed that resin concrete

layers can successfully be used in the tension zone of flexural members

to increase their stiffness and strength. For the materials and beam

geometry chosen, 1-1/2-in.-thick layers of epoxy resin concrete led to a

i0 to 20% increase in the ultimate moment of reinforced beams and up to

a 200% increase in the load-carrying capacity of unreinforced beams.

Thicker (3-in.) layers did not appear to more beneficially affect the

moment capacity, possibly due to increasing internal stresses (shr'nkage

and temperature) in the thicker layers.

95. More important than its influence on strength is the ability of

a properly designed resin concrete layer to provide a noncracking moisture

barrier and corrosion protection for the embedded reinforcement. While

the conventional concrete above the 1-1/2- and 3-in.-thick epoxy concrete

layers developed the usual hairline cracks under relatively moderate loads,

the epoxy concrete Lyer on the bottom of the beam remained uncracked up to

or very nearly up to failure, thus providing a reliable built-in vapor

barrier and corrosion protection.

96. Although the polyester resin concretes chosen for this program

were capable of developing a higher compressive strength than the epoxy

resin concretes used, their modulus of rupture, direc+ tensile strength,

and tensile strain capacity fell consistently below those of the epoxy con-

crete. This together with the higher exotherm and the excessive autoge-

nous shrinkage, made the investigated polyester concrete unsuitable for

the intended purpose. Tes•ts on a few composite beams with polyester resin

concrete layers yi-lded disappointing results.

9'(. In developing high-strength resin concrete mixtures, it should

be kept in mind that routine laboratory strength specimens will not prop-

erly reflect the potential deficiencies of the mixture with respect to

shrinkage, exotheim, th-nnal expansion, creep, sensitivity to envirormiental

factors, etc. Separate tests should, therefore, be conducted to evaluate

these properties before any mixture can be considered suitable for a yr-

ticular practical application, regardless of how good its strength

39

properties inay have been in standard laboratory tests.

98. The effect of aggregate mineralogy on the polymerization of

resins should also be the subject of further investigation.

LITERATURE CITED

1. Jejcic, D., "General Report," RILEM Bulletin No. 28 (Reunion Interna-tionale des Laboratoires d'essais et de Recherches sur les Mat~riauxet les Constructions), Sept 1965, pp 7-1-3.

2. Andries, S., "Les betons de resine," RILEM Bulletin No. 28 (Reunion In-ternationale des Laboratoires d'essais et de Recherches sur lesMateriaux et les Constructions), Sept 1965, pp 15-17.

3. Bares, R., "Les betons a base de resines de furane: le 'berol',RILEM Bulletin No. 28 (Reunion Internationale des Laboratoires d'essaiset de Recherches sur les Materiaux et les Constructions), Sept 1965,pp 19-24.

4. Palencar, Z., "Apercu de la technique et de l'application ces mortierset betons a base de dechets de l'industrie chimique," RILEM BulletinNo. 28 (Reunion Internationale des Laboratoires d'essais et deRecherches sur les Materiaux et les Constructions), Sept 1965,pp 81-86.

5. Sneck, T., Marttinen, P., and Eneback, C., "A Preliminary Investigationon the Properties of Some Furfuralacetone Resin Mortars," RILEM BulletinNo. 28 (Reunion Internationale des Laboratoires d'essais et deRecherches sur les Materiaux et les Constructions), Sept 1965,pp 95-101.

6. Bares, R., "Uber die Technologie der Kunststoff-Betone," Zement-Kalk-Gips, Vol 20, No. 2, Feb 1967, pp 47-60.

7. Delmonte, J., "Current Usages of Epoxies as Grouting and Cement forConcrete," Furane Plastics Bulletin No. EP 59-6, presented to LosAngeles Chapter of ACI, 20 Jan 1959.

8. Welch, G. B. et al., "Epoxy Resin Concrete," Civil Engineering andPublic Works Review, Vol 57, 1962, p 759.

9. Flake, L. S. and Pullar-Strecker, P., "Epoxy Resins for Rerpir ofConcrete," (with comments by M. W. Franke), RILEM Bulletin No. 28(Reunion Internationale des Laboratoires d'essais et de Recherchessur les Materiaux et les Constructions), Sept 1965, pp 25-29.

10. Wakeman, C. M., Stover, H. E., and Blye, E. N., "Glue! For ConcreteRepair," Materials Research and Standards, Vol 2, No. 2, Feb 1962,Pp 93-97.

11. Pepper, L., "Epoxy Resins for Use in Civil Works Projects, Summary ofData Available as of 1 March 1959," Technical Report No. 6-521, Re-port 1, Aug 1959, U. S. Army Engineer Waterways Experiment Station,CE, Vicksburg, Miss.

12. Tremper, B., "Repair of Damaged Concrete with Epoxy Resins," Journal,American Concrete Institute, Vol 82, No. 2, Aug 1960, pp 173-182

41

. 13. '"Epxy Systems for Concrete Repairs," Engineering Research ServiceTM 62-1, Oct 1962, State of New York, Department of Public Works.

14. Okada, K., "Resins for Concrete in Japan," RILEM Bulletin No. 28(Reunion Internationale des Laboratoires d'essais et de Recherchessur les Materiaux et les Constructions), Sept 1965, pp 73-79.

15. Cirodde, R., "Contribution au choix d'un liant pour la confectiondes betons de resine," RILEM Bulletin No. 28 (Reunion Internationaledes Laboratoires d'essais et de Recherches sur les Materiaux et lesConstructions), Sept 1965, pp 31-35.

16. Postl, H., "Klebung einer Verbundbricke," Der Bauingenieur, Vol 37,No. 10, Oct 1962, pp 390-391.

17. Kriegh, J. D. and Richard, R. M., "Epoxy Bonded Composite T-Beams forHighway Bridges," Arizona Transportation and Traffic Institute Report

No. 12, Oct 1966, and supplement, June 1967.

18. Pollet, H., "Le beton a liant plastique, polyester ou araldite, etsa rentabilite," RILEM Bulletin No. 28 (Reunion Internationale desLaboratoires d'essais et de Recherches sur les Materiaux et lesConstructions), Sept 1965, pp 91-94.

19. Stamenov, St., Goudev, N., and Malcev, R., "Betons leger et ordinairede resine de polyester," RIIEM Bulletin No. 28 (Reunion Internationaledes Laboratoires d'essais et de Recherches sur les Materiaux et lesConstructions), Sept 1965, pp 103-107.

20. "Report of the F. P. I. Commission on Prefabrication," Journal, Pre--tressed Concrete Institute, Vol 12, No. 5, Oct 1967, pp 41-53.

21. "Engineering Developments in the U. S. S. R.," Civil Engineer andPublic Works Review, Aug 1965, P 1151.

22. Rees, M., "Some Recent Developments in Structural Precast Concrete,"Civil Engineering and Public Works Review, July 1962.

23. Johnson, R. P., "Glued Joints for Structural Concrete," The StructuralEngineer, Vol 41, No. 10, 1963.

24. Hummel, A., Das Beton-ABC, 12th ed., WilheLm Ernst & Sohn, Berlin,1959.

25. Warren, C. K., "Resin Bonded Concrete," (with comments by G. Periale,

M. W. Franke, and R. Bares), RILEM Bulletin No. 28 (Reunion Inter-nationale des Laboratoires d'essais et de Recherches sur les Materiauxet les Constructions), Sept 1965, pp 109-119.

26. U. S. Army Engineer Waterways Experiment Station, CE, "Handbook forConcrete and Cement," Aug 1949 (with quarterly supplements), Vicksburg,Miss.

27. Collet, Y., "Les matieres plastiques et le Ciment," RILEM BulletinNo. 28 (Reunion Internationale des Laboratoires d'essais et deRecherches sur les Materiaux et les Constructions), Sept 1965,pp 37-43.

42

28. L'Hermite, R., "L'appi4cation des colles et resines dans laconstruction-Le beton a coffrage portant," Annales de l'InstitutTechnique du Batiment et des Travaux Publics, Vol 20, No. 239,Nov 1967, pp 1481-149 8.

* 29. Flandro, A. W., "The Use of Polyester Resin in lieu of Cement inConcrete," Report of 6150th USAR R&D Unit, USAR Center, Fort Douglas,Utah.

30. Modern Plastics Encyclopedia, Issue for 1966, Vol 43, No. 1A, McGraw-Hill, New York, Sept 1965.

31. Reinecke, A., "Kunststoffbelag auf beschaedigten Betonfahrbahnen,"Der Bauingenieur, Vol 39, No. ll, Nov 1964, pp 458-460.

