Phe JBLIC.F Y ¥ wav’ »)@ . Nos { wo, BR ih 10) ail AD D he. [_A JOURNAL OF HIGHWAY RESEARCH was ’ UNITED STATES DEPARTMENT OF AGRICULTURE SZ, BUREAU OF PUBLIC ROADS Wwe Sei — APRIL, 1926 VOL. 7, NO. 2PUBLIC ROADS A JOURNAL OF HIGHWAY RESEARCH U. S. DEPARTMENT OF AGRICULTURE BUREAU OF PUBLIC ROADS CCenrureyer iy delim of the Seertary of Axriulre, the ma foe the pope ontsinod herein ie published ae admiitative information and ix eemsced etin of the publ iste, VOL. 7, NO. 2 APRIL, 1926 H. S. FAIRBANK, Editor | TABLE OF CONTENTS Stresses in Concrete Pavements Computed by Theoretical Analysis ; 5 Tests of Vibrolithie Conerete 36 A Siudy of the Effect of Temperature on the Stability of Asphaltic Pavements... 46 Motor Vehicle Registrations, Revenue, and Gasoline Taxes for the Year 1925 . 49 THE U. s. Willard By DISTRICT No, 1, Oregon, Washington, Manta, and Alaska Box 3000, Portland, Orc. DISTRICT No. 2, Clini, Asizona, and Nevada, | Bay Building, Sam Franciaco, Cabs DISTRICT No, 3, Colorado, New Mexico, and Wyoming, 5301 Customhoute Building. Denver, Clo DISTRICT No. 4, Minacsota, Noth Dakot, South Dakota, and Waconia. 4410 Harm Building, St Poel, Minn DISTRICT No. 5 lowa, Kansas, Missouri, and Nebraska Sth Floor, Saunden-Kennedy Building, Omaha, Nebr. DISTRICT No. 6 Arkansas, Losisiana, Oklahoma, and Tens. 1012 F&M. Bank Building Forth Worth, Tee ng, Washington, BUREAU OF PUBLIC ROADS REGIONAL HEADQUARTERS Bay Building, San Frans co, Call DISTRICT OFFICES DISTRICT No. 7, Ilnss Indiana, Kentucky, and Michigen ‘South Chicago Staion, Chieago, Ile DISTRICT No.8, Alabama, Georgia, Florida, Misi South Carolina, snd Tecan Box J, Montgomery, Als, DISTRICT No.9, Connecticut, Maine, Mastachasats, New Hampshire, [Newserey. New York, Rhode Island, and Vermont Federal Building, Troy, N. Ys DISTRICT No. 10, Delaware, Maryland, Nonth Carolina, Ohio, Penosylvaria, Vieginia, and Wes Virginia, Willard Building, Wathington, D.C. DISTRICT No. 12, Maho and Utah. Fred J. Kievel Building, Ogden, Usa ingitutons other than Si highteay engineering, Government Painting Office, Washington, D. C. Oving to the necessarily limited edition of this publication it will be impossible to distribute it free to any persons or ate and county offcials actually engaged in planning or constructing public highways, instructors in riodicals upon an exchange basis, and Members of buth Hoases of Congres. Public Roads” can do so by sending 10 cents for a single number or $1 per year to the Superintendent of Documents, Others desiring toSTRESSES IN CONCRETE PAVEMENTS COMPUTED BY THEORETICAL ANALYSIS’ [iy M. M. WESTERGAARD, Associate Prfesar of Theoret and Ayelied Mashanic, Universi f Hines Doctor Westergaard, in the following paper prepared for presentation before the. Highway Research Board, summarizes the results ‘of a long period of study’ partly under the auspires of the Bureau of Public Road ‘From aseurmplaons of the eonditions of loading, sup= port, ete of conerete road slabs, which eonform elovely to the actual conliions generally obtaining, he has developed by mathematical analy Wira‘melhod by which the stsss in rod slabs maybe computed imaper the melhod can be appled contcniendy by highway engineer Filing the esa! stresoes tx exiting pavements and ‘By the use of the Jormules, charte and tables which avconspany the for the design of concrete road sas. It also offs a means of tom: ‘mat be used to furnish the anevier tothe question, often propounited, as to the possible eerease in the thickness of « pavement i the operation of the heavier echicles is prohibited, and, vice versa, hat additional Uhiekness £8 required by a giten increase in wheel pressure, One may obtain 1 computation of stresses in con- crete roads by assuming the slab to act as a homoge- neous, isotropic, elastic solid in equilibrium, and by assuming the reactious of the subgrade to be vertical only and to be proportional to the defleetions of the slab, With these assumptions introduced, the analy sis is reduced to a problem of tle mathematieal theor of elasticity. ‘The reaction of the subgrade per unit of aren at any ven point may be expressed a3 a coeflicient, F ho deflection, 2, at the point. This coefficient is a measure of the stiffness of the subgrade, and may be stated in pounds per square inch of area per inch of deflection, that is, in ]b.fin*. The coefficient, kt, will be called ‘the modulus of subyrade reaction. Tt corve- sponds to the modulus of cleasticity of rail support” which has been used in recent investigations of stresses in railroad track.? The modulus, k, is assumed to be constant at cach point, independent of the deflections, and to be the same at all points within the area which is under consideration. Itis true that tests of bearing pressures on soils have indicated a modulus, &, which varies considerably depending upon the area over which the pressure is distributed.? Yet, so long as the loads are limited to ‘a particular type, that of wheel loads on top of the pavement, it is reasonable to assume that some con- stant value of the modulus, k, determined empirically will lead to a sufficiently accurate analysis of the deflec tions and the stresses. One finds an argument in favor of the assumption of a constant modulus, &, for ‘a given section of road by examining the tubles whieh are given below. They show that an increase of k from 50 1b./in* to 200'Ib./in.*, that i i the stiffness of the subgrade in the one, eatises only minor changes of the important stresses. Minor variations of é, therefore, can be of 10 great consequence, and an approximate Single value of & should be sufficient for a quite accurate determi nation of the important stresses within a give! of the road. ‘The modulus, & in the formulas for the deflections of the pavements, and may be determined empirically, accordingly, for a given type ni Bngloeers, Trans. 32,1905, pt soir tee thao et oat cara abe darcy p supporting val of sail induenecd by be bearlog ares" by’ A te"F. Busmrd, PUauie ROADS". 3, Seo teee so0ne—26}—1 of subgrade, by comparing the deflections found by tests of full-sized slabs with the defections given by the formulas. Ti will be assumed for the time being that the thick- ness of the slub is uniform and is equal to h. ‘A certain quantity which is a measure of the stiffness of the slab relative to that of the subgeade occurs re- peatedly in the analysis. [tis of the nature of a linear dimension, like, for example, the radius of gyration It will be ealled the radius of relative stiffness. Denoted is expressed by the formula [ER Viaae whore E is tho modulus of elasticity of the eonerete, and g is Poisson’s ratio of lateral expansion to longi- tudinal shortening. ‘The stiffer the slab, and the less stiff the subgrade, the greater is 7. One may observe that 7 remains constant when E and E are multiplied by the same ratio. ‘Table 1 contains values of T for three different values of & and for different thicknesses of the slab. In computing this table as well as the three tables following, Poisson's ratio, 2, was assumed to be 0.15. ‘This value agrees satisfactorily with the results of tests by A. N. Johnson. ‘The values of 1 given in the table lie between 16 and 55 inches, and about 36 inches may be considered to be a typical average. THREE CASES OF LOADING INVESTIGATED, Figure 1 shows three eases in whieh it is of particular intetest to be able to compute the critical stresses. In Case [a wheel load acts close 10.2 reetangular comer of a large panel of the slab. This load tends to produce a comer break. ‘The critical stress is a tension at the top of the slab. ‘The resultant pressure is assumed to be on the bisector of the right angle of the corner at the small distance « from cach of the two intersecting edges; the distance from the corner. accordingly. is, @=ay2. Tn Case TI the wheel load is at a consider. able distance from the edges. ‘The pressure is sasumed to be distributed uniformly over the arca of a small circle with radius @. ‘The critical tension occurs at tho bottom of the slab under the center of the eirelo. In Case ITT the wheel load is at the edge, but at a eon- siderable distance from any corner. ‘The pressure assumed to be distributed uniformly over the area. of a small semiciicle with the center at the edge and with ‘Dove measurement of Poison's rao fr eget” by A. N. Fahnon, Am. sou be Sta oes eo Part a om26 radius a. The critical stress is tension at the bottom under tho center of the circle. In cach of the three cases the load incntioned is assumed for the time being to be the only load acting. reaction, I, computed from equation (1) MPS petwtb in? | kt00tw Ko. | kav tb Ant ieee: es aa G) eg FR ; Rel 8h) Re Ses ote el Hi Bey For Caso Ls eomputation which may be Jooked upon es.4 fist approximation was proposed by. A. T. Gold. beck. Further emphasis was given to this method by Gifford Older ‘The load is treated as a force con- oC @ case TENSION 0; THE BOTTOM. case r ose at TENSION o TENSION G AT THE TOP ATTHE eOTTON es. 1-~The csr ling, he renin rove roses oe ea centrated at the comer jtself—that is, one assumes @=4,=0. At small distances from the comer the influence of the reactions of the subgrade upon the stresses, will be small compared with that due to the load. ‘The corner portion may be considered, there- fore, to act as a cantilever of uniform strength. At the ‘distance z, mensured diagonally from the corer along the bisector of the right angle of the comer, the bending moment is — Pz. “This bending moment may be assumed to be distributed uniformly over the cross section, the width of which is 2x, ‘Thus one finds the bending moment per unit of width of eross section equal to—f and the Lensile stress at the top equal to 3P omy: a @ eng Ms” by, lord hdr A, Se. Cv ats, ROADS Since the wheel load is distributed over the area of contact between the tire and the pavement, the dis- tanees a and a, ean not be zero. "The greatest. stress ‘occurs, then, at'some distance from the loa tanee will be sufficiently large to make th the subgrade outside the eritical section contribute a notiecable reduction of the numerical value of the bend- ing moment. = ‘An improved approximation has been obtained in the following manner. ‘The origin of the horizontal rec- tangular coordinates x and y is tnken at the corner, the axis of z bisecting the right angle of the corer. | By use of Ritz’s method of successive approximation, whieh el a re 1 and sun odioan ae eit, MAA Lh | t | t orbs sash nae pants LH UL 7 " T SUM Nibsen HL . \ Ei { i | Srogah Ba1S6h L— azn LULL [paisa Pra, 2-Cones ofeguivaet dttbton of presse ip based on the principle of minimum of energy? the following approximate expression was found for the deflections in the neighborhood of the comer: - » (2 14 ose“). r Then the reactions of the subgrade will be expressed with sufficient exactness in terms of this funetion as kz. One may compute, then, the total bending moment, I’, in the