L I B HAHY OF THE U N IVER.5ITY Of ILLINOIS 622 lJL65c hlNib :i,NQ - The person charging this material is re- sponsible for its return on or before the Latest Date stamped below. Theft, mutilation, and underlining of books are reasons for disciplinary action and may result in dismissal from the University. University of Illinois Library E SEP 6 WW WEE! lib 0CT3 p r - ?, I0T71J - • - r L161— O-1096 Digitized by the Internet Archive in 2013 http://archive.org/details/jointmaterialsfo03babb no. 3 ENG ?fNG LIBRARY UNIVtRSITy OF ILLINOIS IKSANA, ILLINOIS CIVIL ENGINEERING STUDIES SANITARY ENGINEERING SERIES NO. 3 JOINT MATERIALS FOR VITRIFIED CLAY PIPES By HAROLD E. BABBITT Professor of Sanitary Engineering These Studies deal with current activities in the department of Civil Engineering. They serve as progress reports of major investigations, and in general as a means for disseminating information not readily adaptable for presentation in formal technical papers or bulletins. December 1, 1951 DEPARTMENT OF CIVIL ENGINEERING UNIVERSITY OF ILLINOIS URBANA, ILLINOIS 6> Sanitary Engineering Series Number 3 JOINT MATERIALS FOR VITRIFIED CLAY PIPES by Harold E. Babbitt, Professor of Sanitary Engineering December 1, 1951 1. Purpose and Sco pe. --This investigation includes a study of the physical qualities of various materials used in joining vitrified clay pipes and the determination of the composition of joint compounds which should give satisfactory service. The tests were made on six-inch vitrified clay sewer pipes and were all performed in the sanitary engineering laboratory. Joint materials tested were bituminous compounds, cement mixtures, and sulphur compounds. Tests included the ability of the joint to remain watertight when subjected to either internal or external water pressure; the resistance of the joint to flexure; various physical properties of the joint material such as specific gravity, melting point, and other properties as listed in Table 7, and. the ability of the joints to resist root penetration. 2« Acknowledgment . --The tests were made possible through funds provided by the Clay Products Association which cooperated with the Engineering Experiment Station in the conduct of the work. Much of the laboratory work was done by R. B. Moorman and H. E. Dew, Special Research Assistants in Civil Engineering. Acknowledgment is made of the work done by Professor J. J. Doland and Mr. H. E . Schlenz who supervised the work during the absence of Professor Babbitt. -2- 3. Pi]^s_andJVi. a ter.iaU,--Al.l joints tested were made on Six-inch vitrified clay pipe supplied by two cooperating manu- facturers; all pipes conforming to standard specifications C 13-24 of the American Society for Testing Materials. Special one-foot lengths of pipe were used for the internal and external pressure tests but standard two-foot lengths were used for the flexure test and the root penetration test. Eight poured bituminous compounds, eight cold bituminous compa inds, four sulphur compounds, one cement-like material, and ,wc different kinds of cement were used in the tests. All of the bituminous compounds, both poured and cold, and two of the sulphur compounds were proprietary articles. Physical analyses of the poured bituminous compounds are given in Table 7. Some of the physical, characteristics of the poured bituminous compounds were as fellows: - *"•"* N °- I - l^ee^r^leu ^orouslf aftaf LS, ! poured 9 Sto joint and continued to do so until cooled enough to thicken. M^U^ - Sof t ;1 th r a c f 3-t v ta St O dn O,. oi Upon d bc,.ng ? heated with the appearance of steam were given oil. Material Mo. 3 - Similar to Material No. 2. Material No. 4 - Similar to Material No. 2. Mate rial Mo. 5 - Brittle and odorless. When heated does not bubble nor fume„ Material No . 6 - Faint burnt odor. Bubbled vigorously when heated ~ m'vinn off white, odorless fumes. Hareencu Sly wnen heat was removed and did not always fill space in joint. Adhered well to sides of pipe. ' -3- Mater ials 6a. 6b . , 6c , and 6d -were similar to Material 6 except that they contained different percentages of bitumen, as shown in Table 7 Mate ri al No. 7 - Hard and gritty a Became thick liquid and gave off white, odorless fumes when heated slightly above melting point. It should be heated to about 235 degrees C before being poured into a joint. Materia l No- 8 - Similar to Material No. 2. The characteristics of cold, bituminous materials used were as follows: Mat erial No. 9 - Finely ground .mixture of asbestos and cement. The material is mixed with a bituminous liquid to form a compound that will harden after a few clays. Material No. 10 - Oil, bituminous substance with plasticity of fresh putty. Material No. 11 " The material is a pre-cast gasket cut to lengths necessary to fit the pipes to be joinedo The gasket consists of a pliable core of asphalt and fibrous filler encased in a heavy fabric which is, in turn, coated with a sticky asphalt. The asphalt core is fairly stiff and may require warming in cold weather before it can be used. Material Mo. 12 - Similar to Material No. 10, except that it is much softer and runs almost as freely as a liquid. Material Mo. 13 - A liquid waterproofing compound rather than a joint compound. M aterial No. 14 - Plastic bituminous material having viscosity about that of thick molasses. Material Mo. 15 - Mixture of Material No, 13 with equal parts, by weight, of powdered limestone to give the material sufficient stiffness to remain in joint, Material Mo. 15 - Mixture of Material No. 14 with powdered lime- stone. No proportions