32. Kreijger, P. C., "Concrete Connected to Concrete by Means of Resins,"RILEM Bulletin No. 28 (Reunion Internationale des Laboratoiresd'essais et de Recherches sur les Materiaux et les Constructions),Sept 1965, pp 45-58.

33. Franz, G. and Bossler, R., "Prifung del Wichtigsten Stoffeigenschaftenvon Giessharzbeton," Betonsteinzeitung, Vol 28, No. 2, Feb 1962,pp 49-59.

34. Liesegang, H., "Kunststoffe and Beton," Betonsteinzeitung, Vol 28,No. 9, Sept 1962, pp 439-444.

35. , "Polyesterbeton in Strassenbau," Kunststoffe, Vol 55,No. 5, 1965, PP 372-374.

36. Kuenning, W. H., "Nature of Organic Adhesives and Their Use in aConcrete Laboratory," Journal, Portland Cement Association Researchand Development Laboratories, Vol 3, No. 1, Jan 1961, pp 57-69.

37. Lee, H. and Neville, K., Epoxy Resins, McGraw-Hill, New York, 1957.

38. ACI Committee 318, "ACI Standard Building Code for Reinforced Con-crete," ACI 318-63, 1963, American Concrete Institute, Detroit, Mich.

39. Sinha, N. C. and Ferguson, P. M., "Ultimate Strength with HighStrength Reinforcing Steel with an Indefinite Yield Point," JournalAmerican Concrete Institute, Vol 61, No. 4, Apr 1964, pp 3997-18.

43

Table .

Epoxy and Polyester Beam Program

Beam Dimensions

Average Average Thickness Area ofBeam Depth Width d* of Resin Type of Reinforcement

No. in. in. in. in. Reinforcement sq in.

Epoxy Resin Series

IA 9.00 3.97 8.00 -- High-strength steel 2(0.2)

2A 9.00 3.63 8.00 1.35 High-strength steel 2(0.2)

3A 8.91 3.92 8.00 1.42 High-strength steel 0.2

4A 9.12 3.93 -- 1.56 None --

5A 9.06 3.87 8.00 1.69 Fiberglass 2(o.18)

6A 8.99 3.81 8.00 2.56 High-strength steel 2(j.2)

7A 9.05 3.98 8.00 2.79 High-strength steel 0.2

8A 8.99 3.72 -- 3.Oh None

9A 9.02 3.98 -- 1.44 None.

11A 8.99 4.O0 -- 3.09 None

Polyester Resin Series

1B 9.1 3.90 8.0 -- High-strength steel 2(0.2)

6B 9.0 3.90 7.60 2.88 High-strength steel 2(0.2)

7B 8.9 3.95 7.90 3.00 High-strength steel 0.2

8B 9.0 3.95 -- .00 None

X Distance from centroid of reinforcement to the top of the beam.

9 1 0H H H-%. H

*- LP. a -aaUH 'oO H 0 - ýI19iI

0 ) 0 0

0o C 0 N\ CQsH rH H H 4

X:~ oo 000ogo 0000088§ 00t- - T 000

0 $4 H H- -4HH HH

w t- C~0 r('P-1 _:I, Ljco DH - N r4-tU\C E H H - I H H~ot

~I4C3.I -4 0 -

C9 04)1

*5 H4 Ur\ L \ 0 C 0 Lr\ C\

4-))

V. .4.-)r-, 0000000 000 0000 008 000 0oý M 0o R

~4-) W\ _:

0

41 Co4)OaN H H Hy HN ') H C'. _:

o) C* C C C C CiC

0~ Ho

E--44 H . (V

0. '---~----.- -

'0,

0 0( (D H rr4

0- *r-a)I Q). r- - ()r4 -

*d a Y) 0 z - r r

4).

tic d~)E4Oa) 01..m

C)4.

a) k r\ Lr m -:. a)

*v-4 0 Y-co -4 (Y) LIr -

o aý )HHH

(4P- \0\ ON(Y 0 CY) C\ co 0 0oY '. \0 \.0 0 H '.0 Cr 0Y t- \0 U'\ U' UN _:r G\ 00 Lý

CY4)~~ a)

d ) rA -3

4Z)(Ytr )L' H1 tr- LC% Nr~U r 0 0 m 0 0\

a) 0H H H HH .

4-))

4-) )

0 W 43 0 CM Cuj _:t .-t- '.0 \. Cuj _zt \,0

z.4 P- a )I H - H - H H - H - H - H - H - H - H - 0H

43 )H o> s- is s- is '- i- t-- L-- t- t.- ts-

4-d CWid Suflj2sonuUcuoo 2u-jpl3.z dv!j to

00

H)I H H H H- Cuj cy

:3 0 Cu H

H H H

H ~O H- H H) Hl 11 H Hv u H H

P4- (4 CY~ N

03 ~ ~ ~ ~ H HN-r _r UN Ul : Y Y

rlu 0 0u

t4 . H H H ;id 4

0) HD \Hmj N C) -r

Eta ~~C ., M - urq N cu 0 l C

H a~ \ CY O'~ cP-1

00

4-0)W 4.4

$:4

0 A .

4.) '

0. 04 080) ~ ~ ~ ~ - -+' i H H H H H

4-)

0,@5

0)0

4) 4 I

cd 0

44.)

4. W. 0 't, ý DI r- I- rI rI rI I- I I I I~ S~to0) :33: H H ~ > ~ -

4J H ~ .

C~2.dH H0

Table 5Portland Cement Concrete Mixture Data and Results of Tests

Mixture Data (1-BaBatchVolume, cu ft Weight

Material (Solid) lb

Epoxy Resin Beam Series

Type I cement 0.479 94Crushed fine limestone aggregate 2.].95 365.1Crushed coarse limestone aggregate 2.109 354.8Water 1.217 75.8Admixture None

W/C ratio: 0.806 Slump: 1-1/2 in.Cement factor: 4.5 bags/cu yd Unit weight freshly mixed: 145 pcf

Compressive strength of 6- by 12-in. cylinders, psi

7 days 3040 3000 2820 295328 days h090 4060 4300 4150

Polyester Resin Beam Series

Type II cement 0.479 94Crushed fine limestone aggregate 2.154 358.33Crushed coarse limestone aggregate 2.070 349.44Water 1.297 80.8Admixtu re None

W/C ratio: 0.86 Slump: 2 + 1/2 in.Cement factor: 4.5 bags/cu yd Unit weight freshly mixed: 144 pcf

Compressive strength of 3- by 6 -in. cylinders, psi

Avg

7 days 1768 1768 1888 19081895 2093 2037

28 days 2709 2560 2723 28303041 3062 2885

wl 0

I d

4I

goH§ctC~

go

9 t- t .0

"0 0

14-

0 M

g R

P, v o

L) 01 0 ýc) 0

w 0 0

1.5

q7j

ill.;!''

4, x,

0 ° 0'2

1 4) o 54

.4 4, '.3 C1

+4

,-I P 5' S•

Ut). " N-'-,

Ft i '0'-

,,, ,,J,,,,

24 4)

Iý a

020

u 4, 4

xx

02) 44)ND4I.

4''

4) N-

$4

Co 022

4' 4)00IV P.4 4,1(N

-k ~ I Hd

0H H

0) co x

-I (NN

> H H

4,4)

4' 4' 1-4

U d 0.S

-)02

.4 0) xfl, R p'A (N0 000 00

I> r4,

N,44,N,4

A,~ ~ ~ ~ ~ ~ ~~~i ro P, too~o~h idawea oprei~

Iti 1'yr't .Oflu o. am \.nri s t ad

I0.0rsi5101 0 1(1 27.V 9: 05 9,6170 1,30,000

I ~ ~ ~ p 0.$(26 9901,,0 1 0,5 1,40000 9092390000b

0 r . OH& 5280 9,770 10,140. 1,33,00O.it OL O.oOkw 61 2f6io ' 8.50 8,20 81:41:000 919,50

V(69, 70 9,720 8,36 990 998,50000

crete~~~2M 9eniuul 9.12) ,5 ,020 1039nd671

'I ON~ I5,826701,00 133,0

crt.(eninosl 3ei) 0. 10.3, 5,251 5,070clidr

o1 .08-0O~P: E4,91 180.05120 5.8 ,1250610

moi7tsso.1 6,096(2o

08 .011131 5,799

o.fl$16 (1.09,36 6,182 6,2020.0817 6,6214

rolyester resin con- 1 0.08614 10,758 3- byt 6-hi.crete (continuously 0.O.01, 0.0551 9,1465 9,667 cylinders,'ra'ied liirastoneo. r,.0814)4 8,779opgregate, 1 2ý O

ester rosin)

1 .08381 11,8570.08A6! o. (,42,8J9 12,14280.081,7 12,577

7 0.08W, 13,33?0.0557 0.O09A 13,558 13,1476om86i 13,5142

~11 0.0880 23,0950.085 0.0878 12,5714 12,9250,0871 13,106

L1 0.08142 12,475O.53(.VA511 12,7C: 12,786

0 .085e1 13,120

2t .0661 13,11420.0852 (1.0a53 12,286 12,7060.05142 12,691

10 '00 0. C; C;~ ~

um %0 n m% o G % U r M r 1c i .