section x= 2, due to the combined influence of the applied load and the reactions of the subgrade. When 4, 1s not too lange, this bending moment will be approxi- mately uniformly’ distributed over the width, 22,, of the ree an ce it is, the bending moment per unit of width becomes --@) ns 2q,' The numerically greatest (W, Rlte, Celle Fura, #1851980,April, 1926 value of fwas found, in this manner, to occur approxi- mately at the distance pesy oie 2h lge% 14560 03054h 03709 0504h 0705h sian PUBLIC ROADS 27 stress may be stated also in the following form ch is derived by substituting the value of ¢ from equation (1): ( Bis 8 a-ak) 3 a 0, the last two equations assume the simpler quation (2) eereeees (4) + With a, form of STRESS NOT GREATLY AFFECTED RY SURGRADE CONDITION ‘fable 2 contains numerieal values of the critical stress 9, for P=10,000 pounds, £=3,000,000 pounds per square inch, and ‘The table shows the influenco of three variables: ‘The thiekness, J; the modulus of subgrade reaction, &; and the distance, a, from the edges to the center of the load. ‘An inspection of the table shows the influence of the variation of the distance, @, to be appreciable, amounting easily to a reduetion of more than 30 per cent as compared with the value found by the first approximation, with The influence of the variation of the modulus, , from 50 to 200 Ib,/in., on the other hand, is not particularly large. In Case IT, that of a wheel lond at a point of the interior, complications arise due to the fact that the load is concentrated within a rather small area. ‘Tho theory of clasticity offers two types of theory of slabs: (025h 05h O75h h 15h 2h ‘io, elation butween th tru radu, a the equialst rai, 8 and the “husks and to be, approximately, eee G) Division by the scetion modulus per unit of width, > leads to the corresponding greatest tensile stress One that may be called the “ordinary theory of slabs,” and the other, the ‘special theory.”” ‘Tho difference may be explained by an analogy with beams. In analysis of beams it is assumed ordinarily that a plane cross section remains plane and perpendicular to the noutral surface during the bending. or beans of ordinary proportions, this assumption leads to satise factory’ results, unless onesis coneemed with the local stresses in the’ immediate neighborhood of a. ef rated load. In the latter ease the assumption of the plane cross section must be abandoned, and a special theory, which takes into aceount the deform tions due to the vertical stresses, is required. In the ordinary theory of slabs it is assuined, correspondingly, OTR. ary a, 4-Deections produ by a concerted iad wich a plat othe intra ats considerable datesPUBLIC ROADS Fi. £-Tuopeati banlag nomen, 4, and ad bending mom bp coocntted ad whic ‘etlava pant ofthe atonor ts eomiderble tan Wom the ees Yn etisalat e odin Pie 4 erate ba gore of ‘The ‘eteioas are food by superpaition of two dlgrate ofthe Ma ahah FpApril, 1926 PUBLIC ROADS ‘io. 7.—Deions poduced by to eau ois ike the Tod in Fis 4 separated by stance of Vie. 8—Detaetins produced bya cancetratd oadPUBLIC ROADS f S00 fa ays aaaiaaaaae ee ee 2 oF THE MOMENTS ALONG TH Oc all oe E 3 ori 1 Tt Tala] al | 1 Sj ia £ 2 TUT al 14 5 : ze 3 z | Ht oe on B I u L{T ee 7 { 3 | i mi f ee | | ra jul q I mia a a Ep freee tend DISTANCES MEASURED ALONG THE £O3E ‘ic. 10—Sonding moments slong de eles fr a eed concentrated ta plat ofthe eg on diagram) ‘Uietbuted unforaiy over neo hac leather a he ede Gower ae Serna aApril, 1026 PUBLIC ROADS 31 a { | i. aa yet a 2 Ea / Fl. ‘| ea elall = Pty i =| i i és oO | a | Cpl nat fend deme Fro, 1--Rending mowers log et af Figure 1, bl for on 0.5 that a straight line drawn through the slab perpendicu | With slabs of proportion as found in pavements, the Jar to the slab remains straight and perpendicular to theory based on these assumptions leads to a satis- the neutral surface. factory determination of stresses at all points except ; in the immediate neighborhood of a concentrated load, Tanux, 2-—Siresses in pounds per square, ingh computed from andl leads to a satisfactory determination of the deflec- ee ed oan Len ea tiara tall ‘AL the point of application of a concentrated this ordinary theory leads toa 10009 prams, F3.90,00 pounds pram inc, n= .35 peak in the diagrams of bending moments, with in- =i finite values at the point of the load itsolf (as indicated crue, (Mga Sessa a in figs. 5, 10, and 11). When the force is applied at Temi ade | the top of the slab, the tensile stresses at the bottom SDD rfl | ang | Shes! aittes | oiSSes are not, in fact, infinite. One may say then that the 2 effect of the thickness of the slab is equivalent to a ig off of the peak in the diagrams of moments. lor to find out to what extent the dingrams are uekes | cagins “anime! Tog” gee ah | te “ae ee Tek | around a concentrated load. = os & | jd off, it is necessary to abandon the assumption 1% il straight lines drawn through the slab remaining: @ sy ight, as applying to the immediate neighborhood «| 3 | s of the lond, and a special theory is required. This Bo te ie special theory rests on only two assumptions: One *| 2| & a 35 that Hooke's law applies, Ue constants being the RB) | a modulus of elasticity, £, and Poisson's ratio, a; th 3| | is other is that the material keops its geometrical’ con 1 3 B is tinuity at all points. As in the ease of beams, the = i ordinary theory is much simpler than the special 2) | me ie theory, and is used, therefore, except in particular gs 2 ceases like the present ‘one, which’ deals with local effectsPUBLIC It is expedient to express the results of the special theory in terms of the ordinary theory in the following, manner: Let the load, P, be distributed uniformly over the ayea of the small circle with radius a. The tensile stress produced by this load at the botiom of tho slib under the center of the circle is denoted by ‘This stress is the critical stress excopt when oe cthe radius, a, is so small that some of the vertical stresses near’ the top become more important; the latter exception need not be considered, however, in nse of a Wi tire, heel lond which is applied through a rubber By use of the ordinary theory one may find the at the same plaée by assuming the load to be distributed over the area of a circle with the same center, but with the radius 6. One finds that this equivalent radius, 6, can bo expressed with satisinctory approximation in’ terms of the true radius, a, and tlie thickness, h, only. In onder to find the relation between %, a, and 8, numerical computations were made in accordance with an anslysis which is due to A. N&dai? The center of the lond P is assumed for the time being to be at the centcr of a circular slab, ‘The slab is supported at the edge in such a manner that the sum of the radial and tangential bending moments is zero at every point of the edge. Compiitations according to Nédai’s analysis, with the radius of the slab equal to 5h, gave ich are represented in Figure 2 in the f equivalent distribution” and in with coordinates « and B. Ap- proximately the same cones and the same curve are obtained for other radii of the slab; and the results may be applied generally to slabs of proportions such as are fount! in conerete pavements, with any kind of support which is not concentrated within’ a small aren. close to the load. that when @ inereases gradually 0, bis at first larger than a; but when @ passes jain’ limit, & becomes smaller than a. For the and the ordinary theory of slabs, accordingly, gives nearly the same results as the special theory: ‘The curve in Figure 3 is fouad to lie close to a hype bola, the equation of which may be written in the following form, which is suitable for numerical com- putations, and which may be used for values of a less than 1.724%: be Va? $F 0.67 5h. anal) For larger values of a, one may use =a, that is, the ordinary theory may be used without, corrections. By the ordinary theory one finds the following ap- proximate expression for the critical stress: oP) Pug, L0.018)... nh CO) With 252,000,900 pounits por square inch and yo 0.15, and with’ substituted from equation (1), this formula takes the form: 3162 P (logy, (08) —4 logs @ — logy. k+ 6.478]. -- (10) 1, NA, ie tieuabeeroomgs veFhaia awh eoeara? hese ae mere ea Piston Cie Hs ps oA ‘The correction to be made in this formula in order to make it agree with the special theory is merely to replace the true radius, a, by the equivalent, radius, b. ‘Thus one finds the following formula, which replaces equation (10) when a is less than 1.724: 3102 f logy (Ht) —4 losy(Y1.Ga? +H =0.675h) —logys k-+6.478].- ------ (11) —log, ‘The stresses given in Table 3 have been computed in accordance with this formula for = 10,000 pounds. Like Table 2, this table shows the influence of three variables: the thickness, A; the modulus of subgrade reaction, k, anda. In Table 3, as in Table 2, one may notice the relatively greater influence of the variation of as compared with the infhience of the variation off. In dealing with Case III, that of 2 wheel load at the edgo, it was assumed that an equivalent radius, 5, may be introduced in the place of the true radius, a, in the same manner as in the preceding ense, and by the samo formula, that of equation (8). This assumption may be justified on the ground of the similarity in the two cases im the distribution of the energy duo to vertical shenr- ing stresses. By introducing the equivalent radius, B, in the place of @ in the formula for ‘the tensile stress, @,, along the bottom of the edge under the center of the circle, as obtained by the ordinary theory, one finds the following expression which, like the analogous equation (11), is based on Z=8,000,000 pounds per square inch and y= 0.15: 0.572 F, lot (h8) ~4 loss (1.6072 ~0.675 h) —log,, k + 5.767] Stresses computed according to this formula are given in Table 4, again for P=10,000 pounds. The influence of the three variables, h, k, and a, is shown in the same manner as in the two preceding tables, and is seon to be of the same nature, the variation of « being of greater importance than that of F (12) TapLe 3.—Stresses in pounds per square, inch computed from ‘oquation (I1) for loud condition az in Case II, Figure 1, for Afferent values of hy &, and a 209009 pounds pr square neh, ws P1000 pounds, B= se tee! ane btsab irae re eat Se ies nag | sang | EE Beet eae | eae | ae | | # : = 8] ! 2 | iw | =| 2| 2 | 2 3 aApri, 1926, ‘Tanum 4—Stresses in pounds per ‘equation (12), for load condition asin ent rabies of h, &, and a square inch compulel from P1000 pounis, £=3.00000 nds Per square inch, w= 018 Modstoe Stes stab ‘Thieepes “ofaabs is eaab aS ere ton ger | per | De ger Teer mo eRe ae 2 eee 8 2 3 | 3 = #8 2 =e z #8 8 g gi 2 g Boe g »! 