could be found that would form a compound that would not flow from the joint. -4- Somo characteristics of other materials usee! were: Material N o,, 17- A mixture of 2 parts sulphur and 1 part sand. The sand had a fineness modulus of 2.78; 99 per cent passed a No. 4 sieve and D,8 per cent passed a No. 100 sieve. Ma terial No» 16 - A mixture of 2 parts of sulphur to 1 part of limestone. Ninety per cent of the limestone was fine enough to pass a 200-mesh sieve. Material No. 1 9- Analysis of this material is shown in Table 1, Mat erial No. 20- Analysis of this material is shown in Table 1 Material No. 21- A common brand of Portland cement with sand and water, mixed in proportions shown in Table 5, Material No. 22 - A pulverized proprietary material, mixed with Material No. 21, in the proportions shown in Table 5, to make the material more workable. Material No. 23 - A common brand of hydraulic cement mixed with sand and water in the proportions shown in Table 4. Making Joints ,— All joints, with a few exceptions, were made with the pipes in a horizontal position,, The two lengths of pipe to be joined were supported on a 2 x 4 timber placed through the pipes. Jute, with the appearance of frayed rope and containing no tar, was calked into the bell by pulling off a length from the rope of jute and pounding it into the annular space to a depth of about 3/4 inch. After the joint was made the complete unit was lifted without disturbing the joint, by means of the 2x4 and hung on the pipe rack. Some cement joints were hung in a moist room where the temperature remained at 72 degrees F, and the atmosphere was saturated with moisture by sprays of water which played constantly through the room. •■' Field conditions were simulated as nearly as possible by making joints under four different conditions, as follows: Condition. A - Clean, dry pipe was used as it would be under the most favorable conditions in the field. Cond ition B - The pipe was wet with clean water just before the joint was made. Condition C - The joint was made with the lower half of the pipe submerged in clean water. Condition D - The pipe was wet with muddy water just before the joint was made. Several joints were made with cold materials under two other conditions: Condition E - The joint material was struck off flush with the bell of the pipe, as shown in Fig. la and lb Condition F - The joint material was placed outside of the bell as shown in Fig. lc. 5 • Making Poured Bituminous Joints and Poured Sulphur Joints ,— After the joints hadajbeen calked with jute they were further prepared for pouring by placing a runner around the bell,-, the runner being formed of a 1-inch rubber hose with wet clay packed around it to assure tightness. Poured bituminous and poured sulphur-containing joint materials were heated in an iron ladle over a gas flame. The temperature was controlled so that the materials were heated slowly and evenly and excessively high temperatures were avoided by stirring the materials and by observ- ing the temperature constantly during the process of heating. The temperatures reached in pouring bituminous joints are recorded in Table 2. As soon as the material had attained the desired consistency it was poured into the joint through a hole in the top of the runner. •■ ' '. -6- 6, Making Cold Bituminous Joints , -"-Because of the different characteristics of the cold bituminous compounds different methods were used in making joints with these materials, The methods used were as follows: Mat erial No. 9 - The mixture used consisted of about three parts of ground material to two parts of liquid asphalt, by weight » The exact proportions were of importance, as too dry a mixture would crack, causing leaks, and too wet a mixture required too much time to harden sufficiently to hold water pressure. After the mixture had been prepared it was placed in the joint by hand with trowel, care being taken to see that the material was well pressed into the joint and to fill the annular space to overflowing. Material No. 10 - This plastic material was calked into the joint by hand. The substance was found to be of no value as a joint material, as it did not harden for two months during which time the joint was preserved for observation. Material No. 11 - The pipes were placed in the position in which they were to be joined and the gasket was forced into the joint by means of a jack. No jute was required with this gasket as the fabric prevents the joint material from flowing into the pipe The joint is easily and quickly made, its most undesirable feature being the stickiness of the gasket. Materials No. 12. 13 T and 14. --These were unsatisfactory as they were so soft that they flowed out of the joints. 7. Making Cement Joints,— Jute was placed in the joint after the pipes had been placed in the position in which they were to be joined. The sand and cement to be used in the joint were measured by bulk volume in a metal beaker, weighed, and poured into the mixing pan. The water to be used was then weighed the weight being the product of the weight of the cement and the water-cement ratio selected for the particular joint being made. (•4 -7- After the cement, sand and water had been placed on the mixing pan, the materials were mixed with a trowel until a uniform mixture was obtained. The mortar was then placed by hand and trowel in the annular space in the joint and was beveled off to the edge of the bell, as shown in Fig. lc Those joints which were to be cured under moist conditions were placed in a moist room under the conditions described in Section 4. The joints remained in this room until tested. Air-cured joints remained in the pipe rack, described in Section 4, until tested. The temperature in the room in which the pipe rack was located remained fairly constant between 70 and 75 degrees F,, with an atmosphere relatively low in humidity. 