44 4 6o

0 CID

o 3 IS' I *1..4 ".

0 'U .0 %0.4 0%(S-.4 .4to

C~j =] CJ Cj =1C31C31 D O

01"< 1,900 0.017 0.,03 0.'13 0.(L' 0.01Ž

'V~~O .0. ~ ~ ..~.13 0 0.03 0.001 0.003 0. 0.009h 104 P.11.• . I,0 . 0.01 1 11 0.I1T 0.03 0.00S 0 0.00"

0,.'0. -.L,11 t.L't14 I Xt'. '.0. t vol, I. j 1 o, 00 0.019 0. 0 .0 31 o ,i.M T

.'P4 0;.'.' 0, 0, .11. 0.02 .r f 1 .3 , 0.0 9 0.0 31 0.030 0. 99 0.018

0., V.0.-'oV7k "IWI.0 S O1o0t 0.01'i 0.0m . 0.0)5 0.039 0.0114"pY, 0..'v L 04f r.1 •, .+0 o.o17 I, FOO 0 O.0.0 0.031

17 p• .*,.r .. ' oo .oo7 o.o0+ O..O•. 13 • 00.030 0.010* 0.01A 0.030 0,0114

s 0..o1•1, oo0 0.0 0.0o 6 o.00ý. 10 03 0.099 0,061 0.0W 0.033WA I.,. 0.01?1 0.0A 0.A41f 0.oWi 0.-1? 1,6 ,.500 0.01. 0.060 0.067 0.05 7 0.

1 (1 3,M00 o.06 0. o.o o5 00 0. 7 6000 r.. 0.066 0.07, 0.063 0.""4 0.02. o.0o6 0.o19 '1.04:4 O.Ok5 0.0:.6 0.0: 0 0.078 0.06 0.01.

V' .',-0oo U,.,1 7 0 =.2, 0.0 0. 0.0o 1 0. 06 46 5,0.50 0.0 h1. 0.0 09 0 0 " 0:.0,6 0.: 941, a•,0o o.0• O ,Orn o0'jj '.0• 0.01 3' 0 C.008 0.410 0.010 0.011 0000o-4 o0.01 0.06 t.019 4.0 00 ,7 OVA iiA2 rs 52 0 .046 0.006 0.03x 0.00 n.9hlI"1 1.,L00 0.0• 0.0 4, 0.0 C0 4 0.0 41..,. &.OO in t ob4erve 3 f,000 0.038 0.0410 0.01 0.037 0.00941. 4,0M6k o.047 0.0'4 11.O'3) 0.0O9 0.01 53. 6I000 0.065 0.067 0.01 0.001 0.0O071). 0.0-.1 0.071 0.06t G.F071n 0.045 & sriw ~ 525 0.01.T 0.070 0.012 0.066 0l.01.344 0 .0 0.04- .o714 0.071 0.(74' 0.0409 I6 (.n300 0.0r1 07 0 0.o o0o01 oloko o.o4o u.037 o.oek

1.( 5,000o 0.•43 0.o87 o.) o.0oj 0.0") 0.04o 54 6 ,o00 0 .0o 0.067 0.0-6 0.083 0.0k1 i por.1an )emst0.!, o.01 0.o0 0.10o 0.071 0.0•5 63 0.097 0.Ob'070 0. 0.080 0.093 0 omorsto"Le. 4,000 0.0oj1. 0.065 0.073 o.078. 0.0o91 5 ( 5,00 0.051 0.09 0.105C 0.071 0.0 FC

67 C . 0.0w 0.60618 0.0 0.00 17 0.01. 6 8:.,00 O.006 0.09" 0.110 0.09e 0.060(. 00.0"2 0.017 0.019 0.017 0.01) 63 0.066 0.0& 0.116 0.018 0.062oi 1.000 0.O98 0.0:6 0.063 0.0FJ5 0.037 750 S,50 0.00T 0.092 0.10. 0.109 0.05661 6,000 0.0+Jt 0.061 0.0010 0.091 0.019 71 9,000 0.076 0.119 0.112 0.11•98 0.0f, 11.012 0.0o1 0,19 0.017 0.019 68 0.077 0.119 0.137 0.119 0.0

18 6 .500 0.061 0.0-8 0.101 0.058 0.063 76 6.000 0.0? 0.09 0.11. 0.099 0.06064 7,00o ,.,o.0 o.:o1 o0.10o o.lo0 o0o.9 71 , o.01.07 0.1.5 0.o81 0.119 0.07071 0 0.069 0.0 0,1 0. 0.09 0 0.096 0.7 0 0.030 0.019 0.016(A 7,5w 0,075 0.018 0.131 0.106 0.076 80 005 0016 O.0 211 0.0 0.0110` 9 ,07 0 0,0010 o.34'!, 0.11.0 0.1l0.0 066 81 3,00 0.06 0.051 0.081 0.090 0.040

74 8,,00 0.00 0.107 0.11 0.107 0.09 9 6,000 0.091 0.081 0.090 0.029 0.01674 7,000 0.075 0.11.5 0.1,31 0.141 0.o076 8D 9.000 0.07. 0.x11. 0.162 0.113 0.0681 0,01 O.OU 0.'11 0.166 0.1126 0.09M5 81, 0.07? P..059 0.13 0.118 0.0701. 6,000 0.070 0.151 0.1146 0.110 0.089 85 9,00 0.001 0841 0.0.92 0.121079 0.07b1 1 ,ooo 0.o05 o.o1. o.r'o 0.144 o.o08 84 10,00o 0.0• 0.12 0.152 0.112 0.o7'

61 0 0^018 0.127 0.019 0.027 o.o21 90 0.088 0.138 0.:19 0.11 0.080417 0.017 0.026 0.018 0.0 60.082 91 10 ."00 0.098 0.113 0.169 0.124 0.08n0 *,N n. 3

1r C ..'O@ .47.'; 0.u..7 0.090 ~ 1IA .04.0 0.1.4 0130 0.1.4 0,09089 •.000 0.0752 0.110 0.12.) 0.109 0.076 6 10,000 0.099 0.1.6 0.1w0 0.1941 O.090

0 90 0.096 0.130 0.109 0.018 0.001 97 11,50 0.103 0.161 0.186 0.160 0.0o

95 0.017 0.02 0o.171 0.1-0 0.102 98 10,000 0.1070Oe 0. .19 0.1697 0.0o967 9, 00 0.10W 0.110 0.12o 0.109 0.107 010 0.9 0e.r17 0o.20 0.173 0.099

98 10,0o0 0.10' 0 o.14 0.189 0.m66 0.11w 102 9,000 0099 0.199 0.180 0.1!1. 008101 0.109 0.171 0.193 0.170 0.116 103 6,000 0.08 0.122 0.113 0.I22 0.071102 10.500 0.113 0.178 0.0o0 o.117 0.120 104 3,000 0.0-59 Oo08 0.101 0.086 0.0W

10! 11,000 0.120 O.18: 0.211 0.188 0.127 105 0 0.021, 0.032 0.042 0.035 0.033106 11,94) 0.121 0.19z) o.'21 0.199 0.135 107 0.020 0.026 0.036 0.029 0.0331 10 12,000 0.112 0.208 0.233 0 oý 0.1.108 3,000 0.01 0.063 0.073 0.061 0.0301.4 0.136 0.211 e.239 0.21[4 0.141. 109 6000 0.615 0.097 0.310 0.096 0.058