3 a) 30: > Zone | ul & B| 3) | Bl 3 #\ 2) 2 a: sie! ahs BALANCED DESIGNS TESTED RY USE OF TABLES From the three tables, for Cases I, II, and TIT, one may obtain suggestions on the question of balanced design, Consider, for example, a pavement with the thickness of 7 inches in the interior portion, and 9 inches at the edges. It may be assumed for the time being that the cuter portions behav with uniform thickness of 9 inches. With the thick- ness diminishing slowly toward the interior, the stresses @,and ¢, would be somewhat larger than with constant. thickness of 9 inches, but tho correction needed for this reason is probably only small. For the time bein only the one wheel load which is considered in each o the three tables will be taken into account. ‘Tho influence of other wheel loads acting on the seme panel, but at some distance, will be considered later; in any ease it is found to berelatively small. With P= 10,000 pounds, k= 50b,fin#, and a=4 inches, the three tables ive the following values: g7¢= 262 Ib, per sq. in., o4=319 1b. per sq. in, 4¢—312 1b, per sq. in In comparing these stresses, their different charac- tors should be considered. ‘The stress, 2, at the corner acts presumably throughout the width of a whole cross section, whereas g; and @, are localized within smaller regions. With equal tendeney to rupture at the three places, 2. then, should be, probably, somewhat smaller than oan ee. ‘The stress, gg. produced under the influence of @ load which is distributed over an area only one-half of that assumed for ¢ situation represented by the smaller area may” occur when a wheel moves in over the edge of the pavement, itis reasonable, for the purpose of a comparative study of the tondeney to rupture, to assume a larger radius of the semi-circle at the edge than for the full circle in tho interior portion. With a=6 inches, for example, at the edge, one finds the stress a¢=276 Ib. per, In comparing this stress with of it should be observed that a, represents a state of equal stresses in all hori- 3992426} —2 q. in. PUBLIC ROADS ee ITT, Figure, for digfer as a large slab a Although the # 33, zontal directions at the point, whereas ¢, is a onc- directional stress. ‘The elongations per unit of length ols) are in the two eases 1 and $f Tt appears to be reasonable, therefore, for the purpose of comparison, to replace 2, by an equivalent oue-directional stress; if in this case the clongation is a direct measure of the tendency to rupture, this equivalent stress should be oi =9; (1a) =319 (1—0.15) =271 Ib, per sq. in. ‘The three values 262, 271,and 276 pounds per square inch point toward the conclusion that the assumed design is suitably. balanced. ‘The suggestion has been made already that one may determine suitable values of k by comparing the de- Aections found by tests of full-sized slabs with those given by the formulas. ‘The following formulas lend themselves to this purpose; they rofer to the three eases shown in Figure 1, and in cach case the load P is the only one act Case I) Equation (3) gives the dofleetion at the ze(1a-08s 9) fp == (13) Case Il, The deflection under the center of the load fers only lightly from the fllowing valve whieh is murate when @=0: _? A> skp (14) Case III. ‘The deflection at the point of application of a fost) force P at the edge is approximately equal to (15) that is, for P 2.70.33 -- (16) ‘Whe quantity KE occurring in each of these formulas may be expressed, according to equation (1), a5 mi Eek Bane When experimental values of the deflections are at hand, one may determine the corresponding vahies of by means of equations (13) to (16). ‘Then equation (17) gives the value of k as 120— PY 7) (as) Figures 4 to 11 are diagrams of deflections and mo- ments. The titles of these figures explain the nature of the diagrams. ‘The deflections and bending mo- ments have been computed by means of the, ordinary theory of slabs, ‘The diagrams, therefore, give inf mation concerning deflections in general, and concei ing bending moments except in the immediate neigh-34 bothood of the concentrated load which produces the bending moments, ‘The diagrams in Figure 4 and Figure 5 have been obtained by an analysis which rests essentially on that given by tho physicist Hertz" in 1881, DETERMINATION OF DEFLECTIONS DUE TO MORE THAN ONE ‘The diagrams in F' in Figures 4 and § may be use in the following way for the purpose of finding the resultant deflectione and stresses due to the combined influe of two or four wheel Jonds, ench acting ata considerable distance from the edges of the slab. Let ench load be 10,000 pounds and let the horizontal reetangular coordinates of the centers of the four loads be as follows: ser | SST , fee z sf o | in Loads 1 and 2 alone may represent the two rear wheels of a four-wheel truck, and the four lords com- bined may represent the four rear wheels of a si wheel truck With h=7 inches, FE. inch, 0.15, and 3,000,000 pounds per square 50 th.fin®,, one Buds hy equations able V ches; k= 66,200 pounds per inch; di and 1-8: 66" in.= 1.8137; distance | 1-4: ‘Then equation (14) as well as Figure 4 gives the following value of the deflection at point 1 aie to Lond No. 1: P __ 10,000 SEE SX 66,200 Furthermore, Figure 4 leads to the following value of the deflection at point 1 due to toad No. 2 alone: P._« pagoy 10,000 p= 0.03921 90 ‘Then, by superposition of the two deflections, one finds’ the deflection at point 1 due to the combined influence of the two rear wheels L and 22 (1) and (17),or by 1236.40. i 12 eam 0.0189 ineh, 242 0.03921 =0.0059 inch. 44,42) 4int244= 0.0248 inch, ‘The deflection nt point 1 due to lod Ni 0.0059 inch. . 3 alone is “pales, Sey, Oeste wine, ‘Seimieg Sab, fr esata, dice, leaded by a singh fer. A. Pep in hs Tech EFS eh dit Wate uae ai Ae eS GESeI careratogcheute wie, Met chil oes rae eer oe See Tne eo Tie art ere tees te mg aoe ot ee tin mn Se rae er cae ree meen, Poor nc eye SPR tia fies as Tema HOSE make raat uicreen sects teat "bsoinee te ote ata te wie ten oui elastiscber Untertage,” Feststhrilt eur Hundertjabeteser der Tech ‘nigchen Hochschule Karlsruhe, 1925. i PUBLIC ROADS Vol. 1, No.2 ‘The deficction at point-1 due to load No. 4 alone is, according to Figure 4 - 244=0.01620 Fe, =0.0024 inch. By superposition of the four deflections due to each separate load, one finds the resultant defleotion cue to the four loads: 0.0331 inch. Evers For the purpose of computing the state of stresses at the bottom of the slab under the center of lond No. 1, it will be assumed that load No. 1 is distributed unie formly over the area of a circle with radius a= 6 inches, "The stresses due to lond No. 1 will be the same in all directions, and they are, according to Table 3: 5 2y=279 pounds per square inch, According to Figure 5, lond No. 2 bending moment dees a radial “V,, in this ease in the direction of 2, equal to M,= —0.0211P = ~211 inch-pounds per inch (or —211 pounds), and a tangential bending moment Mz in this ease in’the direction of y, equal 10 M,=0.0181P =181 pounds. ‘The corresponding stresses are found by dividing these bending moments by the section modulus per unit of idth, that is, by { J?=8.167 in.t, ‘Thus one finds the stresses in the ditections of z andy: au oe 8107 16 pounds per square inch and pounds per square inch, ses are prineipal strosses, that is, one is the maximum, the other the minimum stress,’ and there are no shearing stresses in the directions of 2 and. y. For the ease of the four-wheel truck, one finds, then, by superposition, the following principal stresses due 18 thd tito rear wheels, louds Ne. Land No. 2, these principal stresses being in the direetions of and ys ge 219-26 =253 pounds per square inch, = 279-422 =301 pounds per square inch. DUP, TO SIK-WHREL TRUCK In the ease of the six-wheel truck the effects of loads No. 3 and No. 4 must he included. Load No. 8 con- tributes the same stresses at, point 1 as does load No. 2, only the indices x and are to be interchanged. Con sequently the resultant stresses in the directions of and y dite to the combined influence of loads 1, 2, and 8 become e,~0y=279— 26 +22=275 pounds per square inch Those stresses, again, are. principnl stresses, Since thoy are equal, the horizontal stresses will be the same in all directions, each stress being a principal stress.apr, 1920 Let 2’, y' be a new system of horizontal rectangular coordinates with the‘axis of x” along the diagonal line from point 1 to point 4. Load No. 4 produces a radi pending moment in the direetion of 2" and a tangential bending moment in the direction of y’. According to Figure 5 these bending moments are M, = —0.0186P = 186 pounds and My=0.0058P =58, pounds respectively. The corresponding stresses are found, again, by dividing the bending moments by the section modulus per unit of width, that is, b 81167 in2,and they are g2= —23 pounds per square inch and ey =7 pounds per square inch. ‘These stresses are principal stresses. ‘The resultant principal stresses due to all four loads combined, therefore, are in the directions of 2” and y/, and have the values B 52 pounds per square inch, oy 27547 282 pounds per square inch. One may draw the conelusion that the main part of the state of strasses at a given point is due to a whe lond right over the point. Tn the ease examined, the contribution due to the three additional rear wheels of the six-wheol truck is of less importance than that flue to the one additional rear wheel of the four-wheel true Figure 6 and Figure 7 show deflections due to two wheel-lo mbined, Each of these diagrams was obtained by superposition of two diagrams such as shown in Figure 4 Figures § to 11 show effects of loads at the edge, but ata considerable distance from any corner.” By virtue of Maxwell's thoorem of reciprocal deflec- tions, the deflection at a point B of any slab due to a load P at the point A is the same as the deflection at A duo to a load P at point B. Figures § and 9 may be interpreted, therefore, in a double ianner: First, as diagrams of deflections at any point B due to a load P at the particular point «1 at the edge; secondly, as influenco diagrams, showing the defloction at the particular point «A at the edge due to a toad P at any point. From this reciprocity of deflections one may draw a further conelusion which may be applied to Figures 8 and 9, and which concerns the curve of deflections, curve, which is obtained by intersection of the deflected middle surface by a vertical plane. Two lines Z4 and Lz are drawn parallel to two opposite parallel edges of a slab. ‘Two equal loads are con- sidered, one acting nt a pont A of the line Ls, the other acting at a point B of the lino Ls. The poin A and B are assumed to be sufficiently far from the nwo edges of the slab to perniit the assump- remai SiTig ey by on hme ps weit ay ted Ins Up sy Fa ea ea etme ge Seater Fadathae'e alae Re ie rik PUBLIC ROADS 1 satisfactory for a given ( : 35 tion of zero deformations at. these edges. Then one may conclude that the elastic curve produced along the line LZ» under the influence of the load P at A has exactly the same