8. Joint Te s ts Made . --Joints were subjected to an internal water-pressure test, to an external water-pressure test, to a flexure test, and to a test for penetration by tree roots. Joint materials were analyzed for those physical and chemical character? istics and constitutents which might affect the behavior of the joint. 9. External Water-pressure Test »--A steel cylinder was placed around the joint, as shown in Fig 2. The steel cylinder was divided longitudinally into two equal parts. Before the two parts were joined together the joint to be tested was placed in the lower half of the cylinder after which the upper half was bolted on to make a watertight connection, aided by tightening up on the packing gland shown in the figure. -8- The air vent was then opened and the cylinder was filled with water directly from the University distribution system. When the water appeared at the air vent, the water waste valve was opened and the air vent was closed. Any desired pressure, up to about 33 psi. could be maintained in the cylinder by manipulation of the flow through the water supply valve and the water waste valve. Pressures were measured on a mercury gage connected to the water pipe entering the cylinder. During routine tests pressures were held constant for a period of five minutes for each different pressure, beginning at 5 psi, and increasing in increments of 5 psi up to failure of the joint or until a pressure of 30 psi had been reached and held. In one test, after a pressure of 30 psi had been held for five minutes, the pressure was reduced in increments of 5 psi, each reduced pressure being held for five minutes. The test waa repeated on this joint with increasing and decreasing pressures and the rate of leakage was observed at each pressure. In another series of tests a constant pressure was maintained on the joint and the rate of leakage was observed at regular intervals of time. The temperature of the water used in all of the tests was about 62 to 69 degrees F. 10. Internal Water-p res suro Tests .— "A diagram of the appara- tus used in the internal water-pressure test is shown in Fig. 3. The pipe connections to control water pressure were similar to those used on the external-pressure tests, shown in Fig 2. -9- Unfortunately the air-outlet valve was in the middle of the pipe so that it was not possible to remove air entirely from the pipe. In the application of pressures a pressure of 5 psi. was first applied for five minutes. The pressure was then raised to 10 psi 3 and held for ten minutes, The pressure was then raised to 15 psi and held for fifteen minutes. This procedure of rais- ing the pressure in 5-psi, increments and increasing the time by 5-min e intervals was continued until the joint failed or a pressure of 30 psi, had been held for 30 minutes. This procedure is in accord with the standard test, C 13-24, approved by the American Society for Testing Materials for testing the bursting strength of vitrified-clay pipe. 11. Flex ure Tests.,— Two 2-ft. lengths of pipe were joined by the material to be tested and were supported in a cradle with points of support 44 inches apart, as shown in Fig. 4\with the end of the spigot at the center of the span s A load was applied to a lever arm resting on the top of the bell, as shown on the figure, The lever arm was pin-connected at one end to a rigid support. Weights were hung from the other end of the lever arm, which was of such a length that the load on the pipe joint was double the weight hung on the lever, plus the weight of the lever arm. Failure was said to occur when the deflection at the center of the pipe span was such that the edge of the pipe bell touched + b° base of the cradle support. Loads were added to the end of the lever arm in increments of 25 pounds, giving increments of 50 pounds applied to the pipe. Each increment was allowed to hang for a few seconds before the following increment was added. -10- 12. Root Penetration Test .— 'Fourteen tests were prepared to determine the ability of roots to penetrate joints made with various materials, Four holes were punched in each joint, with the exception of two joints which were left intact. The holes ■■:- were slightly less than 1/8 in. in diameter. Each joint was tested to see that it leaked water through all four holes before the pipe was buried, Each of four joints was made up with various percentages of creosote mixed with the material in the joint. These percentages in the various joints were, respective- ly: 0, 2, 1/2, 5, and 7-1/2; 2-1/2, 5, 7-1/2, and 10; 5> 7 T l/2, 10, and 15; and 7-1/2, 10, 15, and 20. The two lengths of pipe joined together with the desired material, each constituting a test piece, were placed in a box with the pipe axis vertical, and good, black soil was p : •. ' ~ filled around the pipes to a depth of 3 inches below their tops. Two maple saplings and two elm saplings were planted close to each pipe on May 15. A glass cover was erected around and over the box so that no rainfall could reach the saplings. The trees were watered periodically by pouring water into the pipes to a level just above the joints so that the trees could get no water except that which seeped through the previously prepared holes in the joints. The care of the saplings continued