rjfa% Iff? fALE uiniVERSITV OCT 21W20 LIBRARY EXPERIMENTS WITH AVAILABLE ROAD-MAKING MATERIALS OF SOUTHERN WISCONSIN. ELLSWORTH HUNTINGTON. From the Transactions of the Wisconsin Academy of Sciences, Arts, and Letters, Vol. XL {Issued September, 1897.] I I 2 £i$ EXPERIMENTS WITH AVAILABLE ROAD-MAKING MATERIALS OF SOUTHERN WISCONSIN. ELLSWORTH HUNTINGTON. The subject of road-making has been almost entirely neglected in this country in so far at least as its geological features are concerned. Some attention has been paid to engineering prob lems in their relation to roads, but almost nothing has been done to find out the value of different materials in the construction of a cheap yet serviceable surface on our common roads. In most parts of the country the only practicable way of road-making which will be good at all times of the year is to put a crust of crushed stone from six to twelve inches thick on a foundation madeof?" whatever soil the country happens to furnish. The most important conditions which the road-material must satisfy are cheapness. harii&Qss or capacity for resisting the wear of horses and of wheels, readiness in cementing into a compact mass, and ease and cheapness of repair. The necessity of cheapness is of course the most important factor. But it must always be borne in mind that the cost of a road built of any given material includes not only the first cost but also the cost of keeping the road in good repair. The in terest on the first cost of a durable but expensive material has to be compared with the extra annual cost of repairs where a less durable and cheaper stone is used. This necessity of cheapness forbids the transportation of large amounts of material for great distances except for the most important roads. Since the glacial drift and the various bedded rocks of the Silurian are the only available material in southern Wisconsin and in a large number of the other states of the Mississippi Valley, it is prob able that in view of the recent increase of interest in good roads 250 Huntington — Boad- Making Materials. they will be used in the construction of thousands of miles of roadway within the next half century. Under such circum stances it is well worth while to investigate the most econom ical and efficient ways of using the materials at hand. With this end in view a series of experiments was begun last fall in the laboratories of Beloit College, the results of which, as far as they have proceeded, are given below. Experience has shown that almost all limestones are too soft to make good or economical roads. Although they cement readily they quickly wear away producing a disagreeable mud and dust and costing a great deal to keep in repair. The glacial drift, on the contrary, consists in considerable degree of pebbles which are much harder than ordinary limestone but do not ce ment readily. This hardness is due to the fact that in the process of glaciation the soft parts of any rock are ground to powder and only the hardest parts are left as pebbles. In most rocks the hardest parts are those which contain the most quartz, and in the case of limestone the pebbles which remain in the drift contain a great deal of infiltrated silica. While quartz gives hardness to a stone, it is a very poor cementing material. If we can find some way of firmly cementing the drift we shall be able to construct good and cheap roads. In the experiments car ried on at Beloit a few tests were made with Trenton limestone, but in most cases the drift was the basis of work. The pebbles were crushed in the way usually recognized as best for macadam roads, namely into fragments of various sizes, the largest not to exceed one and one -half inches in diameter, and the very fine material screened off. From six to ten pounds of the crushed gravel was taken and to this was added a cement of powdered rock sufficient to firmly bind the mass of gravel. The cement ¦consisted of several kinds of stone or of mixtures of the various kinds. After the crushed pebbles and the fine material were put together in a box and thoroughly mixed, the whole mass was wet down and pounded and rolled and then allowed to dry. Th is process was repeated several times until the whole became one -solid mass. When this was thoroughly dry it was broken to pieces by allowing to fall upon it a weight so shaped as to give .a blow as nearly as possible like that of a horse's hoof. In this Tests Made. 251 way the surface was broken to pieces five or six times and the average number of blows required to break up the different mixtures was taken as representing their relative surface strength, the highest being scaled as one hundred. Then the same process was repeated with the difference that this time the whole depth of the mass, about two and one-half inches, was broken up into its original condition of loose gravel and fine material, and the results scaled in the same way as before. The average of these two ways of testing may be taken as indicat ing approximately the value of the different materials and com binations for use as cementing material on roads. The accom panying table shows the results obtained. Coarse material. Cement. Strength of surface, Strength of whole Value. 10II 12 13 14 Gravel Gravel Gravel Gravel Gravel Limestone Gravel Gravel Gravel Gravel. . . . LimestoneGravel Gravel. Gravel Diabase J£, blue lime stone y% Diabase %, sand % Fine granite J£, blue lime stone y% Fine granite yz, sand J^ . Diabase Blue limestone Fine granite Medium granite J£, blue limestone y% Coarse granite J£> blue limestone J£ Gravel Buff limestone Medium granite Sand Coarse granite 100. 81.6 38.3 48.9 52.0 58.8 51.653.3 100. 75.586.5 75.672.258.8 60.1 55.6 45.3 62.2 43.4 44.0 48.6 38.6 42.4 44.8 45.0 27.8 25.9 28.7 100. 