116 9,000 0.122 0.111 0.212 0.190 0.130 110 9,000 0.081 0.131 0.1:1 0.130 0.078

118 6.000 0.300 0.192 0.168 '1. 11 0.107 111 12,000 0.106 0.167 0.193 0.166 0.096119 1,00 o.o68 0.102 0.111 4 .102 0.077 112 13 D.2041 0.176 0.301120 0 0.027 0.043 n.044 0.013 0.038 111. 12,"00 0:117 0.19 0.22 0.113 0.10h121 o.026 0.012 0.013 0.02 0.037 11 13,000 0.121 0.191 0.220 0.190 0.307125 3,000 0.057 0.085 0.093 0.085 0.060 1l8 0.125 0.198 0.1127 0.197 0.111

126 6,000 0.085 0.132 0.1.6 0.131 0.091 119 13,900 o.12 0.301 o 0.20k 0.115127 9,000 0.111 0.177 0.19 0.176 0.321320 I .1,000 0.131. 0.211 02. 0.211 0.119128 12,000 0.139 0.219 0.2V, 0.219 0.1O.7 123 0.139 0.221 0.256 0.220 0.123132 a. 1460(Q 0.2• 8 0.230 0.155 121 12,500 0.1166 0.232 0.270 O.31 0.129133 32,500 0.18 0.260 0.288 0.255 0.169 126 19,000 0.1"9 0.57 0.300 0.25 0.1 0.4

136 13,000 0.210 0..30 0.500 0.420 0.280 130 15,000 0.2•0 0.1.3 0.590 0.330 0.230140 0.28 0.507 0.596 0.1.M0 0.317 133 15,000 0.610 Crack In PM b hzt.155 11•000 0.720 CourossloMW emaerete 135 13,300 0.800 Compres.ions• eamol t

failure 137 0 0.990 til.utr

0 00 0 0 0 0 0 0 0 0 0 019 500 0.002 O.O8 0.00 0.007 0.003 3 500 0.008 0.010 0.010 0.006 0.00620 1.000 0,010 0.012 0.013 0.01' O.007 1. 1,000 O.O01 0.09 0.019 0.01. 0.01022 0.011 0.013 0.011 0.012 0.007 5 0.011 0.019 0.019 0.01. 0,01021 1,500 0.011. 0.n17 0.018 0.0(15 0.009 6 1.500 0.019 0.021 0.0 0.020 0.011m

25 2,000 0.017 0.021 0.02• 0.019 0.011 2 t,000 0.019 0.0o6 0.007 0.025 0.018,6 0.017 0.021 0.022 0.019 0.012 0.020 0.027 0.029 0.0R26 0.01827 2.500 0.02T 0.025 0.027 0.021 0.014 9 2,900 0.023 0.033 0.039 0.032 0.02229 3,000 0.0W2 0.0=9 0.031 0.02: 0.017 10 3,000 0.031 0.0147 0.052 0.01.9 0.03029 0.021 0.03D 0.032 0.029 0.017 14 0.03h 0.053 0.09 0.090 0.032

(Oo.inutd) (I Or 3 011000)

- -.-----.---.-. - .--- --- -~ - ---- --. --- ii

1, 1,,00 0.W Q '1 o.01, O.01.1 0.0j 0.t(J% "a 0.03? 0.03O OCA.I) 0.03 0.0311 3 ,.00 .oo+ 0 .010 0.09 ('.o004. 29. ,50 o.039 0 .01.o,,0 o.o,.

0.0" 0.0'M .3.009 O.01% 10 5,000 0.0,2 0. 1 0.0) O..041 o.O911 163o0 0.091 0.011 0.0 0. 000 41 0,1o3 . 0. 0.01 0.029

, .= .013 0.o0 0. 0 .. 0'1 ,1 ,.O 0.o, 0. 3 .o,, 0.01A. 0.031

01 .3,% 0.063 O.o0 0 %o.o,, to ,,rtl.md -own& 0. 1 65 0..10 0.090 First , rr,.. to port.% .coo 0104 O 051 0, 04) C.060 0,0el vow-rPto 0.101 0.39 Pojit o.0 0.10. ). nt"a

33 0 k" a 0.66 0. 0.07 1.6 m" O.0 C.6 0.5'4 0.132 0.^5 epox vncrto

N how 02 00.0"6 0o.:0 02 o.03 0 0.033 0 o. 0. ,.oa. o.oes

5.0000 0.o0 0.099 0.311 O.0o9 0.040 -1 0.023 0.041 0.01. 0.015 0.0250.006 0.10o 0.11 t),.103 0o• 5, 3,%, 0.o 0.1 O: o 0.8 0o.0

.31 5,530 0.076 0.1.6 0.11 0.113 0 60000 • 0 .16 0.o101)m0 6,5O 0.070 0.125 0.19 0.121 0.017 0 11. 0.2Ik 0.278 0.242 0.,31'31. 0.04 1 C.'11A 0.11 0,131 0.079 65 0.141 0.2•. 0.293 0.255 0.141

37 .,000 0.040 0.099 0.107 0.091 0.037 67 6,500 0.151 0.281 0.315 0.275 0.15,

36 0 0.015 0.0L 0.027 0.02) 0.01. TO 7,000 0.166 0.332 0.330 0.]02 0.1w 9

313 0 0.011 0.000 0.021 0.015 0.031 0.1JT 0.323 0.366 0.320 0.10039 L OW 0.016 0.0o CO W o.04 o. 0" 0 o 0. o.10 61000 O.O" 0.130 0.145 0.137 0.077 I: 000 0.19 0 0.37 0 0.372 0.210

.08 0.1 0.1% 0.113 0.081 5 04.17 0.1O 0.1.71 0.420 0.231"" 500 . 0.148 0.1 0o.,A5 0.06 C6 .•500 0_30 0.1.29 0. "0 0.441.8 0.26.67,000 0.0" o.:10 o. .1" 0:09• 89 9,ooo Oa.• o. o.:37 o.mi 0.283

.6 0.00 .16 02. 0.12 0096 9 9,0o00 0.•29 0. 0 0.525 0.289k7 7.300 O.ICT 0.140 0.103 0.17-1 0.0 6 6.00 0 22 0. 9 0.45 0.5

1.8 ,1000 0.116 0.216 0.3149 0.215 0.1"1 97 3,000 0.160 0.299 0.337 o.98 0.1660o 8,000 0.21 0.m 0.270 101 0 o.o0 0.111 0.129 0.311 0.060

%2 8,000 0.•0 101 0.061 0.109 0.126 0.112 0.066

33 8.100 1.000 Creek in Opo01 105 3,000 0.114 0.211 0.218 0.2.'" 0.122

5, 0 0.6,0 conervt. 106 6,00I1 0.198 0.S.,66 0.1.5 0.368 0.2105

3..._.A 108 9,000 0.280 0.919 0.595 0.530 0.292ill 9,500o ox"' 0.555 0.638 o.,T 0o.1

0 0 0 0 0 0 0 123 9I,00 . 1 1 0 0.: O 0 0:.3(.%? 23.13

j 0.009 0.0 0.006 0119 .. 0.63 0.711 0.640 0. -141h

1 3'. o.o11 0.013 0.013 o.01 0 121 WOO o.10 o.712 o.67" o..636 0.012 0.013 (0.013 0.011 0.007 a.50 o 0.015 o 0.1, 0.017 0.015 0. 010 12 11,50 0 0.% 012 11,500 O.T33 O .Tm 007 .O

9 3,000 0.018 0.029 0.022 o.019 0.013 12 12,0o 0.767 0.867 0.79 0.29

10 0.01 0.022 0.031 0.019 0.013 129 0.703 0.681 0,0 0.1.23

11 1,0 WO C0.21 0.036 0.026 0.022 0.015 131 9,000 0.715 o.8o1 0:.3 0. 39

13 1,000 0.034 0.010 0.031 0.027 0.01815 0.024 0.00 0.031 0,037 0.018 132 6,000 o.,t, 0.6!5 0.5ft 0.320

133 3,000 0.369 0.:19 0.382 0.20116 1,300 0.018 0.022 0.021 0.019 o,0,, , 0.150 0.171 0o. 00

17 o o.oo3 0.002 0.002 0.003 0.003 o.Ib :.53 0.1147 0.