for a period of 16 months, the box being placed in a warm location to avoid freezing during the winter. ■ !ti 13. Tests Made of Materials . --All tests of materials were made in accordance with the standard methods approved by the American Society for Testing Materials, with the exception of the tests for adhesion, expansion, and absorption. The standard tests made have the following serial numbers: Specific gravity D 71-27 Bitumen content D 4-27 Ductility D113-26T Flash point D 92-24 Penetration D 5-25 Softening point D 36-26 14, Adhesion Tests . --Specimens for the adhesion test were prepared by marking off an area of one inch square on the glazed surface of the inside face of a small piece of 6-in, vitrif ied- clay pipe. A mold was made to cover the surface of the pipe adjacent to the marked area. The mold was made of a mixture of limestone dust and Milarkote of a consistency which would not adhere to the pipe nor to the joint material. After the mold was made and placed it was partly filled with the .'joint material to be tested, A strip of lightweight canvas was promptly placed in the mold touching the top of the material just poured, with long ends of the strip hanging free. The mold was then filled with the joint material so that the canvas was thoroughly embedded in it. The joint material was allowed to harden for 14 days before being tested. A test was made by supporting the piece of pipe with the specimen adhering to it in such a position that whe*n a UNIVERSITY OF ILLINOIS JJfiflARY • -12- load was hung on the canvas strip the load on the joint material was evenly distributed. Ihe load was applied by pouring sand into a buck- et hcnging from a spring balance supported by a canvas strip embedded in the joint material. The sand was poured into the bucket at such a rate that about one-half minute was required to add each increment of load. In one method of loading a weight of five pounds, including the balance and bucket, was first placed on the canvas strip. This weight was held for five minutes. The load was then increased by five pounds, making a total of ten pounds, which was held for five minutes. Each subsequent increment of load was five pounds, and each total load was held for five minutes. This continued to failure of the adhesion or rupture of the joint material. Another method of loading was to suspend a load on the canvas and to observe the time required for the specimen to break. Most of the specimens broke off on the place of contact between the pipe surface and joint material. The results of those few tests in which the specimen .broke elsewhere have been disregarded. Test results are shown in Table 3, 15. Expansion and Contraction ."-When a bituminous material is heated and poured into a joint it contracts upon cooling. The amount of contraction of the various materials could not be measured easily. It was decided, therefore, to find the per* centage of expansion of the material upon heating and to compute the contraction on the assumption that the contraction on cooling -13- was equal to the expansion on heating. This would be true provided the material did not volatilize nor change its specific gravity during heating and cooling. Twenty ml. of each joint material was heated and poured into a calibrated test tube. After the material had cooled the weight of the material and tube were observed and recorded. The temperature of the tubes were then raised in an electric oven at such a rate that the temperature was increased from 212 degree F. to 486 degree F, in 12 hours. Readings of the expansion were taken periodically during the heating period. After the tubes had been removed from the oven and cooled, the weights of each test tube and contents were observed and recorded. None showed any appreciable loss of weight and it was assumed that the speci- fic gravity had remained closely unchanged. Computed percentages of contraction, based on observed expansions, are shown in Table 4. 16. Tests for Absorption of Water by bituminous C om pound s. -• Absorption., of water by bituminous joint compounds was tested by submerging a weighed amount of the compound under water for 30 minutes. The specimen was then taken out of the water, surface water was wiped off with absorbent paper, and the material weighed. In no case was there any appreciable change in weight, indicating that each material tested was non-absorbent. -14- RESULTS OF JOINT TESTS 17. Poured B i tuminous Mat erials .--The percentages of dry joints at different intensities of external pressure on joints made under the best conditions with different poured bituminous materials are shown in Table 5 and 11, and results of internal- pressure tests are shown in Table 6. Results of tests made on joints under unfavorable conditions are shown in Table 11. It was found, in general, that the joint material which would make the best joint under the best conditions would likewise make the best joint under less favorable conditions. It was possible to make the conditions so unfavorable, inso far as mud, dirt, and water in the joint were concerned, that a satisfactory joint could not always be made even with the best material. It is to be noted that the first six materials listed in Table 5 are almost equally satisfactory in so far as the making of good joints is concerned. Although some of the joints showed slight leakage, and hence could not be classed as dry, the leakage was in- sufficient to condemn the joint as worthless. Material No. 6 seems to produce joints which are unreliable and are unsatis-" factory in sewer pipes. Analyses