78.662.462.3 62.1 58.855.9 54.5 53.9 43.743.643.636.427.3 With the exception of the two kinds of limestone all the stones used in the experiments were selected from the glacial drift. The diabase was a compact dark green or nearly black rock. The granite of No. 3 was a fine-grained dark variety containing a large proportion of horn blend, no mica and but little quartz. Most of the crystals were not a sixteenth of an inch in di ameter. The granite of number 8 was a little coarser, with crystals ranging up to an eighth of an inch in diameter. It contained rather more quartz and less hornblend than 252 Huntington — Road-Making Materials. the preceding. The other granite was quite coarse grained and contained a small amount of hornblend with a large pro portion of quartz. The gravel used as cement was taken just as the pebbles happened to come and represents the average gravel of Southern Wisconsin, The sand was taken from the drift and had a little clay intermixed. The limestone was of the Trenton variety, the buff being from the upper layer and of average hardness. The blue limestone was from a shaley layer and is the most compact and durable part of the Trenton lime stone. It is, however, decidedly softer than the average peb bles of the drift. If the materials used as cement in numbers ten to fourteen, viz., gravel with a value of 43.7, buff limestone with a value of 43.6, medium granite 43.6, sand 36.4, and coarse granite 27.3, could be placed on the road at equal cost, it would evidently be folly to use either very coarse granite or sand as far as bind ing is concerned. In many places where coarse granite can be secured cheaply it may, of course, be profitably used as the major part of the macadam if some first-class binding material is added. The other three materials, gravel, quite fine-grained granite, and rather soft limestone show the same cementing value of 43. The limestone is so soft that it can never be used profitably. Granite is usually harder than gravel, and also, for use as the main material of a road, it has another decided ad vantage. The results of a few experiments in which crushed stone was used, and not crushed gravel, bring out this disad vantage of gravel. In ordinary crushed gravel only about two- thirds of the surface consists of fresh fractures, and these alone cement readily. The other third is rounded and worn and not only does not cement readily, but also by its roundness tends to be thrown out and to make the whole mass of road material less firm and solid. This is an evil which cannot be avoided if gravel is used, but it may be lessened by rejecting a greater proportion of the smaller stones. It would be worth while to transport good granite at quite an expense by rail even if good gravel were close at hand. Numbers 8 and 9, which are mixtures of medium granite and limestone, and of coarse granite and limestone, show about Results of the Experiments. 253 what would be expected, their values being 54.5 and 53.9. They prove that with stones like granite, which have a low cementing power, a good quality of limestone may be of con siderable value as a binder. The diabase of number 5 with a value of 62.1, the limestone of number 6 with a value of 58.8, and the fine granite of number 7 with a value of 55.9 show nearly the same value as binders ; but in actual practice lime stone is much inferior to the others because of its low coefficient of wear, which is less than half that of diabase. Being rapidly reduced to a powder, it is washed away by heavy rains or blown away as a disagreeable dust. On this account roads on which it is used require constant repairs and its first cheapness is offset by this later expense. Numbers 1 to 4 are rather a sur prise since they indicate that the cementing power of mixtures is higher than that of stone of a single kind. The cements and values of these four are as follows: diabase and limestone 100, diabase and sand 78.5, fine granite and limestone 62.4, fine granite and sand 62.3. In actual practice, if diabase were used as a binder for the upper part of a macadam road made of gravel, the cementing value would be 85 or 90 and if granite were used the value would be about 62. This means that a road whose surface was half gravel and half diabase would oppose twice as much resistance to the breaking action of horses' feet as would one made wholly of gravel. The results of the experiments which have been described above are merely preliminary to future work and rest upon too small a body of data to be taken as final. Future and more ac curate tests may and undoubtedly will give results which differ widely from those here given. There is urgent need for ex tended investigations along this line. Many towns are begin ning to build macadam roads and if it be true, for instance, as these experiments indicate, that by spending fifty per cent, more for the transportation of basic rocks like diabase for use in con nection with drift material in building the upper parts of the macadam, a road can be built which will last twice as long as one made of pure gravel, it is time the fact were known. The cost of building one mile of gravel macadam road fifteen feet wide and one foot thick is about $3,500 under favorable condi- 254 Huntington — Road-Making Materials. tions. If instead of using gravel alone, we use it only for the lower eight inches and use a mixture of one-half gravel and half dia base for the upper four inches, the cost per mile would not ex ceed $4,800 in any part of Wisconsin. The interest on the ex tra $1,300 at 5 per cent, would be $65, but the surface of the road would be^twice as durable and the cost of repairs instead of being two hundred dollars per mile per year would be only one hundred and there would be an actual saving to the tax payer. Beloit, Wis 1803