19 O.003 0.002 0.001 0.002 0.003 188 3,000 0.255 0.290 0.265 0.1.3

20 1,•500 0.017 0.020 0.090 0.017 0.01221 3,000 0.024 0 0.031 0.03 .027 0.018 169 6,000 0.431. C.497 0.o.1.0 0.P46

190o 9000 0.612 0.691 0.63 0.31•0

33 o.O31 0.031 0.032 0.026 0.018 1 3,000 0.820 0.924 0.848 0.460

1. 3,500 0.037 0.035 0.0 0.032 0.021 19. 0.92936 4,o0o 0..:9 0.039 0. 0.035 0.023 198.9028 0:O 0.049o0o 0.035 0.02330 4150 0o:032 0.0041 0.045 0.040 0.025 200 0.98o

201 1.010

32 5,500 0.032 O).Oh8 0.091 O.O0A O.O: 203 13,000 1.050

31 0.0 0.01. 0.0351 0.037 0.02e 21 13#500 1.080

0 .0 0.006 0.00 0.006 0.00Ou1ooS 0.01 0.0003 0.005 0.00 14,0oo 0..2

16,000 0.2 0.0 63 0.01 0.5 0.0301. 0.0 16. 0.0 6 5 o.o 0.057 0.036 208 14,5W 1.180

209 15,00o 1.260o

Ail 3,O D 0.031 0.06, 0.01A 0.06 0 .024 210 10,037 1.310

45 ~ ~~ ~ ~ 00 0 0 01 66015 opesoa ocee. _____ 0.0 0.00 00050.060.0

S 3,000 0:29 0.030 0.000 0.005 0.002 10 0 0.401 03 0lureSk7 ,0SooD 0 03 0.o61 o.066 0.057 0.036

s5o 0.05 0.064 0.069 0.060 0.o037 9*= (A-

0 00 0 0 0 0 0

500 00 00 0o0 0.003 .006 o.o0T 10.06 0.01o1,0 00.0100 O. 0.010 0.013 0.011 0.0

0 0 0 0 0 0 0 1 0.010 o.013 0.01. 0.012 0.009

W 0.0011 0.007 0.007 0.001 0.009 9 1,500 0.013 0.O33 0.017 0.00 1 0.015

1,m 0.00 •011 0.013 00 .2 0.070.066 0.01 0.011 0.060 0.03? 0.01. 10 0.036 0.033 0.027 0.032 0.03•

T 1,q0 0.013 0.019 0.016 0.013 0.001 11 0.017 0,1R3 *.Ogg 0.0n 0.01,120 01•,00 0.0200O 0.031 0.006 0..0TaI ,eow o.014) 0o019 ow• 0.01? o.ot 1 3,o low m 0 o1 .0o1 0.0" 0.03D 0.000

12 0.010 0.013 0.0%0 0.015 0.01l 11 0.016 0.033 0.037 0.031 0.021

15 ,• 0.031 0.037 0.031 0.03, 0.017 16 1,0 0.001. 0.001. 0.0• 0.oo1 0.00116 0.003 0.001 0. 0.00 0.001.

7 1,3000 0.01 0.016 0.03•1 0.010 0.012 19 1,6500 0.01 0,019 o0.01 0.011 0.012

1 0 0.000 0.003 0.003 0.003 0.003 20 3,000 0.033 0.033 0.036 0.031 0.05

1 0.3 0.0 0.003 0.003 0.00 . . ., 1, 0.013 0.016 oo 0.01 7 0.1, 0.010 33 0.: ` 0o. 033. , Is.0 0 0.01

33 3,00 0.030 0.066 0.0 03 0. 0 02 00 31. 00 0.023 U .327 o.0:03 6 0 .03o1.E5 0.002 0.009 0.00 00.003 0. 100

o IN* 0.31 0.016 0.017 0,3 0010 0.03 0.03 9 0.032 0.03100 ,, "' . O 0.030 o .o 0. 1 oO,,, 4• o.30 oo33 30 00035 o0o39 o:01 23 0.01, 0.0%7

S(Oontitnu4) (I r 3 Aset)

w .4 t.,% 0'6 h .*. OW 0 ..cO 0 'K.100 b00..)00"4 44:'% u "' 0 to U ol" p60

If 1.4,.o o4.o• 0.01% 0O~.1 q 0.01.1 (4"8.i %o0.1 0.1"/9 0.099 0,.0W1 0.0517tl

0 0 (1 0.00 0.01.A 0.049 0.O66 0.05 Q-. U191.1 ;:loot Uit1 4 (1. 0ý11 1,O 4wj . 1 , .01 A mI U.1 9. 0

1. ~ 0014 004A9.7' .t4 010 I ~ 4 I' 01 1 0,1

16 1.0w 0.0'.6 0.0ý -05O 0.0( 0.01 .. 1,0 0.1 0 0 C.OIs5 0 006 u0,1. 0.0 0. 0 000001f..6 w 0.0%r Z 00.006 001503 0.0 0.0 ".05117 9, 0 0.01.9 0.0o 0.ii O .0 0.01.7 I r, v0 0. 0100

Wo onn o (W 10 : M o &0 0;moo, , 0 000 0.1.6 0.100 0. z 3

o 0.05?m0.9^4 o,04 0.0ft 0.10 0.o,0 4 66 7100 0.1.00 0.

41 0 CAI 0.0& 9j.0"1 0.10 .04 0. 1. ~th~1:. ~s1g 0.%7o

61~i 0.1(6 0..00 0.6 1rac 0.e0 0.65 ,crt.t.

64 .. coo.o36 0.~ ,092 0.10 o o.o8 o 0.0o5o%• o•o~ oo61 ,o o.o0 0 0 . 0. 071 0 o.o o .o 0 0 o o 00 .0 0 o 05 V• t.0 0,0 0.0750000 0 0408 .9 0.100 •.

00.01' 0." 22 0.05 0.0,15 W 0,0100 1000 0.01001 0.016C 0.311 0.0"10.) 0!o00 o.05 9 0.0292 O.16 0.0109. 0.10 0. 0.o10 0.011 Fail

70 3100 0.076 0.054 0.099 0.0186 0.0540 0.01' 0.01 0.016 0.02 0.01

74 ..I 0.01.9 0.07 0.1 007 u00'.6 7, 10 0.400 00

57 9,000 0.06o 0.10 0.110 0.10 0.00.10 I.n 0 0.0• 0.0o1. 0.00161 0.06 0,109 0.129 0.106 0.06 5 .' 0. 0. r. 0A

S %.VA 0,012 0.01 3,0 0.01 0..5 0.0 0. 0.6o 1,000 0.00 o.,G- 0.14) 0.100 0019 . 1 5 00 3O6 0.058 0.068 0.0"3 O.O36

91613 0.076 0.129 0.1.6 0.127 0.078•0 3000.3 .•00h .•f00'9U 0,00 0.01 0.0221 O0.02 0.01 2 1 1.500 0.010 0.099 0.05& 0.017 0.026

(A 0 0.011 0.0A 0.0vI 0.019 0.0119 19.000 0.031 l 0.180 0.1. 0.008 lb 6 10. 0.0 0.01- 0.016 0.08)5 0.01r

7o1 6,000 o0o09 0.o79 Oz• o.o?6 v.046 • ,,ooo60.ho.oo,1,: o.:rt

* 90~~~~1• o.e89, 0.11.6 o.1(.)t0o.11.9 0.0911 ,0 .2 .3 .1 .3 .2

99 3,00 0.0 0.05 0.06' 0.03! 0.031•Tj 6,00C0 A 0.1O.1060O.1l0 0.1030O.061 00 1nh . .279 91000 0.070.118 0.-13 0.116 0.09912101000 0.0 0000 0.05. 0.075 0.067 0.036

6 0.100 0.10. 0.10 70.10 23 30 0 0 0 0 099 9.000 0.00. 0.129 0.146 0.1 7 0.090 i. 1,000 0.00$ 0.046 0.01.3 0.0 05 01 * c100 16050 0.015 0.12) 0.150 0.151 0.47 is 01.0.9 0.099 0.110.010 0.011101 0.10c 0.189 0.o(9 0.106 0.01 j7 1,500 0.067 0.031 0.,10 0.013 0.021

1 9 0:,02 0.WO ,000 0.032 0.055 0.11 0.08?1 2.000 0.0197 00 IE0.037 0.033 0.009 31 0. 0O.1102 o.00 0.066 07• o.6 o.0 O98 0 0.06 1 0. 1 9 0.1. 0.100 0.0613 o.i cr

9 ,0000 0.080 n0.'110.15 0.1')30.4 24 4.000 0.000 0.02 09 0.00 0.053

110 12,000 0.107 0.165 0.169 0.129 0.1077 2rae In.O 0:99a

10C~a 0•• 0.-0°-" o°:•o.g 024 O.hoo. 0: 2.°'--1 oo•oosos9OlOoo129 5,poo 0 00 .1 0.155 0.0 O.O 0