of the physical properties of the poured bituminous joint materials are stated in Table 7 A comparison of the results of the tests on joints made of these materials with the physical properties of the materials is presented in Table 8. It is evident from a study of Table 8 that those four ' ■ ! -15- characteristics which arc of importance in affecting the suita- bility of a poured bituminous material for making joints are: adhesiveness, softening point, pouring temperature, and penetra- tion, and of those four characteristics adhesiveness seems the most important. High adhesiveness, a low softening point, a low pouring temperature, and high penetration are desirable characteristics e Since these four desirable characteristics can be secured with different bitumen content, it is evident that bitumen is not the controlling constituent in joint compounds, although it may be an important factor in affecting the four desirable characteristics. The fact that the bitumen content is not the controlling factor in determining the watertightness of a joint is borne out by the data in Table 9. The material used in these joints was made up similarly to Material No* 6, but with different percentages of bitumen as reported by the manufacturer. It is somewhat surprising to note that the amount of contraction of the material on cooling in the joint does not unfavorably affect its qualities as a joint compound. Material with a contraction of 41 per cent on cooling made a satisfactory joint, whereas an unsatisfactory joint was made with material contracting only 25 per cent. However, the best joint material showed the least contraction of the joint compound, the contrac- tion being only 9 per cent. . -Ib- In preparing bituminous joint material for making joints in vitrified clay pipe, it is probable that if its characteristics are similar to the average of six good materials listed in Table 8, a good joint can be made. However, wide variations from these characteristics can be found in satisfactory materials. For example, the average pouring temperature of the best joint material was only 2 degrees higher than the pouring temperature of the poorost material. Other unexpected variations can be found in Table 8 e 18. Cold Bituminous Materials . — Of the eight cold bituminous joint materials tested three, Numbers 9, 10, and 11, were suitable for joint making; but only half of the joints made with the best material, No. 9, were dry at a pressure of 30 psi, external pressure. It was found, in making joints with Material No. 9, that the proportions of the mixture of dry and liquid materials were of importance. If made too dry the material would crack and would not adhere well to the pipe. If made too wet it would push out of the pipe when subjected to pressure. Material No. 10 was among the easiest to handle. Joints could be made satisfactorily either in air or under water and the material remained sufficiently soft to permit slight move- ments of the pipe without injury to the joint. It was easy to make joints under good conditions with either Material 9 or 10, particularly when compared with the care and experience required in making joints with poured bituminous compounds. -17- Joints could bo made quickly and easily with Material No. 11, even under difficult conditions or beneath water. The principal difficulties met were with the stickiness of the gasket and the need for bringing the material to a suitable working temperature without getting it so warm as to be too pliable. If the material is not handled properly in the making of the joint it is possible to leave a portion of the gasket obstructing the inside of the pipe None of the joints made with Material No. 11 was dry under external pressures of 10 psi., or higher. Many of the joints leaked badly at pressures below 5 psi. In making joints with this material it was observed that the gaskets did not fit all pipes and care was taken to see that the proper size of gasket was used for the joint being made. Where pipes made by different manufacturers are available on the same work, as in these laboratory tests, it is necessary to be constantly alert to use the proper size of gasket. Results of flexure tests, some of which are reported in Table 10, show that cold bituminous materials are not suit- able for making rigid joints. This may be a desirable character- istic, however, since Materials 9, 10, and 11 would permit relatively large movements in the pipe without injury to the '■"'■ joint. Movement of the pipe with joints made with Material No,.ll can be faster than when the joint is made with Material No. 9, without causing injury to the pipe or joint because the former is softer and will flow more quickly and easily than the latter. -18- All other cold bituminous materials tested wore un- suited for the making of watertight joints, some of them being so soft that they ran out of the joint before the material could be rejected, 19. Poured Sulphur Compounds . -- Results obtained from Materials No. 17 to 20, inclusive, wore disappointing. Water- tight joints were the exception under an external pressure of 5 psi. and no joints were watertight under an external pressure of 10 psi. or greater. Care and skill were required in making the joints, and overheating resulted in the creation of sulphur fumes and poor joints. All of the joints which were tested were made with care under the best possible conditions. It is probable that joints made with care under less favorable condi- tions would be even less satisfactory. 