shape as the clastic curve produced along the line Le undor the influence of the load P at point B-” In applying this conclusion to Figure 8 or Figure 9, let the line Z4 be the edge shown on the drawing, and let the line Zs be at somie distance from the edge. By the direct use of the diagrains one obtains the elastic curve at any line Ly parallel to the edge, due to a load at the edge. But one may interpret this curve as the elastic curve for the edge produced under the influence of @ load at a point of the line Ly. The curvature of the dofleeted middle surface at point 1 of the edge in the direction of the edge, produced by the load P at any point B at some distance from the edge, is the same, accordingly, as the curvature of the deflected middle surface at point B in a direction parallel to the edge, as obtained in Figure 8 or Figure 9, due to the load P at the point A of the edge. ‘Thus Figures $ and 9 may be used in studying the stresses produced along the edge by a wheel load at some distanco from the edge ‘The following use of the tables and diagrams is suggested. Let it be assumed that « certain pave- ment, has’ been proved by tosis and experience to be of trallic. By the tables and diagrams one may compute, then, the corre- sponding eritical stresses. These’ stresses. may be adopted for the time being as allowable working stresses. With the stresses given, the tables and diagrams, through computations of the kind which has been’ shown, furnish answers to two questions: (1) What additional thicknesses are required if the wheel pressures are increased in a given manner; and, (2) what may he saved in the thieknesses by eliminating, some of the heaviest vehicles. Prof. TI. Agg has called attention to the importance of having an answer to the Intter question, when one attempts to apportion the cost of the pavement to the various kinds of traffic for which it is used, Tn using the tables and diagrams it should be kept n mind that the analysis is based on the assumptions which were stated at the begining of this discussion. By the nature of these assumptions certain influences were left out of consideration, especially the (1) Variations of temperature, and other causes tendency to change of volume;’ (2) the gradual dimin- ishing of tho thickness from the edge toward the interior; (@) local soft or hard spots in the subgrade; (4) horizontal components of the reaetions of the sub- grade; and (5) the dynamic effect, expressed in terms of the inertia of the pavement and subgrade, ‘The horizontal components of the reactions of the sub- grade, which are due to friction, may hme a strength- ening’ influence, especially at some distance from the edges, hy causing a dome action in the pavement As to the dynamic effects, with known values of the maximum pressure developed between the tire and the pavement, the effect of the inertia of the pavement may possibly be expressed approxiinately in terms of un increased value of the modulus, &. "These additional influences are suitable subjects for further analyTESTS OF VIBROLITHIC CONCRETE BY THE DIVISION OF TESTS, U. S. BUREAU OF PUBLIC ROADS potted by L. W. TELLER, Bogincer of Tes HIE Bureau of Public Roads, cooperating with the ‘American Vibrolithic Corporation has recently completed 28-day tests on concrete slabs con structed of the same aaterials in accordance with ordinary methods and the Vibrolithie process patented ‘by, the company. ‘The tests were made for the purpose of obtaining data on the relative strength of specimens equivalent ‘To that ond every effort was made to eliminate all variables except the methods of placing, tamping, and finishing, and to have all operations performed under Frat renriyeimilartworcinp condi tiaralee| pti bias Although it is recognized that the ultimate problem is one of economy, the current investigations have been confined to the study of certain physical properties of slabs made by the two methods: among them, the tensile strength as determined by bending, density, coefficients of expansion, and uniformity of the product. The program of tests includes, in addition to those whieh have been made at the end of 28 days, others that will be made after one year; and the formulation of definite conclusions must await the completion of the latter tosts, ‘There are certain indications, how- ever, which may be noted at this time with the under- standing that they may be modified by the data ob- tained from the remaining tests. Giving due consideration to all features of the investigation, such as the workability of the concrete and the method of finishing the normal specimens the following tentative conclusions may be drawn from the ending tests at the age of 28 days. 1. The YVibrolithie process resulted in 2 more uniform produet. 2, For a given ecment content, the slabs construc- ted by the Vibrolithic” method exhibited greater strength thon the normel concrete 3. The strength of the slabs constructed by the Vibrolithic process, when tested with tension in the bottom, was practically the same as when tested with tension in the top. ‘The strength of the normal concrete, when tested with tension in the bottom, was less than when tested with tension in the top. DIMENSIONS AND CONSTRUCTION OF THE TEST SLABS 30 by 72 inch slabs of each. class were con- structed of the same kinds of materials under condi- ions as noarly uniform as practicable. Designed to be 6 inches in depth, the individual slabs actually varied slightly from this dimension as indicated in Tables Vad 2 The slabs wera constructed on a specially prepared subgrade which was sprinkled and rammed thoroughly with a 20-pound tamper the day before plicing the conerete and sprinkled again the morning the slabs were poured. ‘The 2 by 6 inch oiled, dressed-lumber outside forms set on the subgrade thus propared in- closed a row of live slabs whieh wore formned by 2 b 4 inch separators, the beveled upper edges of ‘whieh were sct 214 inches below the top of the outside forms, as shown in Figure 1 The cement used was a brand of known reputation and predetermined satisfactory quality. All ¢ 36 very particular except the method of construction. 1 ©. PROUDIEY, Assstnt Baginer of Tests for a day's run was thoroughly mixed and. resacked in lots of 47 pounds each and one of these sacks ws tested each day in the laboratory with the results shown in Table 3 and Figure 12. Other physical properties of the cement were as follows: Specific gravit Eines, Telained of 200-meh seve. "The aggregates used 1h all dlabs—Potomnas River sand and limestone in two sizes obtained from Frederick, Md—had the following grading and physieal properties. (ORADING AND PHYSICAL PROPEIETIES OF THLE SAND Passing 14-inch sereen Teich stove 20. Loss by washing ‘Tensile strength rato <2"pounds => “ithmaterial coloration, julking with 4 t9 6 per cent moisture, approximately 20 per cent. ONADING AND PHYSICAL PROPERTIES OF THE STONE peice nite Percentage of wear Harcnass cootficlen ‘Tough Spoeite gravity. Aborptlon 2 Weight per eubie foot, So Crushed? ‘Stual stone. Tange stone Mined 50-5Vol 7, No.2, Apri, 1026, In order to obtain «unit volume of the mixture in equal parts of large and small stone it was found that 12% por cent additional volume of each was required; i. 6, a mixture of 134 eubie fect of each size yielded 2eubic feet. To compensate for the bulking of the sand and the oversize material in it, 134 cubie feet. of damp sand was used volumetrically as the equivalent of 1 cubic foot of dry sand. The small quantity of coarse sand psrasasred etc gravel size did not alfect the volume of stone in the proportions used. Figure 2 shows the grading curves of the aggrogates reel One bag of coment weighing 94 pounds was assumed to be equal to 1 cubie foot; and cubie-foot boxes, carefully marked in quarters, were used for measuring the sand and stone. ‘The volumes of materials act ually used in each batch were as follows: |acuesa) uszss vezeme vase We Mg ig a ae] my mm 3 By M4 Be 3 | gems stiatd Lease deaeae Me 1 u L 2 8% 2 3 3 pred Ba ORE el Ty q° kK E oll LH =° ry T ri VY { : Lyf DETR OF OFENMG=NEHES vo, 2—Orading of agrerates usd in the onerete Tho 10-cubie-foot, drum-type mixer, driven electri- cally at the rate of 20 r. p. m., was charged with the Proper quantitics of stone, sand, and cement in the 9924-2634 PUBLIC ROADS 37 order named, followed by sufficient clear Potomac River water to give the required consistency. ‘The water tank was equipped with a gage giass calibrated to quarters of gallons. As the eonerote was discharged from the mixer a flow determination was made on. a portion of each batch, and as nearly as possible, the consisteney was regulated so as to give a flow of 110 to 115 on the 30-inch flow table, Asa further check on. the consistency the second and sixth batches of every run were tested for slump with the approved slump cone, the requirement being a slump of about 2 5 3e—Pnising Vibro sustoe with steal out pir to bekng ‘The concrete, which was placed in the forms without regard for the Soparators, was tamped and struck off with a straight 2 by 12 inch strike board worked length- wise over each series of fivo slabs, and tho desired finish of the normal conerete was obtained by belting length- ‘wise over the slabs with an S-ineh rubber belt. ‘The Vibrolithie sections were similarly pliced and ruck off, after which they were covered u 1 Frederick limestone of 2-inch to I-inch si the following rates: MM Pounds por square yard On this stone wero placed special racks over which the vibrators were run, according to the patented process. Approximately four minutes of vibration was allowed for each slab. ‘The removal of the racks left an irregular mortar surface which was smoothed down with a long-handled steel float as shown in Figure 3, after which the surface was belted. Dry spots were sprinkled with a little water to facilitate finishing, All slabs, as soon as theit hardness would permit, were covered with wet burlap which was kept damp during the day and thoroughly wet down before loay— ing at night." The following morning it was removed and replaced with a covering of damp carth. ‘The arth was kept damp by daily’ sprinkling until 28 days had elapsed, at which time the top surface was cleaned. After the removal of the earth the slabs were kept damp by means of wet burlap until time for testing. Slabs to be tested at ono year will remain uncovered 10 months, but will again be covered with damp earth for the 28 days immediately prior to testing. Luge Lage ad babs38 “PUBLIC RoaDs ‘Tanne 1—Summary of data of tests on normal concrete slabs | tous oe i | areotvee | pon | ee c HE | cata mie oon | ee P i a ms filer ‘9 i) ae ED sy ee 7 as hao ty us Set qi Seas iq Het uy ma yl | pee q Hpi u uw u a Bt i fl ey de Rims! ; ‘Tyne 2—Summary of daa of tite on Vibralithie concrete slabs Section ‘Modubus or ruptare | ‘Strength ratio: Lenten Sapte Mic i Save intmion Deen MOH) aoa | | ec