112 0.•10 0.17 0.201 0.13 00026 00 3.O.M 0.007 01.19 0.017 0.10.113 2 0 o 03 O.6 07 0.1 0.02 v 1,00 0.01 0.0 0. 0.006AA)o 1 .110 0.0:6 01 0.0o910 . 6 6 1.5000 0.013 0.001 0.020 0.036 0.001. 2 rl 9A116 11,500 0.1 0.06 0.o36 0.211 06 0.016 0.006 0.026 I 0.01

019 1100 0.127 0.213 0.21.c 0.c0t 0.01000 0.01 0.007 0.01112 0.106 0.330 0,201 0.393 0.20, 9 9,500 0.007 0.036 O.i.3 0.0302Ill 0 0.018.4 O.O:5 0.31 10 13000 0.001 n.01 0 0.0 60 0.03712 13,00 0.21 0.01D 0.1670 0.16 1 0.016 0.01 O.O1 0120 0 0.0'2 0.020 0:0011 00.17('18118 13,5M 0.o12 0. 0.236 0o.211 0.0126 , 00 0.011 0.01•. 0.019 0.089 0.050

10 0.12 0.01 004! 0.000 0.o11 00 O.OM7 0'07 00.2 0.016

190.o .• ~• .o1o.85 o eoo ce. a.. 18 0.016 0o01.1 o~o o~ o.o1

12 3,o 0.186 0.o33 0.02 0.039 0.220 6 1,900.03 .O1 0.0% 0.07% 0.066 0.0O112 o.o 0.34) 00 A 0 3,000 0.01. 0.11 0.107 006oo02 45 002 0:7 .e 0.40 0. 01. 0.046 0.02350.109 0.150.08 0."7 ~ t6128 0.V96 0.0005 0.500 0.05 .17 16 3,T ude a3~

10 1,,00 0.001 0.01 0.01 0.000 0.000

12 1,500 0.015 0.017 0.700 0.010 0.011. 00 0' C0- 0.

16 2,00 0. o 0.405 c0.r 0.0f6 9 r00 0.11 0.05 O.O0I 0 0.01o 0.00918 00 0 .026 0.01 1 1,000 0.0o1 0.01o 0.1 0 .n7 0.01 0.019 0,500 0.0 0.026 0.03 0.015 0.020 1 0.019 0.090 0.101 0.100 0.061 *t. at

20 3 00 0.0 6 0 0 1 0. 3 0 .0 0 0. 2 64 1, 0 .0 0.03 0.0 0 0.1326 0.016 e rw o

23 0 .027 0.033 0.039 0.031 0.002116 ,000 0.10 0.01.1 0.o1.. 0.01.0 0.

13 0.010 0.0o 1 o .05 o.0 0.0120 0o5 04411 106 0 0.015 0.017 0.005 0.0•. 0.001 9 2,500 0.1.1 0,068 0.17 0.066 0.01 169 0.00o.o• 0.00 0.00 0.0 0 0 0 0 0

29 1,000 0.017 0.019 0.092 0.021 0.018 15 90 .09 0.015 0.0167 0.0116 0.0091 ~Cis 3,000 0.007 0.03 0.380. 0.02841000.1 .19001 .1 .1319 0 0.022 0.036 0.031 0.025 0.00'.7 150 0.001% 0.11.1.0.0 0.10.3 0.018

20 3.000 o-o2 o6o•? ~o• 0.09 0.o• 0 o.010 0.2 0 O.O Ox" O.0Ol32 3 .00 0.010 .0 0.03 N 0.037 0.024 7 2,1 0.03: 0.0,1 0.01A 0.,4 0.0

5 1,500 0.000 0. 022 o.o P6 0. 022 0. 010 5 ')0, 3 0. 0 . 0

6000060o0. 0.000 12k Orlw meme 0.|10 80000prua ao29 1 o0 .061 0 019 02o2 0001 o0o Olz t 15 0n089 0.158 0:9 0910.9 Cmo

30~~~~~~~~~~ ~ ~ ~ 3.0 0.02 0-03 0-3 :02OOe1 OW001 1 0 011 .0

IU N

Fi IQ040

.9 .94

v -1 W~l A.

::vj om u ~ k 0§%j 0888 a%0000t

01 9 1 N ell:8888

b:! C. *. 0*a CRa ýWo -MOtQ0%9: j*W -(MEm~U~O ~ -.94 1 9

I ;It 4 A 0u * % W 0O.Ife% U ~ %% 0 W I~

Pn*c .o&oc% *...............n.......

* CrW

'0

co V%%D.4 0Al'0i t % % H% -*' 0 0 H 0%tO V% # % W. U'%.* % r4 W% c4.-4

("W ~ ~ (V ýo" 4t o Aý

0 91

co OD*OCo 0-O~00 U00 4

G0§ 0OR§0r0 §

*~ ~ M U .4Nt'k MCt.'O I0-0 POUR' okkkk

'41 fi,0 r4aC CN r-- .4 (~101

'4414

oo

H NV CO H \C' G-

o C C eV ~

ý0 ~ O~tOV 0 -t N f :

Al ID 0 0

A~ ~~ C'0C I 0 t -

4r,

o' C 00000 0 810 8 0§Cj ,ýC3 r 1 1 88gr

00

~1 N HcC ~C COj

CC)'~0 O ~ 0\VC

do. . * .

00

4 HS

0 0 0 F4Ir LA\

u C3Cjcu 4A0N 0

000

4)

0u 00

0l Q? 00

to 4)

000 00!%

0 0 C,

$43 +

0 0o o0 0

0) u 0 01 d d 01 0\CLOOn

H t

03 P0 4- 0E-4 4.' 4-3L Lt\to toNl

0~ 4 )f

O4 0 r+'

to. 0 O0

I Q 0

toc' H toII

*d Id

0 V. cu>

eq 0 $4 4 Hs-

0 r

HO .0 H k4 $4 4 . ,R 4) Q4.-4 k a) 0 0 H 000Q)P44- ý4) .p03oa Ua 0 543U to V 4 -l

4-) V. +3 rj 4)

* 2 Q) 4U) 4-4 OU) 4) C- 0 q) CC)' *rj 4)

Qto ýE ) 7 ,44)0 4-) :-H H)+ 00 a,$4 Eam 0 a0c/0 k I w $4 (00 0oc D()4 Y )C

0 P4-

ýa,to

0-tIpe jpq ý4

I v \ ( "0' - CeN(0 0 0OON.-4 O\ 0 (" On ' (n c'J Mrl ,tt

1 C 0000 00000 0 00000 0r000 c 0o 0 0 0 0000ý

1,0000 0 0000 0 00000 0 0 0 0D000 0 0 c0 0

0 d

4) 4.' ('4C ' M '.Ud L) 0C\J0C4 \ I? H-- 0-'0 0' 0 C0 20O1); In 010 \ . t " l - 0 'ý4 .11) 1-)~ I" \0'.O -1' ro" In . "0 N

Ol 0 OOH-4-4 -4,u (V N ± 0 000 000 HH141 00 H 0000

H ) 00 0 0 HH0 C U 000 0 0 00000 0 00C 00 0 0 0 00

o~~ooo 000 00o ooo

T HI o § 8 o..2 I: C-a a) ON 00 H CU 0 CU H, C' CU ". 1 1 ..0 '0HM U' '

0 0

4-.

14 (7\ 0Cl\ .0 r HC')ul,ý -0\ 8CH ý r\G u*\00 O t -HC'D3'O a) N0 '.rN OO

H F4 O n O\O CL' 8 j (\ - U '.'.OO 0' C O\ 4'0 '.OO --- rO-4 n nI lO o~

00000000 00000 00-) HH H0 H ) 0000 00000

0r :t10 OD 0V H- \0 :rR0 CN ON c . ( )N C'O CC H -I) ,CcaC'" " Ir. 0. iP',-4N) 'N

00 000 HH () ~ 00 r H -HH H ! H ) C'OCJ, x T - ()IHr L- 0000 0 O \OH

0000 00000 00000 00000 00-40 00000 00000o

02

0) 1- n _ 0C l 3 L-ý l ,C OsC - l )ci(\ Y

Ho o 00000ON ocH C'. 04 0'WM C- 0 0 C 01' C 0-:C.0 0 H

0000000 00 0Q0(0 00000 00 0 .j00000 0oo(00

0§ OH § Cu1'r. C-., 0 H(1 CO ) - -

0 -. rL' 0 :Tp IN q 0 HA , H C'i0 H0 L 0 00r r~;0 00000' '000 0000t- tt\ ý.r0 0000 00000c C 00 0c 0000 I00 0

E48 000

Table 15

Moment-Curvature Relations, Series B

r "'c8,9,12,13 m Ec E101I,14,15 " Ecr c Ca Avirage

d a1 a2

Moment -ft-lb 10"6 Radians 10"6 Radians 10"6 Radians 10-6 Radians

Beam lB

(d = 8.00 in.; a1 = 8.10 in.; a 2 = 8.95 in.)