20. Cement Mixtures . --Strong, watertight joints were produced when the proper mixtures of cement, sand, and water were used and the joints were kept moist during the curing process. The great importance of proper curing is emphasized by the data in Table 5 which show that all of the good joints were properly cured and all but a few of the joints that leaked were not properly cured. The most desirable proportion of cement to sand is about 1 to 1^, or 1 to 2, with a water-cement ratio of about 0.6 to 1.0. The poorest joints were generally made with a 1 to 1, or a 1 to 3 mixture of cement to sand, regardless of the water-cement ratio.. -19- It is interesting to note that the addition of a small amount of plastic material greatly improves watertightness and facilitates the making of the joint. A mixture of cement to sand in the proportions of 1 to 1, with a water-cement ratio of 0.7, was unsuitable for making watertight joints without the addition of plastic material, as shown in Table 5, Only a small amount of plastic material is required since almost equally satisfactory results were secured with proportions of plastic material to cement and to sand of 0.1 to 1 to 1, and 0.25 to 1 to 1, although the higher amount of plastic material slightly increaaed the number of watertight joints secured, and improved the workability of the mixture. More watertight joints were obtained under similar conditions with Portland cement than with hydraulic cement, although the difference was only slight. The good effects of keeping the joint made with hydraulic cement moist during curing were not so marked with Portland cement, as is noted by the fact that some of the leakiest joints made with hydraulic cement had been carefully cured under moist conditions, and some Portland cement joints which leaked very little or were dry had been cured in the relatively dry atmosphere of the laboratory,, Cement joints were strong and rigid, as shown by the results of the flexure tests reported in Table 10, Joints which were kept moist while curing were stronger than those which were dry-cured, and the strength of the joints to resist bending -20- strcsscs increased with the ago of the joint. After the joint had cured long enough for the cement to set, usually for one day, the joint would not become more watertight with increasing age. 21. .Miscellaneous Observations. --During the progress of the investigation various observations were made which throw light on the underground conditions which may affect leakage into joints. For example, it was observed that leakage from a joint sometimes decreased as the pressure increased. This was duo either to solid matter in the water working its way into the joint material to fill the openings through which water was passing, or due to the disintegration of the joint material. The material was moving so as to close the passageways for the water rather than to open them. This phenomenon was more noticeable under the external-pressure tests than under the internal-pressure tests. In the external-pressure tests some of the joint materials were forced deeply into the bells, and in some of the internal-pressure tests the material was forced out of the joint. If the holes in a joint through which water is leaking remain constant in size the rate of leakage through the joint should increase in proportion to some power of the pressure against the joint. If the holes increase in size as the pressure increases, then the rate of leakage should increase more rapidly than a constant power of the pressure against the joint, and if the holes decrease in size the leakage should increase less rapidly than this rate. If the relation between the pressure on -21- a joint is plotted logarithmically against the rate of leakage from the joint, the linos should be straight if the holes remain constant in size. The fact that the holes tend to close with increase in pressure is shown by the tendency of the lines in Fig. 5 to curve towards the right. In line 25 the sudden increase in slope indicates a rupture of the joint at a pressure of 10 psi., and in line 21 a crack or other opening has closed at that pressuroo Joints constructed with a beveled edge, as shown in Fig. lc, were more resistant to leakage than those in which the material was left flush with the edge of the bell, as shown in Fig. la. This is indicated by a comparison of the results shown in lines 27 and 32 of Table 5. These four joints were made and tested under the same conditions with mortar and the two reported on line 32 were flush joints, as shown in Fig. la. No leakage occurred from the beaded joints, but both of the other joints leaked. 22. Conclusions .— Watertight joints can be made with any of the proprietary, poured, bitminous compounds tested and with either Portland cement or with hydraulic cement. Cement joints can be improved, both in ease of making and in water tightness, by the addition of certain plastic materials. None of the cold bituminous materials and none of the poured sulphur compounds tested could be relied on to give watertight joints in all cases. Poured bituminous materials which have the properties and constitutents of the best material, as shown in Table 7, can ; • '.