he 2 | arenes | a . = = le ae a inte | See | | | ee . = — 62) mo) 16255 om 4a] 8) 21) 2S\p ee oy #] ga] ze] By js] oe] EH FEW i i) By] eS) 2 Lh 60] iad) ins ie | ‘Lal | Gl} mas! isa% oe a aE HESIRA IL... | Fe P) Siee lee ep Sl gp Rg) gto ,| on| 3 8) 2) te) & al Gs HS Ee aD ee lee eed a Oy eileen wv 63) Hea an 8 1a oy ee) Sle ih o) 2) ce), 2) eG | oP get iv | i] Hi) tel] uy ol gel aes Sie i ae 7 Hl a: 22) Si u S) ga] ge wl on rot 86 | ahi] BST el u Oo} Sales om) eee Se ee fH) es) 8 H Mean. a aApril, 1926, ‘THE METHOD OF TESTING When the slabs were ready to be tested they were lifted from the subgrade by means of a cradle of the special design shown in Figure 4. n_nearly every caso, cracks had formed aver the separators there was no difficulty in raising the individual slat most instaiees a large amount of subgrade with the slabs and had to be removed. Half of the slabs were tested with tension in the top surface, the others with tension inthe bottom, the former condition being effected by inverting the slab in tho testing machine, In ordér to obtain n solid bearing for the knife edges on the under surface of the slabs, flat steel strips [24 inches wide and three-six- teenths inch. thiek were set in plaster of Paris in the proper position. ‘These strips were not needed on the Finished surface, ay the rubber pads on the ki shown in Figures 5 and 6, took up the slight tics which, in general, were merely the oat or belt and were not more than on regular irks of the ixtecnth inch tee Heferring to Figure 7, which shows the elevation of the testing machine, the method of testing may be described as follows: The slab, A, carefully te tered on the lower knife edge, B, and the rocking ro, 4—Liftn sa from sree edge, ( (see also fig. 5). ‘The upper knife edges, D, mounted with the hydraulic jack, E, and the ¢ali- brated head, KC, on the carrier plate, @, was swung into position over the slab so that fonding twas at the third points of the slab. The collar, Z, kept the carrier PUBLIC ROADS 39 plate elevated so that the knife edges would swing clear of the slab until in position, at which time the screw jnck, J, was used tol ost, I, to which the collar was fastened. ‘The collar was lowered until the carrier plate was entirely free. ‘The follower h carried by the plate, If, was then swung about post, 77, and brought into contact with the top of the machine by means of hand serew, N. ‘The initial load Was then put on the slab hy means of the jack handle Fru. &. Laer ile edge showing rocks pi snd aber pa after which the fond was con= timiously and steadily applied by means of the small auxiliary pump, F. 'Phe calibrated beams, Figure 8, were arranged to deflect as simple beams as the load was applied. ‘The dial, read initially before the load s apphed, was watched carefully’ by at least two ves so as to catch the maximum dial reading, and a calibration chart was used for converting th difference in dial readings to total applied lond in pounds. pete broken sections were removed and stacked with their broken faces outward, as shown in Figu 1, and their cross sections were then meastired by two operators each of whom made four depth measure- ments on exch slab. of hydrautie jack, TATION OF MODE ‘The weight of the slab itself, the weight of the upper knife edges and loading device, and the pressure applied by the jack comprised the total load. ‘The uniformly distributed weight of the slab was, converted into a concentrated load which could be added to the weight on the knife edges, ‘Thus, the weight of the 1:24: norma eonerete slabs, 152 pounds per cubic foot, was found W be equivalent to two loads of 410 pounds each concentrated at the knife edges; and the 1:114:3, vibrolithie slabs which weighed 157 1 Toot were equivalent to two one 1 pounds. ‘The dead load of the knife evges, jnck, ‘was found to be 640 pounds, or at each knife edge, wands. From an initial Inad computed in this manner the pressure on the slabs was gradually increased by Ieans of the hydraulic jack, E (lig. 7), until rupture ocurred, at which time the pressure indicated by the calibrated beams added to the initial dead load con stituted the total load, P.40 The extreme fiber tensile stress, §, was then com- puted from the formula —i 1 ¢ S= My in which the bending mo- ment. for the 60-inch span with third-point loading is P yn M = x20 10P. The computed from the me ments of the broken sections, was found to be different in almost every specimen, prin- cipally beesuse of variations in the thickness of the slabs and in the location and num- Der of the separator proje tions on the edges. The projections, or sections of the slabs formed over the intermediate separating fornis, were trapezoidal in cross section, approximately seven cighths inch in width and from 24% to 3 inches in depth. The thicknesses of the several slabs, their section moduli, the (otal loads at rupture, and the computed moduli’ of rupture are shown forthe two classes of cone Tables 1 and 2, respectively. The moduli of rupture are shown also in Figure 12 and ‘Table 3, with other factors indicative of the characteristics of the conercte, These other factors include the tensile strength of 1:3 mortar briquettes made of the cement used in the slabs and Ottawa sand; the modulus of rupture of 1.2 Ottawa sand mortar beams; the consistency of the concrete is indicated by its flow; the watcr-cement ratio; and the ratio of the coment to aggregate by weight for each of the slabs. The water-cement ratio is a volumetric relation and assumes 1 bag of cement to equal 1 cubic foot. Thus, if a batch of concrete were mixed with 74 gallons of water for each bag of cement, the water-cement ratio, astm mothe, & @ 4 PUBLIC ROADS T, would be recorded as 1.00. In computing the re- lation the as found quantity of moisture contained in the sand, by a daily moisture determination, was Evevation 10,8 —Calibrated bas td to sas hold added to the volume of water introduced in mixing to obtain the total value of WV, ‘The cement-aggregate ratio is based on the nominal mix for both Vibrolithic and normal concrete. ‘The top stone of the Vibrolithie conerete is not included ic. 7—Ajyparats fe tating she sabeApril, 1926 PUBLIC with the aggregate in computing the relation. Using the weights per cubie foot of the sand and stone %3 determined by tests, the volumetric proportions were readily converted to weights; and the Weight of cement divided by the total weight of sand and stone gives the ratio, CEMENT-ACCREGATE AND WATER-CEMIENT RATIOS AS STRENGTH Past experience naturally leads to the expectation that the strength of conerete will follow, in general, the cement content. ‘Table 3 and Figures'9 and 12 show that the expected relationship obtained in these tests, the cement content being expressed as the ratio of coment to total aggregate in the mix. "The water-cement theory is also seen to agreo with the strongths of the slabs. Tho leaner mixes were gauged with higher watercoment ratios and con: quently gave lower strength. It is seen, however, that there is practically no difference between the average strongth of the 1:2:3 and the 1:2:34 normal concrete, although the other date do not show the reason for this. Similarly, there is no diforenco between the average strength of the 1:2:4 and the 1:2:414 Vibrolithic con- In this case, however, the crete, is the same for exch of these proportions. ‘This probably accounts, in part at least, for the uniformity in strength. I is also interesting to note in this connection that the quality of the cement, as indicated by the tension tests of mortar, was lower for that used in the 1:2:4 mix than in the 1:2:434 mix, whereas the quality as indi- cated by the eross-bending test of 1:2 Ottawa sand mortar beams is the reverse. It is probable that a, more thorough examination of the slabs as called for by the program of tests will furnish information that will explain some of these deviations. ‘The relation between the quantity of cement used and the strength obtained is shown in Figure 9. For equivalent quantities of cement the strength of the Vibrolithie concrete is higher than the normal conerete. ‘The difference smounts to about 16 per cont for 1:2:3 and 1:134:334 conerete, and about 21 per cont for 1:2:4 eonereto based on the strength of the normal concrete. Considering the two processes from the standpoint of the cement required to give equal strength, a com- Parison at 000 pounds per square inch todulus of Tupture shows that roughly, 19 per cent additional cement is required in normal concrete; or, in othor words, only 84 per cent of the cement required for normal concrete is necessary for equal strength in Vibrolithic work. “At higher strengths the curves (Gg. 9) indicate slighily greater advantages for the Vibro- lithic, and at lower strengths less advantage insofar es saving in cement is concerned. Tante 3—Reeul oe ‘Slab numbers tncusee. eso es Normal eononte J oaenga | nara | ease tao oo asso |e | RoaDs 41 OP AND BOTTOM STRENGTH OF VIBROLITHIC SLAMS NEAL ava c Due to difference in density and, possibly, to other variables as yet undetermincd, the strength obtained when the bottom of the slab was in tension was, in most cases, lower than when the top was tested in tension. ‘The average amount of this. difference. is shown graphically in Figure 12, the shaded bars giving the strength with the top in tension and tho solid bars the strength with the bottom in tension, ‘This is also shown in Tables 1 and 2 we J TI eon ieee eee eat : Bo Ie H =| 500 1 RATIO OF CEMENT TO AGGREGATE - BY WEIGHT Yo. Retain of woah of rapture of normal ant Vitalie coe to It has been suggested that, as a result of excessive tamping and finishing, the strength of the upper sur face of a concrete road might be increased beyond that of the bottom, These tests do not substantiate this, theory. ‘There is no doubt that the Vibrolithie method is the more vigorous finishing treatment; nevertheless, insofar as the uniformity of strength in top and bottom of the slab is concerned, the Vibrolithic is more remark- able than the normal conerete. Expressed numerically, tho average resistance to tension in the bottom of normal concrete slabs is 87.7 per cent of that in the top, and for Vibrolithie it is 98.0 per cent. of teats of slabs, oggregate, aud coment per) Modfenaan ope eo mw ome) we om fue BLM Gm ee eos Boe Sm ee Be va eile aioe abtlon apo! abo | agan | ot Bee ee ee BE BE Ee) Sl Begs 88 8 2 8 sg) 8 842 PUBLIC ROADS Vol 7, No.2PUBLIC Roaps 43 oral coneresihs of various mies showing eros ations a pint of aus i ast44 PUBLIO ROADS Vol. 1, No.2 ls 00 TENSILE PoO STRENGTH leoo #3 MORTAR a BRIQUETTES 100 LBS. PER SIN. 0 MODULUS 600 oF doo RUPTURE t 1:2 MORTAR 200 LBS. PER $Q.IN. "° CONSISTENCY | 10 FLOW TABLE rE Pee cENT WATER CEMENT RATIO L20 00 40 #9 RATIO CEMENT 1p TO AGGREGATE 8y WEIGHT, OF RUPTURE ° OF CONCRETE LBS. PER SQ.IN. sersion irr = TENSION Ww eoTTOM= Bh 8 126] 136] 146] 156 