0 0 0 02,000 62.0 78.1 119.24,000 142.6 207.5 223.86,000 202.8 301.0 310.3

0 25.3 211.2 198.1 NA8,000 273.3 398.0 403.8

10,000 331.6 470.6 488.312,000 481.1 1419.0 1504.6

Beam 6B

(d = 8.00 in.; aI = 8.00 in.; a 2 = 8.85 in.)

0 0 0 0 02,000 42.0 41.1 45.5 42.94,o00 96.3 90.6 lOO.8 95.96,000 159.5 147.3 166.1 157.6

0 21.1 17.9 20.3 19.88,000 248.7 228.6 264.1 247.1

10,000 324.9 300.0 346.0 323.612,000 409.5 366.5 314.5 363.5

Beam TB

(d = 7.90 in.; aI = 7.90 in.; a2 = 8.75 in.)

0 0 0 0 02,000 42.4 37.6 41.7 40.64,000 91.8 86.3 92.9 90.36,000 329.7 600.3 724.8 551.6

0 84.3 320.1 277.5 027.3

Note: NA = not applicable.

Photograph 1. Typical epcxy resin con-crete specimens after testing

Photograph 2. Flexuwr3. test on 2- by 2- by U-l/4-in.prisms

PhotographI 3. Ty-pical. specimen u~sed for direct ten-sion and tensile splitting tests

Photograph 14. Setup for beam tests

Photograph 5. Shrinkage cracks in polyester resin-limestone aggregate concrete

Total load, ~4=0 lb

b. Total load, 6000 lb

c. Total load,, 9000 lb

d. Total load, 11,000) lb

e. Total loads, l3*C0. lb

failure

Pbotograph 6. Crack - .beam IA (refez .-- ce lvesa)

f 4PonnL$o CEM8WET___________CONCRETE ........ C#~mf

a. Total loa', 6500 lb

b. Total load, 80(-0 lb

c. Total load, 10,000 lb

d. Totel load, 14,000 lb

e. Tetal land, 15,00G lb

f Total load, 15,300 lb (udSýte)

Photcograph 7.. Crack pattern, beba, 2A

a. Total load, 3000 lb

b. Total load, o000 ib

c. Total load, 6000 lb

d. Tctal load, 7000 lb

e. Total load, 8100 lb ((".tiate)

Photograph 8. Crack pattern, beam 3A

ý P M P C OM MT DI X ne "

Photograph 9. Crack pattern, beam 4A, after failure at 614O lb

I

a. Total load, 6000 lb

b. Total load, 800 lb

c. Total load, 10,000 lb

e. Total load, 16,000 lb

f. After failure

Fbotograo. 10. Crack pattern, beam 5A

tORTLA,• Cg&MENT EPOXY P,•WN

a. Total load, 9000 lb

b. Total load, 12,000 lb

c. Total load, 13,000 lb

d. Total load, 14,500 lb (ultimate)

Photograph 11. Crack pattern, beam 6A

Pr A .O Total Xa, 50 l

a. Total load,, 55000 lb0

b. Total load, 5000 lb

c. Total load, 5500C lb (ultimate)

Photograph 13. C'rack pattern, beam 7A

a. Total load, 3000 lb

b. After failure

Photograph 14. Crack pattern, beam 9A

a. Total load, 3000 lb

b. Total. load, 3500 lb

7i~1 ý7

c. Total load, ~4000 lb

d. After failure

Photograph 15. Crack pattern, beam 11A

FIIa. Total load, 3000 lb

b. Total load, 4000 1b

c. Total load, 6000 lb

VI'd. Ttal load, 000I

d. Total load, 70000 lb

f. Total load, ii,000 lb

g. Total load, 12,000 lb

h. Total load, 12,280 lb (ultimate)

Photograph 16. Crack pattern, beam lB

1*

a. Total load, 9000 lb

b. Total load, 10,000 lb

C. Total load, 12,000 lb (ultimate)

Photograph 17. Crack pattern, bean 6B

adPORTLAND CEMENT EPOXY RESINCONCRETE frr--

a. Total load, 6000 lb

b. Total load, 6800 lb

c. After failure at 6840 lb

Photograph 18. Crack pattern, beam 7B

PORTLAND CWIMT EPOXY REiMN

ýCO mCRETECOfRT

Photograph 19. Crack pattern, beam 8B, after failure

So

iId

II

SK /

!I / Iw•

L .... .. J a * +e a * * : !D

SA -J L

P I

fl l J [ o + i+•ii M[J~i l I Il " llMI l I • i I I i + la

li I

-- - -. a.-

&w MOM4•

--------------- RE~ L aSO PX EI

--- " ••' •' "T IA MITUE

b E

0UATZSS@RTA-REA-b•. ATAK

j-fiI. Mi!& ELAITI 6 M[I Oi

couSTra" ~aaae. A* (r)* RESUL.TS OF EPOXY RESINGOP TRIAL MIXTURES

QUARTZ- CHERT AGGREGATE

PLATE 2

0.0"0 UNIJ IEj

18,000 - -

I8004-- 11 1 1 -- -11,000- - -

4,000 ---- --

8.500 m PAo

j 1E -1 1-

l~~~m cvw% *OrTOTAL AGAU"T UPow"

# FLEXURAL ELASTICU0iIU

- Auro RESULTS OF POLYESTER6" -- lAD. as U 00M1 @8)4 96 RESIN TRIAL MIXTURES

QUARTZ-CHERT AGGIPE6ATE

PLATE 3

0.0700

SUNIT V4t10T

1110

slowm

ji0.0"0 -

56.0"

-- I

I o

~" 'rr -- - - --.Im11Q..... - • .... & ...

w 1"' WAN

A FLEXURAL ETIC MD S

COW cd,,,,,, W , (,f. RESULTS OF POLYESTERRESIN TRIAL MIXTURES

LPLSTTE AC4WC.ATE

PLATE 4

$0,000

72.000 _ _ _ _ _ _ _ _ _ _ _ _I_ _ _

S4.00011

[ 1 f o fp/U 40.000 4 r

aU 3.500 PSI /Pu 00oU

__ __ f jl ;!E.,s ' -Poos___

I _ _ I _ _

to 000O 00 S00 800 000 ~ A@ '.0

STRESS-STRAIN CURVESOF REINFORCING BARS

PLATE 5

I-w

< 10

0.

PLATE 6

LL.

zi-J

0-s-0

ww

PLA-

I'-

<La.

0 o)Iz 0i- f

w -F LLI(C

0 _4 IX N 4I L

in < I-

0 1

-0

I 0z

Ct

00I

WOz

In

____ - fm_ 0~

ILo

PLATE 8

Z w

ujg

wz

cr

0u-

IRI>

0 0

D ~

Z W I 0OZW

z 0 Z 0 w

WZ-J 0 ~Z N

-10 N mnw- Xw > 0I

0J 0. 0 0

0 0 I.8 ~0I00

U, 0 U

ISd 'Hj!0N3d]lS 3tISS3bldbN0:

PLATE 9

Id0

I I__.%_ _ _ 0 4

w II . ; . --t d - 30

I0 U

I-0 x I'OVi-N

PLAT 10h

4

IAO I

130

IO0

110

lo

.o

LEGENDA 11%b POLVIESTEWI RE(SIN AND (COW. -

TINU, OUSLV GRIADEDO LIIMESTONEiAGGRE(GATE[

0 10% POLYESTERl RE[SIN AND CON-

70,TINUOUSLY GRlADED QUARlTZ-7 0 -- •CNIAT AGGREGATE

a 14%*/ EOXY POLTSULFIDO[ RESINAN UP-GOlADEO QUJART-

CHit4lT AGGmREGATE

so.1a 6 0 Is

TIME, MRH

TEMPERATURE RISEIN RESIN CONCRETEIN I-QUART VACUUM

BOTTLES

PLATE RS

•U

- wla)w

Z U

z0

0

I-

Li

hi

'o-!i0

PLAT E 12

00

--- ... . . __.

oweLIu

IIf-

SI -

. . . . . . .LAT 13? "

--- - + ' + , ++ - • .I

III I-.I -it AT-.-.