-. ■■• -22- probably bo depended on to make watertight joints when carefully made under favorable conditions. Other poured bitminous compounds whose constituents and characteristics differed from those of the "best" material were used successfully to make watertight joints, but the results were not always certain. Joints made with Portland cement in the proportions of 0.25 : 1 : 1.5 of plastic material : cement : sand with a water-cement ratio of 0.6 to 0.7 gave the greatest number of watertight joints, provided the joint was kept moist while curing. The moistening of the cement while curing is important as other- wise the cement may crack and opon the joint. Watertight joints can be made with Portland cement or with hydraulic cement mortar, although more dry joints will be made with the former than with the latter material, Satisfactory mixtures of cement and sand in which plastic materials had not boon mixed were found to be about 1 : 1.5 for Portland cement and hydraulic cement, with a water-cement ratio of 0.6 to 0.8. Joints can be made more easily and less skill is required in making them with cement mortar than with poured materials. The principal objection to cement-mortar joints is their rigidity which permits no movement of the pipe without breaking the joint. Although the joints are strong, when broken they will leak badly, .whereas poured bituminous joints will allow some movement of the pipe before leakage will occur through the joint. -23- Watcrtight joints can be made with proper materials under favorable conditions in the laboratory. Watertight joints can be made in the field with proper materials, but there can be no certainty that any particular joint will be or will remain watertight. In general, that material which will make the great- est number of dry joints under favorable conditions will make the greatest number of dry joints under unfavorable conditions. This is borne out by the data presented in Table 11, which shows that, with the exception of Material No. 2, poured bituminous materials making the largest percentage of dry joints under the best condi- tions also make the largest percentage of dry joints under less favorable conditions; and that Material No. 6, which makes the smallest number of dry joints under unfavorable conditions. Similar results are shown by the records of dry joints made under different conditions with cold bituminous materials. It is evi- dent that the greatest number of dry joints can be made cither under favorable conditions by the use of a joint compound which will make the greatest number of dry joints under the most favorable conditions. The tests on root penetration were inconclusive because no joint tested was penetrated by roots. -- V _D o CD • ! ? jQ O .o c: .^ O u» O "O L. -«= 0) « -0 £ | o a? 5 £ •5 > u~ 5: Z o U- uJ en uJ _j < c ^ 83 CD D ,W z o UJ CO "') u f>LlCD BY V\ /e^.^M la L ■ ( t — / V^/>w. H 30 - ~j\/\s — I -30 Apparatus For I : Test of 8 f* r "S i f ) i O "7 o A — ' i~~- . C* 1 1 >. -- £ LL - ■u LiJ 5 JsT" ; <; J- U 5 ri i LJ, - H • CO CO CO ) rH 3 cr _. O\DOr--C0f^t"-CM o «oooo o s SO s *— % p en w o O 00 00 \0 vO vO if) CM lO • CM <- CD BJ\ CM rH • o o o ON j ^ CD • -p -p -p (0 O ^ CO a i Oi J2 O O O CO iD r^ O CM O vO On o o o O -P -P o -P CD ■— 1 lO o o co co r- o lo cm lO lO CM -st <-H <4-i 4^ -p I C7V-P i— 1 rH rH O o o CO 05 CO X as m c CD -P -H C C C o o o co co r^ o uo cm lO «CM ^ O O O G 4 O G 5 rH rH rH • H H H O (H'H O i (d ox; • ■j Q< •-) CO OOOOOOCMC0 OOCOOOOhCO o *co o o c- lO • '<' ^ > -Jeq.Bjft • CO pue s • • CO • rH CO rH C CD CO CO •H P, CO -H C>.. «H 3 q.uomaQ 3 P 3 rH to -p o o o a c x C -P c -P O «H •H ! "[etjo:}.!?^ f"H i~H r*H rH i—H r*H i™H r~H r— H ^~| c ojnSxj :oqran^ sut-f • UT J 1 • j :oquiri[sI 9uxi ; OrHCMcOsrcOvOt^COONO 1 rHOJco-^-io^or-co ONrHrHrHrHrHrHrHrHrHrHCM T CD C C -H •H\a sJ-CMvOkj- • • lD vD rHvO CO rH D"rH H-< JC CMr-OOOrHOOOCMiDCO co (0 o 8c^ • -P -P D> O O rH sj" \f) f*- '3' rH OrH vO CO CO ^D f- <-H lO \T> iT) \T) cm C OOOOOOOCDOHHOJ^OCOvDiOCMOO COCMr- lDI> p* CD rH cm co ^r in M-vor-rH cm cm CO O -H •H S -PH\ CO —1 rH<"H r-\ CO H > •H CO CO <" CM CO CD S Q s: CO <-H lO O vO vQ O COCO vO o cO • -p -P (J> OrHrHOr-OOl^-COrHrHLO ^OHOOvO AS OOOOOOOOrHCMrH-Nj-vDCMCMinajOOr-CMcO^OCOO o (0 O r-\ rH cm co invo r* t*» o rH r-o o CM i-J ^.UTOf ccad 'x^o rHcoKfvooeoinr-'sfCJco O O O r-l CM CM in ^rH O rH CO vO rH -.]• CO CO rH vo n n ooooooooooOrHrHcocooocoinoo'si-o no o i-HrHrHC\!rHCMcO \i"n n oooo ooor-ot^ooooooooooooooooooo o o OOOvO vD CO rHrHrH • oooo co ooor-or-ooooooooooooooooooo o CO C lO LUOOOO vO >* cd «h CM f^ rHrHrH ^-^ Q 3 O 3 CD CJ) co co oooo >< ooct^or^ooooooooooooooooooo o CO fl <+-< w\ O MOOOO vO o. J o o • CM "S. rHrHrH 3 rH CO en Oi X5 H -p J CD rH oooo 2 ooor~coh-^-oooooooooooooooooo o CD o-p in uj ooo\ocovO\o n c CO CO C rH S rHrHrH 1 ■P -H m c C CO o o J CD 4-> C, o oooo ooooor-r-oooooooooooooooooo o 4 O C 5 rH QOOOOOvOsD lO UD • J c U -H O oox: ^1 **"H i"H r— [i— ■) r~l -p H cu *->(/> h4 o If) r-'sioo H OOOOOOr-OOOiOOOOOOOOOOOOOOO o o v£)rH fiOOOOOOvO iDiDOJiDOif) O ■— 1 rHrHrH rHrH i— i opey q.uauiao -o:aq.2A.\ vor^or^coooocx}novOcor v -oooor'H>i>co -on^oo co OOrHOOrHOrHOOOOOOrHOOOOOOOOOrHO CO C CD CO uon in nuon •H M pues C • • • • • rH rHCMrH CM CO CO CM r-|rH rHrHrH rH CM CO CM CM CM rH CO rHrHCMCO CO CO -p P c x o O -H •H *s; q.uauo3 a £ T5 O rHrHrHrHrHrH rHrH rHiHrHrHrHrH rHrHrH rHrHiH rHrHrH i-HpH H "0 N T^T^^ew rHrHrH CM CM CM CMCMCMCM CM CM CM CM CMCMCMCM CM CM CM CM CM CM CM CMCMCMCM CM UT J equina autq rHCMco^nvor-cooorHCMco^invDr-coooHCMco ^in vo rHrHHrHrHrHrHHHrH CM CM CMCMCMCM CM X oquin^ auxq rHCMoOv^ \D vq [^cooorHOJcosr in-^or-co oorH c\ico ^j u invor- coo o CMCMCMCM cMCMcMCMCMcococococococococO(r)s^^t^4• 5 4'^ , s, , ■^■ , N^^'^ n c •H >, • (T3 0*0 • c if- 1 " H o ftJ a, co O -p c o CO o •H O (DrH CM r- LOcotnt- -f-> 0> OvD O ^HOiOlO CO CO CO IH C o • 0) o oo^v oo o ooov \i- o o ■; '" ^t CO hJ CMiD CO CO lO lO lO lO CM CM CM CM CO ^i- CO CO- ■H CMirVHLO X! ooo rH CMGV0C0 CO >H O • -t-> •. • • • ••#•«••• -P cn o oocoo oo o ooino rH 10 o o o o 4h C COCO lO lO !