17131 [1417 151 | SLAB NUMBERS 130140 | 150 | 160 125 | 135 | 149 155] INCLUSIVE $153 | 1-2-3 | 1-2-33 | 1-2-4 (-2-4 1-2-4} | MIX NORMAL VIBROLITHIC FINISHApt, 1925 Another important consideration not to be ov looked is that of uniformity of the product, It may be seen in the last column of Tables 1 and 2 that the average variation of the individual specimens from the average for the five specimens of the group is a He loss in tho caso of the Vibrolithic than im the normal conerete. ‘The average of all of these varia- tions for Vibrolithie specimens is 5.3 per cent, and for the norinal concrete specimens 6.8 per cent. ‘ho photographs shown in Figures 10 and 11, furnish an explanation of some of the apparont abnormalities in the resulis given in Tables 1 and 2. Ltis reasonable to supposo that, other things being equal, the slab with the minimum void spaces will give groatest strength; also, that the closer voids are to the outer surfaces, tho ‘more serious will be the effect on tho strength ‘The distribution of void spaces in the specimens under consideration may be seen by a carcful inspection of the photographs. Tho 1:13:33 Vibrolithic specimens, Figure 10, show very few Toids. Slab 94, which contains more than any other of this mix and ‘type, shows one group of voids near tho bottom and another a little belew the center; but their effect is apparently negligible. ‘The 1:9:3% specimens of Vibrolithic concrete are also quite dense except for « thin layer of voids oceur- ting in slab 162 about one inch from the bottom, which seems to have had no sotious effect on the strength. ‘Tho specimens of 1:2:4 Vibrolithie concrete exhibit greater percentage of air pockets duc, no doubt, to tho loaner mix. It is interesting to observe the posi- tion of the voids in slabs 124, 131, 135, and especially in 121 and 122, whore they arc quite close to the neutral axis. ‘Tho 1:2: 414 mix of tho Vibrolithie series apparently ineloses a still greater number of voids, particularly in slabs 142, 143, 144, 151, 153, and 155, ‘In 142 and 143 they are principally in the compression side as tested and probably havo little effect except as thoy tend to lower the neutral axis and thus shorten the distanee to the oxtrome fibor. In slab 144 a bad condition exists near the bottom, which is reflected in the unusually Jow stronath obtained. The same is also true of slab 153. Slnh 155 shows considerable void space, at an average distance of about 135 inches from the bottom. ‘The richest mix in the normal concrete slabs, shown in Figure 11, is 1: 134:3. The most noticeable voids in slabs of this mix occur in Nos. 87, $8, $9, 97, 98, and 100. Of these slabs, $8 and 97 wére effected only in the ‘compression side, but the other specimens men- tioned gre ponctrated varying distances up to 234 inches in the tension side, These conditions, no doubt, explain some of the wmusually low strengths ‘obtained in this group. , The next Jenner mix, 1:2:3, also shows many voids extending from the bottom well into the slab.” Slabs 116, 120, 168, and 170 are badly honeyeombed in Uh PUBLIC ROADS 45 tension sido, and an 117 and_119 the voids extend so deeply into the compression side as to catise a decrease in effective depth and consequent decreased strength whea hased on the slab thickness as sneastred In the 1:2:314 mix of normal concrete, slabs 126, 130, 137, and 140 are most notable for their honey combing, although every slab of this group includes an appreciable quantity of voids. The 1:2:4 normal conerete is in particularly poor condition due, very likely, to the leanness and dryness of the mix. ‘Th consistency as measured by the flow table was no le workable, however, than mueh of the conerete used in present-day construction of first-class eonerete pave ment on which a machine finishing is used. Special consideration should be given the method of constructing the normal concrete slabs for this investi- gation, as it is plainly upon this factor’ th of these tests depend ‘The concrete for the normal specimens was dumped into the forms, shoveled into place, and spaded along the edges of the forms. The specimens were tamped across once with the 2 by 12 ineh strike hoard, after which the surface was struck off and belted to a finish. s the consistency was held to the same value used for the Vibrolithie specimens, it was more nearly that which should be used for machine finishing, hand finishing deseribed was not adequate. not apparent until the specimens were turned over at the time of testing, but the result was an unfortunate amount of honeycombing and consequent variation in tho test results. It is believed that, if the concrete in the normal speci- mons had been more thoroughly compacted as would be the ease in machine finishing, less variation in strength would have resulted and generally higher values would have been obtained. ‘This is indicated by the fact that those specimens where the laek of compaction was most apparent showed noticeably lower resistance to cross~ bending. “An intoresting general observation regarding the loca- tion of the void spaces in the specimens tlus far tested is to be made from an inspection of the photographs and much better, of course, by a study of the actual specimens, Through the action of the vibrator on the surface of the concrete in the Vibrolithie process, mortar is apparently worked down to the subgrade, thus leaving the air spaces at some distance from the bottom of the slab. When the ordinary methods of finishing are used the voids which occur at the bottom while the concrete is being deposited remain unfilled as a result of the arch action of the conse aggre~ gate, Advantages of conerete made by the Vibrolithie method, as brought out in this report, may be traced in many instanees to the more favorable location of voids tthe valueTEMPERATURE AS A FACTOR IN THE STABILITY OF ASPHALTIC PAVEMENTS BY THE DIVISION OF ‘TESTS, U. S. BUREAU OF PUBLIC ROADS Rept by WJ EMMONS (Geieas asphaltic paying mixtures, notably those containing vessively high percentages of bitu- men for the aceoinpanying aggregates, are susceptible to displacement under trafic, Such weal ness is generally manifested during the warmer seasons of the year when the bituanen, present in space-filling volumes of considerable magnitude, absorbs sufficient radiant energy from the sun to render it fluid. ‘The Fic, 1-—Geoorl, vw of expurimental Lack surface with asplnltie payments shiwtig ara empertine essreiens satow ha eerste wie pavement then exists as a structure composed of a more or less independently stable aggregate and a viscous fluid of low binding power. When, under these eondi- tions, the pavement is subjected to traffic of sufficien magnitude the resisting power of the plastic mass is soon overcome and the development of ruts and waves ‘cents. Du past thi ¢ Bureau of Public Ro: has conducted a se ests Lo measure the intcrnnl temperatures of asphaltic pavements and to determine in'a general way the effect of such tempera tures upon the resistance to displacement und ‘These tests were condueted in conneetion with two scries of stability experiments which involved the construction of 60 sections of pavement on a circular track or roadway 13 feet in width. ‘The mixtures included coarse-graded asphaltic coneretes and sheet asphalts and were laid 2 inches thick upon a very smooth conerete foundation. For the purpose of determining their relative dis- placement under traffic, screws were driven into the pavement across the line of traffie and referenced to per- Innuent markers set in the conerete. Thermocouples of copper and constantin wire were installed in many of the pavements and connected to a eentral station for the observation of temperatures. Most of the thermocouples were set in the pavements at a depth of one-half inch below the surface, although in several in- stances additional ones were placed at half-ineh intervals down to the conerete foundation. Figure 1 shows the general location of the experimental pavements, the cen- 46 wey Resech Speci a. A. ANDERTON, Chane Eset tral temperature measurement station and the thermo- couple wires leading to it. Observations of tempera~ tures were made at 10 a. m. and 2 p. m, each day when the track was subjected to traffic. It was found that an average of the air temperatures at these times vory closely approximated the average air temperature for the daily period of operation of the test. Moreover, the maximum air temperature during the day was recorded at approximately 2 p. m. The finst seties of stability tests was made on 27 different coarse-graded asphaltic concrete mixtures. Construction was completed Inte in the year and the operation of a loaded -ton truek over the test sections was begun in October. During the autumn, winter, and early epring poner the erry air temperature at 2p. m. was below 65° F’. and all of the pavement mixtures remained rigid. Late in the following May a sudden and decided rise in temperature took place, maintaining for the following four weeks a 2 o'clock average of about 80° F. ‘The effect of this increased temperature was im- mediately evident. Although many mixtures re- mained entirely stable, a number began to shove and rut under the continuing traffic, ‘This series of mixtures was subjected to 50,000 passages of the truck, of which 60 por cont was imposed during the period from October to May, without approciable effect upon the contour of the pavement, the remaining 40 per cent during the season of high temperatures served virtually to destroy several of the weaker sections. DISPLACEMENT GREAT DURING SUMMER SEASON ‘Twenty-vight sheet asphaft and five asphaltic eon- crete mixtures were tested in the second series. ‘The sheet ‘asphults were laid directly upon the smooth (a-ng vewreeniune] Byte # ay + ania “a 4 a ( ne ‘ner aie aa seh ic. 2—Digplucement and texoperatueof experimental asphltle pavementsVol 7, Ne. 2 April, 1026, eonerate foundation without tho customary intermedi- ato binder course. In this test, traffie was begun during the last few days of August and differences in the stability of the several mixtures were immodiately developed.” As the prevailing temperatures decreased the effeot of traffie boeame less and less marked until, during the month precoding November 11, virtually ue += OBSERVATION AT JOA. x= OBSERVATION AT 2 P.M, FA 139} = PUBLIC Koaps 47 between, which measurements of movement were taken, ‘The movement of cach seotion is determined by averaging the total forward shove of two lines of 25 serows each eros the 13-foot rondway. ‘The points plotted on the eurve are tho averages of the 83 scetional movements. The high degree of plasticity of the pavements during the initial trathie interval ending September 2 may be significant. ‘This period game not only in extremely hot weather, but followed immediately after construction, and ii is probable that. the apparent instability ‘may be nseribed in considerable