----- I 711 -1 1

*1 -'�-- 'I IhJz

-- '�-4---- I0

I'' U

¼ / a

- 'v--I-- I

I

-I

-- ii -I- 'a

4 I

- -12 I.-

-- Lq �

-- B -

--p.----- I I.-- V - -

-- -.- / -I .- I

I -I / �iiiii

o11z- ___ _-I -� --I

- � liii!I�h. 0L.ILi,� Liii�J 1

6' �

PLATE 4

II

S,,o@oC! I ____ .Io8,606 -,F~.' I -'! / •• ,

0.000 SAM PA

i"7• I i ." [ . - _-i I .. . +•_1.-I - -" .... AMCW RA

""8000 400. ._ 0-- SO I.OO 0,6 I-- S0 0,-

•" a. EPOXY CONCRErTE SERlES

o0

- - SAm AA

sIM1 5

b-.POYESTE COCEESMC

MOMEN -CRAUE

RLATI ONS

MLATE IS

12.000

J4a0400O *OOO444

0

4.000

Oa 0,2 040.90

MIOSPAtl DEFLECTION. IN

LEGEND

SEAM ISSEAM 6•

BEAM 78SECAM1*1S

MIODSPAN DEFLECTIONS

POLYESTER CONCRfET E SERIES

PLATE 16

-UJ

Law

0

-4- 0L 0

1T 0

II I

0 02

* 0 0 0 00 0 0 0 40 -

0 0 8 0 L CL 0 0 0

0 0 0 0 a a 0 0

0 ft

PLAE1

92,0001-----1-I -

./

s,ooo•: f-----I I I'

6,~000 900

1,000-100 'oo 2,000 3,000 4,000 5,000 6,000

co. BEAM 1B

10,000/ ____ ____ _ __

_ / / _-

6,000 __

0,000 -----

"0 4,000\ .

2,000 - . -

0-2,000 -1,000 0 1,000 2,000 3,000 4,000 5,000 6,000

b. BEAM OB

92,000

90,000 -

6,000 -

6,000-

2,000 -t____01

-2,000 -9,000 0 9,000 2,000 3,000 4,000 5,000 6,000

STRAIN, IN\N. X

c. BEAM 78LEGEND -

- RINFORCIEWNT STRAINS- COMDR ESSIVE STRAINS (1-4)

--- LAI PA L ST PAINS CS - T)TENSILE S'AINS AND 13)

- TENSILE STRAINS (10,11,14, AND IS)

AVERAGESTRAIN READINGS

POLYESTER CONCRETE SERIES

PLATE 18

'4w

r-N-Wu. I

N0 _W -

01 z 0

I- OZ

Q.-J 0

E~hw i~w

~hJWnww 0

).Q

0 -0

00i

0 0 a 0 0N 0

. 38fl±V83drfl3

PLATE 19

APPENDIX A: ELASTIC ANALYSIS FOR CRACKIIM OF RESIN CONCRETE

1. For the sake of simplicity, and due to the large variation of

resin concrete properties, a linear elastic analysis (similar to the Work-

ing Stress Design method in reference 38) was considered satisfactory.

This analysis is based on the following assumptions:

a. Plane sections remain plane (strains are a linear function

of the distance from the neutral axis).

b. All materials are linearly elastic.

c. Portland cement concrete has zero tensile strength.

d. A perfect bond exists between the portland cement concrete,

the resin concrete, and the reinforcement.

e. Sections experience pure axial bending.

For these conditions, equation Al can readily be derived from fig. Al by

fulfilling the plane strain and equilibrium of forces requirements and can

be solved for kd

E aPORTLAND

CEMENT kd

a

RESINLAYER

Fig. Al.

bkd)A + (h - M + h - kc - t)¢co lc T R (d P 7 d 2

(kd) 2E b + kd(E A + E bt) - (ERARd + E hbt - b ! ) 0 (Al)2 7 " p p2

Al

if C Pu is the ultimate resin concrete strain before cracking, we obtain

the cracking moment as follows:

u b 2

M _p h - u-k (kd)2(Ec 7)(7 kd + d - kd)

h - kd bt [(h -kd- t (h t d) +(h t d)t

h kd~a) 2 2

=P b 2 d - ) I t " t kd - t)(h - - L)Mcp h - kd 2 Ec 3 2

- - d)I (A2)3 2j,

where

AR = area of tensile reinforcement

b = width of Ieam

d = distance from extreme compression fiber to centroid oftension reinforcement

E = elastic modulvy! of portland cement concretec

E = elastic modulis of resin concrete (in tension)P

E = elastic modulus of reinforcementR

f = yield strengthyh = height of bean

kd = distance from extreme compression fiber to N.A.

14 = cracking momentcp

M = yield momentyN.A. neutral axis

P = cracking load (per loading point)cpP : ultimate load (per loading point)u

P = yield load (per loading point)yt thickness of bottom resin concrete layer

concrete strain in extreme compression fiber

6 = resin concrete strain in extreme tension fiber

epu ultimate tensile strain in resin concrete at cracking

CR = average strain in tensile reiiwX'rcement

aco = concrete stress in extreme comi., ssion fiber

A2

a = resin concrete stress in extreme tension fiberpo

aR = average tensile strain in reinforcement

Results of this analysis as well as the calculated yield moments (refer-

ence 38, ultimate strength design) are summarized in table Al.

A3

P, :

to 0 H0

0 0

UN .U-\ a .. G\ CM ..* i r\ ' 1-4 1- 14 ~ r H

0, 0ý "'\ 0-.

0~~~~ 0 r)3,

ý8 C-l r- *l D I %

OR 0) no-

CY0)

4 C\)

.2T CT

c C 'j C' j QD ac liX CV

to)

-H r4, -4N .- 4. C'J

C- ý) 0i t-- coH

(I) n r' rn 7 (1 Ie

Unclassifiedspud.cwftw CicauUctoas

DOCrchN CONRO 84A-*6. ORGNATINGACTIOVITYM M-O. s meow S&OIATW 09PORT M6UMGISI CASIICTO

Fia repoertncl eor ~- -

0.. AVMI EEPORf MOISI (An. .ddae mW11 U I apS a.

un~limited.

SC. UPIMPORT ATS7 TOTAL 000. SN O FIM MSAGK A .CT.IormTp

Mepatcen of6- the A 9

Thordcment t larg deectaproens for pulician tenesie crd ack is. The reprt describs

~st mpotan. eginerig ro~etie o±theselecate mixrtures .of incude Arry (R he

sults areport desries ithe results of tant o lretý,ton reeinte thea fewiiithou ofir combniregelayer angh iohprestivel srenulth. ofe study leent tonrthe andlotie uvrinor concilu-strength a. eprx or polyester re: ;in concrete: intor ai' hcmoste. ln.' fhie wol cin-!.brease the beamuse flexral streng~ythz thd int~roemthe ecro prtyectirn foth ~reinfreeto athe.b lar re deletiontby tianth.irg tensluee cracks. The report descibtyes

rsin de crelthe dvlopyernt of VIJ.Fh*i louct resi'n, conctret mixtures* an sur~rmsPir., ec theoprast icpranlpt eninerin fapr-e. .Thez of thy relecte mixtu~res.t b- rncluuetable fhrthiuls ofhet thad-on th loadyoster;jo -re i e nforc aed in ture tn I.er cori1,.ite~ an Iwi~the. a1-1/2s-l,1'e ttn -n-hc aersiof 'remox an pozlyen csterarein. cancritv.y. '1. o Jre-sultsoi~ ares coine withrete fý cituf tt_,t e orf ttworfreL .hnce bean willhu reoi toorncrctelthers adwthn utrtlytial. rh j.ltrtmn1ae -. u led h'r:to t~ em `ilci~v ri-.a! cn 'conlzreeon s: n:at. Pr2r y ler.' r l, !c ai, fc *rs etc.). ,: -. -c fcnce

bewr ~ -anA~G beYS use tA, m. itSSey ýn r-ývt e ,te.-t SIr. er~r-,crt em ,) oip-rleteflxrl 'rnqi tr- t !-G? AýY i 1ncarn s 1, a actr

tw ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ sst , otre, b e:-otn ha lr rfure ,':te alU.G y

ren~n ii),cre c Tzyt~:: , r vde I r_'.;uruýýq rT- x-c p~ ett~o

4

Unclassified

U1wt CoaaIIIati..___________

may WORODS L A L a L

ROL] WIT MOL6 ST MOLS W?

Beams (supports)

Concrete beams

Concretes

Concrete testing

Epoxy resins

Ifigh.-strength concretes

Polyester resins

Unclassified