> CO CO CO 0) CO CO CO CO CM >-l * VJTOf ccad w jib d •H CO vOif) OO O OCMsTCO lO if) o o o o • • • ••••■•••••• O OOCDO OO O OOrHrH CO % v O O O O O rH rH CM O O O r-HJ f-H] i— ^ r— i 40- o? o CO CM o CM tf) rH 111 pd o po H rH X M So o HrH feo W o UrH Q 20 n in c • «H r—if-{r-ir~\ rHrH rH rHrHrHrH CO rH CO -H CO rH rH rHrHrHrH rHrH r-\ rHrHrHrH rH rH rH >H rH rH in lOlOlo CM rH CM CM CM • o » • • O OOOO pa^sai sauxox^ JO ; on •on pexj:eq.ey\i •on 3UT1 ISSrSSQQ 552QCQQQ5^Q5 CO C0C0C0C0 COCO CO COCOCOCO rH CO "si" CM C CM CM CMCMCMCM COCO CO COCOCOCO CO CO CO CO CO CO CM CMCMCMCM CM CM CM CMCMCMCM CM CM CM CM CM CM >-i cMcovj-in or- co c-o-hcm co ^ u in vo i> co in lo inn n intnin invOvQvo ^o vo o vp \o o TABLE 6 RESULTS OF INTERNAL PRESSURE TESTS Condition A jLine No. Material Number "*?" No. of Joints Tested Percentage of Dry Joints at Pressure Shown in lb, per sq. in. 10 15 Materials i j 1 7 3 100 1 100 100 Poured 1 2 3 3 100 100 100 ! 3 6 2 100 100 100 Bituminous I 4 3 4 100 100 75 5 | 3 100 100 67 6 9 7 100 86 14 Cold Bi- 7 11 3 tuminous 8 1 1 1 1 17 4 50 Sulphur and Sand 1 I > i p C -P c 0) C CD £ cu O 3 P -PCM •H O CD OQ O Ct. Aj:oq.ea:oqE7 q. Today s , cca.inq.oeintrei/v ON CO CO CO CO LOCMOO cm co r- r~ CO o CO CO* KM «tf o -J- <-i o UD sTnOsQ vO lO vD -^J- lO ON CO CM o in if) iO • o • n o r- • • ld o in in vo n vo vd vt in -CD p XJ c 03 rf-fc c5 n <+-• o T5 03 O u c*3 CD a B CD (0 > T5 W C TJ CD O CT < CO Cm CO ejinadna: asneo o% paTTnbaa: CO V]* o H o i— i -H ON MO -P 03CDM-I O O -Pert 03 C03 CD H CDCD £ OrH D fHOrH CDftl O CJCmQ> g^gPT3 2 roo3£ OH ^ODI O •P c o o -p O paq.ndwoQ ^ r- o CM in in on ^- \D on o on CM rH I— ) »H - -P •H -H •H -P O Q Aa:oq.e«ioqBT; q. Today o CO o o CM CO ♦ rH CO o in co in in cm co n in co i> o n O CM HHHHO CO CO CO CD £ •H -P CO • •Hfn JZQ) '-P-P P3 4-+H O • O T3CD CV) CD O +H 03 CD MX 03O q-< -p •H -H o > CD 03 CLH CO O Ajoq.ea:oqe7 q-iodey s ^ejcnq-OBinue^ r- no co o oo ^ oo cm o in vc o i> in rH vo rHCMo^r-OsC o rH in in ^j- vo co co ^ co mo ^r • s • » • •••••• rH rH rH r-\ rH rHrHrHrHrHrH + CO CO CM in in vo -st 'd- co «CM vC • vO 'st JequiriN XBTJeq-BW cm co s i - i n 03 J3 OT5 nQ no \0 nq sq r- or ,; >*;■ < • • i i • j TABLE 8 A COMPARISON OF TEST RESULTS ON POURED BITUMINOUS MATERIALS AND THE PHYSICAL CHARACTERISTICS OF THE MATERIALS E Constituent or Characteristic Best Mater- ial No. 8 Six Good Materials Including Only No. 1 2, 3, 4, 7, 8 — f Max. Avg. Min, Poor- est Mater- ial No. 6 Remarks' Bitumen Content Contraction Load in lb. per sq.in. required to break the material in tension Softening Point Centig. Pouring Temp. Centigrade Ductility Specific Gravity Flash point. Fahren. Penetration 57 9 60 99.3 41 104 63.1 23.5 55.7 41.7 9 30 94 246 0.5 1.46 535 6 97 246 3.07 1.605 585 26 93.3 227.6 1.6 1.387 532 11.2 83 202 0.5 1.007 470 6 43.5 25 23 114 244 0.53 1.6 540 5 TABLE 9 External and Internal Pressure Tests on Materials 6a, 6b, 6c, and 6d to Study the Effect of Different Percentages of Bitumen on the Number of Water-tight Joints Produced Material Number Percent Bitumen Mfg Lab. LHeport I Anal IZ3 Percentage n of m Dry,, Joint s Lb. per sq„ in. |jLb./sq # in, External Pressures j Jnternal 10 1 15! 20 25 Pressures 30; ; 5 i 101 15 100 1 100 100 { 67 67! 67 1001100 i 100il00il00il00i 100 100 ; 100 67 ! 67 j 671 67 ! |l00 100 j 100 67. 67! 67: 67 ! j 100! 100 j 100 ioo i ioo j ioo i ioo! ioo j ioo 1 100 ' , TABLE 10 RESULTS ON FLEXURE TESTS on COLD BITUMINOUS AND CEMENT MORTAR JOINTS Best Conditions No. did not fai.. at 800# Note 1 # — All of the joints, with one exception, failed under their own weight. The one exception failed under a weight of eight (8) pounds, ' - . -- ■■■• ■ ' to rH co •H M CD P cO to P o C •H e 4-> •H CO if to o o O • o O o if) o o 1 +> CO CM * CM ! c x: • •H o O c •H if) CM o O CM • • o o o if) CM o o >. CD • Q h H Q CO C O o o O • o o o O o o O H-« U" CM CM • if) •H O 05 • P •H CD p X5 CD CD in o O B o o o|| o o o c co a rH CM a H lo O -P • I CD o to a o o O • o o o o o o 13 S M D H CM O I if) I CO CD O • CU CU 9* if) LO o O • CO o o o lO o o ~ Cv! • CO o o • •H S sq-sei JO *om o CO lO o o CO CO co 1 * a CM <-\ 3 P • •H 0) ( o p- If) l> n o o ° i g o ° 1 XJ P to CO rH I s - rH 9 CM CM if) c p c x • i rH O •H o ° o -o o c lO r- if) r- • o O r- o o O 1*3 •H CM rH o rH • • CM CM ■o If) U) CD P >> CD P O CQ h Q P. co o o if) r^. • O O o t*- o o I CO C •H c C5 CM • rH F» rH • CM CM if) if) 1 CD S O <+* -P •H O CO cr to # • c o •H P -P •H O p if) r~ o o • o O o D- o o rH -H T5 CT) CD rH rH o if) • CM CM 1 1 1 o if) if) CD Con< enta G. to T5 o • r- rH o o • O O E"; c o ■H X5 X) CD D fH O 3 o C 1—1 rH o If) CM CM if) lO T3 O u 3 rH • cu CD O CD CU CU x: o if) r- o o • O O o l> o l> p p \ rH o rH uO CM CM if) o rH sO CO O CD rH C s+sai jo 'oh o O Si" vO o If) if) - o o o 0) CD p CO o 00 00 Co \Q Co if) CM if) C\l rH X I C r; rH CO P •H O •H O c rH CD l~3 •H lO o vO o r- CO r- [> CM o ON o O > CM o CO CO vO vO vO lO CM if) CM 4-' O >> < P CD P rH CO Xi :s CO Q CO C o o vO o r- CO i> r*» CM CM J o o o C <-HP cr CM o CO CO vO •vO ^o io If) CM •H O CO to ■H P •H CD p 13 en CD lO o o o CO if) ["*■ I s - CM o C> ° C CO a rH o o CO CO r- vQ lo CM If) CM O P rH rH o c to CD 73 o C O o o o CO if) l> r- CM o O o H cs rH o o CO CO r — vO lO CM if) CM CD o rH rH cu cu lO o O o o o O cm CO o CO o O o CO o o o i> CO If) ■sj" CM rH rH rH rH rH sisai JO' •on "si" f- If) v3 00 vO r» o vD r- o rH •on te TJ.sq.l3w CO CM -vT r*» co -H if) vO ■ o ... — j rH ■