degree to the faet that the sections had not attained the ultimate compression to which they. wore susceptible under, traffic. In substantiation of this, theory, attention is dirocied to the fact, that virtually equivalent temperatures during the following year resulted in far less displaecment, Furthermore, in a considerable number of the mixtures at least 50 per eent of the total movement recorded during the entire {est oeeurred in this short initial period. By the latter part of the summer of 1925, five sectio whieh carried exeessive amounts of bitumen had de- formed very badly. In tho traveled areas their original internal structure was entirely disrupted and their resistance to displacement at high temperatures was aveneNt TEMPERATURE. (OEGREES £) Cp 30 7 830100 AIR’ TEMPERATURE. (DEGREES F) Pio, 8—Rlaton beeen alr and sbeot asphalt pavement temperature. Ob ‘atasins taken os al inch babe toe cue To no displacement. was apparent. The average daily air temperature for the last period was 66°F. Traffic was suspended during the winter months, but was resumed early in the following May and continued until October 15. During this interval a eomplete eyele of temperature influences was observed, Stability of all mixtures at the low spring temperatures, shoving freee re 4Resrd of exretponlngtumpaaires of and abut ata pavement and rutting of weaker mixtures during the summer months, and oneo more an inercased rigidity as eooler weather began to prevail Figure 2 indieates the average movement por 1,000 passages of the truck of all 33 seotions of the second series with respect to the average daily air and pave- ment temperatures for the period or *Cralfie interval tc, Hourly seo of teraalvesngeratue of est asphalt pavement virtually dostroyed. The condition of these pave- ments is reflected in the rather sharp rise in the eurve of Figure 2 for the period beginning August 3 It should be recognized that this ehart docs not pretend to measure the susceptibility of asphaltic pavements as a class to movement at the indieated Temperatures. In these tests many mixtures remained stable at all temperatures whereas others, several of which were obviously of poor design, deformed very badly. ‘Thus the plotted movements are representit- tivo only of a group of arbitrarily chosen mixtures, PAVEMENT TEMPERATURES HIGHER THAN AIR TENPRRATURES When exposed to direet sunlight the internal tem- ‘ture of asphaltie pavements is probably always than that of the surrounding air. This diffe ence is very likely slight during the winter month: but under summer conditions the bituminous imixture48 accumulates heat rapidly and its temperature rises faster than that of the air during the period of the day when it is subjected to direct action of the sun’s rays. In these tests internal pavement temperatures 25 to 35°F. higher than the air temperatures were of eom- mon occurrence. ‘The highest pavement temperature recoriled was 140° and this was reached only a very few times, Figure 3 shows the relation between ait and pavement temperatures, the latter obtained from the thermocouples placed one-half inch beneath the surface. The 10 o'clock and 2 o'clock readings are re- corded separately. All plotted points represent. data obtained on clear days, although winds of varying intensity and direction’ often prevailed. It will be noted that, in general, the sanie air temperature re- sulted in higher afternoon pavement temperatures indicating that gains by absorption exceeded losses through conduction and radiation, Rain, clouds, or local shade result. in an imn decrease in the temperature of asphaltic pay ‘The effect of the last-named f noticeable in several of the mixtu z afternoons of the Inte summer months, were shacled by a high bank surmounted by trees just west of th test pavement. As the line of shade erossed the lo tion of the embedded thermocouples their temperatures abruptly decreased. In Figure 4 are plotted morning and afternoon observations of air and pavement tem= peratures, the latter being taken by means of the ther- mocouple placed one-half inch below the surface of the shect asplialt mixture. ‘The record extends over a period of approximately four wecks and. illustrates characteristic reactions of pavement temperature to changes in climatic conditions. There appears to be no extteme difference between the temperatures of asphaltic pavements at various depths heneath the surface of the customary 2-incl layer. While the pavement is subjected to direct sunlight the top was found always to be warmest. In these tests, however, differenees in temperature at the one-half-inch depth’and at the 2-ineh depth never exceeded 15°F. On cloudy days the pavement tem- peratures were virtually uniform throughout, while rain, furnishing a medium for rapid losses of heat from the ‘surface, freqnently ted in slightly higher temperatures at the Dotiom. ‘This latter condition seems also to pravail at night. Figure 5 shows a series of 26 consecutive hourly temperatnre measurements taken at four depths in a sheet asphalt pavement. Of particular interest in this figure are the uniform- ity of the temperature throughout the pavement at 8p. am. and 7 a. ny the warmer condition of the protected under layers during the hours of darkness and the quick reversal to normal daytime conditions 1s direct sunlight once more strikes the pavement the following morning. Tn common with all other structures, asphaltic pavements may be so designed that, they’ will prove successful under some conditions but fail under others, Obviously, under no traffic, all 60 of the test sections involved in these experiments would have remained stable regardless of elimatie conditions. ‘Test data also show this practieally to be the ease when traffic was imposed during periods of average air tempera~ odiate nents PUBLIC ROADS © number of cyl tnre below 70° F, Higher temperatures aided traffic in developing instability in certain mixtures, but with lighter traffie it is certain that several of these would have been classed with those which proved satis- factory, | Its elenr, therefor, that pavement behevior frequently explained on the basis of traffic only, should be analyzed also with regard to the prevailing climatic conditions at the time such traffic is imposed. CONCRETE TESTS TO BE MADE IN NEW JERSEY The New Jersey State Highway Department, in Ppoperaliones ii iol Deevnnc RoubLca i orate begun a series of concrete tests for the purpose of studying the relative concrete making properties of crushed stone and gravel used in conerete road con- struction in that State. ‘The tests are being made in tho State Highway Laboratory at Trenton, and in- yolve the fabrication and testing of about 250 concrete heams $ by 8 by 48 inches in size, as well as a large \lers for compression tests. ‘The program calls for three series of tests. In the first series the workability of the concrete is to be kept constant, as nearly as possible, by means of the flow test, and the relative yield and strength of the conerete determined for each of several gradations both of crushed stone and gravel, using concrete proportions as given in the current New Jersey Standard Speci- fications. ‘The object of this series is to determine the lative strength and yield of gravel concrete as com~ pared with crushed stone conerete for several sizes and gradations of coarse aggregate, In the second series an effort will be made to design conerete of a given strength by means of the water- cement ratio theory, for each type and gradation of coarse aggregate. ‘The procedure to be followed in this series is esentnlly as follows, ‘To euch gradation and type of coarse aggregate, fine aggresate will be added in ine following ratios by volume, 0 te 67, 36 to G4 and 40 to 60, To each of the above combinations water and cement in fixed ratio, depending on the strength de- sired, will be added until the desired workability has heon reached. ‘The end point in cach ease will be de- termined by means of the flow test, supplemented by the, Judgment of experienced eonereto operators. Con crete specimens will then be made up in the proportions as determined by the trial method referred to, and the comparative strength, which should be constant, the comparative yield, and the comparative absorption will be determined. In the third series of tests, specimens will be made in which the conerete mixture has been designed in ac- cordance with the fineness modulus theory as. given in “The Design and Control of Concrete Mixtures,” publication recently issued by the Portland Cement Association. The results obtained from this series will be used as a check second series. Assuming constant strength and a constant degree of workability, it is hoped to determine by means of these tests what grading of coarse aggregate and what proportions of fine to coarse will give the greatest. yield of conerote for both crushed stone and gravel. inst the results obtained in theMOTOR VEHICLE REGISTRATIONS, REVENUE, AND GASOLINE TAXES FOR THE YEAR 1925 For 1925 the motor vehicle statisties have been amplified somewhat and are published in two tables; the table on page 50 showing the number and classes of veh fered and the table on page 51 showing the corresponding receipts. ‘This table shows by States the motor vehicle regis- trations, the tax-exempt cars, dealers’ licenses issued, and operators’ and chauffeurs’ permi As far as possible all reregistrations are eli d and to avoid duplication of the recorded cars of other States all nonresident registrations are excluded. The grand total of registered motor ears and trucks has Been divided into tivo classes based on their use, either as passenger-carrying, or eommodity-carrying vehicles, the latter including tractors and ‘motor-driven road equipment. In the first column is the grand total of all registered nd taxed motor ears and trucks, whieh is the total of Jumns 2 and 3. Column 2 shows the passenger cars, cluding all privately owned passenger automobiles, taxis and passenger cars for hire, and busses. A few States (noted in column. 3) have included busses with motor trucks, but nearly all busses are included in column 2. Column 3 shows the commodity or freight carrying ears, which inelude all motor trucks, road tractors, and all motorized road vehicles, not primarily used for the transportation of passengers. “Tractors used for farm purposes have been excluded although these are registered and pay a fee in a few States Only 12 States record road tractors separately, amount ing in 1925 to 2,749 tractors. ‘The other States in- chide road tractors with trucks. Considerable interest, has been shown in the number of taxis and cars for hire and busses. In the absenee of sufficiently complete statisties from registration offices, data have been taken from other sources and shown under the heading “Special list of passenger ears for hire.” The list in column 4 taken from Internal Revenue Bureau records shows the number of taxis and passenger ears for hire having seating capacity of two to seven persons for which an occupational Federal tax of $10 was paid by the owners during the fiseal year ending June 80, 1925. A few States supplied this data as of December 31, and where available, it hhas been used. ‘The list of busses shown in column 5 is taken from the February issue of Bus, Transporta- tion,” and is only partial list. This data has not been used in connection with the official figures sum- marized in column 2, except in a few States, as noted, which record passenger ears for hire and busses sep- arately. 5 Registered trailers and motor eyeles for which fees are paid are shown, ‘The trailers in column 6 cover semitrajlers (two wheels) and trailers (four wheels). In general, trailers are hauled by road tractors and carry commodities or freight. There are some two- wheel trailers used with passenger ears for “eamp kits” but these re not segregated in the records. Moto cycles are listed in column 7 and include eyeles with or without side eats. ‘Tax-exempt official cars are shown for all States which record these ears and trucks. ‘The total number of United States cars and trucks given was obtained from the United States Budget Bureau. Under thi heading is also shown official motor eyeles, whieh arc recorded in less than half the States ‘The next to the last column shows the 1924 grand total with 1,696 tractors added to formerly published figures (in States noted) in order to he comparable with the 1925 grand total registered motor cars and trucks, ‘The first column of this table on page 51 shows the total gross receipts. The next eight columns show: various items which make up the gross receipts as segregated by 33 States and the Distriet of Columbia. ‘The States starred in the column of States are the ones which reported the complete details, and these States have been summarized to make a. subtotal ealled “Detailed total.” The details of registration receipts correspond with the number of veliicles in the table fon page 50 wherever reported. Under “ M receipis” are shown dealer's license fees, chauffeurs’ amd, operators’ permits and other iiscellaneous receipts, the latter covering many items such as reregistrati fees, nonresident registration fees, traflic fines (if these are included in the motor vehicle fund) cert of title, duplicate tags, etc The ‘collection and ‘administration expenses of the motor vehicle license offices are generally deducted Lefore final division is made allocating ecrtain shares to the State highways, to county or other local rural ronds to pay retirement and interest on State rond and bridge bonds, and for other purposes, as noted. Several States as noted pay collection and administration expenses out of State appropriations. In many cases a Stated lump sum is authorized for this expense to be deducted from motor vehicle fund, or else a certain percentage of gross receipts is allowed. Data on gasoline tax rates and collections are shown on page 52: This table shows the gross receipts and the nei receipts after the deduetion of refunds allowed by law. Some States have no provision for refunds and in such cases gasoline used for purposes other than motor vehicles is included. Collection costs are shown where paid out of gasoline tax earningsPUBLIC ROADS Vol. 7, No.2 Motor-rehiele registrations for the year 1926 "special tst of | caer susdamsauatip ik, | SERRE pant eed pene cept stam | wala teas, | pam meters Bhs olay | “oad i ws al | =I ‘Alabama. 194, 580 us| 270 mas 4A | ze i ne | ti] as Midas. 2 att S| ahs | 2.201 do | 2 238) 8] Sas “he. vagina ono | So foe Ein Peer sity Fe) ie Ee oS iets | ion 5 Es | 1h 8) aE isan | 4,408 | va Ae atl gap VTS | PE | a5 Other astred| Terese oil re oto eyes ‘umbels oa ree td ‘nit (sues) Per rapa tata | cet teabtere | Son sdduss, Wa Moje ‘eh gear sae | iat BBE RESSEERoneBEe bs ‘ ae wesoese! “a a Ee Be ae He aca ie a i otal ae ve, WELT | 17,512,855 1a, 34817, 400) 79,520, Vai, s03,677 | a4 pg st foe sere ner tha cei ole, ‘ass an a pe taht ul omnes, Ragen areas ture oe of hese tae fog Bie Reaaporatin.” he abide ae includ is the wand ‘orale Nor (glade the grand cout pvon la ube rst enum. ‘Total trmetypobiie has lem reba to ee ob oud tractors to ve a oll wc he Gt tine alae ond tae )Abreported Ey soar ves nee | Tocadss 72% Pate Serving Corporation vines whieh are tax exempt. ‘helices nd traters ore ent nad. * acl Gass es oporied by Stes. ‘Taciacea with opertor ieee, ‘Gal aa a July 110 Dee 3) reported, 2 Neue wth tg eek se rts Pusey by Mole wb ge a tat dows a" Bue Tasepraton “aig tris of ibe urea a Booger Beprtient vests thApel, 1026 PUBLIC Receipts from motor vehicle registration fers, te, jor the your 1923 Subsison of eistrasion cepts? ‘States ross, sa ah a False pte nae SSE So | oes eae a ey = Tange 6, i a ra a i epee Ha is ee Rae 18 pig "Sera dee 2S ARS te - is ad be # Haas es UR Sts i Bae SBS 255 GSESEES: Ross 5h ino peice SEB tue 2s okt nem ea SO ERT tte) Het | Sas 5 S120 $5,200 +5105, st] Hs 0S], 135,825) #310, 0 ail “a a BO a: eer ii 1g Se ee ES ty ke & iti BG RE RE Da Re Ss a uy ee RSG SRE EA Baas Rta ging fei ne ae BA ERA So tent hate siti tH ae iGimoid rein So is 5a, i rae errs whee Dacated total + [754 02,509 (Grand total... 38 05,583 Tui unis ed tom geen of mares Indain een “nig abe B Staves nn tho District of Column reporedata in full. deta, ffeil dea an See te dea here he eae oa fs Buea an Fee aes 4 Assur for eet tai shure aotied to Sate gener fun. ete wo [Xcel Trae ofa expenses, deduced tora county's srw of et reels ‘esl ine gril ropa Soe Maroy od BR re eee ‘Efpinan hi rotor nei ow necemen 2a Exnenes moter-velce cet apattment M Hsmted ot io pat om Sate appropriation 3 Iosiaed ae motor att. aie 1 Tol edge commas, a es ry tng sin pet rn “i'Fur period of anti, Jay ite December eatin sar bein ab Sail ea FE PRN nth Fe Bern tt yam ate TAS Re Roe aka? Stee 3 fohuis 7 reel by anendiment to law and 54 to Sate soi ela cm a Pentecost ae Heer eee TE SESSLER er, ratamem sre "NehAtmouat eo baaace with gros recelpe. ‘The United States appropriations for aaa aPUBLIC ROADS Vol 7, No.2, April, 1926 Gasoline tazes for the year 1925 | ispastion af total tax araings ‘Taxrates 15 aiauctiae | oiksade Psomptian| Total tax | ra Nexans of| Etited so Segment | adie Tor Sod ueed by | dead aactby -mator vebicles seotoe vehlles ‘Construction an maite- ‘nos of url roads Hes | Dateot ae San Dae ioe | Pas ishrays iB Beste shraleseugen a F a6 822 ee i id id Hd ii i]t i} id dares ta i]? pee ae 5 iG Eeece ge ues = iG ie oS aa a ii Hees Bao Ei ii seer Sti a8 id Nebr: 9 Sian ST ia a) i iia] ee) te bee Bd Res Seep! "Nope... oo Gig | Seo ii SHG if i Ete om si ii Hoses ee Ma 4 ie os o Ha e ce guee ne e ili 1 e ii = ae 2 it = is ii e te {Seas He j yee a8) 188 Ae ius ne tee” ae Wyoming. 40r436 207 1} a Vitus BG HE ‘BE td Teta een Sonn NOOB rao 2 | 7a,70 | TyeeKeE| Av aaa | Ha talign HR cited a moter oi tae at arate of 10 oats per gallo. ‘elade 80 pnomeets pu oouny fo tn ge Bonds "4 Payment io countas State ard igh ways i Ror mwalt ia Hartasn Coutts 1 Far Suto geal und, {Delinem eiasioaestod no tsporsbe tn 18, 1 Aatatenas of rip! stret, Postite tear. {i llelade ingen pat deli ates of omer yore ‘ fscenins potable doliangons toe, 1s Covers pat ¢ Beat four tate of yoo ony, as oo 1 exces Stat goose {a nee to rats eve Pobre H, 1036 tugd fom tin ta fund 1 Far mttcengen oly ‘eels sed nd, ‘agus ota, 1 For matntenane of Dlunore ras 2 fod Hoos for nor ait rement payroests on State rat bonds eth fot oo akg id on State red ns ladon San payment of ton Silo oad bond 3 impeosoaal abd ent of Wesaglon 3 osROAD PUBLICATIONS OF BUREAU OF PUBLIC ROADS Heater Raa A i ANNUAL REPORT Report of the Chief of the Bu of Public Roads, 1924, Report of the Chiet of the urea of Public Roads, 1925, DEPARTMENT BULLETINS. No. 105. ents in Dust Prevention Progress Report of Expe Hal Besryaion, 1 Road Models. Progress eport of Exporiments in Dost Provention ‘tad Road Preserestion, 10 4314. Methods for the Examination’ of Bituminous Road Materials. 10 #317, Methods for the Determination of the Physical Properties of Road-1 The Results of Physical Tests of Road-Building Rock. be. Public Road Mileage and Revenues in the Middle 386. ‘Atlas . 387. Publie Road Mileage and Revenues in the Southern States, 191. 888. Public Road “Mileage and Revenues in the New England States, 1914. 300, Public Road Mileage and Revenues in the United States, 1014, A Sununary. 407. Progress’ Reports of Experiinents in Dust Prevention ‘and Koad Preservatinn, 191 Barth, Sand-Clay, and Giayel Roads, 15. ‘The Bspansion and Contraction of Conerete and Conerete Roads. 10e.. he Healt of Phvsical Rests of Road Bring, Rock pression ‘Tests. Se 1916, Inching all a aaa ec 100. 285 of Road-Buiding Rock Hea A Bea eee eee oe sein aici Sa een [ieee al sugar cn ol hey Bre cen ar See Le Sr SAaL Are casa pices eB comaahs Sr Fada a son, nie for Bitumfnous Road Mate- Rock 1259. Mileage, Ineome and Expenditures, DEPARTMENT CIRCULARS, No. Bt ENT as Blasting Ksplosve, BBL Standard” Specifications for’ Corragated Calvert * Metal Pipe MISCELLANEOUS CIRCULARS 60. Fedoral Legislation Providing for Federal Aid in High- ‘way Construction FARMERS’ BULLETINS. adam Roads. be SEPARATE REPRINTS FROM THE YEARBOOK, S727. Design of Public Roads. 5 #750; Federal Aid to Highways, 19 *Si9, Roads. O11, Highways and Uighway ‘Transportation OFFICE OF PUBLIC ROADS BULLETIN +45, Data for Use in Designt ridges. “(1913.) 1. Culverts and Shortspan OFFICE OF THE SECRETARY CIRCULARS No. 49. Motor Vehicle Registrations and Revenues, 1914. 59. Automobile Registrations, Licenses, and. Htevenues in the United States, 1918, way Milvage aid Expenditures to January 83. 8 1, 1916, Wideh of Wagon ‘Tires Recommended for Loads of Varying Magnitude on Earth and Gravel Roads. So Licenses, and Reventies i ist. Riles and Regulations of the Secretary’ of Agriculture for Carrying out the Feder! Highway “Act and ‘Amondmonts Thereto. REPRINTS FROM THE JOURNAL OF AGRICULTURAL RESEARCH Effect of Controllable Variables Upon ‘the Peueiration Test for Asplialts and Asphalt Cements, Vol. 17, D2. Vol. 5, No. 20, D-s, Apparatus for Meisuring the Wear of Concrete Toads. Vol 5,No.2,D-6. A New Penetration Needle for Use in “bituminous Materials Val. 10, No. 5, D-12. of Grading Augregate Used ment Conerete Road Construction. Vol, 10, No. 7, D-18. Te Aggregates Vol. 11, Ne: 10, Deis: ‘rete Slab Subjected to Eeceatric Con- contrated Loads.UNITED STATES DEPARTMENT OF AGRICULTURE BUREAU OF PUBLIC ROADS STATUS OF FEDERAL AID HIGHWAY CONSTRUCTION AS OF MARCH 31,1926 STATES FI -ARS 1017-12 ISCAL 26 SCAL YEARS 1017-1925 FISCAL YEAR 192 FEDERAL - — —| a Fix, ‘mo}EerS comrteren prion vo | PRosecTe comuerro smc |. ppoyscys yyoen cONSTRUSTION ‘Projects arrzoven ror | AVAILABLE sexes ress PRO nh ‘cosstroenox ‘ror aE RouecTS BALANCE OF | states | | Sihek| Beles! Ba ike. |